Propionyl-L-carnitine Reduces Proliferation and Potentiates Bax-related Apoptosis of Aortic Intimal Smooth Muscle Cells by Modulating Nuclear Factor-κB Activity*

Propionyl-l-carnitine (PLC) has been introduced among the therapeutic approaches of peripheral arterial disease, and more recently, an increase of intimal cell apoptosis has been demonstrated to contribute to its effectiveness in rabbit carotid postinjury myointimal hyperplasia prevention. How PLC mediates these effects on vascular smooth muscle cells (SMCs) remains poorly understood. We investigated the role of NF-κB in PLC-induced arterial remodeling. In vivo, daily PLC treatment 15 days after injury resulted in a reduction of relative rat aortic intimal volume, an increase of apoptosis, Bax up-regulation without changing the Bcl-2 level, and a reduction of NF-κB, vascular cell adhesion molecule-1, monocyte chemotactic protein-1, and survivin in myointimal thickening compared with controls. In the presence of 10% serum, a reduced G1 → S phase progression preceded PLC-induced intimal cell apoptosis; in 0.1% serum cultures, in a dose-dependent manner, PLC rapidly induced intimal cell apoptosis and reduced p65, p50, IAP-1, and IAP-2 expression. Inhibiting NF-κB activation through SN50 increased apoptotic rate and Bax expression in intimal but not in medial SMCs, and successive PLC treatment failed to induce a further increase in apoptotic rate. Bax antisense oligodeoxynucleotide reduced PLC-induced intimal cell apoptosis and cytochrome c release. The PLC-induced attenuation of NF-κB activity in intimal cells was also due to the increase of IκB-α bioavailability, as the result of a parallel induction of IκB-α synthesis and reduction of phosphorylation and degradation. Collectively, these findings document that NF-κB activity inhibition contributes to PLC-induced proliferative arrest and Bax-related apoptosis of intimal SMCs.

Propionyl-L-carnitine (PLC) has been introduced among the therapeutic approaches of peripheral arterial disease, and more recently, an increase of intimal cell apoptosis has been demonstrated to contribute to its effectiveness in rabbit carotid postinjury myointimal hyperplasia prevention. How PLC mediates these effects on vascular smooth muscle cells (SMCs) remains poorly understood. We investigated the role of NF-B in PLC-induced arterial remodeling. In vivo, daily PLC treatment 15 days after injury resulted in a reduction of relative rat aortic intimal volume, an increase of apoptosis, Bax up-regulation without changing the Bcl-2 level, and a reduction of NF-B, vascular cell adhesion molecule-1, monocyte chemotactic protein-1, and survivin in myointimal thickening compared with controls. In the presence of 10% serum, a reduced G 1 3 S phase progression preceded PLC-induced intimal cell apoptosis; in 0.1% serum cultures, in a dose-dependent manner, PLC rapidly induced intimal cell apoptosis and reduced p65, p50, IAP-1, and IAP-2 expression. Inhibiting NF-B activation through SN50 increased apoptotic rate and Bax expression in intimal but not in medial SMCs, and successive PLC treatment failed to induce a further increase in apoptotic rate. Bax antisense oligodeoxynucleotide reduced PLC-induced intimal cell apoptosis and cytochrome c release. The PLC-induced attenuation of NF-B activity in intimal cells was also due to the increase of IB-␣ bioavailability, as the result of a parallel induction of IB-␣ synthesis and reduction of phosphorylation and degradation. Collectively, these findings document that NF-B activity inhibition contributes to PLC-induced proliferative arrest and Bax-related apoptosis of intimal SMCs.
