INTRODUCTION

Recent evidence indicates that arachidonic acid (AA)¹ and/or its metabolites may be involved in regulating cellular proliferation (1-7). Arachidonic acid is released from cellular phospholipids (PL) via phospholipase A₂ (EC 3.1.1.4, PLA₂) hydrolysis of the acyl bond at the sn-2 position (8). Multiple forms of mammalian PLA₂ have been identified. The type IIa 14-kDa PLA₂ is well characterized and known to exist in both an extracellular form in inflammatory fluids (9, 10) and in a cell-associated form (11-14). The cytosolic 85-kDa PLA₂ is structurally distinct and, unlike the 14-kDa PLA₂, exhibits a preference for AA in the sn-2 position of PL and is regulated by physiological intracellular Ca²⁺ concentrations and phosphorylation (15-17). An 80-kDa calcium-independent cytosolic PLA₂ first identified in P388D₁ macrophages (18) and recently cloned from these cells (19) and Chinese hamster ovary (CHO) cells (20) possibly serves as a housekeeping enzyme involved in the remodeling of membrane phospholipids (21). We and others have shown that both the 14- and 85-kDa enzymes are induced by inflammatory cytokines and growth factors (22-29) and that both enzymes influence cellular AA release and subsequent eicosanoid production in a variety of cell types (8, 22, 30-32). However, the contribution of these distinct enzymes in regulating cellular proliferation has not been examined directly.

Proliferation of vascular smooth muscle cells (VSMCs) is implicated in the pathogenesis of hypertension, primary atherosclerosis, and restenosis following interventional revascularization procedures such as balloon angioplasty, arterial stenting, and by-pass surgery (33-35). Evidence suggestive of a role for PLA₂ in VSMC proliferation includes: 1) that exogenous administration of AA increases expression of the early response genes c-myc (36), c-fos (36), and c-jun (37), as well as activity of mitogen-activated protein kinase (38); and 2) that these effects as well as serum- and growth factor-induced proliferation are inhibited by nonselective PLA₂ inhibitors such as mepacrine (1, 37, 39).

VSMCs exhibit both 14-kDa and high molecular mass cytosolic PLA₂ activities (24, 40). Recent descriptions of hydrogen peroxide-stimulated DNA synthesis and expression of proliferation-associated early response genes (36-38, 41), and hydrogen peroxide- (42) and angiotensin II-mediated (43) phosphorylation of the 85-kDa PLA₂ infer that this enzyme may participate in AA release during cellular proliferation. Other data (44, 45) demonstrate that vascular smooth muscle cells possess high affinity Type I PLA₂-specific binding sites and that 14-kDa phospholipase A₂ is present in atherosclerotic plaques (46, 47), suggesting that the 14-kDa PLA₂ serves a function in processes related to vascular disease. But, whether the 14-kDa PLA₂ plays a role in cellular proliferation is not known. The purpose of this study was to directly examine the roles of the distinct 14- and 85-kDa PLA₂ enzymes in VSMC proliferation. Herein is presented the first direct evidence that PLA₂ activity is associated with VSMC proliferation. Neither the 14-kDa PLA₂ nor prostaglandin E₂ (PGE₂) appear to mediate VSMC proliferation. In contrast, the 85-kDa enzyme appears to serve a selective role in the initial production of arachidonic acid, which, through direct action or via action of metabolites other than PGE₂, is critical for VSMC proliferation.

EXPERIMENTAL PROCEDURES

Cell Culture

Human coronary artery vascular smooth muscle cells (hCAVSMCs, Clonetics, San Diego, CA) were plated at 3 × 10³ cells/cm² and grown in monolayer in the optimized culture medium recommended by the manufacturer (Clonetics SmGM2: modified MCDB131 containing fetal bovine serum (5%), insulin (5 mg/ml), fibroblast growth factor (2 ng/ml), epidermal growth factor (0.5 mg/ml), gentamycin (50 mg/ml), and amphotericin  (50 ng/ml)) at 37 °C in an atmosphere of 5% CO₂, 95% air and 95% humidity. Cells were passaged when 70-80% confluent via a 4-min treatment (37 °C) with trypsin-EDTA (0.01:0.02%) in calcium-magnesium-free phosphate-buffered saline containing (in mM): NaCl (137); Na₂HPO₄ (8.1); KCl (2.7); and KH₂PO₄ (1.5), pH 7.4. Cells were reseeded into 24-well plates, 150- or 600-cm² culture flasks at 3 × 10³ cells/cm². The medium was changed every third day, and these asynchronous cultures of randomly dividing cells were allowed to grow for various times as indicated. Our goal was to take advantage of the fact that these cells routinely demonstrate reduced DNA synthetic rate upon reaching confluency and contact inhibition within several days of reaching confluency. All experiments were done using cells from passages 5-7 and were done using two different hCAVSMC lines to rule out donor-specific results. Microscopic examination of cellular morphology was done at 40× magnification with an inverted phase-contrast microscope (Telaval 31, Carl Zeiss, Inc., NY) equipped with a camera port for photographic documentation.

