p21Cip1 Levels Differentially Regulate Turnover of Mature Endothelial Cells, Endothelial Progenitor Cells, and In Vivo Neovascularization
p21Cip1 (p21) controls cell cycle progression and apoptosis in mature endothelial cells (ECs) and regulates size and cycling of the hematopoietic progenitor cell pool. Because circulating endothelial progenitor cells (EPCs) contribute to postnatal neovascularization in addition to mature ECs, we investigated the regulation of ECs and EPCs in p21-deficient mice. Mature aortic EC proliferation was increased in homozygous p21−/− and heterozygous p21+/− mice, in which p21 protein levels are reduced to one third of wild-type (WT). In contrast, apoptosis sensitivity was increased by 3.5-fold only in p21−/−, but not in p21+/− mice. Consistently, in vivo apoptosis of ECs within areas of neovascularization was elevated in p21−/− but not in p21+/− mice. EPC numbers were elevated 2-fold in p21−/− mice compared with WT (P<0.001), and clonal expansion capacity of EPCs was increased from 25±4 (WT) to 57±8 colony-forming units in p21−/− mice (P<0.005). EPC numbers and expansion were likewise increased in p21+/− mice. As the integrative endpoint, in vivo neovascularization reflecting all p21-affected parameters was increased over WT only in p21+/− (P<0.001), but not in p21−/− mice. In conclusion, reduced p21 protein levels of mice lacking one p21 allele are associated with increased proliferation of ECs and EPCs, whereas survival of ECs to apoptotic stimuli in vitro and in vivo is not impaired. Under these conditions, neovascularization was increased. In contrast, complete p21 deficiency did not result in an increased neovascularization despite increased mature EC and EPC proliferation. This may be due to the sensitization of ECs against apoptosis.
Angiogenesis and vasculogenesis contribute as two essentially different mechanisms to blood vessel formation in the adult organism. Angiogenesis is understood as capillary sprouting from preexisting blood vessels, which relies on the proliferation, survival, and migration of mature endothelial cells (ECs; see reviews1,2). In contrast, vasculogenesis is confined to new vessel development arising from circulating endothelial progenitor cells (EPCs), which are released from the bone marrow into the circulation and home to sites of neovascularization.3–6
The cell cycle protein p21Cip1 (p21) regulates cell cycle progression7 and inhibits apoptosis8 of mature ECs, suggesting a potential role for angiogenesis signaling.9 Likewise, p21 maintains the quiescence of hematopoietic stem cells,10 which may serve as common precursors for vascular progenitor cells. Therefore, p21 could also regulate vasculogenesis. p21 regulates the cell cycle concentration-dependently in an ambiguous manner.11 At low levels, p21 is sequestered to cyclin D during mitogenic stimulation facilitating the activation of cyclin D–dependent kinases 4 and 6 and, thus, cell cycle progression,12 whereas at high concentrations, p21 robustly inhibits cyclin-E–dependent kinase 2, leading to cell cycle arrest.13–15
p21-deficient mice develop autoimmune disease16,17 and are characterized by enhanced tumor susceptibility.18 Interestingly, the latter includes in particular the formation of benign hemangiomas,18 which may highlight the dependency of endothelial cell cycle regulation on p21. Besides angioma formation, however, no apparent vascular phenotype of p21−/− mice has been reported. We, therefore, investigated the native adult blood vessel formation by the disc neovascularization model in p21-deficient mice. To understand the cellular mechanisms regulating angiogenesis and vasculogenesis in these animals, we examined the effect of p21 deficiency on both, mature EC growth and survival as well as EPC number and function. Because of the concentration-dependency of p21 effects, we studied heterozygous p21+/− mice in addition to p21−/− mice.
