Oxidized low density lipoprotein and lysophosphatidylcholine stimulate cell cycle entry in vascular smooth muscle cells. Evidence for release of fibroblast growth factor-2.

We have previously shown that oxidized low density lipoprotein (LDL) but not native LDL stimulated DNA synthesis in cultured smooth muscle cells (SMC) and that α-tocopherol (vitamin E) inhibited this proliferative response (Lafont, A., Chai, Y. C., Cornhill, J. F., Whitlow, P. L., Howe, P. H., and Chisolm, G. M. (1995) J. Clin. Invest. 95, 1018-1025). The moiety of oxidized LDL that stimulates DNA synthesis and the cellular mechanism for this potentially mitogenic effect are not known. We now report that lipid fractions containing lysophospholipids from oxidized LDL or phospholipase A2-treated native LDL stimulated SMC DNA synthesis as did palmitoyl lysophosphatidylcholine (lysoPC). Protein kinase C inhibitors and down-regulation of protein kinase C activity by phorbol ester inhibited oxidized LDL- and lysoPC-induced DNA synthesis. A neutralizing monoclonal antibody against fibroblast growth factor-2 significantly inhibited oxidized LDL and lysoPC-induced DNA synthesis in SMC; irrelevant antibodies were ineffective. Vitamin E inhibited the DNA synthesis stimulated by lysoPC, an observation that distinguished this effect from DNA synthesis induced by another detergent, digitonin. These results suggest that oxidized LDL and its lysoPC moiety stimulate SMC to enter the cell cycle via an oxidative mechanism that causes the release of fibroblast growth factor-2 and a subsequent autocrine or paracrine response.

We have previously shown that oxidized low density lipoprotein (LDL) but not native LDL stimulated DNA synthesis in cultured smooth muscle cells (SMC) and that ␣-tocopherol (vitamin E) inhibited this proliferative response ( Invest. 95, 1018 -1025). The moiety of oxidized LDL that stimulates DNA synthesis and the cellular mechanism for this potentially mitogenic effect are not known. We now report that lipid fractions containing lysophospholipids from oxidized LDL or phospholipase A 2 -treated native LDL stimulated SMC DNA synthesis as did palmitoyl lysophosphatidylcholine (lysoPC). Protein kinase C inhibitors and down-regulation of protein kinase C activity by phorbol ester inhibited oxidized LDLand lysoPC-induced DNA synthesis. A neutralizing monoclonal antibody against fibroblast growth factor-2 significantly inhibited oxidized LDL and lysoPC-induced DNA synthesis in SMC; irrelevant antibodies were ineffective. Vitamin E inhibited the DNA synthesis stimulated by lysoPC, an observation that distinguished this effect from DNA synthesis induced by another detergent, digitonin. These results suggest that oxidized LDL and its lysoPC moiety stimulate SMC to enter the cell cycle via an oxidative mechanism that causes the release of fibroblast growth factor-2 and a subsequent autocrine or paracrine response.
Oxidatively modified forms of low density lipoprotein (LDL) 1 have been shown to cause multiple changes in cellular functions distinct from the effects of native LDL. Oxidized LDL induces certain genes (1,2), suppresses others (3)(4)(5), alters cellular lipid metabolism (6), and injures cells (7)(8)(9). Because oxidized LDL has been shown to reside in arterial lesions in animals and humans (10), the alterations in cell function observed in vitro have been suspected of involvement in vascular lesion development (11)(12)(13). Arterial lesions are believed to form after focal "injury" to endothelium that activates or causes dysfunction of endothelium (14) and leads to endothelial cell-mediated entry into the intima of monocytes that engorge with lipoprotein lipids. Subsequent migration into the intima of medial smooth muscle cells and their proliferation further advance lesion development. Each of these events has been linked to actions of oxidized LDL on cells grown in culture (12,15).