Intimal smooth muscle cell (SMC) 2 accumulation plays a crucial role in the pathogenesis of vascular diseases, which include postangioplasty restenosis (1,2), and many studies aimed at explaining the molecular pathways regulating the myointimal hyperplastic process. Restenosis remains the most feared complication following percutaneous transluminal angioplasty (3). Since angioplasty procedures grow exponentially (4), even slightly limiting restenosis can be an enormous socioeconomic benefit. SMC apoptosis plays a relevant role in fetal cardiovascular tissue remodeling (5). In normal arterial vessels, postnatal apoptosis is practically absent (6), whereas an increase is observed in pathological conditions, such as atherosclerotic plaques and restenosis (7)(8)(9)(10). In rat myointimal thickening following vascular injury, a common experimental model of postangioplasty restenosis (11,12), an apoptotic reduction of neointimal SMCs counteracts the excessive proliferation and favors the restoration of vascular homeostasis (6,8). Consequently, vascular SMC apoptosis control appears to be a goal in strategies aimed at limiting the clinical impact of restenosis (13). The NF-B proteins and their IB protein inhibitory subunits make up a group of regulatory transcriptional factors in a variety of physiological functions (14,15), which include cell survival (16,17). In regard to vascular SMCs, in vitro experiments demonstrated that NF-B influences bovine SMC proliferation (18) and variably modulates human and rat SMC survival (19), also depending on culture conditions (20). In postinjury rat aortic intimal thickening, increased NF-B levels are detected (19,21), similar to those observed in human fibroatheromatous plaques (22). Balloon angioplasty-induced NF-B activation seems to contribute to lumen loss in rabbit iliac arteries via induction of an inflammatory response and a decrease in apoptotic rate (23). All together, these data suggest that NF-B-regulated apoptosis plays a critical role in postinjury arterial remodeling. Propionyl-L-carnitine (PLC) (24) is a carnitine derivative that has recently been included among pharmacological approaches to peripheral vasculopathy (25). PLC has a higher affinity for the plasma membrane transport system, being more lipophilic and penetrating better than L-carnitine (26). We previously documented that daily PLC administration reduces thickening and increases SMC apoptosis of rabbit carotid neointima 3 weeks after injury (27). Since propyonic acid is a precursor of some nonsteroidal anti-inflammatory drugs (28), it is still unclear whether the proapoptotic effect derives from a direct PLC-SMC interaction or depends on an aspecific reduction of parietal inflammation. In the present study, we investigated the mechanisms through which PLC modulates intimal SMC apoptosis and influences arterial remodeling in vivo and in vitro. Our results document that PLC specifically induces growth inhibition and an increase in the apoptotic rate of intimal SMCs by inhibiting NF-B activation and regulating IB-␣ inhibitory protein expression.

EXPERIMENTAL PROCEDURES
Rats-Male Wistar rats, weighing 270 -290 g and individually housed, were used according to the experimental distribution reported in Table 1. All experiments were performed according to guidelines compatible with the National Institutes of Health Committee on the Care and Use of Laboratory Animals. Rats were anesthetized with Nembutal sodium (Abbott), 35 mg/kg body weight (bw) intraperitoneally, and the thoracic aorta endothelium was removed by an embolectomy catheter (2F-Fogarty, Baxter, American Edwards Laboratories, Anasco, Puerto Rico), as previously reported (29). In sham-operated (SO) rats, the carotid was ligated without ballooning. Immediately following ballooning, a group of randomized rats received 120 mg/kg bw of PLC (Sigma) dissolved in 1 ml of saline solution as vehicle by oral gavage. The remaining balloonized (CTR) and SO rats only received the saline solution. Treatment was repeated daily. Fifteen days later, the animals received 30 mg/kg bw of bromodeoxyuridine (BrdUrd) intraperitoneally (Sigma). Two hours later, rats were anesthetized, and vessels were washed with saline and perfused with buffered formalin through a cannula inserted in the left ventricle. After aortic tissue sampling (30), 4-m-thick serial sections were stained with hematoxylin and eosin or Movat's pentachrome or employed for immunohistochemistry. For the biomolecular studies, rats were anesthetized and killed by means of cervical dislocation, and their aortas were isolated in sterile conditions. Small thoracic aortic rings were frozen in cooled isopentane for cryostatic sections. Remaining intimal and medial tissues were isolated as previously reported (30).
Morphometry, Immunohistochemistry, Proliferation, and Apoptosis in Vivo-To determine the effect of PLC on intimal hyperplasia 15 days after injury, we calculated the relative intimal volume as the ratio of intimal area/(intimal ϩ medial area) ϫ 100 on sections stained with Movat's pentachrome (31).
To quantify intimal and medial SMC proliferation, an anti-BrdUrd monoclonal antibody (Ylem, Avezzano, Italy) was used, and the percentage of BrdUrd-positive nuclei per total cells (BrdUrd labeling index) was calculated (27).