Materials and Methods
Heterozygous p21+/− and homozygous p21−/− mice were developed in a C57/BL6 background strain of mice and kindly provided by S.J. Elledge.18a Age-matched littermate wild-type C57/BL6 mice were used as controls. The hetero- or homozygotic absence of one or both p21 alleles from the mouse genome, respectively, was confirmed by PCR using the following primers: 5′-tcctggtgatgtccgacctg-3′ (forward p21); 5′-tccgttttcggccctgag-3′ (reverse p21); 5′-gcgaggatctcgtcgtgac-3′ (neomycin cassette forward); 5′-tcatcaatttatgcagac-3′ (neomycin cassette reverse). The present study was performed with permission of the State of Hesse, Regierungspräsidium Darmstadt, according to section 8 of the German Law for the Protection of Animals and conforms to the Guide for the Care and Use of Laboratory Animals.
Isolation of Mature Aortic ECs
Outgrowth of ECs from aortic tissue was performed as previously described.19,20 In brief, aortas from female or male mice were cleaned, longitudinally opened, and placed with the intimal side downward onto a collagen matrix in a medium supplemented with 10% FCS and 100 μg/mL endothelial cell growth supplement (Calbiochem) for 6 days. After digestion of the matrix, the remaining cells were pelleted, seeded on fibronectin-coated 24-well plates, and grown to confluence in endothelial basal medium (EBM; Cell Systems). To identify mature ECs after aortic outgrowth, cells were trypsinized, stained with a phycoerythrin (PE)-labeled anti–Flk-1 antibody (BD Pharmingen) or with an FITC-labeled antibody against CD146 (S-Endo-1, MUC-18), and analyzed by fluorescence-activated cell sorting (FACS Calibur, BD Biosciences).
Measurement of Proliferation
For proliferation analysis, cells were, in addition to PE-labeled anti–Flk-1, incubated with bromodeoxyuridine (BrdU) and subsequently stained with a FITC-labeled anti-BrdU antibody (BD Pharmingen) and with the DNA dye 7-amino-actinomycin D (7-AAD). Then, only Flk-1–positive cells were selectively assessed for cell cycle phases using 3-channel flow cytometry. These data were confirmed by the quantitative measurement of BrdU incorporation using an enzyme-linked immunosorbent assay (ELISA, Roche Diagnostics) after incubation of the cells with 10 μmol/L BrdU for 1 hour.
Measurement of Apoptosis Sensitivity In Vitro
For assessment of apoptosis sensitivity, cells were washed with phosphate-buffered saline (PBS) and incubated in EBM without serum and growth factor supplementation. Then, cells were trypsinized and incubated with a PE-labeled anti–Flk-1 antibody, FITC-labeled annexin V, and 7-AAD. Flk-1–positive annexin V–positive cells detected by flow cytometry were considered as apoptotic mature ECs, with 7-AAD-positivity as an additional marker of late-stage apoptosis.
Detection of Caspase-3 Activation
Mouse aortic endothelial cells were fixed and permeabilized before the addition of a rabbit affinity-purified polyclonal antibody raised against amino acid 163 to 175 of murine caspase-3 (Cell Technology), which is specifically present on the p17 subunit of cleaved caspase-3. A secondary PE-labeled goat anti-rabbit antibody and subsequent flow cytometric analysis was used to quantify caspase-3 activation.
Isolation and Characterization of Progenitor Cells
Total mononuclear cells (MNCs) were isolated from homogenized murine splenic tissue by density gradient centrifugation with Biocoll separating solution (density 1.077; Biochrom AG). MNCs (4×106) were plated on fibronectin-coated 24-well plates in 0.5 mL endothelial basal medium (EBM; Cell Systems) supplemented with 1 μg/mL hydrocortisone, 3 μg/mL bovine brain extract, 30 μg/mL gentamicin, 50 μg/mL amphotericin B, 10 μg/mL EGF, 20 ng/mL vascular endothelial growth factor (VEGF), and 20% fetal calf serum (FCS; Gibco). At day 4, adherent cells were washed with medium and incubated with 2.4 μg/mL 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine–labeled acetylated LDL (DiLDL; Harbor Bio-Products) for 1 hour. Cells were fixed in 4% paraformaldehyde and counterstained with FITC-labeled lectin (Sigma RBI). Two independent investigators counted the number of EPCs in three randomly selected high-power fields.
