Reversal of Senescence in Mouse Fibroblasts through Lentiviral Suppression of p53
2003; Elsevier BV; Volume: 278; Issue: 14 Linguagem: Inglês
10.1074/jbc.c300023200
ISSN1083-351X
AutoresAnnette M.G. Dirac, René Bernards,
Tópico(s)Cancer Research and Treatments
ResumoSenescence is generally defined as an irreversible state of G1 cell cycle arrest in which cells are refractory to growth factor stimulation. In mouse embryo fibroblasts (MEFs), induction of senescence requires the presence ofp19ARF and p53, as genetic ablation of either of these genes allows escape from senescence and leads to immortalization. We have developed a lentiviral vector that directs the synthesis of a p53-specific short hairpin transcript, which mediates stable suppression of p53 expression through RNA interference. We show that suppression of p53 expression in senescent MEFs leads to rapid cell cycle re-entry, is associated with loss of expression of senescence-associated genes, and leads to immortalization. These data indicate that senescence in MEFs is reversible and demonstrate that both initiation and maintenance of senescence is p53-dependent. Senescence is generally defined as an irreversible state of G1 cell cycle arrest in which cells are refractory to growth factor stimulation. In mouse embryo fibroblasts (MEFs), induction of senescence requires the presence ofp19ARF and p53, as genetic ablation of either of these genes allows escape from senescence and leads to immortalization. We have developed a lentiviral vector that directs the synthesis of a p53-specific short hairpin transcript, which mediates stable suppression of p53 expression through RNA interference. We show that suppression of p53 expression in senescent MEFs leads to rapid cell cycle re-entry, is associated with loss of expression of senescence-associated genes, and leads to immortalization. These data indicate that senescence in MEFs is reversible and demonstrate that both initiation and maintenance of senescence is p53-dependent. plasminogen activator inhibitor-1 Friend virus B-strand mouse embryo fibroblast cytomegalovirus green fluorescent protein wild type long terminal repeat Most primary mammalian cells have a limited ability to proliferate in tissue culture (1Hayflick L. Moorhead P.S. Exp. Cell Res. 1961; 25: 585-621Google Scholar, 2Hayflick L. Exp. Cell Res. 1965; 37: 614-636Google Scholar). After a variable number of cell divisions, primary cells will undergo what is believed to be an irreversible form of growth arrest in the G1 phase of the cell cycle and become refractory to further growth factor stimulation (3Seshadri T. Campisi J. Science. 1990; 247: 205-209Google Scholar, 4Drayton S. Peters G. Curr. Opin. Genet. Dev. 2002; 12: 98-104Google Scholar, 5Sherr C.J. DePinho R.A. Cell. 2000; 102: 407-410Google Scholar). In this state of growth arrest, referred to as senescence, cells adopt a typical large and flat morphology and express a number of senescence-associated markers, including senescence-associated ॆ-galactosidase, plasminogen activator inhibitor-1 (PAI-1),1 and p21cip1(5Sherr C.J. DePinho R.A. Cell. 2000; 102: 407-410Google Scholar). The triggers for the induction of senescence differ between mouse and human cells. In cultured rodent fibroblasts senescence is thought to result from stress signals generated in response to the inadequate tissue culture environment. This includes supraphysiological oxygen tension, lack of proper extracellular matrix, and the liberal administration of bovine growth factors (Refs. 6Packer L. Fuehr K. Nature. 1977; 267: 423-425Google Scholar and 7Saito H. Hammond A.T. Moses R.E. Exp. Cell Res. 1995; 217: 272-279Google Scholar and reviewed in Refs. 5Sherr C.J. DePinho R.A. Cell. 2000; 102: 407-410Google Scholar and 8Wright W.E. Shay J.W. Nat. Biotechnol. 2002; 20: 682-688Google Scholar). Indeed, under more gentle and more defined culture conditions, primary mouse cells can be convinced to proliferate for extended periods of time in vitro (9Loo D.T. Fuquay J.I. Rawson C.L. Barnes D.W. Science. 