during mitosis, sister chromatids are aligned on the metaphase plate and held together by cohesin (33). Metaphase to anaphase transition occurs whereby SCC1, a component of the cohesin complex, is cleaved by the separase protease. Separase is released from tonic securin inhibition by activated anaphase promoting complex (APC), which degrades securin (33). Securin was first identified in budding yeast as PDS1 (32), and the mammalian homolog (57) was identified in pituitary adenomas as pituitary tumor-transforming gene (Pttg) (34). Despite the key role for securin in cell division, Pttg^−/− mice are viable, display multiorgan (pituitary, testis, pancreatic β-cell, and spleen) hypoplasia, and are lean (3, 52, 53). The tissue-specific response to global Pttg deletion likely depends on intrinsic tissue properties. Thus, highly differentiated and slowly proliferating tissues (for example, pituitary and pancreatic β-cells) appear to be more sensitive to Pttg absence than rapidly proliferating epithelial cells.

Adult or tissue-specific stem cells maintain and regenerate mature tissues as a function of normal physiology or in response to injury and have been detected in the skin (2, 49), gut (18, 30), brain (8), liver (51), skeletal muscle (12), male germ cells (9, 26), and bone marrow (BM) (11, 17, 37, 38, 41). The BM cavity is a repository for hematopoietic stem cells and adult mesenchymal stem cells (MSCs), which give rise to bone, cartilage, and adipocyte lineages (10, 11, 28, 35, 37). BM may also comprise stem cells with a broader differentiating capability and with more primitive proliferation and differentiation properties. Multipotent adult progenitor cells (MAPCs) reside in the BM, differentiate into ectodermal, mesodermal, and endodermal germ layer lineages (16, 17, 41, 42, 45), and proliferate indefinitely without losing pluripotent characteristics or undergoing transformation. The function of diverse BM-derived pluripotent BM stem cell (BMSC) populations is not completely clear (1, 6, 7, 19, 24, 46, 50), but they have been proposed to migrate from the BM to comprise tissue-specific stem cells (25), and cells with similar features have been isolated from tissues other than the BM (17), including the brain, spleen, liver, skin, muscle, and fat.

To elucidate the role of Pttg in BMSC proliferation and differentiation, we studied Pttg^−/− and wild-type (WT) stem cells derived from the BM under MAPC conditions. We show that Pttg^−/− cells proliferate more slowly than WT cells, and lower cyclin D₁ and elevated p21 levels with enhanced features of cell senescence confirmed these results. The DNA repair system is upregulated in Pttg^−/− BMSCs, as these cells exhibit higher Rad21 and Rad17 levels. These results indicate a role for Pttg in the proliferation and maintenance of BMSCs.

MATERIALS AND METHODS

Primary culture and BMSC culture.

BMSCs were isolated under MAPC conditions (16). Experiments were approved by the Institutional Animal Care and Use Committee. BM was flushed from the femur and tibia of 4- to 6-wk-old C57Bl/6J Pttg^−/− and WT mice of the same genetic background into low-glucose (LG) DMEM (Invitrogen, Carlsbad, CA). Cells were sequentially suspended using a 16-gauge needle and then a 21-gauge needle, passed through a 40-μm strainer (BD Biosciences, San Jose, CA), centrifuged, and resuspended in expansion medium: 60% LG DMEM (Invitrogen) and 40% MCDB-201 supplemented with 1× insulin-transferrin-selenium, 1 mg/ml linoleic acid-BSA, 0.8 mg/ml BSA, 10⁻⁴ M ascorbic acid 2-phosphate (all from Sigma-Aldrich, St. Louis, MO), 1× chemically defined lipid concentrate, 100 U/ml penicillin, 100 U/ml streptomycin (Invitrogen), 2% FCS (Hyclone Laboratories, Logan, UT), and 10 ng/ml each of EGF (Sigma-Aldrich), leukemia inhibitory factor (LIF; Chemicon, Temecula, CA), and PDGF-BB (R&D Systems, Minneapolis, MN). Cells (3–4 × 10⁶) were seeded in six-well dishes, which were coated with 100 μg/ml fibronectin (Roche Applied Science, Indianapolis, IN). The average cell yield was sufficient for seeding 10–12 wells. Half the well medium was replaced every alternate day. Two weeks later, when cells had reached 100% confluency, each well was passaged to a 10-cm fibronectin-covered dish until 80% confluency was achieved. Cells were split 1:2 for about four more passages, and CD45⁺/Ter119⁺ cells were then depleted (Miltenyi Biotec, Sunnyvale, CA), yielding ∼50% of total cells. Single cells were then seeded in fibronectin-covered 96-well dishes. Visible colonies were established in ∼5% of wells; these were expanded and replated in larger wells, and those exhibiting morphological features of MAPC (fibroblast like) were selected for further experiments. Five hundred cells per centimeter squared were routinely grown in expansion medium. Experiments were carried out using three different clones derived from each mouse. MAPCs are routinely grown in 5% O₂-6% CO₂-89% N₂. However, since Pttg^−/− BM cells did not proliferate well under these conditions, they were grouped for either low-oxygen or atmospheric oxygen conditions.