To emphasize SMC apoptosis in vivo, rehydrated sections were stripped from proteins through incubation with 300 units/ml proteinase K (Sigma) for 15 min at 37°C, apoptotic nuclei were revealed by TdT-mediated dUTP-biotin nick-end labeling (TUNEL), and the percentage of positive nuclei per total cells (TUNEL labeling index) was calculated, as previously reported (27). To determine the myocytic nature of proliferating apoptotic cells, double immunohistochemistry was also performed (27).
Proliferation, Immunohistochemistry, and Apoptosis in Vitro-Intimal aortic cells 15 days after injury and medial SMCs from SO rats were isolated by enzymatic digestion and allowed to grow to the fifth passage, as previously reported (29). Cells were plated in sparse conditions (2.5 ϫ 10 3 cells/cm 2 ) and synchronized in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 0.1% fetal calf serum (FCS; Biological Industries, Haemek, Israel) for 24 h. PLC was added at various concentrations in the presence of 0.1 or 10% FCS, and the medium was changed after 2 days. Following the treatments, cells were trypsinized, and the counted/seeded cell ratio was evaluated. Cell viability was calculated using 0.4% trypan blue exclusion in triplicate. [ 3 H]Thymidine incorporation was determined as previously reported (32).
For immunofluorescence staining in vitro, cells growing on glass slides were fixed for 5 min in cold methanol (Ϫ20°C), rinsed twice in phosphate-buffered saline, and incubated with anti ␣-actin, anti-Bax, anti-p65, and anti-p50 monoclonal or polyclonal antibodies (32). Cells were photographed using a Nikon fluorescent microscope and DNA chromatin morphology under UV visualization. Apoptosis of adherent cells was investigated by calculating the percentage of nuclei showing apoptotic features using Hoechst staining in triplicate (32). Assigning a serial number to each slide ensured the objectivity of all measurements.
To determine the role of NF-B in the apoptotic process, SMCs were cultured in the presence of 0.1% FCS and SN50 (Bio-Mol), a peptide that specifically inhibits nuclear translocation of NF-B (33), at a 5 M concentration for 24 h; in some instances, PLC at a 50 M concentration was added 1 h after SN50; alternatively, after 24 h of treatment with PLC, SN50 was added for 1 h. As a control, we repeated the experiments with an equimolar concentration of SN50M (Bio-Mol), an inactive synthetic analog with a mutated nuclear localization sequence (33).
In other experiments, gliotoxin (Sigma) at 5 M was used for 6 h, followed by 24 h of treatment with 50 M PLC.
A phosphorothionate-modified antisense ODN 5Ј-TGCTC-CCGGACCCGTCCST-3Ј, specific for the translation initiation region of rat Bax mRNA (34), and a control scrambled ODN were commercially synthesized (Invitrogen), purified, and transfected into cultures using oligofectamine (Invitrogen) for 45 min and then replaced with a new medium plus ODN alone or in association with 50 M PLC.
Flow Cytometry and Cell Cycle Analysis-Flow cytometry was performed as previously reported using a fluorescence-activated cell sorting scan flow cytometer (BD Biosciences) and a Lysis II program (32). The number of cells in the sub-G 1 (DNA content Ͻ2N), G 0 /G 1 , S, and G 2 /M phases was expressed as a percentage of total events (10,000 cells).
Electrophoretic Mobility Shift and IB-␣ Phosphorylation Assays-To perform the electrophoretic mobility shift assay (EMSA), nuclear extracts were prepared by standard protocols (35). Briefly, 2 ϫ 10 6 cells were collected into 1.5-ml centrifuge tubes with 1 ml of phosphate-buffered saline, pelleted for 1 min, and resuspended in 400 l of cold buffer (10 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 0.5 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride). Samples were incubated on ice for 10 min, vortexed for 10 s, pelleted, and resuspended in 100 l of storage buffer (20 mM HEPES, pH 7.9, 25% glycerol, 420 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.5 mM dithiothreitol, and 0.2 mM phenylmethylsulfonyl fluoride). Samples were then incubated on ice for 20 min and centrifuged. Supernatants were collected for use as nuclear extracts. Protein concentrations were determined (29), and EMSA analysis and supershift assays were performed using an NF-B-specific oligonucleotide (Invitrogen). The sequence was as follows: 5Ј-AGTTGAGGG-GACTTTCCCAGGC-3Ј. The double-stranded nucleotides were end-labeled with [␥-32 P]ATP using T4 polynucleotide kinase; 5 g of nuclear extract was used in each assay for NF-B DNA binding using standard protocols (35). NF-B antibodies used for supershift EMSA were anti-p50 and anti-p65 (Santa Cruz Biotechnology).