1987; 236: 200-202Google Scholar). Tissue culture stress signals induce expression of a number of anti-proliferative genes, including p16INK4A and p19ARF (10Zindy F. Eischen C.M. Randle D.H. Kamijo T. Cleveland J.L. Sherr C.J. Roussel M.F. Genes Dev. 1998; 12: 2424-2433Google Scholar, 11Palmero I. McConnell B. Parry D. Brookes S. Hara E. Bates S. Jat P. Peters G. Oncogene. 1997; 15: 495-503Google Scholar). The induction of p19ARF appears more relevant than the induction of p16INK4A, as mouse embryo fibroblasts (MEFs) genetically deficient for p19ARF are resistant to induction of senescence and readily become immortal (12Kamijo T. Zindy F. Roussel M.F. Quelle D.E. Downing J.R. Ashmun R.A. Grosveld G. Sherr C.J. Cell. 1997; 91: 649-659Google Scholar), whereas p16INK4A-deficient MEFs senesce normally (13Krimpenfort P. Quon K.C. Mooi W. Loonstra A. Berns A. Nature. 2001; 413: 83-86Google Scholar, 14Sharpless N.E. Bardeesy N. Lee K.H. Carrasco D. Castrillon D.H. Aguirre A.J. Wu E.A. Horner J.W. DePinho R.A. Nature. 2001; 413: 86-91Google Scholar). Likewise, MEFs lacking the downstream effector of p19ARF, p53, are immortal (15Harvey M. Sands A.T. Weiss R.S. Hegi M.E. Wiseman R.W. Pantazis P. Giovanella B.C. Tainsky M.A. Bradley A. Donehower L.A. Oncogene. 1993; 8: 2457-2467Google Scholar), whereas MEFs lacking the downstream effector of p16INK4a, pRb, are mortal (16Peeper D.S. Dannenberg J.H. Douma S. te Riele H. Bernards R. Nat. Cell Biol. 2001; 3: 198-203Google Scholar). However, MEFs lacking all three pRb family members, pRb, p107 and p130 are immortal (17Dannenberg J.H. van Rossum A. Schuijff L. te Riele H. Genes Dev. 2000; 14: 3051-3064Google Scholar, 18Sage J. Mulligan G.J. Attardi L.D. Miller A. Chen S. Williams B. Theodorou E. Jacks T. Genes Dev. 2000; 14: 3037-3050Google Scholar). These data indicate that the Rb family proteins not only act upstream of the p19ARF-p53 pathway, through regulation ofp19ARF by E2F (19Bates S. Phillips A.C. Clark P.A. Stott F. Peters G. Ludwig R.L. Vousden K.H. Nature. 1998; 395: 124-125Google Scholar), but also downstream by rendering cells insensitive to p53 signaling (20Rowland B.D. Denissov S.G. Douma S. Stunnenberg H.G. Bernards R. Peeper D.S. Cancer Cell. 2002; 2: 55-65Google Scholar). Expression of oncogenes, such as an activated RAS oncogene, can further enhance tissue culture stress signals and induce rapid onset of senescence, referred to as 舠premature senescence舡 (21Serrano M. Lin A.W. McCurrach M.E. Beach D. Lowe S.W. Cell. 1997; 88: 593-602Google Scholar). Oncogenic RAS stimulates many of the same anti-proliferative genes that are induced by spontaneous senescence, including p19ARF and p16INK4A, and again only ablation of the p19ARF-p53 pathway allows escape from oncogene-induced premature senescence to cause oncogenic transformation (12Kamijo T. Zindy F. Roussel M.F. Quelle D.E. Downing J.R. Ashmun R.A. Grosveld G. Sherr C.J. Cell. 1997; 91: 649-659Google Scholar, 21Serrano M. Lin A.W. McCurrach M.E. Beach D. Lowe S.W. Cell. 1997; 88: 593-602Google Scholar). These observations have led to the suggestion that premature senescence is part of a fail-safe mechanism that protects cells from oncogenic transformation (22Campisi J. Trends Cell Biol. 2001; 11: S27-S31Abstract Full Text PDF Google Scholar). Senescence in human cells differs from senescence in rodent cells in that most primary human cells lack the catalytic component of telomerase, hTERT. As a consequence, in vitro propagation of primary human cells is associated with erosion of the chromosome ends, the telomeres, leading to DNA damage-like anti-proliferative signals when telomeres become critically short (23Bacchetti S. Cell Dev. Biol. 1996; 7: 31-39Google Scholar, 24de Lange T. Science. 