Author + information
- Received December 4, 2009
- Revision received January 14, 2010
- Accepted January 18, 2010
- Published online May 18, 2010.
- Herminia González-Navarro, PhD⁎,
- Yafa Naim Abu Nabah, PhD⁎,
- Ángela Vinué, BSc⁎,
- María J. Andrés-Manzano, BSc†,
- Manuel Collado, PhD‡,
- Manuel Serrano, PhD‡ and
- Vicente Andrés, PhD⁎,†,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Vicente Andrés, Spanish National Cardiovascular Research Center (CNIC), Melchor Fernández Almagro 3, 28029 Madrid, Spain
Objectives The goal of this study was to investigate the role in atherosclerosis of the tumor suppressor protein ARF (human p14ARF, mouse p19ARF) encoded by the CDKN2Agene.
Background Atherosclerosis is characterized by excessive proliferation and apoptosis, 2 cellular processes regulated by CDKN2A. Although recent genome-wide association studies have linked atherosclerotic diseases to a genomic region in human chromosome 9p21 near the CDKN2Alocus, the mechanisms underlying this gene–disease association remain undefined, and no causal link has been established between CDKN2Aand atherosclerosis.
Methods Atherosclerosis-prone apolipoprotein E (apoE)-null and doubly deficient apoE-p19ARFmice were fed an atherogenic diet and sacrificed to quantify atherosclerosis burden in whole-mounted aortas and in aortic cross-sections. Proliferation and apoptosis were investigated in atherosclerotic lesions and in primary cultures of macrophages and vascular smooth muscle cells obtained from both groups of mice.
Results Genetic disruption of p19ARFin apoE-null mice augments aortic atherosclerosis without affecting body weight, plasma lipoproteins, or plaque's proliferative activity. Notably, p19ARFdeficiency significantly attenuates apoptosis both in atherosclerotic lesions and in cultured macrophages and vascular smooth muscle cells, 2 major cellular constituents of atheromatous plaques.
Conclusions Our findings establish a direct link between p19ARF, plaque apoptosis, and atherosclerosis, and suggest that human genetic variants associated to diminished CDKN2Aexpression may accelerate atherosclerosis by limiting plaque apoptosis.
Atherosclerosis and coronary artery disease are the major cause of mortality in Western societies. Atheroma development is a chronic inflammatory process resulting from the interaction between modified lipoproteins, circulating leukocytes, and elements of the arterial wall, including endothelial cells, monocyte-derived macrophages, T cells, and vascular smooth muscle cells (VSMCs) (1–4). The disease in humans and experimental animals is characterized by excessive cell proliferation and apoptosis within the atheroma, 2 processes that affect mainly macrophages and VSMCs (5–9).
The tumor suppressors p15INK4b, p16INK4a, and ARF (p14ARFin humans, p19ARFin mice) constitute one of the main antioncogenic defenses of mammalian organisms (10,11). The CDKN2Agene encodes the nonhomologous proteins p16INK4aand ARF, and the adjacent CDKN2Bgene encodes p15INK4b. p16INK4aand p15INK4bcause growth arrest through the inhibition of CDK4 and CDK6 activity and the ensuing accumulation of the hypophosphorylated form of the retinoblastoma protein (12,13). ARF appears to play a more relevant role as a proapoptotic factor via inhibition of MDM2, a ubiquitin ligase that destabilizes the tumor suppressor p53 (10,11). Furthermore, age-dependent expression of p16INK4aand ARF may limit the regenerative potential of stem cell pools in certain tissues (14–17).