IB-␣ phosphorylation was determined by overnight immunoprecipitation of 250 l of cytoplasmic extracts using 2 g of IB-␣ antibody at 4°C. After washing, agarose beads were resuspended in SDS-PAGE, and Western blotting was performed using an antibody recognizing phosphotyrosine residues (1:500), as reported above.
DNA Isolation and Ligation-mediated PCR-In order to better identify and quantify apoptosis-associated DNA fragmentation, DNA was extracted, and ligation-mediated PCR was performed (32). The nucleosomal ladder was quantified in 1.2% agarose gels stained with ethidium bromide (32).
Statistical Analysis-Results were expressed as the mathematical mean of single experiments Ϯ S.E. Results were statistically analyzed by Student's t test. The differences were considered statistically significant with a value of p Ͻ 0.05.

PLC Reduces Intimal Thickening and Potentiates Intimal Cell
Apoptosis in Vivo-In rat aortas 15 days following endothelial injury, a diffuse neointimal thickening was clearly evident and made up of rounded or elongated cells embedded in an abundant extracellular matrix. Ultrastructural analysis demonstrated abundant cytoplasmic synthetic organelles and few peripheral myofilament bundles in intimal cells of both PLC and CTR groups, characteristic of a dedifferentiated vascular SMC phenotype (37). A morphometric analysis demonstrated that PLC treatment induced a reduction in relative intimal vol-ume and an increase in neointimal but not medial TUNEL labeling index compared with the CTR group (p Ͻ 0.02 and p Ͻ 0.01, respectively; Fig. 1, A and B). The TUNEL labeling index of the PLC and CTR tunica media was extremely low and less than the respective overlying intimal value (p Ͻ 0.01) and similar to that of the SO group. On the other hand, the neointimal BrdUrd labeling index did not differ when comparing PLC and CTR groups, although both did increase compared with the respective underlying and SO tunica media (p Ͻ 0.001; Fig. 1C). The PLC, CTR, and SO tunica media BrdUrd labeling index did not differ from each other. Double immunohistochemistry showed slight cytoplasmic positivity for ␣-actin in most BrdUrd (70.7 Ϯ 2.1 and 72.6 Ϯ 2.2%) and TUNEL (73.3 Ϯ 2.5 and 74.6 Ϯ 2.2%) positive intimal cells in PLC and CTR group, respectively. In order to confirm apoptotic rate differences, ligation-mediated PCR was performed on DNA samples extracted from freshly isolated aortic tissue. Densitometric analysis showed a more evident apoptotic ladder in PLC compared with CTR rat neointimal tissue (p Ͻ 0.001; Fig.  1D). A very small amount of fragmented DNA was observed in PCR products from SO aortic tissue. Finally, scattered groups of endothelial cells were focally observed in re-endothelialized areas of the neointimal surface, but their number and extent did not vary when comparing CTR (20.8 Ϯ 4.2 cells/mm 2 ) and PLC groups (27.1 Ϯ 6.2 cells/mm 2 ).
PLC Modulates Intimal Cell Expression of NF-B, IB-␣, and Apoptosis-related Proteins in Vivo-In order to demonstrate the differences in survival-and apoptosis-related protein expression in vivo, serial aortic sections were stained by immunohistochemistry (Fig. 2). In SO tunica media, p65, p50, IB-␣ and Bax protein immunostainings were practically negative. Extensive cytoplasmic immunopositivity for p65 and p50 was observed in most intimal cells; a smaller percentage of nuclear positivity for p65 was also present in CTR intimal cells (6.8 Ϯ 1.8). It was reduced in PLC neointima (3.0 Ϯ 0.8; p Ͻ 0.01) and absent in SO tunica media. In addition, a more extensive positive cytoplasmic survivin reaction was observed in CTR compared with PLC neointimal cells. The opposite was true for IB-␣, Bax, and ␣-actin immunostaining, their expression being greater in PLC compared with CTR neointimal cells. Moreover, a reduction in immunostaining for VCAM-1 and MCP-1, two NF-B-regulated genes (23), was observed in the neointima of PLC-treated rats. Underlying and SO tunica media immunostainings for p65, p50, IB-␣, survivin, Bax, VCAM-1, and MCP-1 were almost negative.