1998; 279: 334-335Google Scholar, 25Lundberg A.S. Hahn W.C. Gupta P. Weinberg R.A. Curr. Opin. Cell Biol. 2000; 12: 705-709Google Scholar). Consequently, most human cells require expression of telomerase to overcome this barrier to immortality. However, similar to rodent cells, primary human cells (especially those of epithelial origin) also suffer from 舠tissue culture stress舡 and often arrest long before their telomeres are critically short (8Wright W.E. Shay J.W. Nat. Biotechnol. 2002; 20: 682-688Google Scholar, 26Ramirez R.D. Morales C.P. Herbert B.S. Rohde J.M. Passons C. Shay J.W. Wright W.E. Genes Dev. 2001; 15: 398-403Google Scholar). This tissue culture stress response of primary human epithelial cells appears to depend on p16INK4a rather than on p14ARF (4Drayton S. Peters G. Curr. Opin. Genet. Dev. 2002; 12: 98-104Google Scholar, 27Kiyono T. Foster S.A. Koop J.I. McDougall J.K. Galloway D.A. Klingelhutz A.J. Nature. 1998; 396: 84-88Google Scholar). However, several diploid human fibroblasts can be immortalized by hTERT expression, suggesting differential sensitivity of primary human cells to tissue culture stress. It has been proposed that the stress-induced replicative arrest induced by tissue culture stress should be referred to as 舠stasis舡 (for 舠stimulation and stress induced senescence舡), whereas the term 舠senescence舡 should be used for cells that undergo replicative arrest as a result of DNA-damage signals emanating from short telomeres. The different names for both forms of senescence suggest that the mechanisms that underlie both processes are distinct. However, the signaling pathways involved and the phenotypic consequences of both forms of replicative arrest are related. For this reason, the study of rodent cell stasis is likely to be relevant for our understanding of the signaling mechanisms that underlie replicative aging of primary human cells in culture. We describe here a novel vector system to study the genes that are required to maintain senescence. We created a virus that can infect post-mitotic cells and direct the synthesis of short hairpin transcripts that mediate post-transcriptional gene silencing through RNA interference. We used this vector system to ask if senescence is a reversible process. The murine p53-specific short hairpin oligonucleotides were first cloned in pRETRO-SUPER (28Brummelkamp T.R. Bernards R. Agami R. Science. 2002; 296: 550-553Google Scholar). pRETRO- SUPER vector was digested with BglII andHindIII, and the annealed oligonucleotides targeting murine p53, 5′-gatccccGTACATGTGTAATAGCTCCttcaagagaGGAGCTATTACACATGTACtttttggaaa-3′ and 5′-agcttttccaaaaGTACATGTGTAATAGCTCCtctcttgaaGGAGCTATTACACATGTACggg-3′, were ligated with the vector, yielding pRETRO-SUPER-p53. The 19-mer p53 targeting sequence in the oligonucleotide is indicated in capital letters. The lentiviral transfer vector HIV-CS-CG (29Miyoshi H. Blomer U. Takahashi M. Gage F.H. Verma I.M. J. Virol. 1998; 72: 8150-8157Google Scholar) was digested with EcoRI and XhoI to remove the CMV-GFP sequence. The cassette containing the H1 promoter and the p53 target sequence was excised from pRETRO-SUPER-mp53 with EcoRI andXhoI and ligated into HIV-CS to yield pLENTI-SUPER-p53. Wild type Friend virus B-strand (FVB) MEFs, ST.HdhQ111 mouse striatum cells (30Trettel F. Rigamonti D. Hilditch-Maguire P. Wheeler V.C. Sharp A.H. Persichetti F. Cattaneo E. MacDonald M.E. Hum. Mol. Genet. 2000; 9: 2799-2809Google Scholar), and 293T cells were cultured in Dulbecco's modified Eagle's medium supplemented with 107 fetal calf serum. For production of lentivirus, 293T cells were transfected by the calcium-phosphate method using 10 ॖg transfer vector HIV-CS-CG or pLENTI-SUPER-p53, 3.5 ॖg of VSVg envelope vector pMD.G, 2.5 ॖg of RSV-Rev, and 6.5 ॖg of packaging vector pCMVDR8.2 (29Miyoshi H. Blomer U. Takahashi M. Gage F.H. Verma I.M. J. Virol. 