By analyzing a small cohort that included 316 patients and 434 controls, Rodriguez et al. (18) reported lack of association between the risk of suffering myocardial infarction and the single nucleotide polymorphisms (SNPs) rs3731238, rs3814960, and rs3731249 located in the CDKN2Agene. However, recent independent genome-wide association studies that included thousands of controls and patients have identified a number of SNPs located in a region of chromosome 9p21 very close to the CDKN2Agene that are linked to high risk of atherosclerosis and associated cardiovascular disease, including coronary artery disease, myocardial infarction, and ischemic stroke (19–25). Interestingly, Liu et al. (26) recently reported reduced expression of the CDKN2Atranscripts p16INK4a, p14ARF, and the noncoding RNA named ANRIL (antisense noncoding RNA in the INK4 locus) in purified peripheral blood T cells of individuals carrying in homozygosis the atherosclerosis risk G allele of rs10757278. These studies suggest a molecular mechanism by which diminished expression of CDKN2Atranscripts associated to genetic variants in this locus increases the risk of atherosclerosis. However, a causal link between reduced CDKN2Aexpression and atherosclerosis has not been established. Here, we investigated the consequences of inactivating p19ARFon atheroma development in the widely used apolipoprotein E–null (apoE−/−) mouse model of atherosclerosis.
Mice, diets, and metabolic measurements
Care of animals was in accordance with institutional guidelines and regulations. Mice deficient for p19ARF(p19−/−) (27) were backcrossed for more than 8 generations in a C57BL/6J background and then were crossed with apoE−/−mice (C57BL/6J, Charles River, Wilmington, Massachusetts) to generate doubly deficient apoE−/−p19−/−mice. Information for genotyping of mice is provided in the Online Appendix. After weaning, mice were maintained on a low-fat standard diet (2.8% fat; Panlab, Barcelona, Spain). At 2 months of age, male apoE−/−and apoE−/−p19−/−littermates were placed for different periods of time on an atherogenic diet (10.8% total fat, 0.75% cholesterol without sodium cholate, S4892-E010, Ssniff, Soest, Germany).
Plasma lipid levels in mice fasted overnight were measured using enzymatic procedures (WAKO, St. Louis, Missouri). High-density lipoprotein cholesterol (HDL-C) was determined after precipitation of the apolipoprotein B-containing lipoproteins with dextran sulphate/MgCl2(Sigma, St. Louis, Missouri) as previously described (28).
Quantification of atherosclerosis burden
Fat-fed mice were sacrificed and the aorta was removed after in situ perfusion with PBS followed by 4% paraformaldehyde/PBS. Fixation was continued overnight. An operator who was blinded to genotype quantified the extent of atherosclerosis by computer-assisted morphometric analysis (SigmaScan Pro5, Aspire Software International, Ashburn, Virginia) of both whole-mounted aorta stained with Oil Red O (0.2% Oil Red O in 80% MeOH) (Sigma) and hematoxylin/eosin-stained cross-sections of paraffin-embedded aortic root and ascending aorta as previously described (29).
Immunohistochemical analysis of atherosclerotic plaques
Immunohistopathological examination of atheromas performed by a researcher blinded to genotype included the quantification of the content of macrophages, VSMCs, and collagen (Masson's trichrome stain). VSMCs were identified with mouse anti-smooth muscle (SM) α-actin monoclonal alkaline phosphatase-conjugated antibody (1/20 dilution, clone 1A4) and Fast Red substrate (both from Sigma). Macrophages were detected with a rat anti-Mac3 monoclonal antibody (1/200 dilution, clone M3/84, sc-19991, Santa Cruz Biotechnology, Santa Cruz, California), followed by biotin-conjugated goat anti-rat secondary antibody (1/300 dilution, sc-2041, Santa Cruz Biotechnology), streptavidin-HRP (Ref. TS-060-HR, Lab Vision Corporation, Fremont, California), and DAB substrate (BUF021A, AbD Serotec). Specimens were counterstained with hematoxylin.
Cell proliferation in atheromatous lesions from the aortic valves and ascending aorta was estimated by immunostaining with a pre-diluted rabbit anti-Ki67 monoclonal antibody (Clone SP6, MD, Vector Laboratories, Burlingame, California), followed by incubation with a biotinylated anti-rabbit secondary antibody (1/300 dilution, Ref. R0919, Vector Laboratories), streptavidin-HRP, and DAB substrate.
Apoptosis in the aortic valve region was determined using the TUNEL method as recommended by the manufacturer (ApopTag Peroxidase In Situ Apoptosis Detection Kit, Ref. S7100, Millipore Ibérica, Madrid, Spain). Immunocomplexes were detected with DAB substrate and counterstained with hematoxylin.