Progenitor cell function was quantitatively assessed by their ex vivo clonal expansion capacity. For this purpose, 1×105 MNC isolated from splenic tissue were kept in 1.5 mL methylcellulose agar (Methocult GF M3434, Cell Systems), which discriminates between cells with true progenitor potency giving rise to cell colonies.21 In order to induce clonal expansion, the agar was supplemented with granulocyte-monocyte colony-stimulating factor (GM-CSF; 20 ng/mL) and with murine VEGF to induce EC formation (100 ng/mL; both from Cell Concepts). After 2 weeks of incubation, formed colonies were microscopically evaluated with regard to their morphology. Endothelial colonies were representatively confirmed by DiLDL and lectin double-staining.
Human Umbilical Vein Endothelial Cell Culture
Pooled human umbilical vein endothelial cells (HUVECs; Cell Systems/Clonetics) were cultured in EBM supplemented with hydrocortisone (1 μg/mL), bovine brain extract (3 μg/mL), gentamicin (50 μg/mL), amphotericin B (50 μg/mL), epidermal growth factor (10 μg/mL), and 10% FCS until the third passage. After detachment with trypsin, cells (3.5×105 cells) were grown on 60-mm culture dishes for at least 18 hours.
For small interference RNA (siRNA) oligonucleotide transfections, scrambled siRNA (5′-UCGUACGGUCAUCGUCAUC-3′) or siRNA (5′-GAUGACGAUGACCGUACGA-3′; each 360 pmol) corresponding to the human p21 mRNA sequence were transfected into HUVECs (3.5×105 cells per 6-cm well) using GeneTrans II (Mobitec) according to the instructions of the manufacturer.
For overexpression of p21, aortic ECs from wild-type or p21−/− mice were transiently transfected with a c-myc–tagged p21 plasmid encoding wild-type full-length p217 using Superfect transfection reagent (Qiagen).
Western Blot Analysis
Cell were lysed with 200 μL buffer (20 mmol/L Tris [pH 7.4], 150 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L EGTA, 1% Triton, 2.5 mmol/L sodium pyrophosphate, 1 mmol/L β-glycerophosphate, 1 mmol/L Na3VO4, 1 μg/mL leupeptin, and 1 mmol/L phenylmethylsulfonyl fluoride) for 15 minutes on ice. After centrifugation for 15 minutes at 20 000g (4°C), protein content was determined according to the Bradford method. Homogenates (40 μg per lane) were separated on 12% SDS-polyacrylamide gels and transferred to polyvinylidene difluoride membranes, which were then incubated with monoclonal antibodies to p21 (BD Pharmingen) or tubulin (NeoMarkers). Then, films were scanned and semiquantitatively analyzed, and the protein ratio of p21 versus tubulin was calculated.
Disc Neovascularization Model
According to the disc angiogenesis system introduced by Fajardo et al,22 polyvinyl alcohol sponge discs (11 mm in diameter and 1 mm in thickness; Rippey) covered at each side with nitrocellulose cell-impermeable filters (Millipore) were subcutaneously implanted into the back of 3-month-old anesthetized C57/BL6 wild-type, p21+/−, or p21−/− mice. Two weeks later, space-filling fluorescent microspheres (0.2 μm; Molecular Probes Inc) were injected into the left ventricle to assess functional perfusion of neovascularization of the discs before discs were explanted.23 After explantation, both the vascularized area of the disc invested by vascular ingrowth and the vessel density were quantified by fluorescence microscopy.24
In order to determine EC apoptosis during adult vessel formation, we implanted additional discs into wild-type, p21+/−, or p21−/− mice for 10 days, whereupon discs were explanted, paraffin-embedded, sectioned, and stained against the endothelial-specific epitope MECA32 (BD Pharmingen), and apoptosis was visualized by DNA nick-end labeling according to the TUNEL procedure (Roche Diagnostics Inc).
Data are expressed as mean±SEM or as stated otherwise. Two treatment groups were compared by the independent sample t test, and three or more groups by one-way ANOVA followed by post hoc analysis adjusted with an LSD correction for multiple comparisons (SPSS). Results were considered statistically significant when P<0.05.