Putative proliferative effects of lipoproteins were reported many years ago (16 -18); these preceded the early studies of the cellular effects of oxidized LDL (7,19,20). More recently multiple effects of oxidized LDL, related conceptually to cell proliferation, have been reported including the induction of platelet-derived growth factor-AA production and platelet-derived growth factor receptor expression in vascular smooth muscle cells (SMC) (21), induction of various growth factors in endothelial cells (1), and stimulation of increased DNA synthesis in macrophages (22) and smooth muscle cells (23)(24)(25). In some studies oxidized LDL-induced increases in cell number were also reported; however, in some, mitogenic levels of serum were also present (26) and in others cells were not previously rendered quiescent (23), leaving ambiguous the role of oxidized LDL as a mitogen.
In the present studies we sought to elucidate further the oxidized LDL-induced stimulation of DNA synthesis in vascular SMC by testing whether the DNA synthesis led to cell cycle entry, by identifying a lipoprotein moiety responsible for the induction of DNA synthesis, and by determining the mechanism by which this effect is brought about. Our results suggest that oxidized LDL does indeed cause these cells to enter the cell cycle, that lysophosphatidylcholine, a component of oxidized LDL, is a potential contributor to the proliferative stimulation, and that the mechanism does not appear to require grossly detectable cell injury, but it depends at least in part on an autocrine or paracrine response elicited by the release of fibroblast growth factor (FGF-2). LDL Preparation and Modification-LDL (1.019 g/ml Յ solvent density Յ 1.063 g/ml) was prepared by sequential ultracentrifugation from pooled, citrated human plasma according to a modification (8) of the method of Hatch and Lees (27). Oxidized LDL was prepared by dialyzing native LDL against isotonic saline (pH 7.0 -8.0) containing 5 M CuSO 4 at room temperature until the characteristic color change occurred from golden to yellow to translucent without color (9). The oxidation was stopped, and the lipoprotein bound metal ion was removed (28,29) by dialyzing against isotonic saline containing 0.5 mM EDTA at 4°C with at least three changes of dialysate. LysoPC was measured in two representative preparations of oxidized LDL according to the method of Nakamura and Handa (30) and found to be 20 and 25 g/100 g of oxidized LDL cholesterol.

Materials
To hydrolyze LDL phospholipids, EDTA was removed by dialysis from native LDL (4.2 mg cholesterol/ml), which was then incubated with 45 mM Tris, pH 8.5, 2.5 mM CaCl 2 , and 14 units/ml of phospholipase A 2 at 37°C for 2 h according to a modification of the method of Quinn et al. (31). To stop the reaction, 50 mM EDTA, pH 8.0, was added to the mixture, and excess EDTA was removed by alternatively washing and concentrating the lipoprotein preparation using centrifugation with Centricon 10 filters (Amicon, Beverly, MA).
Lipid Extraction and Fractionation-Lipid extraction from both oxidized LDL and phospholipase A 2 -LDL were carried out as described (32) using methanol and chloroform as solvents. The lipid extract from native LDL was used as a control. Lipids were fractionated using silica cartridges and collected as described previously (31). Fractions were dissolved in acetone:ethanol (1:1) before adding to cells at solvent concentrations Յ0.4% of medium volume.
Cultured Rabbit Vascular SMC-Vascular SMC (passages 3-10) were obtained from explants of rabbit aorta using procedures described previously (33). SMC were grown in DMEM/F-12 medium with 10% fetal bovine serum and incubated at 37°C, 5% CO 2 in air. Cells were plated into 24-well plates at 20,000 -25,000 cells/well and allowed to reach 90% confluence before making them quiescent. In some experiments, quiescence was achieved by replacing the medium with DMEM/ F-12 medium containing 0.25% fetal bovine serum for 2 days; in others, serum-free DMEM/F-12 medium was used.