Normal growth curves were determined by allowing the cells to grow for 14 days, which, based on prior experience in our laboratory, was expected to span the entire proliferative phase of these cell lines under the present culture conditions. At various times, the cultures were washed with calcium-magnesium free phosphate-buffered saline, harvested with trypsin-EDTA, diluted with culture medium, and counted by hemacytometer. In parallel wells, relative rates of DNA synthesis were assessed via [³H]thymidine incorporation (4-h pulse) into trichloroacetic acid precipitable material as described previously (48). [³H]Thymidine incorporation was adjusted for cell density at the time the pulse was given.

The effects of the 14-kDa PLA₂ inhibitor SB203347 (synthesized by SmithKline Beecham medical chemistry; 0-10 µM), the preferential 85-kDa inhibitor arachidonyl trifluoromethyl ketone (0-10 µM) (AACOCF₃, Cayman Chemical Co. (Ann Arbor, MI)), an iPLA₂ inhibitor haloenol lactone suicide substrate E-6-(bromomethylene)tetrahydro-3-(1-naphthalenyl)-2H-pyran-2-one (0-10 µM) (HELSS, Biomol Research Laboratories, Inc., Plymouth Meeting, PA), and the cyclooxygenase inhibitor indomethacin (1 µM) (Sigma) were determined by cell counts. Drugs or vehicle (ethanol, 0.03%) were added with fresh culture medium on day 3 after passage, and cells were incubated with the vehicle or drug for 3 days prior to determination of cell number. Triplicate wells were counted for each condition per each experiment.

For antisense studies, hCAVSMCs were plated in SmGM2 at 5 × 10⁴ cells/well in 24-well culture plates and allowed to adhere overnight. Lipofectin-mediated transfections with phosphorothioate oligonucleotides SB7111 (3-TACAGTAAATATCTAGGAATG-5, directed against the initiation site, 0-10 µM), SB9030 (5-ATGTCATTTATAGATCCTTAC-3, sense control, 0-10 µM), SB7222 (5-TTACCGCGCCGTAGACGGGCA-3, scrambled nonrelevant control, 1.0 µM), or Lipofectin alone (5 µg/ml, Life Technologies, Inc.) were done in 1 ml of growth factor-free medium as described previously (31, 49). After 18-24 h, 1 mL of 2 × SmGM2 (final concentration, 1×) was added to the transfection medium to stimulate cellular proliferation. In some studies, cells received AA (5,8,11,14-eicosatetraenoic acid, 20 µM, Sigma). Additional wells were used to examine the effect of AA alone, and the effects of 3 µM AACOCF₃ (positive control) ± 20 µM AA. After 3 days of continual incubation, cell viability (trypan blue exclusion) and cell number were determined as described above.

Smooth Muscle Subcellular Fractionation

Preconfluent (day 3; i.e. actively proliferating), near or at confluent (~day 8), and postconfluent (days 10-14) cells were harvested by trypsinization and centrifugation. The smooth muscle cell pellet (2.6-5.7 × 10⁷ cells) was resuspended to 1 × 10⁸ cells/ml of homogenization buffer containing 0.34 M sucrose, 10 mM HEPES, pH 7.4, 1 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 200 µM leupeptin, 20 µg/ml soybean trypsin inhibitor, and 20 µg/ml aprotinin at 4 °C. Inclusion of EGTA localized the 85-kDa PLA₂ predominantly in the cytosolic fraction (50, 51). The cell suspension was disrupted by nitrogen cavitation (450 p.s.i. for 15 min at 4 °C), and the homogenate was centrifuged at 400 × g for 10 min at 4 °C to remove unbroken cells and debris. The resulting supernatant fraction was centrifuged at 100,000 × g for 60 min at 4 °C to obtain a supernatant (cytosolic) and particulate (microsomal) fraction. The microsomal pellet was resuspended in homogenization buffer, and both fractions were flash frozen with liquid N₂, and stored at 70 °C until analysis.

Preparation of Purified Human Phospholipase A₂ Enzymes

Recombinant human Type II 14-kDa PLA₂ (rh Type II 14-kDa PLA₂) was cloned from a human placenta library, expressed, and purified as described previously (52, 53). Recombinant human cPLA₂ (rh 85-kDa PLA₂) was prepared using a U937 85-kDa PLA₂ cDNA subcloned into the baculovirus vector pAcCL29 and expression in Spotoptera frugiperde (SF21) cells as described previously (54).

Phospholipase A₂ Enzyme Assay

Phospholipase A₂ activity of hCAVSMC subcellular fractions (50-100 µg of protein/assay) was measured by the acylhydrolysis of [³H]AA-Escherichia coli (NEN Life Science Products) or [1-¹⁴C]palmitoyl-2-arachidonyl phosphatidylcholine ([¹⁴C]AA-PC, Amersham Corp.) as described previously (14). The assay was initiated by addition of substrate, and assays were incubated at 37 °C for a time predetermined to be on the linear portion of a time versus hydrolysis plot. Specific activity (picomoles of free fatty acid hydrolyzed/minute/milligram) was determined from percent hydrolysis values.

