Perlecan Mediates the Antiproliferative Effect of Apolipoprotein E on Smooth Muscle Cells

Apolipoprotein E (apoE) is known to inhibit cell proliferation; however, the mechanism of this inhibition is not clear. We recently showed that apoE stimulates endothelial production of heparan sulfate (HS) enriched in heparin-like sequences. Because heparin and HS are potent inhibitors of smooth muscle cell (SMC) proliferation, in this study we determined apoE effects on SMC HS production and cell growth. In confluent SMCs, apoE (10 μg/ml) increased 35SO4 incorporation into PG in media by 25–30%. The increase in the medium was exclusively due to an increase in HSPGs (2.2-fold), and apoE did not alter chondroitin and dermatan sulfate proteoglycans. In proliferating SMCs, apoE inhibited [3H]thymidine incorporation into DNA by 50%; however, despite decreasing cell number, apoE increased the ratio of35SO4 to [3H]thymidine from 2 to 3.6, suggesting increased HS per cell. Purified HSPGs from apoE-stimulated cells inhibited cell proliferation in the absence of apoE. ApoE did not inhibit proliferation of endothelial cells, which are resistant to heparin inhibition. Analysis of the conditioned medium from apoE-stimulated cells revealed that the HSPG increase was in perlecan and that apoE also stimulated perlecan mRNA expression by >2-fold. The ability of apoE isoforms to inhibit cell proliferation correlated with their ability to stimulate perlecan expression. An anti-perlecan antibody completely abrogated the antiproliferative effect of apoE. Thus, these data show that perlecan is a potent inhibitor of SMC proliferation and is required to mediate the antiproliferative effect of apoE. Because other growth modulators also regulate perlecan expression, this may be a key pathway in the regulation of SMC growth.

Heparin and HS are biologically active glycosaminoglycans (GAG) composed of alternating residues of uronic acid (glucuronic acid in HS and iduronic acid in heparin) and glucosamine (12). HSPGs have several vasoprotective effects; the best characterized among these is their ability to inhibit SMC proliferation (13)(14)(15). Although the antiproliferative effect of apoE has been realized for many years (16 -18, 10), how apoE inhibits cell proliferation is not known. Our previous studies showed that apoE increased HSPG production in endothelial cells (11). Thus, it raises the possibility that apoE-mediated inhibition of SMC proliferation may be due to its ability to induce HS production in cells. In the present study, we show that apoE stimulates SMC production of perlecan HSPGs, which mediates the antiproliferative activity of apoE. MATERIALS (19) by heparin-agarose chromatography. ApoE2 and apoE4 isoforms were from Calbiochem. Perlecan antibody was from Zymed Laboratories Inc. (South San Francisco, CA).
Cells-Rat and human aortic SMCs were kindly provided by Dr. L. Rabbani (Department of Medicine, Columbia University) (20). Data with rat SMCs are presented below. Initial experiments were also performed with human SMCs, and similar results were obtained. SMCs were grown in basal medium supplemented with growth factors, basic fibroblast growth factor and epidermal growth factor (Clonetics, San * This study was supported by a grant-in-aid and investigatorship from the American Heart Association, New York City Affiliate, by grants HL56984, HL62301, and HLK14 -03323 from the NHLBI, National Institutes of Health, and by a faculty research award (to S. P.) from the Long Island Jewish Medical Center. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Metabolic Labeling-PGs were radiolabeled with either [ 35 S]sulfate or [ 3 H]leucine for the indicated time periods. Medium PGs were collected and purified by DEAE-cellulose chromatography (see below). Cell associated PGs were assessed by extracting cells with 50 mM Tris buffer, pH 7.4, containing 4 M urea, 1% Triton X-100, 0.1 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride. To study the effects of apoE, confluent SMCs were incubated in culture medium containing 35 SO 4 and the indicated concentrations of apoE for 24 h. Cell and medium PG levels were assessed.