Determination of DNA Synthesis and Protein Synthesis in SMC-[ 3 H]Thymidine incorporation into DNA was determined by adding 1 Ci/ml [methyl-3 H]thymidine for the last 6 h of a 26-h exposure to agents being tested for stimulation. After removing radioactive medium, cells were washed twice with ice-cold 10% trichloroacetic acid. Trichloroacetic acid-insoluble material was hydrolyzed by 0.25 N NaOH, and radioactivity was assayed in a liquid scintillation counter. To examine the effects of vitamin E or various antibodies on DNA synthesis, cells were treated with these agents for 2 h before as well as after adding stimuli. A neutralizing antibody against FGF-2 was used that according to the supplier does not cross-react with acidic FGF (FGF-1). SMC were pretreated with phorbol myristate acetate (200 nM) for 48 h or inhibitors of protein kinases, H-7, HA-1004, calphostin C, H-89, and GF 109203X for 30 min to explore the role of protein kinase C in the observed DNA synthesis increases.
To evaluate cell protein synthesis as an indirect indicator of cell injury, [ 3 H]leucine (2 Ci/ml) was added 20 h after of addition of lysoPC, inhibitors of protein kinases, or other test agents. After another 6 h, cells were processed and radioactivity incorporated into cells was determined as described above for DNA synthesis.
Assessment of Cell Permeabilization-As another assessment of cell injury, cell permeabilization was determined by the specific release of 14 C-labeled products from cells previously loaded with [ 14 C]adenine as described (34). Briefly, quiescent SMC were exposed to [ 14 C]adenine (0.2 Ci/ml) for 16 h. Before treating cells with putatively injurious agents, radioactive medium was removed, and cells were rinsed twice with DMEM/F-12 containing 0.25% fetal bovine serum. After a 26-h incubation, medium was sampled from each well and assayed using a liquid scintillation counter. The maximal 14 C release (100%) was obtained from cells lysed with Triton X-100; the minimum release (calculated to be 0%) was from untreated (control) cells. The specific release, expressed as a percentage, was calculated as (cpm released from treated cells Ϫ cpm released from untreated cells) divided by (cpm released by Triton Ϫ cpm released from untreated cells) (35). To examine the effect of vitamin E, cells were treated with vitamin E for 2 h before as well as after adding the agents indicated.

H]Thymidine Autoradiography in Rabbit SMC-[ 3 H]Thymidine
(1 Ci/ml) was pulsed for the last 6 h of a 26-h incubation with serum or lysoPC. After removing medium, cells were processed for nuclear autoradiography by coating with Kodak type NTB-2 emulsion after fixing cells with cold 100% methanol as described previously (36). RESULTS We have previously shown that oxidized LDL increased [ 3 H]thymidine incorporation into DNA in quiescent smooth muscle cells and that vitamin E blunted this response but not the response to serum (25). We wished to determine the oxidized LDL entity(ies) responsible for the effect. We performed silica column fractionation of oxidized LDL lipids according to a previously published protocol (31), in which various solvents were used to elute four general classes of lipids, containing, respectively, cholesteryl esters and triglycerides, cholesterol, free (and oxidized) fatty acids, and phospholipids. These were collected, dried under nitrogen, redissolved as indicated under "Experimental Procedures," and added to rabbit SMC at concentrations indexed to the original lipoprotein. The most active among the oxidized LDL fractions in enhancing 3 H-thymidine incorporation was the fourth fraction as shown in Fig. 1. This fraction was previously shown (31) and confirmed by TLC (30) in our laboratory to include phospholipids and lysophospholipids. Lysophospholipids are produced as LDL oxidizes (37), and we tested whether the stimulating activity of oxidized LDL could be in part due to lysophospholipids. The lipids of