Colony-forming unit assay.

After the third passage, 1,000 cells were seeded in 12-well dishes in triplicate, grown for 10 days in expansion medium, fixed for 5 min in ethanol, and then stained with Giemsa. Colonies (with >20 cells) and cell numbers were then counted.

Cell markers.

The following antibodies were obtained from BD Biosciences: CD34-phycoerythrin (PE), CD45-FITC, Ter119-PE, lymphocyte antigen (Ly)6A/E-FITC, stem cell antigen-I (ScaI)-allophycocyanin, FLK1-PE (VEGF receptor 2), CD117-allophycocyanin (c-Kit), CD13-allophycocyanin, Ly6G-FITC, Ly6C-FITC, and CD31-FITC (platelet endothelial cell adhesion molecule). BMSCs cultured in expansion medium were lifted with trypsin-EDTA and counted. Cells (∼2 × 10⁵) were aliquoted, pelleted by centrifugation for 1 min at 450 g, and resuspended in 1 ml PBS. FITC-conjugated antibodies were added at a concentration of 2 μg/ml, and PE- or APC-conjugated antibodies at 2.5 μg /ml. Samples were incubated with gentle shaking at room temperature for 20 min, pelleted, washed twice with PBS, and cells analyzed by flow cytometry (Cytomics FC 500 FACS; Beckman Coulter, Miami, Florida, USA).

Senescence.

WT and Pttg^−/− cells (2 × 10⁴) were plated on 10-cm dishes and grown in 6% CO₂-5% O₂. Twenty-four hours later, cells were fixed and stained for β-galactosidase activity using the Sigma Senescence Detection kit (Sigma-Aldrich).

Cell cycle analysis, viability, and apoptosis.

After synchronization by serum starvation for 24 h, the cell cycle phase was assessed with propidium iodide after methanol fixation. Apoptosis was assessed by annexin V staining (Pharmingen, BD Biosciences).

Gene array.

An Oligo GEArray Mouse Cell Cycle Microarray kit (SuperArray, Bethesda, MD) comprising 112 genes involved in cell cycle regulation was used. These genes included cyclin-dependent kinases (CDKs), CDK-modifying proteins, cyclins, CDK inhibitors, and CDK kinases as well as genes comprising Skp1-cullin-F box protein (SCF) and anaphase-promoting complex (APC) ubiquitin-conjugation complexes (see www.superarray.com for details). WT and Pttg^−/− cells (2 × 10⁴) were plated in 10-cm dishes, and, 24 h after being plated, cells were serum starved for 24 h. Total RNA was isolated with the RNeasy Mini Kit (Qiagen, Valencia, CA) 6 h after the addition of replete medium, and 2 μg RNA was used as a template to generate biotin-16-UTP-labeled cRNA probes. cRNA probes were hybridized at 60°C with the SuperArray membrane, which was washed and exposed with a chemiluminescent substrate. X-ray film was scanned and imported into Adobe Photoshop as a TIFF file. The signal intensity was compared in membranes using the GEarray analyzer program ( http://www.superarray.com, SuperArray) and normalized by background subtraction of the average intensity value of empty spots. Six housekeeping gene spots were set as baseline values for a comparison of signal intensities.

In vitro cell differentiation.

