Cell Cycle Inhibitors in Normal and Tumor Stem Cells

Cell cycle inhibitors in normal and tumor stem cells

Tao Cheng*,1

1Department of Radiation Oncology, University of Pittsburgh School of Medicine and University of Pittsburgh Cancer Institute, Pittsburgh, PA 15213, USA

Emerging data suggest that stem cells may be one of the cell types have been or are being defined, the hemato- key elements in normal tissue regeneration and cancer poietic stem cell (HSC) remains one of the best-studied development, although they are not necessarily the same stem cell types and therefore data have been largely entity in both scenarios. As extensively demonstrated in obtained from HSCs partially in comparison with the the hematopoietic system, stem cell repopulation is other stem cell types such as the neural stem cell (NSC). hierarchically organized and is intrinsically limited by The distinct impacts of different CKIs in stem cell the intracellular cell cycle inhibitors. Their inhibitory populations underscore the crucial role that cell cycle effects appear to be highly associated with the differentia- inhibitors play in stem cell regulation and offer new tion stage in stem/progenitor pools. While this negative insights into the mechanisms of cancer development, regulation is important for maintaining homeostasis, especially in light of the concept of ‘tumor stem cell’ especially at the stem cell level under physiological cues (TSC) (or described as ‘cancer stem cell’ in general or or pathological insults, it constrains the therapeutic use of ‘leukemic stem cells’ in leukemias by some other adult stem cells in vitro and restricts endogenous tissue investigators in the field). The biochemical pathways repair after injury. On the other hand, disruption of cell of general cell cycle regulation will not be detailed in the cycle inhibition may contribute to the formation of the so- current review, since they have been extensively called ‘tumor stem cells’ (TSCs) that are currently reviewed elsewhere (Pardee, 1989; Sherr and Roberts, hypothesized to be partially responsible for tumorigenesis 1999; Sherr, 2000). and recurrence of cancer after conventional therapies. Therefore, understanding how cell cycle inhibitors control stem cells may offer new strategies not only for therapeutic manipulations of normal stem cells but also Stem cell proliferation with a limited rate for novel therapies selectively targeting TSCs. Oncogene (2004) 23, 7256–7266. doi:10.1038/sj.onc.1207945 By definition, stem cells have the ability to reproduce themselves indefinitely while possessing the potential to Keywords: adult stem cell; hematopoietic stem cell; differentiate into multiple lineages (or a single lineage in tumor stem cell; tissue regeneration; self-renewal; cell a few tissue types) via transit amplifying cells (progeni- cycle; cyclin-dependent kinase inhibitor; p21; p27; p18; tors). However, the proliferative rate of stem cells in vivo p16; Rb protein under physiological conditions is much slower compared to that of intermediate progenies, including the progenitor cells and proliferating precursors. In the hematopoietic system, an increase in stem cell divisions can take place (Pawliuk et al., 1996; Oosten- dorp et al., 2000) under activating conditions, such as Introduction after transplantation or the depletion of cycling cells using the S-phase toxin (5-fluorouracil or hydroxyurea) The therapeutic efficacy of stem cells largely relies on (Harrison and Lerner, 1991; Uchida et al., 1997). their ability to replicate. Therefore, strategies to However, the relative quiescence of the HSC pool manipulate stem cells require an understanding of their appears to be essential to prevent premature depletion cell cycle control. In this review, the distinct cell cycle under conditions of physiologic stress over the lifetime kinetics of adult stem cells will be introduced and briefly of the organism (Cheng et al., 2000a, b). The highly followed by the current understanding of general cell regulated proliferation of HSC occurs at a very limited cycle regulation in mammalian cells. The focus will be rate under homeostatic conditions. It had once been placed on the specific impact of the cell cycle inhibitors, hypothesized that the stem cell quiescence reflects a namely the roles of cyclin-dependent kinase inhibitors complete cell cycle arrest of the majority cells in the stem (CKIs) on adult stem/progenitor cells. While many stem cell compartment, termed the ‘clonal succession model’ (Kay, 