Phenotypic Heterogeneity Influences Proliferation and PLCinduced Apoptosis of SMCs-In order to determine the role of phenotypic differences in the proliferative and apoptotic behavior of SMCs in vitro, aortic intimal cells obtained 15 days after ballooning were compared with SO medial SMCs. Intimal cells appeared characteristically epithelioid with a tendency to grow in small groups. When confluent, they conserved their epithelioid appearance and grew in a single layer (Fig. 3A), differently from medial SMCs, which were spindle-shaped with a characteristic "hill and valley" pattern (Fig. 3B). When cultured in the presence of 10% FCS, intimal cells proliferated more compared with medial SMCs (p Ͻ 0.01; Fig. 3C). The 6-day PLC treatment (Fig. 3D) resulted in a dose-dependent reduction of the counted/seeded cell ratio in intimal (p Ͻ 0.01 for all examined concentrations) but not in medial SMCs. [ 3 H]thymidine incorporation correlated with the growth curves of the two populations (Fig. 3E).
Cytometric analysis (Table 2) demonstrated that in the presence of 10% FCS, the percentage of medial SMCs in G 0 /G 1 was higher compared with intimal cells. The 2-day 50 M PLC treatment inhibited intimal cell G 1 3 S progression, as shown by the higher percentage in the G 0 /G 1 phase, and only a slight increase of the percentage of subdiploid intimal cells was observed. Cell viability was also unchanged at this time. After 4 days of treatment, the percentage of subdiploid apoptotic cells also increased in intimal compared with medial SMCs (p Ͻ 0.01). In 0.1% FCS cultures, the 2-day PLC treatment induced a rapid increase of subdiploid cells in the intimal population compared with control and medial SMCs (p Ͻ 0.01). PLC did not modify the cell cycle in medial SMC cultures. Cell viability corresponded with the cytometric analysis. To confirm the presence of apoptotic cells, we calculated the percentage of condensed or fragmented nuclei in 0.1% FCS sparse adherent cultures. Two days later, the percentage of Hoechst-stained apoptotic cells  was higher in 50 M PLC-treated intimal (6.9 Ϯ 0.8) than in medial SMCs (1.2 Ϯ 0.5; p Ͻ 0.01).
NF-B Inhibition Regulates Intimal Cell Apoptosis-In order to determine the role of NF-B in the apoptotic cascade of intimal cells, Western blotting and EMSA were performed in cytoplasmic and nuclear extracts, respectively. Intimal cells cultured for 4 days with 10% FCS or for 2 days with 0.1% FCS showed a higher constitutive and active NF-B expression than medial SMCs (Fig. 4, A and B).
To confirm whether NF-B plays a role in intimal cell apoptosis, we cultured the cells for 24 h in 0.1% FCS and a nontoxic concentration of SN50, a synthetic peptide that specifically blocks nuclear NF-B translocation (38). Cytometry (Fig. 5A) and Hoechst staining (Fig. 5B) showed that SN50, which inhibits NF-B nuclear translocation, induces an increase in the percentage of apoptosis in intimal (7.9 Ϯ 0.9; p Ͻ 0.01) but not in medial SMCs (2.0 Ϯ 0.5) compared with respective controls, confirming that the reduction of NF-B activity is an important prerequisite for intimal cell apoptosis. The inactive SN50M analog did not induce intimal cell apoptosis. Ligation-mediated PCR confirmed differences in SN50-induced apoptotic response (Fig. 5D). Densitometric blot analysis also demonstrated that adding SN50 resulted in a Bax up-regulation (188 Ϯ 16.5%), cytochrome c release (180 Ϯ 20%), and reduced surviving expression (60.2 Ϯ 10.5%) in the cytosolic fraction of intimal cell cultures.

NF-B Inhibition Associates with PLC-induced Intimal Cell
Apoptosis-In order to determine the role of NF-B in the PLCinduced apoptotic cascade of intimal SMCs, EMSA and Western blotting were performed. In a dose-dependent manner, PLC reduced NF-B activity in intimal cells (Fig. 4C). At 50 M concentration, PLC induced a significant reduction in both constitutive and active NF-B intimal cell expression (p Ͻ 0.01; Fig. 4, A and B). These changes were not evident in medial SMC cultures.