DEAE Cellulose Chromatography of PGs-To determine changes in PGs, DEAE-cellulose chromatography was performed as described previously (21,22). A DEAE-cellulose column was equilibrated with 50 mM Tris buffer, pH 7.4, containing 4 M urea, 0.1 M NaCl, 0.1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 1% CHAPS. The column was washed with the same buffer and with buffer containing 0.25 M NaCl, and PGs were eluted with the same buffer containing 0.5 M NaCl. Fractions containing radioactivity ( 35 SO 4 ) were pooled and dialyzed against minimal essential medium overnight and counted. To determine the relative proportion of HSPGs and chondroitin and dermatan sulfate PGs, an aliquot of the pooled fraction was incubated in 50 mM sodium acetate buffer, pH 5.2, with 1 unit/ml each of heparinase and heparitinase or with 0.5 units of chondroitin ABC lyase for 16 h at 37°C. The reaction mixture was precipitated either with 0.5 volumes of 1% cetylpyridinium chloride or with 3 volumes of ethanol to precipitate undigested GAG. Radioactivity in the supernatant and pellet was determined.
SMC Proliferation-To determine the effects of apoE or HS on SMC proliferation, cells were plated at low density (8 ϫ 10 4 /well) and cultured for 24 -48 h in the presence or absence of apoE or HSPGs. Cell number was counted with hemacytometer, and net growth was determined (15). Alternatively, SMCs were cultured in the above conditions; cells were then labeled with [ 3 H]thymidine for 6 h, and radioactivity incorporated into the DNA was determined by trichloroacetic acid precipitation of the cell lysate.
Determination of Perlecan Protein and mRNA-To determine changes in perlecan protein, control and apoE-treated (10 g/ml) SMCs were labeled with [ 3 H]leucine for 24 h (steady state). PGs were isolated from SMC medium and purified by DEAE-cellulose chromatography. Purified PGs were immunoprecipitated with an anti-perlecan antibody (100-fold diluted), and immunoprecipitates were analyzed by 5% SDS-PAGE. Perlecan (molecular mass, Ͼ550 kDa) was identified by autoradiography.
For Northern blotting, a 497-base pair polymerase chain reaction product representing domain I of perlecan (forward and reverse primers with sequences 5Ј-GCTGAGGGCCTACGATGG-3Ј and 5Ј-TGCCCAG-GCGTCGGAACT-3Ј, respectively) was generated by reverse transcription-polymerase chain reaction of endothelial cell RNA. Northern blotting of total RNA from control and apoE-treated (5 g/ml for 24 h) SMCs was performed using 32 P-labeled perlecan probe. RNA load was normalized by determining 18 S RNA.
Data Analysis-Results are expressed as mean Ϯ S.D. Experiments were done in triplicate and repeated at least once. Statistical analyses were performed by Student's t test to determine the significance of change. A significance difference was considered for p values equal to or less than 0.05.

ApoE Increases Sulfate Incorporation into SMC HSPGs-
Previous studies showed that addition of apoE increased HSPG production in endothelial cells but not in macrophages, which predominantly synthesize chondroitin and dermatan sulfate PGs (15). Similarly, HSPGs represent only ϳ25% of total PGs synthesized by SMCs. To determine whether apoE increased HSPGs in SMCs, confluent monolayers of SMCs were incubated with apoE (10 g/ml) for 16 h, and PGs in cellular and secreted pools were determined following DEAE-cellulose purification. In different experiments, the total PG level in cells was increased by 15-22% and in medium by 25-30% in apoEtreated cells (Fig. 1A). Because HSPGs in the media can act as inhibitors of cell proliferation, we determined HSPGs in control and apoE media. The heparinase-sensitive radioactivity, rep-resenting HSPGs, was increased by 108% in apoE-treated cells (Fig. 1B). The amounts of chondroitin and dermatan sulfate PGs, which constitute ϳ75% of total medium PGs, were not altered by apoE-treatment. These data show that apoE treatment of SMCs results in an increase specifically in HSPGs. As in endothelial cells (15), this increase was found to be primarily due to an increase in synthesis (not shown).