1965). While the retrovirus-based clonal marking *Correspondence: T Cheng, Hillman Cancer Center, Research studies revealed few active stem cell clones at a given Pavilion, Office suite 2.42e, 5117 Center Avenue, Pittsburgh, PA time, which supports the clonal succession theory 15213-1863, USA; E-mail: [email protected] (Lemischka et al., 1986), this view has been challenged Cell cycle inhibitors in stem cells T Cheng 7257 by the competitive repopulation model (Harrison et al., to be one of the hurdles in the context of the in vitro 1988) or 5-bromodeoxyuridine (BrdU) incorporation in expansion necessary for stem cell transplantation and the defined stem cell pools with a simulation methodol- gene therapy. Methods for inducing stem cell prolifera- ogy (Cheshier et al., 1999). Using BrdU labeling tion have long been sought as a means to expand the experiments, cell cycle lengths were estimated at population of cells capable of repopulating the bone approximately 30 days in small rodents (Bradford marrow of ablated hosts and to render stem cells et al., 1997) and it was found that only about 8% of transducable with virus-based gene transfer vectors. the cells are in cell cycle daily (Cheshier et al., 1999). Although great effort has been made to directly expand Similar analyses using population kinetics estimated stem cells using different combinations of hematopoietic that stem cells replicate once per 10 weeks in cats growth factors (cytokine cocktails), no culture system (Abkowitz et al., 1996). In non-human primates, the rate has been applied successfully in clinical settings. This is of cell replication in the stem cell pool was estimated to due, in part, to the lack of proof that any of the culture be once per year (Mahmud et al., 2001). In contrast, the conditions support expansion of a long-term repopulat- essential feature of the hematopoietic progenitor cell ing HSC in humans (Verfaillie, 2002). Although data (HPC) population is that it irreversibly develops into suggest that under certain specific conditions murine maturing cells, and in the process undergoes multiple HSCs may divide in vitro, net expansion is achieved in rapid cell divisions. The progenitor cell pool essentially limited fashion and is associated with and often operates as a cellular amplification machine, generating dominated by cellular differentiation (Ema et al., 2000; many differentiated cells from the few cells entering the Uchida et al., 2003). It remains unclear which combina- system (Potten, 1997), and it is therefore directly tion of hematopoietic cytokines is specifically selective responsible for the number of terminally differentiated for stem cell proliferation without differentiation. So cells. For this reason, the progenitor cell pool is also far, there is no cytokine combination that can achieve termed the transit, amplifying cell pool. The differences the effect of stem cell expansion by the HoxB4 protein between the stem and progenitor cell populations have (Antonchuk et al., 2002). A recent promising study been regarded as phenotypic distinctions marking the demonstrated a potent effect of purified Wnt3a protein stage of a cell within the hematopoietic cascade. in the expansion of mouse HSC in vitro. But the effect However, an alternative model was recently put forward appeared to require a strong survival element since the proposing that the specific position in a cell cycle net expansion could be only achieved on the HSCs from determines whether a primitive cell functions as a stem the Bcl2 transgenic mice (Willert et al., 2003). or a progenitor cell (Quesenberry et al., 2002). In this In short, the dichotomy of the relative resistance to hypothetical model, stimuli received at distinct positions proliferative signals by stem cells and the brisk respon- in the cell cycle provoke proliferation/differentiation or siveness by progenitor cells is generally believed to be a not, and thereby yield either HSC or HPC outcomes. central feature of tissue maintenance, although the This seems to contradict the traditional view of phenotypic distinctions between stem and progenitor ‘hierarchy’ within the hematopoietic differentiation. cells in many nonhematopoietic organs have not been Nevertheless, in either model, ‘stemness’ is associated fully defined. Nevertheless, the functional difference in with the limited rate of the cell proliferation. proliferative response between stem cells and progeni- The slow cycling feature seems to be a common tors represents