For osteocyte differentiation (35), 10⁶ WT and Pttg^−/− cells were seeded on fibronectin-coated six-well dishes in expansion medium. One day later, the medium was changed to differentiation medium: LG DMEM (Invitrogen) supplemented with 10% FCS (HyClone), 2 × 10⁻³ mM glutamine (Invitrogen), 10⁻⁷ M dexamethasone, 3 × 10⁻⁴ M absorbic acid, and 10⁻¹ M β-glycerophosphate (all from Sigma-Aldrich) and changed twice weekly for 3 wk. Cells were fixed with 4% paraformaldehyde for 15 min and stained with Alizarin red (pH 4.1, Sigma-Aldrich) for 20 min. Adipogenic differentiation was carried out as previously described (35). WT and Pttg^−/− cells (10⁶) were seeded on fibronectin-coated six-well dishes in IMDM supplemented with 10% FCS, 10% horse serum (HyClone), 100 U/ml penicillin, 100 μg/ml streptomycin, 1.2 × 10⁻² M l-glutamine, 5 μg/ml insulin, 5 × 10⁻⁵ M indomethacin, 10⁻⁶ M dexamethasone, and 5 × 10⁻⁷ M 3-isobutyl-1-methylxanthine (IBMX) (all from Sigma-Aldrich) and changed twice weekly for 3 wk. Cells were fixed with 10% formalin for 20 min and stained with 0.5% Nile red (Sigma-Aldrich) in DMSO for 20 min. Hepatocyte differentiation was performed as previously described (16): 22,000 cells/cm² were plated on Lab-Tek chamber slides with Permanox (Nalge Nunc, Rochester, NY) covered with 1% Matrigel (BD Biosciences) in expansion medium. One day later, medium was replaced with differentiation medium: expansion medium without FCS, EGF, PDGF-BB, and LIF and instead supplemented with 10 ng/ml each of HGF and FGF-4 (R&D Systems). Medium was replaced every alternate day for 2 wk, and cells were then either fixed for immunohistochemistry with 4% paraformaldehyde or RNA was extracted.

Western blot analysis.

Cells were lysed in ice-cold lysis buffer: 1% Nonidet P-40, 1% sodium deoxycholate, 0.1% SDS, 1.5 × 10⁻¹ M sodium chloride, 2 × 10⁻³ M EDTA, and 1 × 10⁻² M sodium phosphate (pH 7.2) and supplemented with protease inhibitor cocktail (Roche). Protein (50 μg) was separated on 4–12% bis-Tris gels (Invitrogen) to detect proteins except ataxia telangiectasia (ATM) and separase, which were detected on 3–8% Tris-acetate gels (Invitrogen). Proteins were transferred to polyvinylidene difluoride (PVDF) membranes, blocked with 5% skim milk in PBS-Tween 20 for 1 h, and incubated for 1 h with primary antibody in blocking buffer. The following antibodies were used: anti-cyclin D₁, anti-cyclin D₃, anti-p21, anti-p27, anti-Bcl2 (all from Santa Cruz Biotechnology, Santa Cruz, CA; 1:200), anti-p53 (BD Biosciences; 1:100), anti-ATM (Abcam, Cambridge, MA; 1:500), anti-separase (Novus Biologicals, Littleton, CO; 1:1,000), and anti-β-actin (Sigma-Aldrich; 1:10,000). Membranes were washed three times with PBS-Tween 20 and incubated for 45 min with the appropriate secondary antibody, and proteins were detected by enhanced chemiluminescence (Amersham-Pharmacia Biotech, Piscataway, NJ). Lane quantification was performed using ImageJ 1.37 ( http://rsb.info.nih.gov).

Metabolic labeling.

WT and Pttg^−/− BMSCs were grown in 10-cm culture dishes, and, 6 h before cells were labeled, the expansion medium was replaced by phosphate-free DMEM. For metabolic labeling, cells were incubated with 0.1 mCi/ml inorganic [³²P]orthophosphate (Amersham, Arlington, IL) for 6 h. Cells were washed three times with PBS, lysed [5 × 10⁻² M Tris·HCl, 0.5% Nonidet P-40, and 1.4 × 10⁻¹ M NaCl with protease inhibitor (Roche) and phosphatase inhibitor cocktails (Sigma-Aldrich)], and immunoprecipitated with anti-separase (Novus Biologicals). Immunocomplexes were precipitated with protein A/G (BD Biosciences) and resolved on 3–8% gels, followed by a transfer to PVDF membranes for radiographic exposure. Two weeks later, the same membrane was immunoblotted with anti-separase to verify the 240-kDa band identity.