phospholipase A 2 -treated native LDL were also fractionated and collected according to the same method used for oxidized LDL lipids. Again, the most potent fraction in stimulating DNA synthesis was that containing lysophospholipids (data not shown). In addition, as shown in Fig. 2A, synthetic palmitoyl lysoPC at concentrations approximately within the range expected to be present in the levels of oxidized LDL for which we observed activity (38) also increased DNA synthesis. As expected, at toxic concentrations of lysoPC (e.g. 30 M), there is no enchanced DNA synthesis. LysoPC thus appeared to be one lipid formed during LDL oxidation that is capable of contributing to the increase in DNA synthesis. This assertion was strengthened by the fact that vitamin E blunted the enhanced DNA synthesis caused by lysoPC, as shown in Fig. 2B, at vitamin E concentrations shown previously to blunt this effect by oxidized LDL (25). The inhibition was also observed after overnight treatment of cells with 50 M vitamin E and extensive washing (three times with medium) prior to adding lysoPC (data not shown). We then sought whether the increased DNA synthesis evoked in SMC by oxidized LDL or lysoPC was, in fact, related to cell cycle progression or whether the enhanced [ 3 H]thymidine incorporation was due to a distinct, nonproliferative response. Several approaches were taken. Cells made quiescent by 48 h of serum deprivation were exposed to oxidized LDL, lysoPC, or serum and [ 3 H]thymidine incorporation was measured as a function of time. Although as expected the response of cells to oxidized LDL or lysoPC was not of the same magnitude as that of cells exposed to serum (9), the time course of thymidine incorporation was the same as after serum stimulation; that is, the peak thymidine incorporation occurred at approximately 28 h from the addition of oxidized LDL, lysoPC, or serum (data not shown). Thus, cells responded temporally to oxidized LDL or lysoPC as they respond to known mitogens. As a second approach, autoradiography was performed on cell cultures 26 h after exposure to lysoPC. The number of labeled cells after exposure to lysoPC was significantly higher than that in unstimulated controls (p Ͻ 0.01), and the labeling was clearly nuclear (data not shown). Third, we tested pharmacologically whether the enhanced [ 3 H]thymidine incorporation may have been due to DNA repair after lysoPC-induced DNA damage. Inhibitors of poly(ADP-ribose) polymerase were added to cell cultures in concentrations shown to inhibit this DNA repair enzyme in other cell systems after exposure to levels of H 2 O 2 that damage DNA (39). Toxicity of the inhibitors, as discerned from 14 C release from [ 14 C]adenine loaded cells, was also measured to avoid misinterpreting blunted DNA responses secondary to cell injury as being due to inhibited DNA repair. Nontoxic levels of 3-aminobenzamide and nicotinamide, up to 5 and 2.5 mM, respectively, did not blunt lysoPC-or FGF-2-induced (as control) increases in [ 3 H]thymidine incorporation (data not shown). In control experiments, we verified that these agents inhibited H 2 O 2 (1 mM) induced toxicity in our cell system, as reported by Schraufstatter et al. (39) in P388D 1 cells. (They showed that inhibition of H 2 O 2 toxicity was secondary to blocking of DNA repair by poly(ADP-ribose) polymerase.) Finally, we arrested cells late in G 1 , approximately 2 h prior to the beginning of S phase by exposing the cells to 10 ng/ml TGF-␤ 1 (40). As shown in Fig. 3, increased DNA synthesis in TGF-␤ 1 -arrested cells was inhibited after exposure to either lysoPC or serum (as control), suggesting further that the increased DNA synthesis observed after lysoPC addition was due to cells entering S phase. Collectively, these four experimental results are consistent with the concept that the increased thymidine incorporation following oxidized LDL or ly-soPC exposure reflects transit into S phase of the cell cycle.