To confirm that NF-B inhibition contributes to PLC-induced apoptosis, we cultured intimal cells for 1 h with SN50 at 5 M concentration. The sequential treatment with 50 M PLC for 24 h failed to induce a further increase of intimal cell apo-ptosis by cytometry (Fig. 5A) and Hoechst staining (7.0 Ϯ 1.0%). Similarly, SN50 did not further significantly increase the 50 M PLC-induced intimal cell apoptosis (8.1 Ϯ 1.0% by Hoechst staining). These data strongly suggest a link between the proapoptotic action of PLC and NF-B inhibition. (Fig. 4A) showed that p65 expression reduction (28.3 Ϯ 4.1%) was parallel to the increase of Bax expression (169.0 Ϯ 16%) after 2 days of PLC treatment, whereas the Bcl-2 level was almost unchanged compared with control intimal cells. Bax, Bcl-2, and cytochrome c expression in medial SMCs was low. A PLC-induced increase in cytochrome c and caspase 3 expression was also detected in intimal (190.5 Ϯ 15 and 165 Ϯ 12%, respectively) but not in PLC-treated medial SMCs. Densitometric analysis confirmed the lower level of ␣-actin (20.5 Ϯ 4%) in intimal compared with medial SMCs (32). PLC induced a slight increase of ␣-actin expression in intimal (146.2 Ϯ 6%; p Ͻ 0.02) but not in medial SMCs (118.5 ϩ 12%) compared with respective controls. RT-PCR showed a dose-dependent reduction of IAP-1 and, less markedly, IAP-2 in PLC-treated intimal cells compared with controls (Fig. 4D). RT-PCR also confirmed the PLC-induced dose-dependent reduction of the transcripts for VCAM-1 and MCP-1 in intimal cells in vitro.

NF-B Mediates Bax Up-regulation and Cytochrome c Release in Apoptotic Intimal
SMCs-To determine if Bcl-2 family member proteins are NF-B-regulated, an antisense ODN against Bax up-regulation was added to intimal cell cultures. Bax antisense ODN, added to intimal cell cultures, blocked Bax up-regulation in a dose-dependent manner (Fig. 4E), whereas Bcl-2 was unchanged. Control scramble ODN had no effect. The absence of Bax up-regulation paralleled the absence of cytochrome c release in the cytosol, suggesting that NF-B controls cytochrome c release through Bax protein expression. Bax antisense ODN also prevented, at least in part, both SN50 and PLC-induced apoptosis and Bax up-regulation, as documented by flow cytometry (Fig. 5E).

TABLE 2 Cell cycle analysis and viability of intimal and medial SMCs
Experiments were performed in the presence of 10 and 0.1% FCS. Cells were plated in sparse conditions (2.5 ϫ 10 3 cells/cm 2 ) and synchronized in 0.1% FCS for 24 h. 50 mM PLC was added or not to the medium and changed after 2 days. Following the treatments, cells were trypsinized, fixed, and stained with propidium iodide. Apoptosis (ϽG 0 /G 1 ) and cell cycle were analyzed by flow cytometry (10,000 events). Cell viability was calculated using 0.4% trypan blue exclusion in triplicate, and results are expressed as mathematical mean Ϯ S.E.  FEBRUARY 16, 2007 • VOLUME 282 • NUMBER 7

PLC Enhances Intimal SMC Apoptosis by Reducing NF-B Activity
PLC Inhibits IB-␣ Phosphorylation and Degradation-To determine if the NF-B inhibitory activity of PLC was due to the inhibition of I〉-␣ degradation, we performed Western blotting on cytoplasmic extracts of intimal cells. After 24 h at 50 M PLC concentration in the presence of 0.1% FCS (Fig. 4A), the IB-␣ level increased compared with control cultures (189 Ϯ 10%; p Ͻ 0.01). Western blotting also revealed a parallel decrease of phosphorylated IB-␣ (Fig. 4F). Altogether, these data suggest that PLC inhibits NF-B nuclear translocation and transcriptional activity by preventing phosphorylation and degradation of IB-␣.
To better elucidate the mechanism by which PLC reduces NF-B activation in intimal cells, we also investigated the dose-dependent effect of PLC on IB-␣, p65, and p50 mRNA levels. Our results showed an induction of IB-␣ mRNA (Fig.  6A, 50-fold versus control at 100 M PLC after 6 h of treatment; p Ͻ 0.01). In contrast, inhibition of both p65 and p50 (55 and 75% inhibition versus control at 100 M PLC after 24 h of treatment, respectively; p Ͻ 0.01) was found.