a challenge for the specific manipulations feature in most adult stem cell types if not all (Potten, of the stem cells for therapeutic purposes. While a 1997). In the central nervous system (CNS), evidence complex array of extracellular signals and intracellular suggests that the proliferative pools of adult neural transduction pathways certainly participate in the progenitors are derived from a quiescent multipotent distinct response, the cell cycle machinery, as a final neural precursor or NSC (Bonfanti et al., 2001; Palmer step, must communicate with the specific regulatory cues et al., 2001). For example, if the proliferative zone (Steinman and Nussenzweig, 2002) and cell cycle containing the lineage-committed neuronal progenitor regulators must play key roles in this process (D’Urso cells (NPC) is ablated, it can be repopulated from a G and S, 2001). small number of quiescent NSCs (Morshead et al., 1994; Doetsch et al., 1999). Perhaps largely owing to this quiescence, endogenous NSCs do not produce complete Cell cycle regulation and CDK inhibitors recovery in cases of severe injury, though they do participate in self-repair after brain damage (Horner The molecular principles of cell cycle regulation have et al., 2000). In the mouse dermal stem cell population, been defined largely in yeast and in orthologous systems there are about fourfold fewer cells in S-G2/M phases in applicable to the mammalian cell cycle (Hartwell and the stem cell population compared with the progenitor Weinert, 1989). A number of surveillance checkpoints pool, though both cell populations constantly proceed monitor the cell cycle and halt its progression, mainly through the cell cycle (Dunnwald et al., 2003). In via the p53 pathway (Vogelstein and Kinzler, 1992), addition, epithelial stem cells in mouse cornea (Cotsar- when DNA damage occurs. In mammalian cells, the cell elis et al., 1989) or mouse hair follicle (Morris et al., cycle machinery that determines whether cells will 2004) appear to be slowly cycling as well. continue proliferating or will cease dividing and The relative quiescence of stem cells may prevent their differentiate appear to operate mainly in the G1 phase premature exhaustion in vivo, but it has been considered (Figure 1). Cell cycle progression is regulated by the

Oncogene Cell cycle inhibitors in stem cells T Cheng 7258 p16 INK4a reading frame (ARF) (Sherr, 2001). Intriguingly, ARF protein is able to activate p53 function by removing the p15 INK4b p21cip1 inhibitory effect of Mdm2 on p53 protein (Sherr and p18 INK4c p27kip1 Weber, 2000). Therefore, the unique p16 locus serves as a bridging point between the two most frequently p19 INK4d p57kip2 targeted networks in the cells of the majority of cancers: RB and p53 pathways. Both CKI families have been

CDK4,6 G1 CDK2 shown to be essential in arresting cell cycle progression Cdc25A in a broad spectrum of cell types (Morgan, 1995; Sherr, Cyclin D Cyclin E 1996). Studies using antisense of p21 or p27 have been

Rb E2F E2F Cyclin A able to release the cells from Go stage into the cell cycle C-Myc (Nakanishi et al., 1995; Rivard et al., 1996). p27 and p18 Rb- p … have been proposed as intrinsic timers regulating animal G1 organ size in an autonomous manner (Conlon and Raff, 1999). G0 M S While we have a wealth of knowledge about the G2 biochemical roles of CKIs in a variety of model systems Figure 1 Cell Cycle Control of G1 Phase in Mammalian Cells. (Sherr, 1996; Sherr and Roberts, 1999; Sherr, 2000), few Description about each regulator during the G1 phase is detailed in experiments have been performed in the defined stem the text. The CDK inhibitors are indicated in the black boxes cell populations. There appears to be a distinct cell cycle control operating in stem cells in order to maintain their ‘stemness’. Mouse embryonic stem (ES) cells were shown to have a ‘defective’ Rb pathway and a sequential activation and inactivation of CDKs (Sherr, nonresponsive p53 pathway (Aladjem et al., 1998; 1994; Sherr and Roberts, 1995). In somatic cells, Prost et al., 1998; Burdon et al., 2002). In adult tissues, movement through G1 and into the S phase is driven stem cells and progenitor cells share many common by the active form of the Cyclin D1, 2,3/CDK4, 6 cytokine receptors (Berardi et al., 1995; Becker et al., complex and the subsequent phosphorylation retino- 1999; Roy and Verfaillie, 1999; Shen et al., 1999), it is blastoma (Rb) protein (Classon and Harlow, 2002). likely that the distinct cell cycle