In order to gain insights into the mechanism, we tested whether stimulation by lysoPC or oxidized LDL followed cell signaling pathways used by well characterized polypeptide growth factors. We examined if the inhibition of PKC activation blunted S phase entry in response to oxidized LDL or lysoPC. Down-regulation of PKC was achieved by a 48-h exposure of cells to phorbol myristate acetate (200 nM) (41), after which cells were exposed to oxidized LDL or lysoPC. Down-regulation of PKC blunted the DNA synthesis response of both agonists (Fig. 4). Consistent with these findings, concentrations of PKC inhibitors that were verified to be nontoxic for these smooth muscle cells, were shown to inhibit DNA synthesis. (Cytotoxicity of the inhibitors was measured using [ 3 H]leucine incorporation into cell protein measured between 20 and 26 h after exposure of cells to the inhibitors, as well as specific 14  We then tested whether a polypeptide growth factor might be involved as an intermediate, autocrine, or paracrine mediator of the proliferative response to oxidized LDL or lysoPC. We reasoned that the absence of a delay in DNA synthesis between oxidized LDL or lysoPC and serum reported above suggested that release of a stored growth factor was more likely involved than induction of de novo growth factor synthesis and subsequent release. Because FGF-2 is known to be stored in SMC (42), we stimulated cells with oxidized LDL, lysoPC, or FGF-2 in the presence and the absence of a neutralizing monoclonal antibody to FGF-2, as well as two control antibodies, a neutralizing monoclonal antibody to TGF-␤ 1 , and an irrelevant monoclonal IgG preparation. The antibody to FGF-2 but neither of the control antibodies significantly inhibited the DNA synthesis responses to oxidized LDL by 50%, lysoPC by 80%, and FGF-2 by 100%, as shown in Fig. 5.
We verified that FGF-2-induced DNA synthesis in SMC is a PKC-dependent pathway as reported by other groups (43), because at subtoxic concentrations, GF 109203X (10 M) and H-7 (20 M) inhibited the mitogenic effect of FGF-2 (data not shown). Interestingly, vitamin E (50 M), which we showed to be capable of inhibiting lysoPC and oxidized LDL-induced DNA synthesis, did not inhibit the FGF-2-induced response in SMC (data not shown). This suggested that the vitamin E inhibitable step in the oxidized LDL or lysoPC induction was prior to stimulation by FGF-2.
To test further the participation of FGF-2, we used an im- mortalized rabbit SMC line that overproduces FGF-2 that we obtained from Dr. Gene Liau. He and co-workers recently characterized this cell line, validating the overproduction of FGF-2 using RNA gel blot hybridization, reverse transcription polymerase chain reaction, and Western blotting (44). The experiment shown in Fig. 6 demonstrated that lysoPC markedly enhanced DNA synthesis in these SMC but to much higher levels than in our typical cultured rabbit SMC. The response was inhibitable by vitamin E and a neutralizing antibody to FGF-2, indicating that the mechanism for increased DNA synthesis by lysoPC is the same in these cells as in typical cultured SMC but that the response is much higher in magnitude. LysoPC added at 15 M to quiescent cells resulted in a 40% increase in cell number over unstimulated cells in this SMC (p Ͻ 0.02). A preparation of oxidized LDL, added at a concentration (80 g/ml) that did not cause a detectable increase in DNA synthesis when added to typical quiescent SMC, caused a 5-fold increase in DNA synthesis in the FGF-2 overproducing cell line over the unstimulated cells of this line (data not shown).
Others have suggested that FGF-2, which lacks a signal peptide for secretion, may be released in response to cell injury (45). Both oxidized LDL and lysoPC are known to injure cells, and vitamin E has been shown to inhibit the injury induced by oxidized LDL (9) and lysoPC. 2 We thus tested whether the DNA synthesis in response to lysoPC was secondary to and whether it required cell injury or whether vitamin E inhibition of cell cycle progression was secondary to protection against cell injury. Cell injury was inferred from decreased cellular protein synthesis and from increased specific 14 C release from cells preloaded with [ 14 C]adenine, a measure of cell membrane permeabilization. Cells were exposed to lysoPC at concentrations ranging from 2.5 to 50 M with and without treatment with 50 M vitamin E, which was added 2 h prior to the addition of lysoPC and was also present concurrently. Fig. 7A indicates that very slight (Ͻ10%) cell permeabilization was detectable at 15 M but none was detected at lower concentrations; 14  cell protein synthesis measured 20 -26 h after lysoPC addition (Fig. 7B). Increased DNA synthesis was clearly detectable at the subtoxic lysoPC concentrations of 10 M and 15 M (Fig. 2). Stimulation of [ 3 H]thymidine incorporation at these subtoxic levels is also FGF-2-mediated, because the neutralizing antibody against FGF-2 inhibited this effect (Fig. 5).