1998; 72: 8150-8157Google Scholar). Lentiviruses were harvested 24 and 48 h after transfection and filtered through a 0.45-ॖm filter. ST.HdhQ111 cells were shifted to 39 °C 14 days prior to lentiviral infection. WT MEFs were cultured to passage 9–10 whereupon cells were counted every 3–4 days 14 days prior to lentiviral infection. The senescent phenotype was also investigated by acidic ॆ-galactosidase staining at the time of infection (31Dimri G.P. Lee X. Basile G. Acosta M. Scott G. Roskelley C. Medrano E.E. Linskens M. Rubelj I. Pereira-Smith O. et al.Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9363-9367Google Scholar). 1.8 × 105 senescent WT MEFs in 6 cm dishes were infected with lentivirus for at least 12 h in the presence of 0.8 ॖg/ml polybrene and were then allowed to recover for 48 h before reseeding for colony formation assays and growth curves. 0.5 × 105 or 1 × 105 cells were seeded in 10 cm dishes for colony formation assays. Cells were fixed and stained with superstain (507 methanol, 107 acetic acid, 0.17 Coomassie Blue) 16 days after seeding. For growth curves 1.5 × 103 cells were seeded per 3.5-cm dish, at 3-day intervals cells were fixed with 0.57 formaldehyde, stained with 0.17 crystal violet, followed by re-solubilization in 107 acetic acid. The OD590 was quantified as a relative measure of cell number. Whole cell extracts were separated on 127 SDS-PAGE gels and transferred to polyvinylene diflouride membranes (Millipore). Visualization was done using enhanced chemiluminescence (Amersham Biosciences, Inc.) Antibodies used were M-156 (Santa Cruz) against p16INK4a, ab80-50 (Abcam) against 19ARF, F-5 (Santa Cruz) against p21, Ab-7 (Oncogene) against p53, and P30620 (Transduction Laboratories) against PAI-1. 5 × 104 senescent MEFs were seeded in 3.5-cm dishes and infected with lentivirus. Time-lapse microscopy was initiated 34 h after infection in a temperature and CO2-controlled chamber using 10× phase contrast. Frames were taken every 20 min over a period of 38 h. We have recently described a vector, pSUPER, which mediates highly specific and persistent RNA interference through stable expression of short hairpin RNAs (28Brummelkamp T.R. Bernards R. Agami R. Science. 2002; 296: 550-553Google Scholar). We generated a lentiviral derivative of this vector by cloning the H1 RNA short hairpin gene expression cassette targeting murine p53 from pRETRO-SUPER (32Brummelkamp T.R. Bernards R. Agami R. Cancer Cell. 2002; 2: 243-247Google Scholar) into the self-inactivating lentiviral vector pHIV-CS (29Miyoshi H. Blomer U. Takahashi M. Gage F.H. Verma I.M. J. Virol. 1998; 72: 8150-8157Google Scholar). We named this vector pLENTI-SUPER-p53 (Fig. 1A). As a control, we used a lentiviral vector that expresses GFP (HIV-CS-CG (29Miyoshi H. Blomer U. Takahashi M. Gage F.H. Verma I.M. J. Virol. 1998; 72: 8150-8157Google Scholar)). Loss of p53 in primary mouse embryo fibroblasts is associated with acquisition of an immortal phenotype (15Harvey M. Sands A.T. Weiss R.S. Hegi M.E. Wiseman R.W. Pantazis P. Giovanella B.C. Tainsky M.A. Bradley A. Donehower L.A. Oncogene. 1993; 8: 2457-2467Google Scholar). To test whether the lentiviral p53 knockdown vector was capable of inducing a functional inactivation of p53 in MEFs, we infected early-passage primary MEFs with LENTI-SUPER-p53 virus or with control GFP lentivirus and asked whether p53 knockdown caused immortalization. Some 30–407 of control lentivirus-infected cells were GFP-positive, indicating that the primary MEFs were efficiently infected by the lentiviral vectors (data not shown). Fig. 1, B and C, show that infection with LENTI-SUPER-p53, but not with control GFP lentiviral vector, caused efficient immortalization of the infected primary MEFs, indicating that the LENTI-SUPER-p53 virus mediates functional inactivation of p53 expression (see also Fig. 3A). We next asked whether suppression of p53 expression by lentiviral gene transfer in senescent cells would allow re-entry into the cell cycle. We employed two cell systems to address this question. First, we used conditionally immortalized STHdhQ111 neuronal cells derived from mouse embryonic striatum. These cells are conditionally immortalized due to the presence of a temperature-sensitive allele of SV40 T antigen (30Trettel F. Rigamonti D. Hilditch-Maguire P. Wheeler V.C. Sharp A.H. Persichetti F. Cattaneo E. MacDonald M.E. Hum. Mol. Genet. 