Immunoblot Analysis

Human CAVSMC fractions (20-50 µg of protein as indicated) and recombinant enzymes used as standards were analyzed by SDS-polyacrylamide gel electrophoresis (10% gels, Bio-Rad). Proteins were transferred to nitrocellulose paper, incubated with rabbit polyclonal antiserum (NS1; 1:500-1:1000) against rh 85-kDa PLA₂ (30, 54), and then incubated with donkey anti-rabbit IgG conjugated to horseradish peroxidase (1:5000) (Boehringer Mannheim). Immunoreactive bands were detected via chemiluminescence (ECL Western blotting system, Amersham International, Arlington Heights, IL).

Quantification of Type II 14-kDa PLA₂ by ELISA

Mouse anti-rh Type II 14-kDa PLA₂ monoclonal antibodies were prepared as described previously (30, 31) and demonstrated no cross-reactivity with either type I pancreatic 14-kDa PLA₂, 85-kDa PLA₂, albumin, or various inflammatory mediators such as tumor necrosis factor, interleukin-1, or platelet activating factor. Microtiter plates (Nunc Immuno Plate Maxisorp F96, Roskilde, Denmark) were coated with the monoclonal antibody SK088-3C6.16.2 (100 µl, 2 µg/ml) in 50 mM sodium phosphate, 150 mM NaCl, 0.02% NaN₃, pH 7.4, for 18 h at 4 °C. The plates were washed 4 times with 10 mM Tris, 150 mM NaCl, 0.02% NaN₃, 0.05% Tween 20, pH 7.4) and then blocked with 200 µl of 1% bovine serum albumin in 50 mM Tris, 150 mM NaCl, 0.02% NaN₃, pH 7.4, for 5 min at 37 °C. Smooth muscle subcellular fractions and the corresponding standards (50 µl) were diluted in assay medium (SmGM2) and coincubated with 50 µl of conjugate (2 µg/ml biotinylated monoclonal antibody SK097-1E8.5.2) for 1 h at 37 °C. The plates were washed 4 times with wash buffer, followed by the addition of 100 µl per well of the streptavidin-alkaline phosphatase conjugate (diluted 1:2000 in streptavidin buffer, 0.5% bovine gamma globulin, 50 mM Tris, 150 mM NaCl, 0.02% NaN₃, 1 mM MgCl₂, pH 7.4) and incubated for 30 min at 37 °C. The plates were washed again, followed by the addition of 100 µl/well of substrate (p-nitrophenyl phosphate, 1 mg/ml) and incubated for 30 min at 37 °C after which the plate was read at 405 nm (M_r 7000 plate reader) (Dynatech Laboratories Inc., Chantilly, VA). Purified rh type II 14-kDa PLA₂ was used to generate a standard curve (0.1-4 ng/50 µl). Standard curves and results were calculated using Delta Soft v2.12 (Biometallics Inc., Princeton, NJ).

RNA Isolation and Northern Blot Analysis

Total RNA was isolated from 150-cm² flasks by the method of Chomczynski and Sacchi (55) and resuspended in diethyl pyrocarbonate-treated sterile water. RNA concentrations were determined by spectroscopy (A₂₆₀), and samples were stored at 80 °C. Northern blots were prepared by capillary transfer and UV fixation of electrophoretically separated (1.2% agarose/formaldehyde gel) RNA (10 µg total RNA per lane) to nylon hybridization membranes (Amersham Hybond N). Blots were hybridized (10% dextran sulfate, 1 M NaCl, 0.1% SDS, 0.1 mg/ml denatured herring sperm DNA, 65 °C) overnight with 10⁶ cpm/ml of the indicated probe and washed sequentially as follows: 20 min at room temperature in 1 × SSC, 1 mM EDTA, 0.1% SDS, 0.01 M NaHPO₄; 20 min at room temperature in 0.2 × SSC, 1 mM EDTA, 0.1% SDS, 0.01 M NaH₂PO₄; 2 × 30 min at 65 °C in 0.1 × SSC, 1 mM EDTA, 0.1% SDS, 0.01 M NaH₂PO₄. The washed blots were exposed to phosphorscreens (Molecular Dynamics, Sunnyvale, CA) overnight. Probes were labeled by random priming (Promega, Madison, WI) and incorporation of radiolabel (Redivue [-³²P]dCTP, 3000 Ci/mmol, Arlington Heights, IL) to a specific activity of 1 × 10⁹ cpm/µg of cDNA. The 85-kDa PLA₂ probe was the full-length cDNA (2.87 kilobases). The glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probe was generated by CLONTECH (Palo Alto, CA) PCR primers.

