|Home | About | Journals | Submit | Contact Us | Français|
Genetic analysis of TP63 indicates that ΔNp63 isoforms are required for preservation of self-renewing capacity in the stem cell compartments of diverse epithelial structures; however, the underlying cellular and molecular mechanisms remain incompletely defined. Cellular quiescence is a common feature of adult stem cells that may account for their ability to retain long-term replicative capacity while simultaneously limiting cellular division. Similarly, quiescence within tumor stem cell populations may represent a mechanism by which these populations evade cytotoxic therapy and initiate tumor recurrence. Here, we present evidence that ΔNp63α, the predominant TP63 isoform in the regenerative compartment of diverse epithelial structuresm, promotes cellular quiescence via activation of Notch signaling. In HC11 cells, ectopic ΔNp63α mediates a proliferative arrest in the 2N state coincident with reduced RNA synthesis characteristic of cellular quiescence. Additionally, ΔNp63α and other quiescence-inducing stimuli enhanced expression of Notch3 in HC11s and breast cancer cell lines, and ectopic expression of the Notch3 intracellular domain (N3ICD) was sufficient to cause accumulation in G0/G1 and increased expression of two genes associated with quiescence, Hes1 and Mxi1. Pharmacologic inhibition of Notch signaling or shRNA-mediated suppression of Notch3 were sufficient to bypass quiescence induced by ΔNp63α and other quiescence-inducing stimuli. These studies identify a novel mechanism by which ΔNp63α preserves long-term replicative capacity by promoting cellular quiescence and identify the Notch signaling pathway as a mediator of multiple quiescence-inducing stimuli, including ΔNp63α expression.
Cellular quiescence is implicated in maintenance of adult stem cells, and evidence indicates that defective quiescence leads to exhaustion of the stem cell pool.1–7 Prolonged tissue stasis is achieved by coordinated regulation of regenerative hierarchies initiated by asymmetric division of an adult stem cell to produce mitotic offspring fated to retain or forfeit self-renewing capacity. While adult stem cells retain proliferative capacity, accumulating evidence indicates that they utilize cellular quiescence to restrict the number of divisions they undergo and to resist differentiation.8–10 Pulse labeling with nucleotide analogs has identified long-term label-retaining cells that have subsequently been shown to co-enrich with adult stem cells.4,11–16 Similarly, inducible expression of a GFP-histone2B fusion protein has enabled isolation of cells based on label retention and the subsequent demonstration that these cells possess potent stem cell activity.17–19 Slow-cycling or non-cycling cells within tumor populations selectively exhibit chemo-resistance and tumor-initiating capacity, suggesting that quiescence is a common feature among tumor stem cell populations20–23 and implying that pharmacologic disruption of stem cell quiescence in the setting of adjuvant therapeutics might reduce rates of cancer recurrence. Quiescence is an active process involving overlapping programs of gene regulation in response to distinct quiescence-inducing stimuli.11 Studies using fibroblasts identified serum deprivation early response genes (SDERGs) that were not coordinately repressed by serum stimulation, indicating a unique transcriptional response to serum deprivation.24 Among the SDERGs, Notch3 was activated in less than one hour following serum deprivation. The canonical Notch target gene, Hes1 is activated in response to quiescence and is sufficient to maintain the reversible nature of quiescence.25 These studies coupled to functional analysis of Notch signaling in MaSCs26 implicate Notch signaling in adult stem cell quiescence.