profile in stem cells is Once Rb is phosphorylated, the critical transcription mediated by either distinct upstream intracellular factor, E2F-1, is partially released from an inhibited mediators or unique combinatorial relationships of state and it turns on a series of genes including cyclin A common biochemical mediators that limit the intensity and cyclin E that form a complex with CDK2 as well as of signals to enter into cell cycle. Defining the cdc25A phosphatase. Cdc25A is able to remove the mechanisms in the stem cell response requires the inhibitory phosphates from CDK2 and the resultant analysis of individual cell cycle regulators and ultimately cyclin E/CDK2 complex, then it further phosphorylates a systematic approach to define how these cell cycle Rb, leading to a complete release of E2F and the regulators interact with one another and intersect transcription of a series of genes essential for S-phase various signaling pathways. progression and DNA synthesis. In parallel, the c-Myc pathway also directly contributes to the G1/S transition by elevating the transcription for cyclin E and cdc25A Roles of CKIs in stem cell regulation (Bartek and Lukas, 2001) (Figure 1). CDK activity is strictly dependent on cyclin levels that Roles of CKIs in stem and progenitor cells have been are regulated by proteosome-mediated proteolysis. shown in several species since Dipio, an analogue of p21/ Upon mitogenic stimulation, cyclin D serves as an p27, was reported to control embryonic progenitor essential sensor in the cell cycle machinery and interacts proliferation in Drosophila (de Nooij and Hariharan, with the CDK4/6-Rb-E2F pathway. In addition to 1995; de Nooij et al., 1996). However, further study of regulation by cyclins and phosphorylation/dephosphor- their roles in stem cells from other higher species by ylation of the catalytic subunit, CDKs are largely conventional experiments in molecular biology is controlled by CKIs (Sherr and Roberts, 1999). Two hampered due to the extremely limited availability of families of low molecular weight CKIs, Cip/Kip and the primary stem cells. Fortunately, genetically engi- INK4, have been identified as capable of interacting neered animal models and cell-sorting technologies with CDKs to suppress progression through G1. The allow us to vigorously test the gain or loss of gene Cip/Kip family, which includes p21Cip1/Waf1,p27kip1 and function in the defined HSC and HPC populations in p57Kip2 (p21, p27 and p57 hereafter), may interact with a the hematopoietic system. CKI knockout rodent models broad range of cyclin–CDK complexes, while the INK4 have been particularly useful in providing a feasible family, which includes p16INK4A, p15INK4B, p18INK4C and approach for studying the roles of cell cycle inhibitors in p19INK4D (p16, p15, p18 and p19 hereafter), specifically stem cell biology (Brugarolas et al., 1995; Nakayama inhibits CDK4 and CDK6 kinases. Of note, p16 locus et al., 1996; Franklin et al., 1998). In the hematopoietic also encodes a structurally unrelated protein (p14ARF in system, most CKI family members have been found to humans or p19ARF in mice), through an alternative be differentially expressed in human CD34 þ cells

Oncogene Cell cycle inhibitors in stem cells T Cheng 7259 (Taniguchi et al., 1999; Tschan et al., 1999; Yaroslavskiy hippocampus and subventricular zone after brain et al., 1999; Marone et al., 2000; Cheng et al., 2001), ischemia. The hippocampal precursors migrated and indicating their distinct effects in hematopoiesis. In the differentiated into a higher number of neurons post following paragraphs, the HSC will be most used as an injury. In an earlier study, p21 null also promoted the example to describe the roles of CKIs in stem cell skin stem cell potential in mice (Missero et al., 1996). regulation unless other adult stem cell types are Therefore, p21 may serve as a common molecular specified. regulator restricting proliferation among stem cell pools from distinct tissue types, despite the distinct kinetics p21: a gatekeeper for quiescent stem cells between different tissue types. Paradoxically, the role of p21 has been noted to p21 mRNA is abundant in quiescent human HSCs, positively affect proliferation following cytokine stimu- while it is reduced in progenitor populations (Ducos lation in progenitor cell pools (Liu et al., 1996). This et al., 2000; Stier et al., 2003). Functional assessment in may be due to the requirement for p21 to promote the hematopoietic cells been carried