Flow Cytometry

Apoptosis was measured by TUNEL (56, 57), using the ApopTag kit from Oncor (Gaithersburg, MD). In brief, the enzyme terminal deoxynucleotidyltransferase extends the DNA fragments with digoxigenin-containing nucleotides, which are then detected with a anti-digoxigenin antibody carrying fluorescein to allow detection by fluorescence (494-nm excitation, 523-nm emission). Propidium iodide was used as a counterstain to measure total DNA content and to determine distribution of cells in G₀/G₁, S, and G₂/M phases of the cell cycle. Flow cytometric analysis was performed on a Becton Dickinson FACScan instrument using CellQuest software.

Eicosanoid Measurement

PGE₂, PGF₂, leukotriene B4, and leukotriene C4 levels in cell-free medium were directly measured using enzyme immunoassay (Cayman Chemical Co.). Sample or standard dilutions were made with SmGM2 and analyzed in triplicate.

Protein Determination

All protein concentrations were determined by Bradford protein analysis kits (Bio-Rad).

Data Analysis

Data are expressed as mean ± S.D. or standard error of the mean (S.E.) as indicated. Each data point represents n = 2-3 unless otherwise stated. Individual statistical comparisons of paired data were evaluated by Student's t test with p < 0.05 representing significance.

RESULTS

Normal Growth of Human Coronary Artery Vascular Smooth Muscle Cells

Cells were plated at 3 × 10³ cells/cm² (6000 cells/well) on day 0 and maintained in normal growth medium as described under "Experimental Procedures." On days 3 (sparse), 8 (near confluent), 11 (confluent), and 14 (postconfluent) cells were harvested for determination of cell number. As shown in Fig. 1, absolute cell number increased as a function of time. Analysis of the data revealed that cell number increased 2.2-fold, 2.5-fold, and 1.2-fold between days 0 and 3, 8 and 11, and 11 and 14, respectively, reflecting the expected reduction in cellular proliferative rate in postconfluent cells. Parallel wells were pulsed with [³H]thymidine to measure relative rates of DNA synthesis. Incorporation of the radiolabeled thymidine was normalized to the number of cells present during the radioactive pulse. [³H]thymidine incorporation was highest in sparsely seeded, robustly proliferating cells (day 3) (Fig. 1) and decreased with time in culture, with little or no [³H]thymidine incorporation evident in postconfluent (day 14) cells. Thus, as expected, DNA synthesis decreased prior to the reduction in cellular proliferative rate, and ultimately proliferation slowed in postconfluent cells. These observations are consistent with postconfluent contact-inhibited reduction in cellular growth. Subsequent studies were done using hCAVSMCs cultured to these same timepoints unless otherwise stated.

[View Larger Version of this Image (14K GIF file)]

Eicosanoid Measurements

The hCAVSMCs used in these studies produced PGE₂ and PGF₂ as determined by its presence in the culture medium (ELISA). PGE₂ levels increased from day 0 to day 3, but did not change over days 3-14 of the culture period: pg/ml/cell, 2.08 ± 0.25 (n = 2, day3); 2.18 ± 1.01 (n = 2, day 8); and 1.76 (n = 1, day 10). Indomethacin (1 µM) produced 87.3 ± 2.04 percent reduction in PGE₂ synthesis (mean ± S.E., n = 3 in duplicate). The levels of PGF₂ were comparable to the levels of PGE₂ detected in the same samples. Given that PGE₂ synthesis was inhibited by indomethacin, one can assume that indomethacin had equivalent effects on PGF₂ synthesis as has been previously described in a number of cell systems (58). These cells did not produce detectable levels of leukotriene B4 or leukotriene C4 (data not shown).

Analysis of 85-kDa PLA₂ and 14-kDa PLA₂ as a Function of Culture Period

Human CAVSMCs were harvested and homogenized, and subcellular fractions were made from preconfluent (day 3), near or at confluent (days 7-8), and postconfluent (days 14-15) cultures as described under "Experimental Procedures." Cell fractions were assayed for acylhydrolytic activity of 85-kDa PLA₂ activity via using [1-¹⁴C]palmitoyl-2-arachidonyl phosphatidylcholine as described under "Experimental Procedures." As shown in Fig. 2, cytosolic 85-kDa PLA₂ activity was highest on day 3 commensurate with the highest proliferative rate. Activity steadily decreased from day 3 to day 14 coincident with the reduction in the relative rate of DNA synthesis over this culture period. The 85-kDa PLA₂ activity in the smooth muscle microsomal fraction was very low and remained constant over the length of the culture, indicating that the reduced cytosolic 85-kDa PLA₂ activity was not due to migration of protein to microsomes. 14-kDa PLA₂ activity measured by acylhydrolysis of [³H]AA-E. coli did not change (specific activities (pmol/mg/min) in a representative experiment were 201.05 and 200.76 (microsomal fraction) and 8.57 and 12.37 (cytosolic fraction) for days 3 and 14 in culture, respectively).