Templates for probes and Northern blot analysis.

Probes for murine (m)Pttg1 were generated as previously described (53). The β-actin DECA probe was a 1.076-kb fragment of the mouse β-actin gene (Ambion, Austin, TX). RNA extraction was performed using TRIzol reagent (Invitrogen). In brief, 10–20 μg of total RNA were electrophoresed on a 1% agarose-6.4% formaldehyde gel, transferred to a Hybond-N+ membrane (Amersham), and UV cross-linked. Probes were labeled with [α-³²P]CTP using the Prime-It Random Primer Labeling Kit (Stratagene, La Jolla, CA). Micro Bio-Spin Chromatography Columns (Bio-Rad, Hercules, CA) was used to purify probes. Membrane prehybridization and hybridization were performed using QuickHyb Solution (Stratagene) and then exposed to Hyperfilm MP (Amersham) at −70°C.

Generation of mPttg-directed short interfering RNA.

mPttg-directed short interfering (si)RNAs were planned and generated using the RNAi oligo retriever website ( http://katahdin.cshl.org:9331/RNAi/html/rnai.html) and the “shagging PCR protocol.” Three siRNAs primers were designed, each directed against a 29-bp sequence in the mPttg and U6 promoter reverse primer sequence. All were found to be specific for mPttg by BLAST search, and their full sequences were as follows: 1) siRNA-1 (against bases 440–468), 5′-AAAAAAGCAACAGCTCCTCCAACCCTCCCTCTTCACAAGCTTCTGAAGAGAGAGGGCTGGAGAAGCTGCTGCGGTGTTTCGTCCTTTCCACAA-3′; 2) siRNA-2 (against bases 381–409), 5′-AAAAAAGGGAGAAGTAAGATCCGGTGCCCCTCAAGCAAGCTTCCCTGAGGAGCACCAGATCTCACTTCTCCCGGTGTTTCGTCCTTTCCACAA-3′; and 3) siRNA-3 (against bases 518–546), 5′-AAAAAAGGGCACTGGAAGAAGAGTACAACGAATCACAAGCTTCTGATCCGCTGTACTCTCCTCCCAGTGCCCGGTGTTTCGTCCTTTCCACAA-3′.

The primers were used together with a SP6 primer to clone the U6 promoter, generating a PCR product containing both the U6 promoter and the mPttg-directed siRNA sequence. The PCR product was inserted into the pCR2.1 vector using a TA cloning kit (Invitrogen). A scrambled siRNA was designed by the same method and was used as a control.

Quantitative RT-PCR.

The following primers were used: mOct4, forward 5′-CCAATCAGCTTGGGCTAGAG-3′ and reverse 5′-CCTGGGAAAGGTGTCCTGTA-3′; murine thiosulfate sulfurtransferase-1 (mSSEA-1), forward 5′-TATTCCAGGAGCGATCCAAC-3′ and reverse 5′-CTCGTTCCA GTTGCTCACAA-3′; forkhead box A1 (FOXA1), forward 5′-TTCTAAGCTGAGCCAGCTGCA-3′ and reverse 5′-GCTGAGGTTCTCCGGCTCTTTCAGA-3′; p21, forward 5′-AGGGGAGTTTACGGGAATGC-3′ and reverse 5′-GCTGGGGTCTCAGACACAG-3′; cyclin D₁, forward 5′-CAGAAGTGCGAAGAGGAGGTC-3′ and reverse 5′-TCATCTTAGAGGCCACGAACAT-3′; cyclin B₁, forward 5′-AAGGTGCCTGTGTGTGAACC-3′ and reverse 5′-GTCAGCCCCATCATCTGCG-3′; checkpoint homolog 2 (Chk2), forward 5′-CTCGGCTATGGGCTCTTCAG-3′ and reverse 5′-CTTCTCAACAGTGGTCCATCG-3′; hypoxia-inducible factor-1α (Hif1a), forward 5′-ACCTTCATCGGAAACTCCAAAG-3′ and reverse 5′-CTGTTAGGCTGGGAAAAGTTAGG-3′; Rad51, forward 5′-TGTTGCTTATGCACCGAAGAA-3′ and reverse 5′-AGCTGCCGTGGTGAAACCC-3′; sestrin2, forward 5′-AGTGTTCTTACCTGGTGGGTT-3′ and reverse 5′-GTAACTTGTTGACCTCGCTGA-3′; and mPttg1, forward 5′-GATCCGCTGTACTCTC-3′ and reverse 5′-ATATCTGCATCGTAACAA-3′.