out using p21À/À mice association of CDK4 with D-type cyclins. LaBaer et al. (Cheng et al., 2000a, b). In the absence of p21, HSC (1997) demonstrated that low concentrations of p21 proliferation tends to increase under normal homeo- promote assembly of active kinase complexes and static conditions. Exposing the animals to cell cycle- thereby entry into cycle, whereas higher concentrations specific myelotoxic injury resulted in a higher rate of are inhibitory. The stoichiometry of p21 and cyclin– mortality due to hematopoietic cell depletion. Therefore, CDK complexes appears to be crucial in determining the p21 governs cell cycle entry of stem cells, and its absence relative effect on movement of the cell through late the leads to increased proliferation of the primitive cells. G1 into S phase. This was further confirmed in a study However, self-renewal of HSCs was impaired in serially showing that p21 and p27 are essential activators of transplanted bone marrow from p21À/À mice (Cheng cyclin D-dependent kinases in murine fibroblasts (Cheng et al., 2000b), suggesting that restricted cell cycling is et al., 1999). Mantel et al. (1996) noted that bone crucial to prevent premature stem cell depletion and marrow progenitor cells from p21-null mice proliferated hematopoietic death under conditions of stress. Such poorly and formed fewer colonies with thymidine results might not be solely attributed to the stem cell treatment except when transduced with a p21-encoding effect and other possible defects in differentiation or retroviral vector (Braun et al., 1998). Similarly, these apoptosis program, especially in the downstream of the defects were complemented by a transient rise in p21 hematopoiesis, cannot be completely ruled out based on immediately following release of cell cycle arrest in the the study. Nevertheless, the increased HSC cycling that murine myeloid progenitor 32D cell line (Cheng et al., results from p21 absence in mice has recently been 2001). Therefore, as observed in other systems, p21 has a extended to human cells and to a nondevelopmental dual function in the hematopoietic system, depending on context. Using postnatal CD34 þ CD38À human cells, it the differentiation stage and CDK complex type and was shown that interrupting p21 expression with status. In addition, complex roles of p21 in apoptosis lentivectors ex vivo resulted in expanded stem cell (Wang and Walsh, 1996) or differentiation (Steinman number validated with the transplantation assay in et al., 1994, 1998) may also participate in stem cell irradiated NOD/SCID mice (Stier et al., 2003). This regulation, though these functions are yet to be study demonstrated a proof-of-concept for an alter- thoroughly investigated. With regard to cell cycle native paradigm in which HSC numbers can be regulation, it is unclear why the positive role of p21 in increased by releasing the brake on cell cycle entry forming the complex of CDK4 or CDK6 with D-type rather than focusing on combinations of pro-prolifera- cyclins dominates its inhibitory effect on CDK2 activity tive cytokines that can also induce differentation. These in the progenitor cell pools. data further supported the notion that postnatal human It is also unclear why p21 expression is elevated in stem cell proliferation can be uncoupled from differ- quiescent HSCs. Two direct upstream regulators of p21 entiation in ex vivo settings. have been assessed. WT1 is known to induce p21 Given that relative quiescence is a common feature of transcription, and overexpression of WT1 results in different tissue stem cell types and that p21 maintains altered stem cell cycling and differentiation in primary HSC quiescence, it was hypothesized that p21 might hematopoietic cells (Ellisen et al., 2001). Also, it is well also participate in limiting NSC reactivity in vivo. documented that p21 is also transcriptionally regulated Neural precursor cells from adults have exceptional by p53 and serves as a downstream mediator of cell cycle proliferative and differentiative capabilities in vitro, yet arrest (El-Deiry et al., 1992; el-Deiry, 1998). It would be respond minimally to in vivo brain injury due to reasonable to expect the HSC phenotype of the p53À/À constraining mechanisms that are poorly defined. The animal to be similar to that of p21À/À. In fact, in the role of p21 in the regenerative response of neural tissue absence of p53, HSC function has been reported to be is evaluated in mice deficient in p21 following ischemic significantly enhanced under stress conditions (Wlo- injury (Qiu et al., 2004). While steady-state conditions darski et al., 1998; Hirabayashi et al., 2002) in a manner revealed no increase in primitive cell proliferation in opposite to that found in the absence of p21 during p21-null brain, which is different from that found in the long-term engraftment. Since one of the major cellular hematopoietic system, a significantly larger fraction of functions of p53 is to initiate cell death upon genotoxic quiescent neural precursors was activated in the stress, the enhanced function of HSCs in the absence of