[View Larger Version of this Image (11K GIF file)]

To examine if the reduction in 85-kDa PLA₂ activity resulted from lower protein levels, the microsomal and cytosolic fractions were subsequently subjected to SDS-polyacrylamide gel electrophoresis and probed with anti-85 kDa PLA₂ rabbit serum. As shown in Fig. 3 (panels A and B), the hCAVSMC cytosolic fractions from two separate studies possessed a high molecular weight protein that was recognized by rabbit anti-85 kDa PLA₂ serum and co-migrated characteristically with the rh 85-kDa PLA₂ protein at 110 kDa. The immunoreactive bands decreased in intensity from day 3 through day 14 indicating a decrease in protein over the culture period. Microsomal fractions contained immunoreactive 85-kDa PLA₂ bands, but band intensity did not change as a function of culture period. In a separate study (Fig. 3C) the 85-kDa PLA₂ protein was evaluated in cells cultured between 3 h and 10 days. Western analysis of cytosolic fractions demonstrated that the 85-kDa PLA₂ protein was present in cells that had been incubated for only 3 h, that the 85-kDa PLA₂ protein level increased between 3 h and day 1, and was further increased on day 3. Levels subsequently decreased by day 10 in agreement with the data shown in Fig. 3, A and B. Northern analysis of hCAVSMCs in parallel culture dishes is shown in Fig. 4. RNA from two independent studies was fractionated on the same gel. Hybridization indicates that the 85-kDa PLA₂ message is present at day 3 and that it decreases with increased time in culture. This supports that the reduction in 85-kDa PLA₂ protein level from day 3 to day 14 was due, at least in part, to a decrease in 85-kDa PLA₂ message.

[View Larger Version of this Image (66K GIF file)]

[View Larger Version of this Image (38K GIF file)]

Analysis of the subcellular fractions by ELISA indicated the presence of low levels of type II 14-kDa PLA₂ immunoreactive protein in the cytosol and approximately 10-fold greater levels in the particulate fraction, but no significant change in the 14-kDa PLA₂ protein was observed over the culture period (Table I).

Table I. Quantification of cell-associated type II 14-kDa PLA2 from cultured hCAVSMC homogenates following 3, 8, and 14 days in culture Smooth muscle cells were collected from 1-5 (600-cm2) flasks and pooled. Cells were homogenized by acid extraction as described under "Experimental Procedures" for analysis by ELISA. 14-kDa PLA2 immunoreactive protein levels do not change over the culture period. Representative results from one study are shown; data are expressed as mean ± S.D. of triplicate determinations. Culture time Type II 14-kDa PLA2 Microsomal Cytosolic ng/300 µg protein Day 3 2.44  ± 0.43 0.34  ± 0.21 Day 8 3.64  ± 1.91 0.25  ± 0.20 Day 14 3.21  ± 1.18 0.32  ± 0.03

Effect of PLA₂ Modulation on hCAVSMC Proliferation

To further examine the nature of the association of 85-kDa PLA₂ with cellular proliferation and to examine whether 85-kDa PLA₂ may have an active role in regulation of VSMC proliferation, we evaluated the effects of 85-kDa PLA₂, 14-kDa PLA₂, and iPLA₂ inhibitors on cellular proliferation. AACOCF₃, a trimethyl ketone analogue of arachidonic acid recently shown to inhibit human recombinant 85-kDa PLA₂ and cell-associated 85-kDa PLA₂ (59, 60) was added to day 3 cultures at different concentrations (0-10 µM) and further incubated for 3 days. Relative to the vehicle (0.03% ethanol) control cells, this compound dose dependently inhibited hCAVSMC proliferation (Fig. 5). Recall that these cells were seeded at a density of 6000 cells/well on day 0 and that by day 3, the cell number had increased 2.2-fold to 13,209 ± 200 cells/well (Fig. 1). Absolute cell number following the 3-day incubation period with 10 µM AACOCF₃ was 12,472 ± 788.2 (mean ± S.D.) indicating that this concentration of AACOCF₃ completely inhibited cellular proliferation, and that the apparent IC₅₀ of AACOCF₃ in our system is 1 µM (Fig. 5). In contrast, neither the selective 14-kDa PLA₂ inhibitor, SB203347 (0-10 µM) (59, 61), nor the cyclooxygenase inhibitor, indomethacin (1 µM), had significant effects on hCAVSMC proliferation. The long doubling time of these cells required that long incubation times be used to ensure a sufficient window over which to observe changes in proliferation. A possible concern is that the lack of an effect of the SB compound on hCAVSMC proliferation may be due in part to a possible breakdown of the drug over the 3-day incubation period used in these studies. However, 1 and 2 days of exposure to SB203347 (10 µM) produced no reduction in cell number (97.9 and 93.3% of control, respectively). Thus, the lack of an effect of SB203347 reported in Fig. 5 is not likely due to compound instability. Morphologic examination (Fig. 6) demonstrated that the various drug treatments produced no obvious changes in cell structure, and trypan blue exclusion methodology determined that cell viability in AACOCF₃- and SB203347-treated cells (10 µM) was >95% and was equivalent to vehicle (ethanol) control cells. Thus, AACOCF₃-mediated reduction of cellular growth was not due to cellular toxicity. Higher (30 µM) concentrations of AACOCF₃ were toxic as suggested by gross morphological changes and lack of trypan blue exclusion. Evaluation of a possible role of the iPLA₂ in hCAVSMC proliferation was performed in a separate series of experiments via examination of the effect of the potent and selective iPLA₂ inhibitor HELSS (21) in hCAVSMC proliferation. HELSS (0-3.0 µM) had no effect on proliferation as assessed by cell counts via hemacytometer. Expressed as percent of control, cell number was 102.0 ± 2.25, 98.1 ± 1.20, and 98.9 ± 2.65 (mean ± S.E.) for 0.3 µM, 1.0 µM, and 3.0 µM, respectively (n = 2; each in triplicate). Observed gross morphological changes indicated apparent cellular toxicity with 10 µM HELSS (data not shown).