The following primers for housekeeping genes were used: β-actin, forward 5′-TGTTACCAACTGGGACGACA-3′ and reverse 5′-GGGGTGTTGAAGGTCTCAAA-3′; and GAPDH, forward 5′-CATGGCCTTCCGTGTTCCTA-3′ and reverse 5′-CTGGTCCTCAGFGTAGCCCAA-3′.

PCR efficiency tests, standard curves, and melting curves were optimized for each respective primer pair. The Fail-Safe SYBR green real-time PCR system (Epicentre Biotechnologies, Madisom, WI) was used. PCRs were run in agarose gels stained with SYBR green to identify amplification products; 96-well plates were run in the 7700 Sequence Detection system (Applied Biosystems, Foster City, CA) using 50°C for 1 min, 95°C for 4 min, and 40 cycles of 95°C for 15 s and 60°C for 1 min as thermal cycler conditions.

RESULTS

BM-derived stem cell isolation and characterization.

BMSCs derived from Pttg^−/− and WT mice were isolated under MAPC conditions. Differences in BM cells flushed from WT and Pttg^−/− tibia and femurs were observed as soon as 3 wk after primary culture. Colony-forming unit-fibroblast assays performed after 3 wk (3rd passage) showed that WT cells formed more colonies than Pttg^−/− cells (8 ± 2.5 vs. 3 ± 1.2 cells of 1,000 cells, P < 0.05). Moreover, WT colonies were composed of more cells than Pttg^−/− colonies (107 ± 31 vs. 28 ± 15 cells, P < 0.01). Colonies also differed in morphology (Fig. 1, A–D). WT colonies had the classic appearance of mMSC colonies, comprising larger cells with smaller cell patches. In contrast, Pttg^−/− colonies did not contain smaller cells. After 6 wk, cells were depleted of CD45- and Ter119-positive cells, and single cells were selected and grown for 45 population doublings, at which time MAPCs gain the ability to differentiate into the three germ layers. Pttg mRNA levels were measured after 45 population doublings and confirmed that recombination had not occurred (Fig. 1E). Pttg^−/− and WT BMSC karyotypes were both normal, with no detectable differences (data not shown), and WT and Pttg^−/− cell marker phenotypes were similar after 45 population doublings. Cultured BMSCs were negative for CD117, CD45, Ter119, Flk1, SSEA-1, and Oct4 but expressed high ScaI levels and low CD13 and CD34 levels (Table 1). Because this phenotype is not characteristic for MAPC, we termed these cells BMSCs and not MAPCs.

Lineage differentiation.

Both WT and Pttg^−/− cells were differentiated into adipocyte-like, osteocyte-like, and hepatocyte-like cells, respectively, consistent with the MAPC property for differentiation into mesodermal and endodermal lineages. For osteocyte-like differentiation, cells were grown to 100% confluency, treated with differentiation medium, and, 3 wk later, stained with alizarin red to confirm calcium deposits (Fig. 2A); no differences between Pttg^−/− and WT cells were detected. Adipocyte differentiation (Fig. 2B) was reflected by Nile red staining, with no observed differences in oil droplet formation levels or the time course of accumulation between WT and Pttg^−/− cells. For hepatocyte-like differentiation, WT and Pttg^−/− cells were grown to 100% confluency in the presence of FGF-4 and HGF. After 2 wk, hepatocyte-like cells were evident, as demonstrated by albumin immunostaining (Fig. 2C), and FOXA1 mRNA expression, as detected by quantitative RT-PCR (Fig. 2D).

Proliferation.

After 45 population doublings, Pttg^−/− BMSC (seeded at 300 cells/cm²) proliferation was 40% lower as measured by cell counting (P < 0.01) and attenuated by 27% (P < 0.05) as measured by MTT incorporation after 8 days (Fig. 3A). Marked differences in cell growth were evident when cells were cultured under different oxygen concentrations after the fifth passage (Table 2); whereas 7 of 10 WT cultures grew successfully under normoxic and hypoxic conditions, 6 of 10 Pttg^−/− cultures grew in normoxia, and only 3 cultures were viable in a hypoxic environment. After culture establishment, Pttg^−/− BMSCs were transferred from 21% to 5% O₂, a change that, although not lethal to the culture, further attenuated Pttg^−/− cell proliferation (45% difference during the linear growing phase compared with WT cells, P < 0.01; Fig. 3B). Interestingly, HIF-1α, a transcription factor that responds to hypoxia, was elevated in Pttg^−/− cells, even at baseline prior to hypoxia exposure. Upon 1-h exposure to 3% O₂, HIF-1α levels increased further (Fig. 3C; P < 0.005 at each time point).