In order to assess the effect of p21 on mature EC proliferation and apoptosis, ECs were cultivated out of aortic tissue as previously described.19,20 After 10 days in culture over 2 to 3 passages, 65±3% or 56±8% of the cells were positive for the EC-specific surface markers Flk-1/KDR/VEGF-R2 receptor tyrosine kinase or CD146 (S-Endo-1; MUC-18), respectively (Figure 1A). Western blot analysis of whole cell lysates confirmed the absence of p21 protein from ECs derived from homozygous p21−/− mice, whereas p21 protein levels were significantly reduced in cells from heterozygous p21+/− mice (Figure 1B).
Mature EC Proliferation and Apoptosis
We then analyzed the proliferation of mature ECs by BrdU/7AAD staining followed by flow cytometric measurement. Only Flk-1–positive cells were evaluated for cell cycle progression. The percentage of cells in S-phase was increased in both heterozygous p21+/− as well as in homozygous p21−/− mice when compared with wild-type animals (Figure 2A). In parallel, cells from both heterozygous p21+/− and homozygous p21−/− mice revealed significantly higher BrdU incorporation than ECs from wild-type mice as assessed by ELISA (Figure 2B). In contrast, apoptosis sensitivity toward serum/growth factor withdrawal determined by flow cytometric analysis of annexin binding to Flk-1–positive cells was enhanced only in cells derived from homozygous p21−/− mice, whereas apoptosis induction was not significantly different between mature ECs from heterozygous p21+/− and wild-type mice (Figure 2C). In order to assess apoptosis induction by an additional, specific parameter of apoptosis, we measured caspase-3 activation by flow cytometric detection of active caspase-3. In parallel with the enhanced increase in annexin V binding in p21−/− cells following serum starvation, the percentage of cells with activated caspase-3 was higher in ECs from p21−/− mice compared with both wild-type and p21+/− cells (Figure 2D). Likewise, depletion of p21 in cultivated human ECs by siRNA oligonucleotide transfection induced a 2-fold increase in basal apoptosis when compared with control oligonucleotides (Figure 2E). In order to investigate, whether the increased apoptosis susceptibility of p21−/− cells is specifically due to the lack in p21, we overexpressed p21 by transient transfection of a plasmid encoding wild-type p21 (Figure 2F, inset). When compared with control, transfection of p21 completely abrogated the increased sensitivity toward apoptosis induction after serum starvation for 18 hours in ECs from p21−/− mice. (Figure 2F). These data suggest that the enhanced apoptosis sensitivity of ECs from p21−/− mice can be rescued by p21 overexpression and, thus, is indeed specifically attributable to the deficiency in p21.
Endothelial Progenitor Cells
To determine the number of EPCs, mononuclear cells were isolated from splenic tissue and characterized by dual positivity for DiLDL uptake and lectin staining as previously described.25 The number of EPCs was significantly increased in both heterozygous p21+/− as well as homozygous p21−/− mice compared with wild-type mice (Figure 3A). We confirmed these data by a methylcellulose colony-forming assay. Endothelial cell–specific colony-forming units (CFUs) were increased in both heterozygous p21+/− and homozygous p21−/− mice in comparison with wild-type animals (Figure 3B). Likewise, both heterozygous p21+/− and homozygous p21−/− mice both showed an increase in CFUs (Figure 3B).
Neovascularization and EC Apoptosis In Vivo
We then assessed the in vivo relevance of p21 protein concentration for neovascularization using the disc angiogenesis model. Native neovascularization was assessed by quantification of the vascularized area and by measurement of caliber and density of the vascular network. The vascularized area was increased only in heterozygous p21+/− mice, whereas it was not significantly different between homozygous p21−/− and wild-type mice (Figure 4). Likewise, in heterozygous p21+/− mice, the fluorescence intensity reflecting caliber and density of the vascular network was increased by about 3.5-fold, whereas in homozygous p21−/− mice the vessel density showed a nonsignificant 2-fold increase compared with WT mice (Figure 4).