For double TUNEL-Mac3 and TUNEL-SM α-actin staining, TUNEL-stained slides were incubated with the anti-Mac3 or anti-SM α-actin antibodies, respectively (see the previous text). The TUNEL-Mac3–stained specimens were incubated with biotin-conjugated goat anti-rat secondary antibody and streptavidin-alkaline phosphatase (1/50, Ref. 551008, BD Pharmingen, Madrid, Spain). In both cases, immunocomplexes were detected with Fast Red substrate (Sigma). Apoptotic macrophages and VSMCs were quantified as the percentage of TUNEL-positive cells that were also Mac3-positive or SM α-actin–positive, respectively. Images were captured with an Olympus CAMEDIA-C5060 wide zoom digital camera (Olympus España, Barcelona, Spain) mounted on an Axiolab stereomicroscope (Carl Zeiss Meditec Iberia, Tres Cantos, Spain).
Primary cell cultures
Bone marrow-derived macrophages (BMDMs) were obtained from femoral bone marrow suspensions plated at 3 × 106cells/ml and differentiated for 7 days in the presence of DMEM/10% FBS/10% macrophage colony-stimulating factor (29). VSMC cultures were prepared from thoracic aorta harvested from 3- to 5-month-old animals after 2 digestions in HBSS/Fungizone medium as previously described (29). Briefly, the aorta was first digested with type II collagenase (175 U/ml, Worthington Biochemical Corp., Lakewood, New Jersey) to remove the adventitia, and cell suspensions were obtained after a second digestion with type II collagenase (175 mg/ml) (Worthington) and type I Elastase (0.5 mg/ml) (Sigma). Cultures were maintained at 37°C in a humidified 5% CO2atmosphere. VSMCs were cultured in 20% FBS/DMEM/Fungizone and used until passage 10. The identity of the cells as VSMCs was confirmed by immunohistochemistry using anti-SM α-actin antibody (see the previous text).
In vitro cell proliferation
For proliferation studies, asynchronously growing cultures of BMDMs were plated on glass coverslips and incubated with 50 μM bromodeoxyuridine (BrdU) for 18 h. Cells were fixed for 30 min with 4% paraformaldehyde/PBS, permeabilized with 0.5% Triton X-100/2 M HCl, washed extensively with sodium borate buffer (pH 8.5), and incubated with a mouse monoclonal anti-BrdU antibody (1 h, room temperature, 1/200, clone MoBu-1, 11-286-c100, Exbio, Vestec, Czech Republic) followed by an Alexa Fluor 488–conjugated goat anti-mouse secondary antibody (1 h, room temperature, 1/500, A11029, Invitrogen, Carlsbad, California). Cell nuclei were stained with Hoechst (1/500, 15 min, Sigma) for total cell count, and coverslips were mounted with slow-fade gold antifade reagent (S36936, Invitrogen) and analyzed by fluorescent microscopy. For VSMC FACS-based cell-cycle analysis, cells were synchronized in G0/G1 by 72-h serum-deprivation (DMEM–0.1% FBS) and then restimulated with DMEM–20% FBS for different times. Cells were trypsinized, washed with PBS, collected by centrifugation (5 min, 600g), fixed with 80% ethanol (1 h, −20°C), and labeled with propidium iodide (50 μg/ml, 30 min at room temperature, containing 0.25 mg/ml RNase A). Cells were analyzed using a BD FACSCanto Flow cytometer (BD Biosciences) and DNA histograms were fitted into cell-cycle distributions using the ModFit 3.0 software (Verity Software House, Topsham, Maine).