2000; 9: 2799-2809Google Scholar). STHdhQ111 cells proliferate indefinitely at the permissive temperature (32 °C), but rapidly and synchronously become post-mitotic and adopt a senescent morphology when shifted to the non-permissive temperature (39.5 °C) at which T antigen is inactive (33Brummelkamp T.R. Kortlever R.M. Lingbeek M. Trettel F. MacDonald M.E. van Lohuizen M. Bernards R. J. Biol. Chem. 2002; 277: 6567-6572Google Scholar). We used STHdhQ111 cells that had been maintained at 39.5 °C for 2 weeks to assure that the entire population was senescent and then infected the senescent cells with the LENTI-SUPER-p53 virus or control GFP lentivirus and maintained the infected cells at 39.5 °C for 2 weeks. Fig.2A shows that knockdown of p53 led to re-entry into the cell cycle and allowed continued proliferation, indicating that the senescence-like growth arrest of STHdhQ111 cells at the non-permissive temperature can be reversed by suppression of p53. Next we asked whether p53 knockdown would allow cell cycle re-entry in senescent primary MEFs. We cultured primary MEFs of FVB genotype until the cells no longer proliferated (Fig. 2D) and expressed high levels of the senescence-associated markers acidic ॆ-galactosidase, PAI-1, p21cip1, p19ARF and p16INK4a (Figs. 2E and3A). All cells in the culture showed a flat senescent morphology and stained intensely for acidic ॆ-galactosidase (Fig. 2E), indicating that these cells were quantitatively senescent. This notion is also supported by the growth curves of these late-passage MEFs, which showed a constant decline in cell number over time (Fig. 2D), indicative of the absence of spontaneously immortalized cells in the culture. Fig. 2,B and C, show that lentiviral knockdown of p53 in these senescent primary MEF cultures triggered a marked degree of proliferation. Importantly, cell cycle re-entry was associated with loss of expression of several of the senescence-associated markers, including PAI-1, p21cip1, and acidic ॆ-galactosidase (Fig. 3,A and B) and senescence-reverted cells continued to proliferate for several weeks without any signs of senescence, suggesting that they had become immortal (Fig. 2B and data not shown). In principle, the observed proliferation following lentiviral knockdown of p53 could originate from cells that were not truly senescent in the culture. It was therefore important to follow the cultures of senescent MEFs in time after lentiviral infection. Fig.4 shows a series of time-lapse photomicrographs of senescent MEFs after lentiviral knockdown of p53, which together indicate that cells with a completely flat and senescent morphology round up and divide within 48 h after infection with the p53 knockdown virus (Fig. 4, cells marked by black arrows). However, not all cell divisions are productive as many cells divide initially, but die by apoptosis during division or just after completion of cell division (Fig. 4, cells marked by white arrows). Assessed by time-lapse photography and colony formation efficiencies (Fig. 2B), ∼0.5–17 of infected cells divide successfully. A complete movie of the senescent MEFs after infection with the p53 knockdown vector is provided as supplementary material. No division or apoptosis could be observed following infection with control lentivirus encoding GFP (data not shown). We conclude that cells with all the hallmarks of fully senescent cells rapidly re-enter the cell cycle after p53 knockdown. We conclude that p53 is not only required to initiate senescence, but is also required, at least in MEFs, to maintain senescence. Using a lentiviral vector system that silences gene expression, we provide evidence that suppression of p53 expression in senescent MEFs leads to a reversion