Oncogene Cell cycle inhibitors in stem cells T Cheng 7260 p53 suggests that increased survival may predominate p27 appears to accumulate at points in which signals over the accelerated proliferation of HSCs in some for mitosis affect cell cycle regulators, and it has been settings. One alternative explanation is that p21 may shown to serve as an important regulator at a restriction affect the stem cells in a p53-independent pathway. This point for mitogenic signals in many cell types (Coats latter possibility is indirectly supported by a recent study et al., 1996). As progenitor cells are highly responsive to where an increase of neuronal regeneration in p21À/À growth factors, though in a tissue-specific fashion, p27 but not p53À/À brain was observed after the ischemic must be a critical cell cycle mediator of many cytokines injury (Qiu et al., 2004). in progenitor cells. p27 was originally cloned in the TGF-b1-treated cell line as a TGF-b1 responsive gene p27: a progenitor-specific inhibitor to the repopulation (Polyak et al., 1994). In the hematopoietic system, TGF- efficiency b1 has been extensively characterized as a dominant- negative regulator of hematopoietic cell proliferation Unlike p21, p27 mRNA is present ubiquitously in all (Hatzfeld et al., 1991; Cardoso et al., 1993; Fortunel hematopoietic cell populations and p27 protein is et al., 1998; Cashman et al., 1999). Antisense TGF-b1or primarily regulated by the degradation mechanisms neutralizing antibodies of TGF-b1 have been used to (Carrano et al., 1999). Direct flow cytometric analysis augment retroviral gene transduction in conjunction shows p27 expression in primitive cells (Tong and Srour, with downregulation of p27 in human CD34 þ subsets 1998) and in more mature progenitors (Taniguchi et al., (Hatzfeld et al., 1991, 1996; Dao et al., 1998). Therefore, 1999; Yaroslavskiy et al., 1999). Using p27À/À mice in it has been hypothesized that p27 mediates the effect of conjunction with the hematopoietic reconstitution as- TGF-b1 in HPCs (Fortunel et al., 2000). In an attempt say, it has been reported that p27 does not affect HSC to further investigate the coordinated regulation of number, cell cycling, or self-renewal, but markedly alters TGF-b1 (Cheng et al., 2001), p27 was analysed in the HPC proliferation and pool size as assessed by the cytokine-responsive 32D cell line. Despite the marked colony-forming cell (CFC) assay (Cheng et al., 2000a). antiproliferative effects of TGF-b1, the expression of When competitively transplanted, p27-deficient HSC p27 was not altered. Further, HPCs derived from mice generated progenitors that eventually dominated blood engineered to be deficient in p27 were inhibited by TGF- cell production. Thus, modulating p27 expression in a b1 to an equivalent degree of HPCs from the wild-type small number of stem cells without necessarily expand- littermate controls. These data indicated that TGF-b1 ing the cells may translate into effects on the majority of exerts its inhibition on cell cycling independent of p27 in mature cells, thereby providing a strategy for potentiat- HPCs. Therefore, the exact upstream regulators for p27 ing the impact of transduced cells in stem cell gene in a variety of progenitor cells are still somewhat elusive therapies. While this paradigm is yet to be further and perhaps depend on the tissue type. Given the defined in a regulatable gene targeting system in HSCs, presence of mRNA in both stem and progenitor cell it has been supported indirectly by improved retroviral populations and the selective inhibition of p27 to transduction following knockdown of p27 with an progenitors, it is likely that the ubiquitin–proteasome antisense oligonucleotide in human CD34 þ CD38À cells machinery plays a key role in the degradation of p27 (Dao et al., 1998). Therefore, the functional role of p27 protein in these cell types. It would be interesting to look on