These results suggested that the mechanism for the proliferative effect of lysoPC did not require mortal cell injury but did not rule out a more subtle, transient membrane destabilization. Increased DNA synthesis due to exposure to lysoPC has been previously reported in rat vascular SMC and attributed to a "detergent" effect on cell membranes by analogy with the effect of the detergent digitonin (46). We confirmed that [ 3 H]thymidine incorporation was also enhanced in our smooth mucle cells by digitonin (1 M), but unlike the effect of lysoPC, the enchanced DNA synthesis by digitonin was not inhibited by vitamin E (Fig. 8), suggesting distinct mechanisms of action. DISCUSSION Our results indicate that enhanced DNA synthesis by oxidized LDL is linked to cell cycle entry, that one of the oxidized LDL components that can induce this response is lysoPC, and that the mechanism, at least in part, involves an autocrine or paracrine action of the oxidized LDL-or lysoPC-induced release of FGF-2. The phenomenon may have biological relevance in cell populations in which lipid oxidation products are formed, such as inflammation sites, early vascular lesions, or other situations in which lysoPC is generated, for example, as a result of phospholipase A 2 action.
Our conclusion that lysoPC was one of the components of oxidized LDL that induced cell cycle entry is based on multiple findings. The fractions from silica column elutions known to contain phospholipids and lysophospholipids were the most active fractions of oxidized LDL and phospholipase A 2 -treated LDL. (Intact phospholipase A 2 -LDL did not increase DNA synthesis, consistent with the fact that this preparation would have sufficient vitamin E to inhibit the response.) Palmitoyl lysoPC, reported to be the principal lysophospholipid formed during LDL oxidation (37), was effective in increasing DNA synthesis at concentrations shown by other laboratories (38) and confirmed by us using quantitative TLC to be present on oxidized LDL. Furthermore, as with oxidized LDL, the DNA synthesis increase by lysoPC was inhibited by vitamin E and a neutralizing antibody to FGF-2. Proliferative responses to ly-soPC have been reported previously but may be cell-specific. Palmitoyl lysoPC alone did not have a mitogenic effect on macrophages (47) and T-lymphocytes (48); however, it stimulated DNA synthesis in rat vascular SMC, and this effect appeared to be related to Ca 2ϩ influx (46). That lysoPC is not the sole mediator of the increased DNA synthesis in SMC by oxidized LDL is suggested by the recent report that a plateletactivating factor-like activity present in oxidized LDL, distinct from lysoPC, is also capable of stimulating DNA synthesis increase in SMC (24). Because of structural similarities between platelet-activating factor and lysoPC, one might speculate that mechanisms of stimulation may be similar for the two, but this awaits further experimentation. Increases in [ 3 H]thymidine incorporation induced by oxidized LDL or lysoPC could theoretically be due to cell responses distinct from the DNA synthesis phase of the cell cycle, for example, DNA strand breaks leading to DNA repair. The following lines of evidence from our results suggest that the DNA synthesis induced by oxidized LDL or lysoPC was, in fact, cell cycle-related. (a) One could expect that if the [ 3 H]thymidine incorporation in oxidized LDL-or lysoPC-treated smooth muscle cells were due to effects other than stimulation into cell cycle, the peak of DNA synthesis would not be contemporaneous with that following known growth factors. Our results reveal that the time course for DNA synthesis, including that the time of peak incorporation of labeled thymidine was the same for oxidized LDL, lysoPC, serum, and FGF-2. (b) Inhibitors of DNA repair enzymes would be expected to block lysoPCinduced DNA synthesis related to the repair of DNA damage but would not be expected to block growth factor-induced DNA synthesis. Neither 3-aminobenzamide nor nicotinamide, given at the concentrations shown to inhibit poly(ADP-ribose) polymerase in other cell systems (39) blocked lysoPC-or FGF-2induced DNA synthesis. (c) Autoradiography showed that the number of labeled cells after exposure to lysoPC significantly exceeded that in unstimulated wells and that the labeling was in fact in the nuclei of smooth muscle cells. (d) TGF-␤ 1 is known to block cell cycle progression, including in SMC (40) in late G 1 , approximately 2 h prior to cell entry into S phase. TGF-␤ 1 will thus inhibit growth factor-induced [ 3 H]thymidine incorporation, but it would not be expected to inhibit [ 3 H]thymidine incorporation in nonmitogenic events. TGF-␤ 1 inhibited ly-soPC-induced DNA synthesis, as well as that stimulated by serum. In addition, the inhibition of protein kinase C activation, a step in known mitogenic signaling pathways in smooth muscle cells (49) also inhibited oxidized LDL or lysoPC-induced DNA synthesis. Taken together, our results link the enhanced DNA synthesis with cell cycle entry.