[View Larger Version of this Image (17K GIF file)]

[View Larger Version of this Image (184K GIF file)]

Previous work in this laboratory utilized 85-kDa PLA₂ initiation site-directed antisense oligonucleotides in human monocytes (31, 59) and human synovial fibroblasts (49, 62) to assess the role of this enzyme in prostanoid and leukotriene production. To more specifically examine the nature of the regulation of cellular proliferation by 85-kDa PLA₂, hCAVSMCs were seeded (day 0), allowed to adhere overnight and then incubated (18-24 h, 37 °C) in growth factor- and antibiotic-free medium in the presence of increasing concentrations (0.01-1.0 µM) of antisense to 85-kDa PLA₂ (SB7111), complementary sense oligonucleotide (SB9030, 0.01-1.0 µM), a scrambled nonrelevant oligonucleotide control (SB7222, 1 µM), or Lipofectin vehicle as described under "Experimental Procedures." Following this preincubation, growth medium (SmGM2) was added. For comparison, some cells received AACOCF₃ (3 µM). To evaluate the role of AA in 85-kDa PLA₂-mediated effects on proliferation, parallel wells were incubated with AA (20 µM) alone, AA together with antisense oligonucleotides or with AACOCF₃. Following a 3-day incubation period, cell viability and cell number were determined. As shown in Table II, Lipofectin alone had no effect on cell number. SB7111 concentration dependently inhibited growth medium-induced proliferation with both 0.1 and 1.0 µM reaching statistical significance at p < 0.05. Recall that these cells were seeded at a density of 5 × 10⁴ cells/well prior to transfection and the 3-day exposure to growth medium to stimulate proliferation. Absolute cell number following this 3-day stimulation in the wells treated with 1.0 µM SB7111 was 4.47 ± 0.9 × 10⁴ (mean ± S.D.) indicating that this concentration of the antisense oligonucleotide completely inhibited cellular proliferation. The sense control (SB9030) and scrambled nonrelevant oligonucleotides had no effect (Table II). Addition of 20 µM AA itself directly and significantly stimulated cellular proliferation (Table II). Exogenous AA attenuated the anti-proliferative effect of SB7111 as evidenced by only 10 and 20% reductions in cell number in the presence of both AA and SB7111 (0.1 and 1.0 µM, respectively) versus 37 and 55% reductions in the presence of SB7111 alone. In agreement with other data presented herein (Fig. 5) AACOCF₃ (3.0 µM) inhibited growth factor-induced proliferation by 47%. This was significantly attenuated by addition of AA as indicated by only 9% inhibition of proliferation during co-incubation of 20 µM AA and AACOCF₃. Trypan blue exclusion and close visual inspection of cell morphology indicated that the observed decreases in cell number following exposure to SB7111 or AACOCF₃ were not due to cytotoxicity.

Table II. 85-kDa PLA2 antisense dose dependently reduces proliferation of hCAVSMCs Cell counts (hemacytometer) for each condition calculated as percent of control (SmGM2 growth medium in the absence of AA). Data are expressed as mean ± S.E. Each data point for a given experiment was obtained by duplicate counts of triplicate wells. The data set obtained in the absence of AA represents the averaged data from three experiments; for co-incubation experiments (+AA), n = 2.  Condition % control (AA)  AA +AA 20 µM Control 100 109.0  ± 4.10a Lipofectin 95.9  ± 1.88 112.6  ± 5.95a,b SB7111   0.01 µM 95.8  ± 2.59 105.2  ± 6.15   0.10 µM 62.7  ± 7.45a 89.5  ± 7.40b   1.00 µM 45.1  ± 5.24a 80.6  ± 6.65a,b SB9030   0.01 µM 102.4  ± 2.10 ND   0.10 µM 95.3  ± 1.21 ND   1.00 µM 85.2  ± 7.30 ND Scrambled 1.00 µM 104.3  ± 7.40 ND AACOCF3 3.00 µM 53.1  ± 6.39a 91.0  ± 1.35b a p < 0.05 vs. control (no AA). b p < 0.05 vs. paired data in the absence of AA. ND, not determined.