of the senescent state and causes immortalization. Several lines of evidence support the notion that the MEFs were fully senescent at the time of infection with the lentiviral p53 knockdown vector. First, the cells had stopped proliferating in the presence of growth factors, indicating that they were senescent and refractory to growth factor stimulation, rather than quiescent and still responsive to growth factors (Fig. 1D). Second, they uniformly manifested a senescent morphology and expressed the senescence-associated markers acidic ॆ-galactosidase, PAI-1, p21cip1, p19ARF and p16INK4a (Figs.2E and 3A). When cells emerged from senescence as a result of p53 knockdown, the cells behaved phenotypically as p53 null MEFs in that they were immortal and had low levels of p21cip1and high levels of p19ARF (10Zindy F. Eischen C.M. Randle D.H. Kamijo T. Cleveland J.L. Sherr C.J. Roussel M.F. Genes Dev. 1998; 12: 2424-2433Google Scholar, 12Kamijo T. Zindy F. Roussel M.F. Quelle D.E. Downing J.R. Ashmun R.A. Grosveld G. Sherr C.J. Cell. 1997; 91: 649-659Google Scholar, 15Harvey M. Sands A.T. Weiss R.S. Hegi M.E. Wiseman R.W. Pantazis P. Giovanella B.C. Tainsky M.A. Bradley A. Donehower L.A. Oncogene. 1993; 8: 2457-2467Google Scholar). Importantly, the cells that emerged from senescence by p53 knockdown maintained high levels of p16INK4a (Fig. 3A). As p16INK4a expression is induced during senescence in a p53-independent fashion (10Zindy F. Eischen C.M. Randle D.H. Kamijo T. Cleveland J.L. Sherr C.J. Roussel M.F. Genes Dev. 1998; 12: 2424-2433Google Scholar), these data indicate that the signaling pathways that led to the induction of senescence are still operational in senescence-reverted MEFs. This provides further evidence that the cells that re-entered cell cycle by p53 knockdown were indeed fully senescent at the time of infection with the p53 knockdown virus. Our data are in agreement with earlier experiments performed in senescent human diploid fibroblasts. Thus, ablation of p53 function by microinjection of p53 antibody in primary human fibroblasts allowed at least temporary reversal of senescence and re-entry into the cell cycle (34Gire V. Wynford-Thomas D. Mol. Cell. Biol. 1998; 18: 1611-1621Google Scholar). However, inactivation of p53 in human fibroblasts delays, but does not abrogate, replicative senescence, indicating that p53 inactivation alone is not sufficient to mediate stable reversion of senescence in primary human fibroblasts and requires also induction of hTERT expression (35Shay J.W. Pereira Smith O.M. Wright W.E. Exp. Cell Res. 1991; 196: 33-39Google Scholar, 36Itahana K. Dimri G. Campisi J. Eur. J. Biochem. 2001; 268: 2784-2791Google Scholar). An essential feature of the lentiviral vector system described here is that suppression of gene expression is persistent, allowing the study of long term consequences of gene inactivation in post-mitotic cells. The LENTI-SUPER vector should therefore be a useful tool to investigate which genes are continuously required to maintain a post-mitotic state in cells that have exited the cell cycle. We thank Dr. Inder Verma for the self-inactivating lentiviral vector, Lauran Oomen for invaluable assistance with the time-lapse microscopy, Thijn Brummelkamp and Roderick Beijersbergen for discussions, and Katrien Berns for critical reading of the manuscript. eyJraWQiOiI4ZjUxYWNhY2IzYjhiNjNlNzFlYmIzYWFmYTU5NmZmYyIsImFsZyI6IlJTMjU2In0.eyJzdWIiOiIwZDg1NDBmNjU5ZWJiNmEzNGM0YmI4MjJiMzMwNWU2NCIsImtpZCI6IjhmNTFhY2FjYjNiOGI2M2U3MWViYjNhYWZhNTk2ZmZjIiwiZXhwIjoxNjgyMjc2OTM0fQ.TyI9LvI301GRTbeyr47A6yO8P1yH4r2GILVIDw-0_Tb-GJMMSPNR9ZNu6fT7Au5mzVG18lce11HNPa1mgiuV8y_1uuQx3ki5zw_57Hjh8MqYLaxT3At6fvpA-Q18rpXBIVzTHnX9V2CSkj5Cx8bMnQqTGN6oYPnuGRZMUCmZHYw1sMhlLYOwTBgn5_5EJkBffyW5sa4Csry3Bq0BXWkceUsSkhNxQAUTXS1ZzLlC6rq8PilaTdlx4VY802oOHKDEUl1esJVBLjJq71iaJbbyJWQt8VvZwb40Oi3AL_ucNQhc2u_XIWV1Zx5f9KJ6g0dUKOEerEEfkCgfuOrsWvm5BA Download .mp4 (3.27 MB) Help with .mp4 files
Referência(s)