DISCUSSION
We previously documented (27) that PLC induces a reduction in myointimal thickening and an increase in intimal SMC apoptosis in rabbit carotids 3 weeks after injury. In the present study, we demonstrated that PLC-induced NF-B inhibition in intimal but not in medial SMCs determines an upregulation of Bax-related apoptosis, along with a survivin reduction and an IB-␣ increase. The present results clarify some of the biomolecular mechanisms through which PLC influences vascular SMC survival and document the close link between phenotypic heterogeneity and apoptosis during postarterial injury remodeling. Previous reports describe PLC as specific for cardiac and skeletal muscle and the positive anaplerotic mitochondrial effect of this molecule (40,41). Successively, an anti-ischemic effect of PLC on endothelial cells (42) and an endothelium-dependent vasodilatation in small intact arteries that vanishes in de-endothelialized ones have also been described (43). These pharmacological effects are confirmed from the improved walking capabilities in PLC-treated peripheral arterial disease patients (25,44). Although these findings suggest a beneficial more than a cytotoxic effect of PLC on endothelial cells (43), we did not observe a difference in the extent of the re-endothelialized surface between PLC and CTR aortas. In rabbits, the PLC-induced reduction in restenosis percentage was not accompanied by significant changes in the carotid mean diameter, excluding a vasodilatative effect (27). As a consequence, PLC seems to exert a direct effect on vascular SMCs.
Our study documents that a PLC-induced down-regulation of constitutive and active NF-B fuels intimal SMC apoptosis, as observed in other cell types (16). This strongly suggests that the switching function of NF-B between vascular survival and apoptosis in response to microenvironmental changes is critically regulated by a defined gene subset expression that characterizes different vascular SMC phenotypes (45). In fact, intimal SMCs isolated 15 days after injury show a dedifferentiated phenotype, with low ␣-actin and high constitutive NF-B and Bax levels (32,37), and maintain an epithelioid appearance and high proliferative capacity in vitro (29,46). The same is not observed in medial SMC cultures, which display a contractile and elongated phenotype with low NF-B and Bax levels in vivo and in vitro (20,21,32). Intimal SMCs, unlike those of tunica media, show an increased susceptibility to apoptosis induced by alltrans-retinoic acid in vivo and in vitro (32,47). Two months after injury, rat aortic intimal SMCs revert to a fully differentiated phenotype (48). After this phenotypic switch, intimal SMCs are similar to medial SMCs and show low NF-B expression along with reduced susceptibility to apoptotic stimuli (29,32). These results highlight the crucial role of phenotypic heterogeneity and help to explain apparently contrasting data concerning the role of NF-B in arterial SMC pathobiology (17,18,20,49). The target cell phenotype critically influences the effects on proliferation and apoptosis induced by NF-B modulation (50), similar to what has been reported for other variables (20).
Our results document that the apoptotic rate increase following PLC-induced NF-B inhibition in intimal cells was accompanied by Bax and cytochrome c up-regulation. SN50induced NF-B inhibition also determined Bax up-regulation and successive treatment, whereas PLC failed to induce a further increase of apoptotic rate, suggesting a similar apoptotic pathway. The Bax to Bcl-2 ratio was increased in intimal cells by Bax overexpression, as reported in human radial artery intimal hyperplasia (51). However, as for other cell types (52), Bax overexpression per se is not lethal but, on the contrary, compatible with proliferation in vitro. We also documented that the PLCinduced increase of the apoptotic rate is associated with a reduction of survivin expression in vivo. Survivin belongs to the family of genes known as apoptosis inhibitors and counteracts a constitutive pathway that induces apoptosis during mitosis (53,54). Serum or platelet-derived growth factor-AB stimulates SMC survivin expression and prevents caspase activation (53). It is worth noting that intimal SMCs showed high constitutive IAP-1 and IAP-2 levels, and their PLC-induced reduction was accompanied by an increase in apoptotic rate. IAP-1 and IAP-2 are members of an antiapoptotic protein family that binds and inhibits caspase 3, 7, and/or 9 (55). IAP-1 and -2 reductions seem to favor intimal SMC apoptosis following NF-B inhibition and Bax up-regulation. The relevance of NF-B in the apoptotic pathway of intimal SMCs was sup-ported by the observation that NF-B down-regulation does not involve changes in p53 expression, a transcription factor involved in apoptosis of other cell types (56). Since serum deprivation favors intimal SMC apoptosis, it is likely that a PLC-induced arrest of DNA synthesis in the G 0 -G 1 phase is a prerequisite for the enhanced susceptibility of intimal cells to in vitro apoptosis. As a matter of fact, other reports document high constitutive NF-B expression as essential for vascular SMC proliferation (49), and its inhibition reduces proliferation and migration in vitro (18).