HSC is indirect and appears to be quite distinct from at how differently the F-box Skp2 functions as the that of p21. substrate-specific subunits of an ubiquitin ligase enzyme The specific role of p27 in neural progenitors has been for p27 protein in stem cells in response to upstream also well defined. In adult or developing mouse brain, stem cell regulators (Carrano et al., 1999). p27 protein is associated with the NPCs (Legrier et al., 2001). Expression of p27 in mouse cerebellar granule cell p18: a strong inhibitor for stem cell self-renewal precursor inversely correlates with BrdU incorporation in the cells, and disruption of the p27 gene yields a larger p18, an INK4 family member, is expressed in multiple precursor pool (Miyazawa et al., 2000). p27 was also tissue types including hematopoietic cells (Tschan et al., documented as an intrinsic timer during the prolifera- 1999) and neurogenic cells (Legrier et al., 2001). tion of oligodentricyte cells in vitro (Gao et al., 1997; Further, p18 has been suggested to be involved in the Burton et al., 1999). In the absence of p27, those symmetric division of precursor cells in developing cultured NPCs cells would divide for one more round. mouse brain (Tschan et al., 1999). Recent studies In one elegant in vivo study involving the use of more demonstrated an intriguing role of p18 in HSC self- defined NSCs versus NPCs in adult mouse subventri- renewal, whose roles are quite distinct from other CKIs cular zone (SVZ), the authors found that loss of p27 had (Cheng et al., 2000a, b; Yuan, et al., 2004). The absence no effect on the number of NSC but selectively increased of p18 causes enhanced stem cell renewal, leading to an the number of NPC concomitantly with a reduction in increased stem cell pool. In the setting of transplanta- the number of neuroblasts (Doetsch et al., 2002). The tion, p18-deficient HSCs replicate with stronger compe- similar advantage of p27-null cells, as demonstrated in titiveness and permit sequential transplantation hematopoietic reconstitution model during cell engraft- strikingly beyond that of wild-type cells. Unlike ment, was reported in liver regeneration mode, suggest- previous demonstrations in p27-deficient mice, in which ing a similar role of p27 in liver progenitors (Karnezis the pool size and cell cycling of committed HPCs but not et al., 2001). HSCs are markedly increased (Cheng et al., 2000a), the

Oncogene Cell cycle inhibitors in stem cells T Cheng 7261 engraftment advantage in the absence of p18 is not due I II III to predominant outgrowth of less restricted cell cycling Specific Oncogenic in committed HPC pools and more differentiated cell modulation insults populations or the outgrowth of a single lineage. Stem cell Further analysis indicates an increase of the highly enriched HSC phenotype but not the lineage-committed G1 G1 G1 M S M S M S progenitor pools. Therefore, the study suggests a G2 G2 G2 relatively specific effect of p18 loss on HSCs and early multipotent HPCs. ion ion iat iat nt Interestingly, while both p18 and p21appear to affect ent re fer ffe Self-renewal Differentiation Dif Di kinetics in HSCs, they have very distinct phenotypes: al al new ew -re n p21 / HSCs undergo premature exhaustion (Cheng elf -re À À S elf et al., 2000b) but p18À/À HSCs tend to self-renew S during long-term engraftment. Without overwhelmingly Possible manipulations nonspecific proliferation in other cell populations, Homeostasis for tissue regeneration Tumorigenesis increased regeneration of p18À/À HSCs suggests that Figure 2 Possible Outcomes of Increased Stem Cell Self-Renewal the balance of differentiation to self-renewal in the Caused by Downregulation or Disruption of Specific Cell Cycle absence of p18 favors self-renewal. These two contrast- Inhibitors. Model I: under normal physiological conditions, stem ing models also suggest that relative stem cell quiescence cells are inhibited by abundant cell cycle inhibitors such as p21 (Cheng et al., 2000b; Cheng et al., 2001) and stem cell self-renewal or slow cycling is not necessarily the countering force for is balanced with differentiation. Model II: under a specific stem cell self-renewal and, in fact, it might be required manipulation such as targeting p18 in early G1 phase, stem cell for stem cell renewal at least in vivo under homeostatic self-renewal can be substantially increased while stem cells remain conditions. While this is a highly speculative point, it normal differentiation potential (Park et al., 