The Notch signaling pathway executes context-dependent cell-fate decisions in diverse tissue types during embryonic patterning, stem cell regulation, proliferation, differentiation and apoptosis.27–31 Notch activity is implicated in numerous malignancies and has been shown to be either an oncogene32,33 or a tumor suppressor28,34 in distinct cell types. It also governs aspects of development and tumorigenesis in the mammary gland.27 Disruption of Notch signaling in the mammary gland via conditional deletion of RBPj (CBF1), causes expansion of the basal/myoepithelial cells and a concomitant loss of luminal epithelial cells, indicating that Notch signaling was important for luminal cell fate determination.35 Additionally, suppression of CBF1 in enriched fractions of mammary stem cells (MaSCs) resulted in increased proliferation and mammary regenerative activity.26
Together, these results indicate that Notch signaling in the mammary gland contributes to at least two distinct cell fate decisions, luminal vs. basal/myoepithelial differentiation and quiescence vs. activation in MaSCs. Constituative Notch signaling in the mammary gland results in a developmental blockade associated with decreased elaboration of alveoli during pregnancy, and this phenotype is believed to underlie tumorigenesis in this model.36 Other studies have associated Notch signaling with the oncogenic activity of the Wnt signaling pathway in the mammary gland.37
Genetic analysis clearly indicates that ΔNp63 products of TP63 are required for preservation of self-renewal in diverse epithelial structures.38,39 More recent studies have indicated that ΔNp63α is a potent blockade to cellular senescence;40 however, it is unclear if this activity accounts for retention of proliferative capacity in adult stem cells. Here, we provide evidence that ΔNp63α promotes cellular quiescence via the induction of Notch3 expression and activity. ΔNp63α is sufficient to promote cellular quiescence, and Notch3 expression is induced by ectopic ΔNp63α and other quiescence-inducing stimuli in HC11 cells. Other data indicate that ectopic activation of Notch signaling is sufficient to restrict proliferation in a manner that is independent of ectopic ΔNp63α, and that genetic and pharmacologic repression of Notch signaling is sufficient to subvert cellular quiescence induced by ectopic ΔNp63α or other quiescence-inducing stimuli. Our findings identify Notch signaling as a mediator of cellular quiescence and provide a novel mechanism by which ΔNp63α mediates cellular quiescence and preservation of replicative capacity in adult stem cells.
Previously, we reported a genetic interaction between ΔNp63α and hedgehog signaling, demonstrating that ΔNp63α preserves long-term replicative capacity via promotion of cellular quiescence.10 To further test this conclusion and characterize the mechanism(s) underlying ΔNp63α-mediated quiescence, we adopted the HC11 cell line, which is an immortalized model of MaSCs possessing mammary regenerative capacity. Ectopic ΔNp63α caused a significant reduction in cell number (Fig. 1A) with no observable increase in PARP cleavage (Fig. 1B), indicating that ectopic ΔNp63α results in proliferative arrest and not apoptosis. To determine if the proliferative arrest was cellular quiescence, we monitored cell cycle progression of HC11s with ectopic GFP or ΔNp63α. An 18-h thymidine blockade resolved populations that had or had not traversed the G1/S checkpoint. Following this blockade, cells were released and allowed to progress to a nocodazole block in G2. Cell cycle distribution analysis indicated that ectopic ΔNp63α, but not GFP, produced a fraction of cells that failed to progress to and traverse the G1/S boundary (Fig. 1C). This result indicates that ΔNp63α was sufficient to arrest cells in the 2N state under conditions in which there was sufficient mitogenic stimulation for GFP-expressing cells to progress to S phase. A common feature of cellular quiescence is decreased RNA biosynthesis, which can be detected by staining cells with pyronin Y.41 To determine whether the ΔNp63α-mediated accumulation in G0/G1 was the result of increased cellular quiescence, cells expressing ΔNp63α or GFP were stained with the DNA dye Hoechst-33342 and pyronin Y and analyzed by flow cytometry. Results (Fig. 1D) indicate that ectopic ΔNp63α resulted in accumulation of a pyronin Ylow subset of 2N cells. These results demonstrate that ΔNp63α is able to promote cellular quiescence, suggesting a cellular mechanism by which it preserves long-term replicative capacity.