Flow Cytometry

Fluorescence-activated cell sorter (FACS) analysis was performed to evaluate the effect of AACOCF₃ on the cell cycle. Cells incubated with 10 µM AACOCF₃ (the highest concentration used in the proliferation studies described above) showed no evidence of morphological changes and no overt cytotoxity but had reduced cell number as expected (42% of non-AACOCF₃-treated controls). Cells were collected and stained with propidium iodide and TUNEL after 1, 2, and 3 days of exposure to the drug. Table III provides data from one representative of two experiments that shows that 10 µM AACOCF₃ caused no remarkable change in the distribution of cells in G₀/G₁, S, and G₂/M phases over 3 days. The left panel of Fig. 7 shows the counts versus propidium iodide intensity and graphically depicts the numerical data for the cell cycle stages given in Table III; again, no difference was observed in the distribution of the cells in AACOCF₃-treated cells. The right panel of Fig. 7 shows that 3 days of AACOCF₃ exposure produced no DNA fragmentation (e.g. no apoptosis) as evidenced by no movement of signal into quadrant R3. A shift of cells into the R3 quadrant is typically observed in our laboratory with anti-FAS antibody-induced apoptosis in Jurkat cells and is routinely reported for other apoptotic cells (57).

Table III. The 85-kDa PLA2 inhibitor AACOCF3 (10 µM) produces no phase-specific arrest of the cell cycle Cells were trypsinized, seeded into 150-cm2 culture flasks, allowed to adhere for 4 h in normal growth medium (SmGM2, control) before addition of AACOCF3 (10 µM), and then incubated in the presence of a drug for 1, 2, or 3 days. Effect of AACOCF3 on hCAVSMC cell cycle was examined by FACS scan analysis of propidium iodide-labeled cells. The data are from one representative of two experiments and are expressed as percent of total cells in the particular phase of the cell cycle. Reduction in cell number (cell counts, hemacytometer) obtained from parallel cultures verified AACOCF3 activity as detailed under "Results." Cell cycle phase Day 1 Day 2 Day 3  Control AACOCF3 Control AACOCF3 Control AACOCF3 G0/G1 92.2 92.4 83.9 84.4 82.1 86.4 S 8.3 8.4 10.6 8.7 11.5 8.0 G2/M 2.8 2.3 4.4 5.8 2.5 1.3

[View Larger Version of this Image (31K GIF file)]

DISCUSSION

Although previous studies demonstrated that nonselective PLA₂ inhibitors reduce VSMC proliferation (36-39, 42, 43), the present study represents the first direct examination of the two distinct PLA₂ enzymes in VSMC proliferation. Moreover, it is the first study that cleary demonstrates in any cell type that the 85-kDa PLA₂ and not the 14-kDa PLA₂ is critical for cellular proliferation. Herein it is confirmed that vascular smooth muscle cells possess both the 14-kDa PLA₂ and 85-kDa PLA₂ enzymes. By taking advantage of contact inhibition of cellular growth despite continual presence of growth medium, it was first demonstrated that 85-kDa PLA₂ activity, protein level and mRNA expression (but not 14-kDa PLA₂ activity or mass) correlated with DNA synthesis. 85-kDa PLA₂ levels were highest early in culture and rose to a maximum level and activity at day 3, which corresponds with the peak in [3H]thymidine uptake, indicating an association between 85-kDa PLA₂ and cellular proliferation and suggesting an important role of this protein in robustly proliferating hCAVSMCs.

To further and more directly evaluate the role of 85-kDa PLA₂ in hCAVSMC proliferation, PLA₂ inhibitors and antisense oligonucleotides were used. In both AACOCF₃- and SB7111-treated cells, replication assessed by cell counts was concentration dependently reduced relative to vehicle control. Indeed, at the highest concentration of SB7111 cell numbers at day 3 did not vary from the initial seeding density suggesting that SB7111 produced complete cessation of cell cycle activity and thereby completely inhibited cellular proliferation. Again, neither SB203347, HELSS, nor SB9030 affected proliferation. Taken together, these data further support an important role of the 85-kDa PLA₂ but not the 14-kDa PLA₂ nor iPLA₂ in hCAVSMC proliferation.

FACS scan analysis suggests that the reduction in cell number following inhibition of 85-kDa PLA₂ results from generalized attenuation of cellular proliferative rate rather than phase-specific cell cycle arrest. Furthermore, the AACOCF₃-treated cells showed no indication of entering programmed cell death. Similiar results were observed in HL-60 cells exposed to AACOCF₃ (63). The present data may offer an explanation for a previous observation (64); increased PLA₂ activity is readily observed following stimulation of proliferating cultures of murine embryo palate mesenchymal cells with the calcium ionophore A23187, but confluent cultures showed no such response. Based on our findings, we hypothesize that the confluent murine embryo palate mesenchymal cultures were expressing very low levels of 85-kDa PLA₂, whereas the actively dividing cells had relatively increased levels of 85-kDa PLA₂ that could respond to the calcium stimulus by the usual mechanisms (e.g. phosphorylation) to be reflected as higher enzyme activity.