Apoptosis and senescence.

To study the mechanisms leading to decreased proliferation, we compared WT and Pttg^−/− cell viability and also assessed apoptosis (Fig. 4A). Propidium iodide staining showed that 32% of Pttg^−/− cells were not viable compared with 26% of WT cells. Increased Pttg^−/− cell death was not due to increased apoptosis (28% WT apoptotic cells vs. 26% Pttg^−/− apoptotic cells) but could be ascribed to the number of dead cells with no apoptotic staining (4.9% WT cells vs. 9.1% Pttg^−/− cells, P > 0.05; Fig. 4A). We therefore studied the features of cell senescence as evidenced by β-galactosidase staining to further understand the mechanisms for reduced proliferation and accelerated cell death. β-Galactosidase staining was enhanced in Pttg^−/− cells (15 ± 6% compared with 3 ± 5%, P < 0.01; Fig. 4B). Moreover, Pttg^−/− BMSCs have the typical enlarged and flattened appearance of senescent cells.

Cell cycle analysis.

To test whether the slower rate of Pttg^−/− BMSC proliferation and increased senescent features were reflective of aberrant cell cycle progression, propidium iodide was added after serum starvation at 0, 6, and 24 h after serum addition (Fig. 5A). Upon serum starvation, most Pttg^−/− cells were arrested in the G₂/M phase (49% ± 12% compared with 28% ± 7% of WT cells), whereas most WT cells were arrested at the G₁/G₀ phase also (51% ± 16% vs. 39% ± 11%, P < 0.05). After 24 h, no differences in S phase cells were seen (35% of WT cells and 37% of Pttg^−/− cells), albeit at every time point more Pttg^−/− cells were in the G₂/M phase than WT cells.

DNA array of cell cycle genes.

As Pttg^−/− cells arrested upon starvation during the G₂/M phase, we assessed cell cycle-related gene expression patterns. WT and Pttg^−/− cells were serum starved, and, 6 h after the addition of replete medium, RNA was extracted and hybridized with a SuperArray Mouse Cell Cycle-Related Gene membrane (Fig. 5B and Table 3). Of 112 cell cycle genes, 2 gene groups were found to be affected by Pttg deletion: cell cycle regulators and genes related to DNA damage and stress. Among the cell cycle regulators, cyclin E₂, cyclin D₁, and Myb were downregulated in Pttg^−/− cells (Pttg^−/−-to-WT ratio: 0.02, 0.41, and 0.14, respectively), whereas p21, a cell cycle inhibitor, was upregulated (1.5-fold). Two cyclins, cyclin A₂ and B₁, expressed at the G₂/M phase, were upregulated in Pttg^−/− cells (2.6- and 1.7-fold increase in Pttg^−/− cells), reflective of Pttg^−/− G₂/M arrest. The array results were confirmed by quantitative RT-PCR measurement of cyclin D₁, p21, and cyclin B₁ mRNA (4- fold decrease and 3.4- and 1.7-fold increase, respectively, data not shown) and Western blot analysis of cyclin B₁, cyclin D₁, and p21 levels (Fig. 5C). Several DNA damage and stress genes were upregulated, including Rad21, Rad17, Rad51, and sestrin 2 (4.4-, 7.7-, 2.1-, and 2.8-fold increase, respectively). Quantitative RT-PCR analysis showed 16.5- and 2-fold increases in Rad51 and sestrin2 expression, respectively. p53 and ATM mRNA expressions were similar in WT and Pttg^−/− cells, as revealed by gene array. However, Western blot analysis showed that these specific proteins were elevated in Pttg^−/− cells, but differences in p27, cyclin D₃, and Bcl2 mRNA and protein levels were not detected in WT and Pttg^−/− cells.

Separase expression and phosphorylation.