The previous result, coupled to studies linking Notch to TP6342,43 and to quiescence in fibroblasts24 and MaSCs,26 suggested that Notch signaling may contribute to ΔNp63α-mediated quiescence. To test this, we sought to determine the effects of ectopic ΔNp63α on expression of all four Notch family members. Results indicated that in HC11 cells, Notch3 expression increased in response to ectopic ΔNp63α, while Notch1 and Notch2 were unaffected, and Notch4 expression declined (Fig. 2A). Induction of Notch3 by ectopic ΔNp63α was also observed in MCF7 cells (Fig. 2B), suggesting that regulation of Notch3 by ΔNp63α may be a common event in mammary epithelium. These results indicated a regulatory relationship between ΔNp63α and Notch3, which further suggested that ΔNp63α may contribute to Notch signaling. Protein gel blot analysis of Notch3 in HC11 cells overexpressing either GFP or ΔNp63α indicated that ectopic ΔNp63α caused increases in both full-length Notch3 and the truncated intracellular domain (Fig. 2C). Similarly, ectopic ΔNp63α was sufficient to increase expression of Hes1 (Fig. 2D). Together, these results indicate that ΔNp63α increases expression of Notch3 and activation of the Notch signaling pathway. These results coupled to studies linking ΔNp63α to quiescence10 and Notch to quiescence11,24,43 suggest that Notch3 expression is a common feature of cellular quiescence. To test this, HC11s cells were cultured under normal monolayer conditions or under three distinct quiescence-inducing conditions, serum deprivation (SD), low-binding culture (LB) and contact inhibition (CI). QPCR-based analysis indicated increased Notch3 expression in response to each of these quiescence-inducing stimuli (Fig. 2E). Together, these results indicate that ectopic ΔNp63α and other quiescence-inducing stimuli enhance expression of Notch3. They also suggest that expression of Notch3 may mediate the effects of ΔNp63α and other quiescence-inducing stimuli.
The previous data coupled to studies implicating Notch signaling in the governance of MaSCs proliferation26 predict that Notch3 expression and signaling mediates ΔNp63α-induced quiescence, and that activation of Notch signaling in HC11 cells will promote quiescence. To test this, gain-of-function studies were performed by infecting HC11 cells with an adenovirus programmed to express the intracellular domain of Notch3 (N3ICD) or GFP. Overexpression of N3ICD(Fig. 3A) but not GFP was sufficient to activate a Notch signaling reporter consisting of four CBF1-binding elements fused to the SV40 minimal promoter.44 In similar studies, N3ICD had a potent anti-proliferative effect on HC11 cells (Fig. 3C) and was sufficient to induce Hes1 mRNA levels approximately 10-fold (Fig. 3C). The latter result indicates that ectopic N3ICD was sufficient to activate expression of a canonical Notch signaling target gene. Additionally, Hes1 is required to preserve the reversibility associated with quiescence.25 In addition to Hes1, N3ICD also induced mRNA levels of Mxi1, a negative regulator of c-myc activity recently identified as a serum deprivation early response gene (SDERG) and shown to be essential for quiescence.24 These studies demonstrate that Notch signaling is sufficient to arrest the growth of HC11 cells in a manner that is consistent with cellular quiescence.
The previous data indicated that ectopic Notch signaling reduced proliferation and increased expression of genes associated with cellular quiescence. This predicts that disruption of Notch signaling in HC11 cells might subvert quiescence. To test this, we sought to measure the effects of Notch disruption on BrdU incorporation rates following quiescence-inducing stimuli. Pharmacologic inhibition of Notch signaling with the γ-secretase inhibitor DAPT doubled the rate of BrdU incorporation following growth factor reduction (Fig. 4A), which supports the assertion that Notch-mediated growth arrest is reversible and consistent with cellular quiescence. Similarly, shRNA-mediated suppression of Notch3 (Fig. 4E) resulted in increased proliferation under growth factor-reduced conditions (Fig. 4B). These results indicate that disruption of Notch signaling or Notch3 expression is sufficient to confer resistance to quiescence-inducing stimuli and are consistent with previous studies indicating that suppression of Notch signaling in MaSCs results in mitotic expansion.26 They also predict that suppression of Notch signaling will promote expansion of self-renewing subpopulations within HC11 cells. To test this, the mammosphere-forming capacity of HC11 cells was measured in the presence of DAPT or a vehicle control. Quantification of mammospheres indicated that inhibition of Notch signaling with DAPT caused a statistically significant increase in mammosphere initiation (Fig. 4C). Additionally, shRNA-mediated suppression of Notch3 resulted in greater mammosphere initiation relative to scrambled shRNA controls (Fig. 4D). These results demonstrate that Notch signaling exerts an anti-proliferative effect on self-renewing populations within the HC11 cell culture system.