Because ECs from p21−/− mice were significantly more sensitive to ex vivo apoptosis induction, we investigated whether the lack of a significant neovascularization increase in p21−/− mice is associated with increased EC apoptosis in vivo. Therefore, we determined EC apoptosis rates of newly formed vessels within discs. In order to monitor apoptosis selectively in ECs, we performed a counterstaining with the endothelial marker MECA32. Homozygous p21−/− but not p21+/− mice displayed a significantly elevated percentage of apoptotic ECs in disc vessels (Figure 5), indicating that indeed increased EC apoptosis in p21−/− mice is associated with reduced neovascularization in p21−/− mice compared with p21+/− mice.
In this study, we investigated the concentration-dependent effects of the cell cycle protein p21 on growth and survival of mature ECs as well as on the number and function of EPCs. In addition, we determined the concentration-dependent effects of p21 on adult neovascularization as an associated biological function in a mouse model of p21 deficiency. Our data indicate that loss of one p21 allele causes decreased p21 protein levels in aortic mature ECs, increased proliferative capacity of mature ECs, and maintained protection of these cells against apoptosis in vitro and in vivo. In addition, the number and clonal expansion capacity of EPCs is significantly augmented, and adult blood vessel formation is significantly increased. In contrast, lack of both alleles of p21, and thus, complete deficiency in p21 protein, is associated with increased apoptosis susceptibility of mature aortic ECs in vitro and increased apoptosis rates during neovascularization in vivo. Although the proliferative capacity of both mature ECs and EPCs is increased, p21−/− did not show increased neovascularization.
Low levels of p21 in mice lacking one p21 allele were still sufficient to prevent EC apoptosis, whereas ECs from mice with a homozygous p21 deficiency displayed increased apoptosis in vitro and in vivo, consistent with data on increased T-cell apoptosis in p21−/− mice.17,18 The mechanism, by which p21 protects ECs against apoptosis, is still unclear. The apoptosis-suppressive capabilities of p21 were hypothesized to be linked to the induction of growth arrest by p21.26 However, because the unchanged protection against apoptosis observed in heterozygous p21+/− mice coincides with accelerated cell cycle progression of mature ECs, growth arrest at p21-sensitive cell cycle check-points is likely not to be attributed to the prosurvival role of p21. In accordance, the antiapoptotic effect of p21 was shown to occur independent of cdk inhibition in cardiac myocytes.27 Beyond cell cycle regulation, p21 may also affect the mitochondrial pathway of apoptosis in colon cancer cells28 and inactivate caspase-3 by complex formation in hepatoma cells.29 In addition to the current knowledge about the prosurvival effects of p21, our data indicate that low levels of endogenous p21 as determined in p21+/− mice are essential to maintain apoptosis resistance in ECs.
Our data also show that EPC numbers and endothelial cell–specific as well as granulocyte/monocyte colony-forming units are increased in homo- and in heterozygous p21 knockout mice. In accordance with these results, complete p21 deficiency was previously shown to induce increased hematopoietic stem cell numbers and proliferation of these cells.10 This increase in stem cell cycling resulted in a higher mortality of the animals after myelotoxic injury associated with premature stem cell pool exhaustion, suggesting that p21 is required for the maintenance of stem cell quiescence and for the protection of the stem cell compartment.10 In our model, in which we used no pharmacological stimulus to induce the release of progenitor cells from the bone marrow to stimulate disk neovascularization, we did not observe any signs of impaired bone marrow stem/progenitor cells. In contrast, colony-forming capacity of the monocytic and endothelial lineages was increased in both homozygous and heterozygous p21 knockout mice. Future studies are needed to assess the effects of p21 haploinsufficiency or complete deficiency on postnatal blood vessel development in aged animals.
In conclusion, we show that decreased p21 protein levels in mice lacking one p21 allele increases the proliferative capacity of aortic mature ECs, whereas protection of these cells against apoptotic stimuli is still maintained. In addition, haploinsufficiency of p21 enhances number and clonal expansion capacity of EPCs and augments adult blood vessel formation in vivo. In contrast, complete p21 protein deficiency in mice lacking both alleles of p21 sensitizes ECs to apoptosis induction in vitro and results in increased EC apoptosis during neovascularization, which is not enhanced in p21−/− animals despite increased proliferation of both mature ECs and EPCs. Our data therefore suggest that the protection against apoptosis is essential to ensure that an increase in neovascularization results from accelerated cycling of mature and/or endothelial progenitor cells.