The mechanism by which 85-kDa PLA₂ influences cellular proliferation remains to be determined. PGE₂ was produced by hCAVSMCs between day 0 and day 3, but PGE₂ levels did not change further with changes in proliferation. This is consistent with the coordinate early rise in 85-kDa PLA₂ protein during the same time frame and suggests a possible role of PGE₂ in the initial phase of proliferation. However, indomethacin had no effect on cellular proliferation under the conditions of our experiments, indicating that the 85-kDa PLA₂ dependent effect on VSMC proliferation is not mediated through PGE₂. This is in agreement with previous reports that cyclooxygenase inhibition by indomethacin had no effect on VSMC proliferation (1, 39), AA-, or hydrogen peroxide-induced c-jun expression in VSMCs (37) or AA activation of mitogen-activated protein kinase (38). When cells were treated with exogenous AA, the reduction in cell number produced by either AACOCF₃ or 85-kDa PLA₂ antisense oligonucleotides (SB7111) was prevented or significantly attenuated, suggesting that the 85-kDa PLA₂ may influence cellular proliferation largely through its generation of free AA. Interestingly, 20 µM AA by itself stimulated hCAVSMC proliferation, increasing the cell number 9% above untreated controls. The exact role of AA is not known, but as discussed above, conversion to PGE₂, LTB₄, and LTC₄ does not seem to be obligatory for proliferation to proceed. This raises the possibility that AA is itself an intracellular signal and/or that it is converted to an unidentified eicosanoid product important in regulation of cellular proliferation. It is possible that the present demonstration of a direct proliferative effect of exogenous administration of AA to hCAVSMCs and attenuation of AACOCF₃- and 85-kDa PLA₂ antisense-mediated reductions in proliferation by AA may be related to the effect of AA to stimulate c-fos, c-jun, and/or mitogen-activated protein kinase (36-39, 42, 43).

The mechanism by which 85-kDa PLA₂ influences cellular proliferation may be related to previous demonstrations that inhibition of the lipoxygenase-cytochrome P₄₅₀ system reduces VSMC proliferation or proliferation-associated events (1, 37-39) and/or that 85-kDa PLA₂ is activated by oncogenic ras (65, 66). Alternatively, or additionally, recent reports indicate that the release of AA by PLA₂ is accompanied by the formation of biologically active lysolipids, and that lysophosphatidic acid, lysophosphatidylinositol, and lysophosphatidylcholine stimulate mitogen-activated protein kinase activity and proliferation in a variety of cell types (67-71). Furthermore, these lysolipids serve as precursors to a new class of biologically active lipid derivatives, the glycerophosphoinositides, which show evidence of being accumulated specifically in ras-transformed cells. Given that the present data indicate that it is the 85-kDa PLA₂ that selectively regulates the initial release of AA to influence cellular proliferation, it would be interesting to examine the effects of AACOCF₃ or 85-kDa antisense on accumulation of these lipids in ras-transformed cells.

Rao et al. (37) report that nonselective PLA₂ inhibition by mepacrine either had no effect on c-jun expression induced by platelet-derived growth factor, partially attenuated epidermal growth factor-stimulated c-jun expression, or significantly inhibited hydrogen peroxide-induced c-jun expression in VSMCs suggesting a potential growth factor-specific nature to the role of PLA₂ in VSMC proliferation. To avoid complications associated with specific growth factors, the present studies were designed to take advantage of contact inhibition of cellular growth despite continual presence of growth medium to define whether the 14- and 85-kDa PLA₂ enzymes are involved in regulation of cellular proliferation. Thus, the present study routinely involved stimulation of hCAVSMC proliferation by a defined optimized medium supplied by the manufacturer that contained fetal bovine serum (5%), insulin (5 mg/ml), fibroblast growth factor (2 ng/ml), and epidermal growth factor (0.5 mg/ml). Accordingly, we cannot comment on whether the 85-kDa PLA₂ serves a particular role in the proliferative response to these specific growth factors. Examination of this issue provides the basis for future additional studies.

The present finding that 85-kDa PLA₂ is apparently required for cellular proliferation may be relevant to pathologies associated with VSMC hyperplasia. In the normal noninjured vessel, smooth muscle cells have a low mitotic rate (72). The VSMC hyperplasia evident in atherosclerotic and restenotic vessels is believed to be part of a sequelae of responses to vascular injury by numerous agents including viruses, caustic by-products of cigarettes, hypercholesterolemia, and physical injury. PLA₂ activation has been associated with cellular injury in kidney, heart and liver (71). It remains to be determined if cellular injury specifically alters 85-kDa PLA₂ expression and activity and if the 85-kDa PLA₂ is up-regulated in atherosclerotic or restenotic vessels. Nevertheless, the present data support the hypothesis that selective inhibitors of 85-kDa PLA₂ may prove effective therapeutics for treatment of vascular disease associated with aberrant VSMC proliferation.