Data presented here support a model in which ΔNp63α promotes quiescence by increasing Notch3 expression and activity. This model predicts that suppression of Notch3 expression will disrupt ΔNp63α-mediated cellular quiescence. We sought to compare the effects of ectopic ΔNp63α on proliferation by HC11 derivatives programmed to express a scrambled shRNA or a Notch3-directed shRNA. Results (Fig. 5A) indicate that suppression of Notch3 subverted the anti-proliferative effects of ΔNp63α. Additionally, suppression of Notch3 expression disrupted ΔNp63α-mediated accumulation of cells in the 2N state (Fig. 5B). Finally, suppression of Notch3 significantly reduced ΔNp63α-mediated accumulation of pyronin Ylow-staining cells. Together, these results demonstrate that suppression of Notch3 expression (Fig. 5C) is sufficient to subvert ΔNp63α-mediated cellular quiescence. These data strongly support a model in which the ability of ΔNp63α to increase expression and activity of Notch3 is functionally linked to the ability of ΔNp63α to promote cellular quiescence.
Studies indicate that label-retaining cells co-purify with tissue-specific adult stem cells, suggesting a physiologic role for quiescence in preservation of self-renewal within regenerative hierarchies that govern development, stasis, aging and cancer. Central to this model is the functional asymmetry of stem cell division that yields mitotic siblings with distinct fates. While one sibling forfeits self-renewing capacity and enters a stage of transient amplification, a second retains it and enters a state of quiescence. Doing so enables retention of proliferative capacity and evasion of the negative effects of excessive cell division, including telomeric erosion, accumulation of reactive oxygen species and increased risk of mutation. Quiescence is also a potent blockade to differentiation,11 suggesting a role in developmental potency. Despite the critical role of quiescence, the molecular and genomic events associated with entry into and maintenance of quiescence are incompletely understood. Work presented here describes a regulatory relationship between ΔNp63α and Notch3 that governs quiescence and demonstrates for the first time that ΔNp63α promotes quiescence and suggests a mechanism by which ΔNp63α promotes stem cell longevity.
While the hypotheses surrounding the role of adult stem cells in cancer initiation and etiology remain controversial and unproven, there is abundant evidence indicating that diverse tumors possess a subpopulation of cells that are uniquely tumorigenic and able to self-renew. Multiple studies have demonstrated that this subpopulation displays broad-spectrum resistance to cytotoxic chemotherapeutics and ionizing radiation,45–47 thereby implicating this subpopulation in cancer recurrence. Other studies demonstrate a correlation between label retention and chemoresistance in cancer models, suggesting that cellular quiescence may confer resistance to therapeutics that target proliferating cells.23 Consistent with this are studies indicating that subversion of quiescence in leukemic stem cells renders these cells sensitive to chemotherapeutics.48 Therefore, targeting genetic pathways governing stem cell quiescence in the setting of adjuvant therapeutics represents a promising strategy to reduce cancer recurrence. Here, we present data indicating that disruption of Notch signaling subverts quiescence in a cell culture model with features of MaSCs. Additionally, we demonstrate that Notch signaling may mediate a cellular response to diverse quiescence-inducing stimuli, including ΔNp63α activity, suggesting a fundamental role in cellular quiescence that may apply to multiple cancer stem cell models.