As a role for the separase-securin complex in DNA damage has been shown in fission yeast (31), we studied the expression pattern of separase in BMSCs. Four polypeptide bands were detected by Western blot, of which three (220, 150, and 65 kDa) represent full-length and autocleaved separase products (4), and all three were downregulated in Pttg^−/− BMSCs (Fig. 6, A and B). An additional band corresponding to a 240-kDa polypeptide was more intense in Pttg^−/− cells (Fig. 6, A and B). To test whether separase phosphorylation serves to inactivate separase in the absence of securin, cells were metabolically labeled with [³²P]orthophosphate and immunoprecipitated with separase antibody (Fig. 6C). A 240-kDa phospho-band corresponding to the separase band was detected in Pttg^−/− cells but not in WT cells (Fig. 6D). To confirm that separase phosphorylation was associated with Pttg deletion, we targeted Pttg expression in a pituitary mouse folliculostellate cell line (TtT/GF) with three different siRNA constructs. When Pttg mRNA was decreased by 50% (Fig. 6E), separase phosphorylation was enhanced (Fig. 6E).

DISCUSSION

Adult stem cells participate in the process of tissue generation and regeneration, and their disruption may contribute to abrogated tissue maintenance (39, 44). We show here that BMSCs are affected by Pttg deletion: Pttg^−/− BMSCs proliferate slower than WT cells and express higher p21, p53, and DNA repair gene levels. Moreover, Pttg^−/− BMSCs exhibit senescent features.

BMSCs were isolated, proliferated, and differentiated under MAPC conditions and exhibited phenotype characteristic of both MSCs and MAPCs. Similar to MAPCs, these cells have extensive proliferation capacity, up to 100 population doublings, with no detectable karyotype alterations (16). On the other hand, the expressed phenotypic markers are similar to those of MSCs, with high ScaI and no Oct4 or SSEA-1 expression (1, 35). Cells were induced to differentiate into bone or adipocyte cells, consistent with the MSC phenotype; they were also differentiated to hepatocyte-like cells, resembling MAPCs. Nevertheless, we did not wholly achieve the rigorous MAPC phenotype. BMSCs were grown as single cell colonies, with the potential to differentiate to both endodermal and mesodermal lineages, thus suggesting multipotency. However, we cannot exclude the possibility that some colonies arose from two or more coisolated progenitor cells.

Although viable and fertile, Pttg^−/− mouse breeding resulted in reduced litter sizes, demonstrating the embryonic requirement for Pttg (53). However, the proliferation of Pttg^−/− embryonic stem cells was not compromised (27), suggesting that Pttg primarily affects later embryonic stages. Indeed, adult Pttg-null mice display fully differentiated organs yet exhibit organ-specific hypoplasia affecting the spleen, pituitary, testis, and, interestingly, pancreatic β-cells, which is seen only at 6–10 mo (53). These phenotypes are therefore reflective of a mature tissue-specific requirement rather than embryonic stem cell requirement for Pttg. We show here that Pttg^−/− adult BMSCs exhibited reduced proliferation and increased senescence. These observations suggest that the BMSC cell population may also be affected by Pttg deletion. BMSCs are assumed to migrate from the BM to comprise tissue-specific stem cells (25) and to regenerate injured skin keratinocytes (15), and cells with similar features have been isolated from many tissues other than the BM (17, 29), including the brain, spleen, liver, pancreas, kidney, lung, skin, muscle, and fat. BMSC dysfunction might disrupt the maintenance of these tissues.

Pttg^−/− BMSCs exhibit increased sensitivity to hypoxia, as evident by the reduced ability to form colonies and proliferate under hypoxic conditions. However, Pttg^−/− BMSCs expressed higher basal HIF-1α levels. HIF-1α-deficient CHO cells were markedly more sensitive to a combined hypoxic and hypoglycemic environment (54), and HIF-1α also induces stress response genes including VEGF and glucose transporter 4 (48, 54). Sestrin2, a protein involved in antioxidative cellular responses and reflective of oxidative stress, is upregulated in Pttg^−/− BMSCs (as seen in the gene array and confirmed by quantitative RT-PCR). As hypoxia mediates rat MSC migration from the BM into the circulation more readily than normoxia (43), and exposure of MSC to hypoxia increases the expression of CX3CR1 and CXCR4 chemokine receptors as well as increased xenotypic grafting into chick embryos (14), Pttg^−/− BMSCs may respond to hypoxia by defective migration to receptive organs, leading to specific organ hypoplasia.