In vitro apoptosis
Cells were either irradiated with ultraviolet (UV) light (BMDM: 80 J/m2, VSMC: 120 J/m2) and cultured an additional 24 h (BMDM) or 48 h (VSMC), incubated with free-cholesterol at 100 μg/ml (BMDM: 24 h, VSMC: 48 h), or incubated with the nitric oxide donor S-nitrosoglutathione (GSNO) (2 mM, Sigma) (BMDM: 24 h, VSMC: 48 h). For apoptosis analysis by flow cytometry, control and treated cells were collected by trypsinization at 37°C, rinsed with PBS, collected at low speed (700g, 5 min), and fixed for propidium iodide staining as described earlier or for cleaved caspase-3 staining as follows. Cells were fixed 10 min at 37°C with 2% to 4% paraformaldehyde/PBS, permeabilized with prechilled 90% methanol (30 min on ice), and stained with a rabbit polyclonal anti-cleaved caspase 3 (Asp175) antibody (1/200, Cell Signaling Technology, Beverly, Massachusetts), followed by Alexa Fluor 488-conjugated goat anti-rabbit IgG secondary antibody (F[ab′]2 fragment, 1 h at room temperature, 1/500, A-11070, Invitrogen). Apoptotic cells were identified as the sub-G0 peak subpopulation after propidium iodide staining or as cleaved caspase-3 (Asp175)-positive cells.
Data are presented as mean ± SEM. Differences among groups were evaluated by the Student ttest, 2-way ANOVA with Fisher post hoc test (Statview, SAS institute, Cary, North Carolina) or nonparametric Mann-Whitney Utest (GraphPad Prism Software, La Jolla, California). Outliers identified by Grubb's test were not considered for quantification.
Genetic disruption of p19ARFaccelerates atherosclerosis in apoE−/−mice
To assess the role of p19ARFon atherosclerosis, 2-month-old apoE−/−mice and apoE−/−p19−/−siblings were fed a high-fat cholesterol-rich diet for 9 weeks. As shown in Figure 1A,body weight increased in both groups of mice during the period of fat feeding, reaching statistically significant differences in apoE−/−mice at 4 and 9 weeks versus 0 and 1 week, and in apoE−/−p19−/−mice at 4 weeks versus 0 week and at 9 weeks versus 0, 1, and 4 weeks (see p values in Fig. 1A). No differences were observed between apoE−/−and apoE−/−p19−/−mice at any time point analyzed. Likewise, circulating levels of total cholesterol and triacylglycerides increased significantly in both groups of fat-fed mice compared with their pre-diet values (Fig. 1B), but no differences were seen when comparing apoE−/−and apoE−/−p19−/−mice. Thus, body weight and circulating lipids in control and fat-fed apoE-null mice remained unaffected by genetic disruption of p19ARF. We next quantified atherosclerosis burden by planimetric analysis of whole-mounted oil Red-O–stained aortic arch and thoracic aorta and by analyzing cross-sections through the aortic root and ascending aorta. Compared with apoE−/−counterparts, apoE−/−p19−/−mice exhibited a significant increase in atheroma size in the aortic arch (1.5-fold, p < 0.012) (Fig. 1C) and aortic root (1.7-fold increase, p < 0.018) (Fig. 1D), 2 highly atherogenic vascular beds in this animal model. In contrast, p19ARFinactivation did not affect plaque size in the less atherogenic thoracic aorta (Fig. 1C) and ascending aorta (Fig. 1D).
We also examined the consequences of inactivating p19ARFon plaque composition. Analysis of the aortic root and ascending aorta revealed no differences in lesional (neointimal) content of Mac3-immunoreactive macrophages, SM α-actin–immunoreactive VSMCs and collagen between apoE−/−p19−/−and apoE−/−mice (Fig. 2).Likewise, lesion cellularity in the aortic root was similar in both groups (apoE−/−: 6.21 ± 1.02 cells/mm2, n = 11; apoE−/−p19−/−: 6.07 ± 1.13 cells/mm2, n = 10).