Cellular and developmental context are critical determinants of Notch signaling output that may account for the diverse cellular responses to perturbations in Notch signaling.27–30 This diversity is best illustrated by abundant and compelling evidence that Notch signaling can be either oncogenic or tumor suppressive in distinct cellular contexts. A model that may account for this variability holds that Notch signaling instructs mutually exclusive cell fates upon Notch donor and recipient cells via lateral inhibition;42 however, the molecular determinants of context specificity are largely unidentified. Within the context of governance of stem cell quiescence, our study suggests that ΔNp63α may be one such determinant of Notch signaling output. Additionally, it is unclear whether the four Notch family members mediate identical Notch signaling outputs. In this manuscript, we reported that while only Notch3 was responsive to ΔNp63α, both Notch1 and Notch2 were expressed at higher levels. This may suggest that effects of DAPT are due to inhibition of signaling from multiple Notch receptors. In this manuscript we present data indicating that shRNA-mediated ablation of Notch3 was sufficient to prevent quiescence following growth factor reduction. Our data indicate that Notch signaling is anti-proliferative and promotes expression of genes associated with quiescence and support a model in which activation of Notch by ΔNp63α represents a mechanism by which stem cell quiescence is maintained. Consistent with this model is the remarkable finding that the mammary glands of MMTV-Notch1ICD and MMTV-Notch3ICD undergo a developmental blockade that disrupts lobulo-alveolar development in pregnant mice.36 This is in contrast to several other MMTV-based breast cancer models characterized by precocious lobulo-alveolar development and is consistent with a model in which Notch signaling suppresses activation of the mammary regenerative hierarchy.
Two recent studies have also implicated p53 in the governance of stem cell activity. One identified necdin as a gene that is regulated by p53 in the absence of any cellular or genotoxic stress and showed that necdin was necessary to maintain the ratio of long-term hematopoetic stem cells to short-term hematopoetic stem cells.49 A second study showed that p53-/- mammary epithelial cells had greater mammosphere-forming capacity and greater mammary regenerative capacity than mammary epithelial cells from wild-type counterparts.50 Importantly this study showed that the increased regenerative activity that resulted from p53 ablation was neutralized by DAPT, suggesting strongly that Notch signaling was activated in response to p53 suppression. This study also showed that long-term BrdU retention was compromised in p53-/- mice. Collectively, these studies indicate that a complex relationship between ΔNp63 isoforms and p53 may underlie the governance of quiescence vs. activation in adult stem cell populations.
Our conclusion that ΔNp63α promotes cellular quiescence coupled to studies indicating that it is required to avoid cellular senescence40,51,52 suggests a dynamic model in which ΔNp63α balances quiescence and senescence to preserve long-term replicative capacity and a prolonged life span. While there is abundant evidence that p53 is a potent inducer of cellular senescence, more recent studies have indicated that it can also promote quiescence and, in so doing, prevent senescence. Importantly, this study demonstrated that the ability of p53 to induce quiescence is the result of p53-mediated suppression of senescence.53 While the cellular and molecular mechanisms underlying this paradox are incompletely understood, another recent study has implicated the status of the mTOR signaling pathway in the p53-mediated outcome.54 Additionally, ΔNp63α has been shown to be a transcriptional target of p53.55 This coupled to the fact that ΔNp63α is expressed in a highly cell-type dependent manner suggests a model in which cells that are capable of p53-dependent regulation of ΔNp63α may be prone to quiescence, while those in which p53 is present but ΔNp63α expression is repressed may be prone to senescence.