Apoptosis also depends on the balance between NF-B and IB inhibitory proteins that bind and maintain NF-B in an inactive cytoplasmic form (21), preventing its release and nuclear translocation (15,21,57). Consequently, IB levels regulate NF-B bioavailability (58). Specific NF-B activity inhibition by either proteosome inhibitors that prevent IB degradation, by antisense and decoy oligonucleotide, or by IB-␣ overexpression promotes apoptosis induced by physical or chemical agents (59). Our results indicated that PLC acts, at least in part, by increasing IB-␣ bioavailability in intimal cells. More in detail, PLC determined an increase of synthesis and a parallel reduction of IB-␣ degradation, an early event in injured vessels (21,49), in part by reducing its phosphorylation. Therefore, the IB-␣ expression increase contributed to PLC-induced inhibition of NF-B activity. This is indirectly confirmed by the observation that pretreatment with gliotoxin, which specifically inhibits NF-B activation by reducing IB degradation (39), prevents further increase of PLC-induced apoptosis. A similar process has been reported for other drugs interfering with spontaneous or pharmacologically induced NF-B activation (60 -62). The PLC effect coincides with previously reported data showing that IB-␣ overexpression reduces lumen loss in a rabbit iliac artery restenosis model (23). As for SMCs, it has been previously reported that in low density or highly proliferating conditions, NF-B inhibition by 1-chloro-3-tosylamido-7-amino-2-heptanone, a proteolytic IB-␣ degradation inhibitor, results in proapoptotic stimulus in arterial FIGURE 6. PLC-induced NF-B inhibition is mediated by IB-␣, p65, and p50 modulation. A, PLC reduces NF-B activation by modulating IB-␣ and subunit-p65 and p50 mRNA levels. Intimal cells were treated with different PLC concentrations for 6 (left column) and 24 h (right column). IB-␣, p65, and p50 mRNA levels were measured using Q-PCR as described under "Experimental Procedures." The results are expressed as means Ϯ S.E. of two independent experiments, each performed in triplicate (*, p Ͻ 0.02; **, p Ͻ 0.01). B, gliotoxin specifically inhibits the degradation of IB-␣. Intimal cells after serum starvation were treated for 6 h with 5 M gliotoxin in the presence of 0.1% FCS, followed by 50 M PLC for 24 h. Flow cytometric analysis shows that PLC failed to further increase the apoptotic rate.
SMC cultures, whereas in high density or low proliferating conditions, this effect was not detected (20).
Our results strongly support PLC as a possible therapeutic adjuvant in preventing postangioplasty restenosis. For the latter, various approaches have been suggested (63), including adenoviral gene-based (23) and oligonucleotide-based molecular therapy (64) that targets the platelet-derived growth factor receptor or Rho-kinase (65). In general, one limit to the pharmacological approach of preventing restenosis is the trouble in maintaining a local or tissue concentration able to reproduce in humans those beneficial effects documented in experimental models without high or toxic drug concentrations. A desirable pharmacological agent would have to be anti-inflammatory, be able to inhibit SMC proliferation, be tolerable, and be free from significant adverse effects (66). PLC offers many of these advantages, including safe clinical practice (25,44). The effects on myointimal thickening we describe here follow other previously reported protective properties of PLC for blood vessels (64). Future trials should aim to determine PLC effectiveness as an adjuvant systemic pharmacological approach or in drugeluting stents to prevent human restenosis.
In conclusion, our findings provide new insights into the positive arterial remodeling induced by PLC. Moreover, we provide new data on the susceptibility of SMCs from different layers of the arterial wall to apoptosis, reinforcing the main role of phenotypic heterogeneity in the apoptotic cascade of vascular SMCs.