This study was supported by grants from the Deutsche Forschungsgemeinschaft (SFB project B6 and Forschergruppe FOR 501/1 Di 600/6-1 to S.D.). We are grateful to Christiane Mildner-Rihm, Christine Goy, Sylvia Rhiel, and Rebeca Salguero Palacios for expert technical assistance.
↵*Both authors contributed equally to this study.
Original received March 11, 2003; revised resubmission received January 22, 2004; resubmission received December 8, 2003; accepted January 22, 2004.
Dimmeler S, Zeiher AM. Endothelial cell apoptosis in angiogenesis and vessel regression. Circ Res. 2000; 87: 434–439.
Asahara T, Murohara T, Sullivan A, Silver M, van der Zee R, Li T, Witzenbichler B, Schatteman G, Isner JM. Isolation of putative progenitor endothelial cells for angiogenesis. Science. 1997; 275: 964–967.
Shi Q, Rafii S, Wu MH, Wijelath ES, Yu C, Ishida A, Fujita Y, Kothari S, Mohle R, Sauvage LR, Moore MA, Storb RF, Hammond WP. Evidence for circulating bone marrow-derived endothelial cells. Blood. 1998; 92: 362–367.
Asahara T, Masuda H, Takahashi T, Kalka C, Pastore C, Silver M, Kearne M, Magner M, Isner JM. Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological and pathological neovascularization. Circ Res. 1999; 85: 221–228.
Rössig L, Jadidi AS, Urbich C, Badorff C, Zeiher AM, Dimmeler S. Akt-dependent phosphorylation of p21Cip1 regulates PCNA binding and proliferation of endothelial cells. Mol Cell Biol. 2001; 21: 5644–5657.
Chavakis E, Dimmeler S. Regulation of endothelial cell survival and apoptosis during angiogenesis. Arterioscler Thromb Vasc Biol. 2002; 22: 887–893.
Cheng T, Rodrigues N, Shen H, Yang Y, Dombkowski D, Sykes M, Scadden DT. Hematopoietic stem cell quiescence maintained by p21cip1/waf1. Science. 2000; 287: 1804–1808.
Sherr CJ, Roberts JM. CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev. 1999; 13: 1501–1512.
Cheng M, Olivier P, Diehl JA, Fero M, Roussel MF, Roberts JM, Sherr CJ. The p21Cip1 and p27Kip1 CDK “inhibitors” are essential activators of cyclin D-dependent kinases in murine fibroblasts. EMBO J. 1999; 18: 1571–1583.
Santiago-Raber ML, Lawson BR, Dummer W, Barnhouse M, Koundouris S, Wilson CB, Kono DH, Theofilopoulos AN. Role of cyclin kinase inhibitor p21 in systemic autoimmunity. J Immunol. 2001; 167: 4067–4074.
Martin-Caballero J, Flores JM, Garcia-Palencia P, Serrano M. Tumor susceptibility of p21Waf1/Cip1-deficient mice. Cancer Res. 2001; 61: 6234–6238.
Hoffmann J, Haendeler J, Aicher A, Rossig L, Vasa M, Zeiher AM, Dimmeler S. Aging enhances the sensitivity of endothelial cells toward apoptotic stimuli: important role of nitric oxide. Circ Res. 2001; 89: 709–715.
Mayani H, Dragowska W, Lansdorp PM. Cytokine-induced selective expansion and maturation of erythroid versus myeloid progenitors from purified cord blood precursor cells. Blood. 1993; 81: 3252–3258.
Hauck L, Hansmann G, Dietz R, von Harsdorf R. Inhibition of hypoxia-induced apoptosis by modulation of retinoblastoma protein-dependent signaling in cardiomyocytes. Circ Res. 2002; 91: 782–789.
Javelaud D, Besancon F. Inactivation of p21WAF1 sensitizes cells to apoptosis via an increase of both p14ARF and p53 levels and an alteration of the Bax/Bcl-2 ratio. J Biol Chem. 2002; 277: 37949–37954.