FIG. 1. A, apoE stimulates SMC PGs. Confluent monolayers of SMCs in 24-well plates were incubated with apoE (10 g/ml) in growth medium (basal medium containing 5% serum and growth factors) containing [ 35 S]sulfate (50 Ci/well) for 24 h under culture conditions. Sulfatelabeled PGs in the cells and media were determined after purification by DEAE-cellulose chromatography. Values represent mean Ϯ S.D. B, apoE increases SMC medium HSPGs. Purified 35 S-PGs from control and apoE-treated SMC media were treated with 1 unit/ml each of heparinase and heparitinase for 6 h at 37°C and precipitated with cetylpyridinium chloride. 35 S-Radioactivity in precipitate and supernatant were determined. The digested GAG in the supernatant represent HSPGs and the undigested GAG in the precipitate represent chondroitin and dermatan sulfate PGs. C, apoE stimulates [ 3 H]glucosamine incorporation into SMC medium HSPGs. To determine whether apoE increased HS GAG, PGs were labeled as in A, but with [ 3 H]glucosamine and incorporation into medium HSPGs was determined. Values represent mean Ϯ S.D. Compared with control, apoE increased [ 3 H]glucosamine incorporation into PGs by 3-fold, suggesting increased HS GAG. However, compared with data in B, the increase in HS GAG was higher than the increase in sulfation, suggesting that medium HSPGs are relatively under-sulfated in apoE-treated cells.
We also determined the effect of apoE on [ 3 H]glucosamine incorporation into PGs. ApoE increased [ 3 H]glucosamine incorporation into HSPGs by about 3.5-fold (Fig. 1C). Thus, the ratio of [ 3 H]glucosamine to 35 SO 4 in HSPGs was increased approximately by 1.75, suggesting that although HS GAG were increased by apoE, these HS are relatively under-sulfated.
ApoE Inhibits SMC Proliferation-The above experiments were done on confluent SMCs. We next determined apoE effects on proliferating SMCs. Previous studies showed that apoE inhibits SMC proliferation stimulated by serum or plateletderived growth factor (10). The growth medium in the current experiments contained serum, basic fibroblast growth factor, and epidermal growth factor. In different experiments, the addition of apoE to the medium inhibited cell proliferation by 45-55% (both cell number and [ 3 H]thymidine incorporation; Fig. 2A) in 24 h. This inhibition was greater than that previously observed with 25 g of apoE (10).
We examined whether apoE altered HSPGs in proliferating cells. Because the cell number is decreased by apoE, by comparing the ratios of 35  HSPGs from ApoE-treated Cells Are Potent Antiproliferatives-We first examined whether apoE-treated cells contained antiproliferative substances in the medium. Conditioned medium was collected from control and apoE-treated SMCs, and apoE was removed by immunoprecipitation (Fig. 3, inset). These media were then added to subconfluent SMCs, and cell growth was determined after 24 h (Fig. 3A). Control conditioned medium (CCM) inhibited SMC growth by 18% compared with control medium. Conditioned medium from apoE-treated cells (ECM) inhibited cell proliferation by 51%, suggesting that apoE treatment stimulated the production of antiproliferative substances.
We next determined whether HSPGs in apoE-conditioned medium mediated inhibition of cell proliferation. PGs from control and apoE-conditioned media were purified by DEAEcellulose chromatography. This procedure also removed any remains of apoE from the medium (not shown). Equal amounts ( 35 SO 4 cpm) of purified PGs were then added to subconfluent SMCs in growth medium, and cell growth was determined. PGs from apoE-treated cells (Fig. 3B, E-PG) inhibited SMC proliferation to an extent similar to that of apoE. PGs from control cells (Fig. 3B, C-PG), although at a similar level, inhibited SMC proliferation by only 18%. These data suggest that HSPGs from apoE-treated SMCs are antiproliferative.
ApoE Inhibits SMC Proliferation Stimulated by Lysolecithin-We previously showed that lysolecithin and oxidized low density lipoprotein treatment decreases extracellular HSPGs (23,24), and others have shown that these agents stimulate SMC proliferation (10,25). We therefore determined whether apoE ability to increase HSPGs would block lysolecithin effects on SMC proliferation. Incubation of SMC with lysolecithin decreased 35 SO 4 incorporation into PGs by 36% (not shown). Concomitant with this decrease, lysolecithin increased [ 3 H]thymidine incorporation into DNA (Fig. 4). Lysolecithin effects on SMC proliferation were completely abolished in the presence of apoE. These data suggest that agents that modulate HSPGs influence cell proliferation.