HC11 cells (a kind gift from Sergei Tevosian) were cultured in RPMI-1640 with L-glutamine (Mediatech Inc.) supplemented with 10% FBS, 5 ug/ml Insulin, 10 ng/ml EGF, 100 units/ml Penicillin, 100 ug/ml Streptomycin, and MCF7 cells were cultured in DMEM supplemented with 10% FBS at 37°C and 5% CO2. Serum deprivation studies involved culturing cells in a 100-fold dilution of normal media for 14 h. Low-binding assays were performed in 24-well Ultra Low Cluster plates (Corning) with normal media. Sphere counting was done using ImageJ software based on three low power images for each sample. Following growth to confluence, media was replaced and cells were maintained for two additional days to allow for contact inhibition. Cell counting experiments were done following trypsinization and counting by hemocytometer.
For cell cycle analysis based on propidium iodide or Hoechst/pyronin staining followed by flow cytometry, cells were treated as described, detached by trypsinization, resuspended in PBS, fixed in a final concentration of ice-cold 70% EtOH and stored overnight at −20°C prior to analysis. For propidium iodide staining, cells were washed in PBS, resuspended in PBS with 0.1% Triton X-100, 0.2 mg/ml DNase-free RNase A and 40 ul of 500 ug/ml propidium iodide stock and incubated at 37°C for 30 min. For Hoechst/pyronin staining, cells were washed in HBSS with Ca2+ and Mg++ and resuspended in HBSS containing 1.2 ug/ml Hoechst 33,342 and 2 ug/ml pyronin Y. Flow cytometry was performed on a BD FacScan or BD FacsAria for propidium iodide or Hoechst/pyronin Y, respectively. For synchronization, cells were treated with 2 mM thymidine or 100 ng/ml nocodazole. To assess S-phase fraction, cells were pulsed with 10 uM Bromodeoxy Uridine (BrdU) for 20 min followed by fixation with CytoRich Red (Thermo Fisher Scientific). BrdU was detected with mouse anti-BrdU (BD Bioscience) and goat anti-mouse AlexaFluor 568 (Invitrogen). Number of BrdU-positive cells per total cells from four high power fields were counted for each sample.
ΔNp63α and Notch3 intracellular domain were sub-cloned from pcDNA3.1 into pShuttleCMV and further recombined through the AdEasy Adenoviral production system using HEK-293Ad cells. Adenovirus was used as 1,000x for all reactions. pLKO.1-Notch3 shRNAs (OpenBiosystems, TRCN0000075571 “shRNA1,” TRCN0000075572 “shRNA2”) were cotransfected with packaging and envelope containing plasmids into HEK-293T cells. Virus-containing media was used to infect HC11 cells, and stable expressing populations were selected and grown in the presence of puromycin. Data from shRNA1 expressing cells are shown unless otherwise specified.
Cleavage of Notch receptors was blocked by inhibition of γ-secretase with 10 µM DAPT (N-[N-(3,5-Difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester, Sigma).
DNA fragment coding for EGFP was removed from CBFRE-EGFP (Addgene) and replaced with DNA fragment coding for luciferase to produce CBF-luc plasmid. Cells were treated as described and transfected with CBF-luc and renilla for 24 h. Luciferase activity was assessed with Dual-Luciferase Reporter Assay System (Promega).
RNA was collected with the RNEasy kit (QIAGEN), and cDNA was prepared using the iScript cDNA Synthesis kit (BioRad). QPCR was performed with oligonucleotide primers specific for each target with iQ SYBR Green Super mix (BioRad) and analyzed with Bio-Rad CFX Manager software. Normalization with GAPDH was done using the 2−ΔΔCt method. Protein gel blotting of protein lysates prepared with NET-N lysis buffer following PAGE was performed with antibodies detecting PARP (rabbit, Cell Signaling Inc.-), p63 (mouse, 4A4, Abcam), Notch3 (goat, M-20, Santa Cruz) or β-actin (mouse, 8H10D10, Cell Signaling Inc.).
This work was supported by grants from the National Cancer Institute (RO1CA108539) to J.D.R. and the Elsa U. Pardee Foundation to J.D.R. S.K. was supported by a fellowship from the Rosaline Borison Memorial Predoctoral Fellowship.
No potential conflicts of interest were disclosed.