Effects of p19ARFinactivation on cell proliferation and apoptosis in atherosclerotic lesions and macrophage and VSMC cultures
We next sought to investigate the mechanisms underlying the atheroprotective action of p19ARF. Since a key event in lesion development is the uptake of modified low-density lipoproteins (LDLs) by macrophages (1), we analyzed this process in cultures of peritoneal macrophages obtained from apoE−/−and apoE−/−p19−/−mice, which exhibited similar uptake of acetylated LDLs (data not shown). We also examined the effect of disrupting p19ARFon proliferation and apoptosis, 2 important cellular processes during atherosclerosis (5–9,30). Asynchronously growing cultures of BMDM obtained from apoE−/−and apoE−/−p19−/−mice exhibited similar proliferative activity measured as BrdU incorporation (Fig. 3A)and flow cytometric analysis (not shown). In contrast, apoE−/−p19−/−VSMCs entered earlier into the S and G2/M phases of the cell cycle compared with apoE−/−controls, as determined by flow cytometric analysis of starvation-synchronized cultures that were restimulated with serum (Fig. 3B). These studies demonstrated a significantly higher percentage of cells in G1/G0 in apoE−/−versus apoE−/−p19−/−VSMCs at all time points analyzed, and significantly higher percentages of cells in S (16 and 24 h) and G2/M (8, 16, and 24 h) in apoE−/−p19−/−versus apoE−/−cultures. Moreover, differences in the percentages of apoE−/−cells in G1/G0, S, and G2/M phases only reached statistical significance after 24 h of serum restimulation compared with 0 h (p < 0.0001). However, in apoE−/−p19−/−VSMCs, differences in G1/G0, S, and G2/M phases reached statistical significance at 8, 16, and 24 h after serum restimulation compared to 0 h (8 h vs. 0 h and 16 h vs. 0 h: p < 0.05; 24 h vs. 0 h: p < 0.0001). Taken together, these findings suggest that serum-restimulated apoE−/−p19−/−VSMCs re-enter the cell cycle faster than apoE−/−cells.
Despite the increased proliferative capacity of cultured VSMC lacking p19ARF, cell proliferation was similar in aortic root cross-sections from apoE−/−and apoE−/−p19−/−mice fed the atherogenic diet for 9 weeks (Fig. 3C), as estimated by the neointimal content of Ki67-positive cells. Likewise, quantification of neointimal Ki67-immunoreactive cells in the ascending aorta did not reveal statistically significant differences between both groups (apoE−/−: 52.9 ± 5.6% Ki67-positive cells, n = 11; apoE−/−p19−/−: 46.9 ± 4.8% Ki67-positive cells, n = 10, p > 0.05). Neointimal cell proliferation was also undistinguishable in apoE−/−and apoE−/−p19−/−mice fed the atherogenic diet for 4 weeks, which exhibited incipient lesions (Online Fig. S1). Interestingly, the mRNA level of the cytostatic gene product p16INK4awas markedly increased in the aortic arch of apoE−/−p19−/−mice fed the atherogenic diet for 4 weeks (∼17-fold increase vs. apoE−/−, p < 0.05) and 9 weeks (∼23-fold increase vs. apoE−/−, p < 0.0005) (Online Fig. S2A). This compensatory mechanism does not occur in cultured VSMCs, since no statistically significant differences in p16INK4aexpression were detected when comparing cells of both genotypes (Online Fig. S2B). We also observed that expression of p15INK4bin both aortic tissue and cultured VSMCs is unaffected by disrupting p19ARF(Online Fig. S2). Thus, p19ARFablation increases significantly the proliferative capacity of cultured VSMCs in vitro but does not affect neointimal cell proliferation at different lesional stages, possibly due to a compensatory up-regulation of p16INK4ain aortic tissue.
To assess the role of p19ARFin plaque's apoptosis, aortic root cross-sections were examined using the terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) method. We found a significant reduction in the percentage of TUNEL-positive cells in atheromas from apoE−/−p19−/−compared with apoE−/−mice, both when quantified as percentage TUNEL-positive cells and as absolute number of TUNEL-positive cells per square millimeter of lesion (Fig. 4A).We also carried out double-staining experiments to identify apoptotic neointimal macrophages (TUNEL/Mac3) and VSMCs (TUNEL/SM α-actin). In both groups of mice, ∼70% and <4% of apoptotic cells in atherosclerotic lesions were macrophages and VSMCs, respectively (Fig. 4B, results expressed as percentage of cells doubly positive for TUNEL and Mac3 or SMα-actin relative to the total number of TUNEL-positive cells within the atheroma).