ApoE Increases Perlecan Production in SMCs-We next characterized the antiproliferative HSPGs in apoE-stimulated cells. Perlecan is the major HSPG secreted by vascular cells (26). To determine whether apoE increased perlecan secretion,

FIG. 2. ApoE decreases SMC proliferation (A) but increases
HSPGs per cell (B). Subconfluent SMCs (ϳ35-40%) were incubated with control medium containing 35 SO 4 (50 Ci/ml) or medium containing apoE (10 g/ml) and 35   DEAE-cellulose-purified, [ 3 H]leucine-labeled (core protein-labeled) HSPGs from control and apoE-treated cells were immunoprecipitated by anti-perlecan antibody and analyzed by SDS-PAGE and autoradiography (Fig. 5). The radioactivity associated with a protein of M r ϳ550,000 (perlecan has a core protein of ϳ400,000 containing three HS chains of M r ϳ50,000 -70,000) was increased by apoE. Concomitant with protein increase, apoE also increased perlecan mRNA by greater than 2-fold. These data suggest that the antiproliferative HSPG in SMC medium is perlecan.
Effects of ApoE Isoforms-We next studied the effects of apoE isoforms to determine whether antiproliferative activity correlated with increase in perlecan HSPGs (Fig. 6). ApoE3, the most common isoform of apoE, showed maximum stimulation on perlecan production and inhibition on cell proliferation (45%). ApoE2 and apoE4 did not significantly increase perlecan HSPGs or inhibit cell proliferation. These data further show that the antiproliferative effect of apoE correlates with its ability to stimulate perlecan HSPGs.
The Antiproliferative Effect of ApoE Requires Perlecan-We next determined whether the antiproliferative effect of apoE is mediated by perlecan. Subconfluent SMCs were incubated with control medium or apoE medium containing nonspecific antibody or anti-perlecan antibody (Fig. 7A). Perlecan antibody did not affect cell growth under control conditions. ApoE inhibited [ 3 H]thymidine incorporation into DNA approximately by 48%. In the presence of perlecan antibody, this inhibition was reduced to about 9%. We performed the same experiment with human vascular SMCs (Fig. 7B). ApoE inhibited SMC proliferation by 71%. Perlecan antibody, however, under control conditions stimulated SMC proliferation by 30 -35%. The effect of apoE was completely reversed by perlecan antibody. These data suggest that perlecan mediates the antiproliferative effect of apoE both in rat and human SMCs. DISCUSSION HSPGs are thought to be important for blood vessel homeostasis, blood clotting, atherogenesis, and atherosclerosis. Atherosclerotic vessels have reduced HSPGs, and previous studies have shown that apoE-HDL increased endothelial HS, which in turn could decrease the occurrence of events related to atherosclerosis (11). ApoE was able to increase secretion of HSPGs in both endothelial cells (11) and SMCs (present study). Because subendothelial matrix HSPGs produced by apoE-treated endothelial cells showed strong inhibition of SMC growth, we postulated that actions of vascular apoE would regulate SMC proliferation in the subendothelial space (11). Our current data show direct effects of apoE on SMC HSPGs and thus identify a mechanism for the known antiatherogenic effect of apoE.
The present data strongly suggest that the antiproliferative effect of apoE is due to induction of perlecan HSPGs in SMCs. 1) HSPGs isolated from apoE-stimulated cells inhibited proliferation better than those from control SMCs. 2) ApoE countered the effects of lysolecithin, which is known to decrease extracellular HSPGs. 3) The antiproliferative effect of apoE isoforms correlated with their ability to stimulate perlecan HSPGs. 4) An anti-perlecan antibody blocked the inhibitory effect of apoE on SMC growth, showing that perlecan is required for the antiproliferative effect of apoE. Moreover, apoE did not inhibit proliferation of endothelial cells, which are not sensitive to HSPG inhibition (27) ([ 3 H]thymidine: control, 4780 Ϯ 270 cpm; apoE, 5540 Ϯ 330 cpm). Our data are also consistent with the observation that both apoE (28) and HSPGs (29) inhibit mitogen-activated protein kinase, a key signaling pathway in cell growth.