Also consistent with our finding that oxidized LDL and ly-soPC induced increases in DNA synthesis reflecting cell cycle entry is our observation that this stimulation is mediated by FGF-2. Mechanisms of release of FGF-2 are unknown. FGF-2 lacks a signal peptide required for the cellular release of other secretory proteins (50). Results from a number of studies have suggested that cell injury can release FGF-2 (51), including a report that cell damage is required for FGF-2 secretion in cultured smooth muscle cells (45). However, it was reported that FGF-2 can be released from SMC without discernible injury by 25-hydroxycholesterol (52). Recently, another report showed that transient mechanical stimuli increased DNA synthesis in human vascular SMC, which was partially due to the release of FGF-2 in the absence of an increase in lactate dehydrogenase activity (53). It was also shown that a monoclonal antibody to FGF-2 inhibited most of the mitogenic effect of thombin in rat vascular smooth muscle cells; cell damage was believed unlikely in this effect (54).
Injury was possible in the present context because in addition to their proliferative effects, lysoPC and oxidized LDL have been shown to injure cells at higher concentrations (7,55). We attempted to quantify a role for injury by assessing the release of 14 C from [ 14 C]adenine-loaded cells and measures of cell protein synthesis between 20 and 26 h after the addition of lysoPC. Our results showed clearly that both 10 and 15 M lysoPC stimulated DNA synthesis, yet no cell injury was discernible at 10 M lysoPC. At a concentration of 15 M, lysoPC caused mild permeabilization by a detectable 14 C release but without any decrease in cell protein synthesis. This suggested that lysoPC may transiently influence membrane permeability, an event that may be related to release of FGF-2, but that mortal cell injury is not required for the FGF-2 release and subsequent stimulation of cells into the cell cycle. A permeabilization or mild, reversible form of injury could have been related, for example, to a detergent action of lysoPC as suggested by a report showing that lysoPC and the detergent digitonin similarly increased DNA synthesis (46). However, our results showed that vitamin E did not inhibit digitonin-mediated increases in DNA synthesis, indicating that these occur by at least partly different mechanisms. Furthermore, lysoPC effectively increased DNA synthesis at concentrations well below its reported critical micelle concentration, 50 M (56).
Vitamin E is reported to have cellular actions other than acting as antioxidant; however, it is presumed that vitamin E was not inhibiting DNA synthesis via its putative ability to inhibit PKC activity (57), because FGF-2-induced proliferation, which could be blunted by PKC inhibition, was not inhibited by vitamin E. Our results support the concept that vitamin E blocks lysoPC-induced release of FGF-2 but does not block the effect of the growth factor after its release. Our results are consistent with a scenario in which sublethal levels of oxidized LDL or lysoPC can inflict an antioxidant-inhibitable oxidant stress on cells, which causes FGF-2 release by an unknown means and a FGF-2-mediated proliferative response. However, we cannot rule out certain alternative mechanisms. For example, lysoPC may be releasing FGF-2 from matrix rather than or in addition to releasing it from cells. In addition, the increased DNA synthesis we observed in response to oxidized LDL could be due to multiple mechanisms, only one of which is lysoPCand FGF-2-mediated. Neutralizing antibody to FGF-2 and down-regulation of PKC by prior treatment of cells with phorbol myristate acetate did not totally block the effects of oxidized LDL.