To further investigate the role of p19ARFin the control of macrophage and VSMC apoptosis, we prepared primary cultures of BMDMs and VSMCs from apoE−/−and apoE−/−p19−/−mice and exposed them to several proapoptotic stimuli, including UV, the nitric oxide donor GSNO, and free cholesterol. As shown in Fig. 5A,p19ARFdeficiency significantly reduced apoptosis in BMDMs treated with all 3 stimuli, as measured using 2 flow cytometry approaches (propidium iodide staining to identify sub-G0 cells and cleaved caspase-3 immunostaining). Similarly, p19ARFinactivation in apoE−/−VSMCs significantly limited apoptosis induced by UV and GSNO (Fig. 5B). Free cholesterol, which induced apoptosis in macrophages (Fig. 5A) (31), failed to provoke apoptotic cell death in apoE−/−and apoE−/−p19−/−VSMCs compared with untreated controls (not shown). This lack of apoptotic response in free-cholesterol–treated VSMCs correlated with the absence of intracellular cholesterol vesicles, which were abundant in BMDMs exposed to this lipid (not shown).
The major finding of the present study is that deficiency of the p19ARFprotein, one of the tumor suppressors encoded by the CDKN2Agene, accelerates atheroma development in the atherosclerosis-prone aortic arch (en face study) and aortic root (analyzed in tissue cross-sections) of hypercholesterolemic apoE-null mice (Fig. 1). This phenotype is not observed in less atherogenic regions of the same mice (e.g., ascending and thoracic aorta), suggesting that p19ARFis not involved in the initiation of atherosclerosis but regulates atheroma development in more advanced stages of the disease. Notably, apoptosis is reduced in apoE−/−p19−/−compared with apoE−/−plaques (Fig. 4), and p19ARFinactivation attenuates apoptosis in primary cultures of macrophages and VSMCs exposed to different pro-apoptotic stimuli (Fig. 5). Of note, recent genome-wide association studies in humans have reported an association between several SNPs in a region of chromosome 9p21 adjacent to the CDKN2A/CDKN2Bloci and the risk of atherosclerotic diseases (19–25). Moreover, it has been shown that expression of transcripts encoded by these genes (e.g., p15INK4b, p16INK4a, ANRIL, and p14ARF, the human homolog of murine p19ARF) is lower in purified peripheral blood T cells of individuals homozygous for the atherosclerosis risk G allele of one of these SNPs (rs10757278) (26). Thus, our findings demonstrating a direct link between p19ARF, plaque apoptosis, and atherosclerosis burden in apoE−/−mice suggest a molecular mechanism by which diminished expression of ARF associated to genetic variants in the CDKN2Agene may increase the risk of atherosclerosis by limiting plaque apoptosis.
Because p19ARFexpression has been shown to promote cellular senescence to avoid proliferation of aberrant, potentially hazardous cells in some pathophysiological settings (e.g., aging and cancer) (10,11), we evaluated aortic cell senescence in apoE−/−(n = 10) and apoE−/−p19−/−(n = 9) mice fed the atherogenic diet for 9 weeks. In both groups of mice, we found essentially negligible senescence-associated beta-galactosidase staining in cryosections through the aortic valves, with some scarce, faintly positive cells within the neighboring cardiac tissue (data not shown). Thus, reduced aortic cell senescence does not appear to contribute to increased atherosclerosis in apoE−/−p19−/−mice. We also investigated the consequences of disrupting p19ARFon the expression of the proto-oncogen Myc, since both proteins have been shown to directly interact, and p14ARFdown-regulates Myc-activated transcription (32). Quantitative polymerase chain reaction (qPCR) analysis demonstrated increased c-myc expression in mouse aortic VSMCs lacking p19ARF(Online Fig. S3A), which, together with the lack of compensatory up-regulation of p16INK4a(Online Fig. S2B), might contribute to the higher proliferative capacity of cultured apoE−/−p19−/−VSMCs (Fig. 3B). Similarly, p19ARFinactivation augmented significantly c-myc expression in the aortic arch of apoE−/−mice fed the atherogenic diet for 4 weeks (Online Fig. S3B), but this response did not correlate with changes in cell proliferation within the atheroma (Online Fig. S1). Moreover, c-myc expression was similar in apoE−/−and apoE−/−p19−/−mice fed an atherogenic diet for 9 weeks (Online Fig. S3B). Thus, the effect of p19ARFon c-myc expression in the aortic arch appears to depend on lesional stage (e.g., increased expression in small lesions at 4 weeks of fat diet and no effect in more advanced lesions after 9 weeks of diet). Hence, further research appears warranted to elucidate the complex functional connections between c-myc, p19ARF, and proliferation at different stages of atherosclerosis.