ApoE treatment increased both perlecan mRNA and protein. Perlecan, the major HSPG of endothelial cells and SMCs (26), consists of a core protein of M r ϳ450,000 to which three HS  5. ApoE increases perlecan protein and mRNA. SDS-PAGE of perlecan. Control (C) and apoE-treated (10 g/ml) SMCs were labeled with [ 3 H]leucine for 24 h. PGs were isolated from SMC medium and purified by DEAE-cellulose chromatography. Purified PGs were immunoprecipitated with anti-perlecan antibody and analyzed by 5% SDS-PAGE and autoradiography. A single band with a molecular mass slightly higher than 550 kDa (apolipoprotein B (molecular mass, ϳ550 kDa) was used as a marker) was observed, the intensity of which was increased in apoE-treated SMCs. No other bands were seen on SDS gels. ApoE increases perlecan mRNA. A 497-base pair polymerase chain reaction product representing domain I of perlecan was used for Northern blotting. Total RNA was isolated from control and apoE-treated (10 g/ml, 24 h) SMCs, and Northern blotting was performed using 32 Plabeled perlecan probe. A single band with a M r of ϳ15 kb was observed. Densitometric analysis (bars) showed that perlecan band intensity, expressed as a ratio of perlecan to 18 S RNA, was increased by apoE by more than 2-fold.
FIG. 6. Effects of apoE isoforms on perlecan production and proliferation. Subconfluent SMCs were incubated with medium alone or medium containing 10 g/ml of apoE isoforms (E2, E3, and E4) for 24 h. [ 3 H]Thymidine incorporation was determined. In another experiment, confluent SMCs were incubated with apoE isoforms in medium containing 35 SO 4 , and incorporation into medium PGs was determined as described in Fig. 1. chains with a molecular mass of ϳ70 kDa are attached at one end of the molecule. The core protein consists of five consecutive domains with homologies to molecules involved in control of cell proliferation, lipoprotein uptake, and cell adhesion. Perlecan core protein can mediate cell adhesion and interact with other matrix proteins, and it plays a critical role in matrix assembly. HS chains of perlecan can bind growth factors. Although in vitro all isolated HSPGs are effective inhibitors of SMC proliferation, the identity of the antiproliferative HSPGs in vivo is not known. Cell surface HSPGs are required for the mitogenic activity of several growth factors (12) and thus are unlikely to inhibit cell growth. Extracellular HSPGs, on the other hand, either by sequestering the growth factors or by other signaling mechanisms, can inhibit cell proliferation. Our data for the first time identify direct effect of perlecan on SMC growth and its requirement to mediate the antiproliferative effect of apoE.
Perlecan was shown to negatively correlate with SMC proliferation (30), and it was shown to inhibit Oct-1, a growthrelated transcription factor (31). In certain cell types, however, blocking perlecan production via antisense DNA inhibited cell growth (32)(33)(34). It is conceivable that perlecan under normal conditions is required for matrix assembly and cell growth; however, excess perlecan in medium that is not deposited into the matrix may block growth factor binding and activity. Support for this comes from the observations that perlecan core protein can bind cell surface integrins and support cell growth (35) and that serum induces perlecan production at early time periods but decreases at later time periods (36).
It is not clear why excess perlecan remained in the medium. It is conceivable that the amount of perlecan in the matrix is saturating, leading to accumulation in the medium. Alternatively, perlecan produced by apoE-stimulated cells is different. The data shown in Fig. 1C suggest that HS chains in perlecan are under-sulfated. Perlecan interaction with surrounding matrix proteins, such as laminin and collagen, requires both core protein and HS chains (37). It is conceivable that reduced sulfation affects perlecan HS interactions with laminin or other perlecan molecules, thereby reducing its ability to incorporate into matrix.