Similar to Pttg^−/− mice, animals lacking either cyclin D₁, cyclin D₂, cyclin D₃, or CDK4/6 also exhibit tissue-specific hypoplasia, as evident in the spleen, pancreas, and pituitary. In these mice, only subpopulations of hematopoietic stem cells are affected (22, 23, 40). These results and ours suggest that tissue-specific hypoplasia occurs as a combination of stem cell alteration as well as intrinsic tissue properties.

ATM is a cell cycle checkpoint kinase activated in response to DNA damage and stress. It secures damaged cell restraint by activating p53, which, in turn, induces p21 and proliferation arrest. In hematopoietic progenitor cells, ATM/p53/p21 is activated upon hypoxic stress, and this activation also drives senescence (56), whereas ATM downregulation reduces cell senescence and directs cells toward apoptosis. When grown under hypoxic conditions, senescence is seen in Pttg^−/− BMSCs as evidenced by increased senescence-associated β-galactosidase activity, cell morphology, and reduced cell numbers. Apoptosis levels, as measured by annexin V and Bcl2, were similar in WT and Pttg^−/− cells. We show here that ATM, p53, and p21 are elevated in Pttg^−/− BMSCs, implying that Pttg protects BMSCs from senescence. In the absence of Pttg, this pathway is activated, resulting in decreased BMSC proliferation. Mechanisms for ATM activation in Pttg-null cells are not yet clear.

Securin/Pttg exerts a protective role in radiation-induced damage, and yeast securin mutants show reduced ability to repair UV, X-ray, and γ-ray DNA damage (31) and decreased repair of double-strand breaks (5). Therefore, Pttg deletion could result in double-strand break accumulation and activation of DNA damage ATM/p53/p21 signaling.

Reduced separase expression could also contribute to elevated DNA damage-repair genes in Pttg^−/− BMSCs. In this study, as well is in others, Pttg deletion was associated with separase downregulation (31, 36). Securin/separase complex downregulation in fission yeast led to impaired DNA damage repair (31). The evidence therefore suggests that the securin-dependent downregulation of separase seen in Pttg^−/− BMSCs could lead to impaired DNA damage repair followed by DNA repair gene elevation and ATM activation.

Interestingly, DNA repair genes were upregulated in Pttg^−/− BMSCs, with no apparent chromosomal defects, as evidenced by the normal karyotype and no tumor formation upon injection into mice (data not shown), all expected with accumulated mutations. Moreover, in other models, Pttg^−/− cells exhibit chromosomal aberrations, reflective of the crucial Pttg role in cell division (53, 55). In thyroid and colorectal cells, Pttg overexpression increases chromosomal instability, and, in colorectal cells, the instability inhibits double-stranded DNA repair activity (20, 21).

The crucial role of securin in cell division indicates that a compensatory mechanism for regulating chromatid separation is likely operative in Pttg-null mice cells (27, 52, 53). Phosphorylation maybe an alternative mechanism for inactivating separase. High in vitro CDC2 activity inhibits separase activity by phosphorylation at Ser¹¹²⁶ and Ser¹⁵⁰¹ (47). Furthermore, nonphosphorylated separase, mutated at Ser¹¹²⁶, in securin-deleted mouse embryonic stem cells led to compromised proliferation and failure to maintain sister chromatid cohesion in response to spindle microtubule disruption (13). We observed a shift of separase in Pttg^−/− BMSCs, suggesting that in vivo separase phosphorylation occurs as a compensatory mechanism for Pttg deletion.

We show here that Pttg is required for appropriate BMSC proliferation and for recruitment upon challenge, as evidenced by the poor proliferative capacity of Pttg^−/− BMSCs in both atmospheric oxygen and under hypoxia. Low Pttg-null BMSC proliferation and enhanced senescent features may imply that tissue-specific stem cells contribute to the observed Pttg^−/− phenotype of proliferative arrest associated with splenic, pituitary, testicular, and pancreatic β-cell hypoplasia.

GRANTS

This work was sponsored by National Institute of Cancer Grant CA-75979 (to S. Melmed).

FOOTNOTES

• The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

• Supplemental material for this article is available at the American Journal of Physiology-Cell Physiology website.