Both cell proliferation and apoptosis are prevalent in atherosclerotic lesions (5–8). Therefore, changes in the balance between hyperplastic growth and apoptosis of neointimal cells are of major importance in atherosclerosis development. Indeed, atheroma progression is affected in mice with altered expression of several tumor suppressors that modulate cell proliferation and apoptosis. Unlike p21Cip1, whose genetic ablation unexpectedly protects from atherosclerosis in apoE-null mice (33,34), the absence of p53 (35–38), p27Kip1(39,40), and retinoblastoma (41) proteins aggravates atherosclerosis in different murine models (apoE-null, LDL receptor-null, and apoE#x002A;3-Leiden transgenic mice). Excessive proliferation of macrophages and VSMCs has been suggested as a mechanism contributing to the acceleration of atherosclerosis in mice lacking p27Kip1, retinoblastoma, and p53. We also found increased atherosclerosis burden in apoE-null mice lacking p19ARF; however, this phenotype did not appear to correlate with increased cell proliferation within the plaque, possibly due to a marked up-regulation in aortic tissue of the INK4 family growth suppressor p16INK4a(Online Fig. S2A). This compensatory mechanism does not appear to occur in cultured VSMCs since apoE−/−p19−/−VSMCs exhibited increased proliferative capacity compared with apoE−/−controls, and no statistically significant differences in p16INK4aexpression were detected between cells of both genotypes (Online Fig. S2B).
In agreement with the well-established proapoptotic function of p19ARF, our in vivo studies reveal a 2-fold reduction in the content of apoptotic cells in apoE−/−p19−/−compared with apoE−/−plaques (Fig. 4A). Notably, we find that the relative proportion of apoptotic neointimal macrophages and VSMCs is unchanged in apoE−/−p19−/−and apoE−/−mice (∼70% and <4%, respectively) (Fig. 4B), suggesting that lack of p19ARFprotects from apoptosis to the same extent in both cell types. Future studies are required to investigate whether apoptosis of other cell types present in the atheroma might be differentially affected by p19ARFablation. Consistent with our in vivo results, p19ARFinactivation also reduces apoptosis by 30% to 57% in BMDMs and VSMCs cultures exposed to different proapoptotic stimuli (Fig. 5). The finding that GSNO-induced apoptosis is attenuated in apoE−/−p19−/−cells is in concordance with previous studies demonstrating that efficient nitric oxide-dependent macrophage apoptosis requires the activation of p19ARF(42).
Consistent with the possibility that decreased apoptosis is a mechanism by which p19ARFinactivation aggravates atheroma progression, a reduction in apoptotic cell death, accompanied or not by changes in cellular plaque proliferation, has been previously associated with increased atherosclerosis in several genetically modified mouse models. For example, ablation of p53, a downstream target of p19ARF, reduces apoptosis and increases atheroma formation in some of the models studied (37,38). Moreover, the unexpected aggravation of atherosclerosis in apoE−/−compared with apoE−/−p21−/−mice also coincides with reduced neointimal apoptosis (33), in agreement with the reported antiapoptotic role of p21Cip1(43). However, decreased macrophage apoptosis might be beneficial in advanced stages of atherosclerosis by reducing the risk of plaque rupture (44).