The antiproliferative effect of perlecan is likely due to the HS chains. Although it is not entirely clear how HS inhibits cell proliferation, several mechanisms have been proposed (29, 38 -40). We are surprised, however, by the observation that perlecan antibody, which reacts with domain III of perlecan, completely blocked perlecan effect. Domain III is thought to mediate cell adhesion (26), and attachment to the matrix and spreading is a key part of cell growth. Perlecan antibody, as shown in Fig. 7 in control conditions, either did not affect or stimulated SMC growth. However, when added during the seeding of SMCs, perlecan antibody inhibited SMC growth by Ͼ40% (not shown). We propose that perlecan in matrix is required for cell growth and that excess unincorporated perlecan may engage the SMC surface molecules involved in cell attachment and spreading. Studies have shown both adhesive and antiadhesive functions for perlecan (26,41,42), and the current studies may offer an explanation why this occurs.
Vascular cells produce a variety of growth promoters and inhibitors (43). Physiologically relevant agents that stimulate SMC proliferation include platelet-derived growth factor, thrombin, oxidized low density lipoprotein, and lysolecithin. Vascular cell derived growth inhibitors include transforming growth factor-␤, nitric oxide/cGMP, and apoE. Going through published literature on perlecan regulation, we identified an interesting possibility that perlecan may be the key for modulation of SMC growth. Platelet-derived growth factor (44), thrombin (45), serum (36), oxidized low density lipoprotein, and lysolecithin (23,24), which stimulate SMC growth, decrease perlecan. In contrast, the antiproliferative agents transforming growth factor-␤ (46), apoE (present study), and even heparin (47) stimulate perlecan expression. Thus, we postulate that modulation of perlecan is key to regulation of cell growth.
The observation that apoE up-regulates perlecan may have implications in other physiological and pathological processes. Perlecan is known to modulate angiogenesis (48). It remains to be determined whether apoE could be an angiogenic factor in vivo. ApoE induction of perlecan may also have implications in the pathogenesis of Alzheimer's disease. Brains of patients with Alzheimer's disease accumulate deposits of ␤-amyloid protein. The ␤-amyloid protein-containing deposits in the vessel wall are primarily associated with SMCs, endothelial cells, and the surrounding matrix, and studies showed that perlecan is associated with the ␤-amyloid protein deposits (49). It remains to be determined whether apoE can induce perlecan production in neuronal cells, and if this occurs, it is conceivable that production of soluble perlecan may compete for ␤-amyloid protein binding to matrix perlecan. Soluble HS-like molecules can inhibit amyloid progression in mice (50).
How apoE stimulates perlecan and what cell surface molecule(s) mediates apoE actions remain to be determined. Based on previous studies, both HSPGs and receptor-associated protein-sensitive pathways may mediate apoE effects on perlecan (11). ApoE-␤-very low density lipoprotein, which binds poorly to HSPGs (51), does not inhibit DNA synthesis (52), and RAP at high concentrations could affect apoE binding to HSPGs (53); thus, it is conceivable that cell surface HSPGs directly mediate apoE effect (54). It should be noted that demonstration of requirement for cell surface HSPGs in mediating the antiproliferative effect of apoE is difficult as agents that interfere with cell surface HSPGs, such as heparinase, heparin, and chlorate, independently inhibit cell proliferation (55,56). Cell surface syndecan is beginning to be recognized as a signaling receptor (57). Alterations in the phosphorylation state of syndecan may affect cell growth. It remains to be determined whether apoE alters syndecan phosphorylation and whether this is required for its antiproliferative effect. ApoE is also known to stimulate nitric oxide and cGMP production (9), which are antiproliferative (58). Although the role of this pathway in SMC proliferation was not determined (28), preliminary results showed that nitric oxide donor and cGMP can increase HSPG production in SMCs (not shown). Perlecan promoter has cAMP responsive elements (46). Thus, it is conceivable that increased cGMP or cAMP will induce transcription of perlecan mRNA through activation of specific transcription factors.
In summary, our data show that the antiproliferative effects of apoE are mediated by perlecan. We postulate that modulation of perlecan is a key step in regulating SMC growth. Factors that increase perlecan inhibit cell growth, whereas those that decrease perlecan stimulate cell growth.