Genetic inactivation of p19ARFin atherosclerosis-prone apoE-null mice limits apoptosis of neointimal macrophages and VSMCs, and worsens atherosclerosis without affecting neointimal cell proliferation, possibly due to a compensatory up-regulation of p16INK4a. These findings suggest that interventions aimed at increasing the expression of ARF may lessen plaque progression. Our results also support the notion that decreased plaque apoptosis might contribute to increased risk of atherosclerosis in individuals homozygous for the G allele of rs10757278, who exhibit reduced levels of not only ARF, but also p15INK4b, p16INK4a, and ANRIL transcripts in blood-borne T cells (26). Thus, it will be of interest to generate additional mouse models to determine whether atherosclerosis is also aggravated upon deficiency of p15INK4b, p16INK4a, and ANRIL, and if combined inactivation of 2 or more CDKN2A/CDKN2Bgenes potentiates atherosclerosis more than does single-gene ablation. Moreover, further studies are warranted to ascertain whether other atherosclerosis risk variants near the CDKN2A/CDKN2Bloci also lead to diminished expression of INK4/ARFtranscripts.
For an expanded Methods section and supplementary figures, please see the online version of this article.
Supported by grants from the Spanish Ministry of Science and Innovation (MICINN)and European Regional Development Fund(SAF2007-62110), and from Instituto de Salud Carlos III (ISCIII)(RECAVA: RD06/0014/0021). Dr. González-Navarro, Dr. Naim Abu Nabah, and Ms. Vinué received salary support from CSIC's JAE-Doctor program, ISCIII, and MICINN, respectively. The CNIC is supported by the MICINNand the Pro-CNIC Foundation.
- Abbreviations and Acronyms
- apolipoprotein E
- bone marrow-derived macrophage
- low-density lipoprotein
- smooth muscle
- single nucleotide polymorphism
- terminal deoxynucleotidyl transferase dUTP nick-end labeling
- ultraviolet light
- vascular smooth muscle cell
- Received December 4, 2009.
- Revision received January 14, 2010.
- Accepted January 18, 2010.
- American College of Cardiology Foundation
- Andrés V.
- Walsh K.,
- Isner J.M.
- Fuster J.J.,
- Fernández P.,
- González-Navarro H.,
- Silvestre C.,
- Abu Nabah Y.N.,
- Andrés V.
- Helgadottir A.,
- Thorleifsson G.,
- Manolescu A.,
- et al.
- McPherson R.,
- Pertsemlidis A.,
- Kavaslar N.,
- et al.
- Broadbent H.M.,
- Peden J.F.,
- Lorkowski S.,
- et al.
- Matarin M.,
- Brown W.M.,
- Singleton A.,
- Hardy J.A.,
- Meschia J.F.
- González-Navarro H.,
- Nong Z.,
- Amar M.J.,
- et al.
- Gonzalez-Navarro H.,
- Vinue A.,
- Vila-Caballer M.,
- et al.
- Kockx M.M.,
- De Meyer G.R.,
- Muhring J.,
- Jacob W.,
- Bult H.,
- Herman A.G.
- Li Y.,
- Gerbod-Giannone M.C.,
- Seitz H.,
- et al.
- Merched A.J.,
- Chan L.
- Andrés V.
- Merched A.J.,
- Williams E.,
- Chan L.
- van Vlijmen B.J.,
- Gerritsen G.,
- Franken A.L.,
- et al.
- Mercer J.,
- Figg N.,
- Stoneman V.,
- Braganza D.,
- Bennett M.R.
- Díez-Juan A.,
- Andrés V.
- Díez-Juan A.,
- Pérez P.,
- Aracil M.,
- et al.
- Boesten L.S.,
- Zadelaar A.S.,
- van Nieuwkoop A.,
- et al.
- Zeini M.,
- Traves P.G.,
- Lopez-Fontal R.,
- et al.
- Gartel A.L.,
- Tyner A.L.
- Gautier E.L.,
- Huby T.,
- Witztum J.L.,
- et al.