2003). If a highly would be a reasonable argument, especially given the controllable or reversible system is developed, this model may have important implications for therapeutic use of stem cells in tissue facts that HSCs can rapidly divide in vitro but they lose regeneration. Model III: Additional disruption of cell cycle self-renewing potential after a few divisions (Ema et al., inhibitors such as p16 or p15 in stem cells coupled with impaired 2000). This notion of favored self-renewal in the absence differentiation or apoptosis may contribute to the formation of of p18 is indirectly supported by the data from others TSCs, thereby leading tumorigenesis (Park et al., 2003). The size of dark area in G1 phase indicates levels of the cell cycle inhibitors demonstrating that p18-expressing cells have an increase and the irregular shape in Model III indicates irreversible in asymmetric division (Phelps et al., 1998; Phelps and disruption of specific cell cycle inhibitors toward tumorigenesis Xiong, 1998; Cunningham and Roussel, 2001). The mechanism of favored HSC symmetric division may also be associated with the increase of the absolute rate of (Austin et al., 1997; Reya et al., 2003; Willert et al., HSC cycling. This possibility cannot be ruled out based 2003). The studies on possible links of upstream on the current study (Yuan et al., 2004). But the net regulators of stem cell function and p18, along with HSC expansion would argue against it as a sole other CKIs, will therefore begin to emerge and provide a mechanism for the increased HSC self-renewal. It is better understanding of the larger regulatory network in believed that critical decisions of cell fate are made the stem cells. during the G1 phase (Pardee, 1989). Given the distinct effects of these two CKI members acting on different The p16–Rb pathway in stem cell regulation CDK targets in stem cell regulation (Cheng et al., 2000a, b; Park et al., 2003), we may propose a model in One of the best-studied pathways in cell cycle regulation which modulation of a distinct CKI or its class at a is that of Rb, which directly interacts with Cyclin D and specific position of the cell cycle may be an important the INK4 proteins in early G-1 phase and serves as a mechanism for balancing self-renewal and differentia- critical and initial interface between mitogenic stimuli tion in stem cells (Figure 2). and cell fate commitment following cell division. Mice The regulation for p18 gene and protein in stem cells deficient in Rb are not viable and show defects in is unclear at this moment. Given the striking outcome of multiple tissue types including the hematopoietic lineage p18 absence on stem cell renewal, it would be of great (Clarke et al., 1992; Jacks et al., 1992; Lee et al., 1992). appeal to specifically look for the link of p18 with Deficient hematopoiesis in RbÀ/À mice indicated that several major signaling pathways controlling stem cell this protein might be critical to stem cell function. self-renewal. Studies have demonstrated that a subset of However, more definitive studies in the stem cell osteoblastic cells in the bone marrow serve as niches to compartment were largely precluded due to the early regulate the HSC numbers (Calvi et al., 2003; Zhang lethality of the Rb-null embryo. Instead, INK4 proteins et al., 2003) and those osteoblastic niches act on the stem closely associated with Rb have been studied in the cell by activating the Notch1 pathway (Varnum-Finney context of stem cell biology. These studies include a et al., 2000; Stier et al., 2002), leading to stem cell recent report indicating that Bmi-1, an upstream expansion. Moreover, another potent stem cell mitogen, inhibitor of p16 and p19ARF (note that it is not the the Wnt3a signal acting through the b-catenin pathway, p19INK4D) expression, is critical for HSC self-renewal can upregulate the Notch1 pathway and HoxB4, a (Park et al., 2003). In the absence of Bmi-1, self-renewal powerful transcription factor for stem cell self-renewal of HSC or NSC is diminished, an effect which is

Oncogene Cell cycle inhibitors in stem cells T Cheng 7262 dependent on the expression of p16 (Molofsky et al., HSCs or HPCs (Passegue et al., 2003). However, the 2003; Park et al., 2003). Notably, from the study, p16 conclusion regarding direct formation of TSC at the only partially mediated the effect of Bmi-1 protein in progenitor cell level does not necessarily exclude the NSCs, thereby suggesting that other CKIs may partici- possible contribution of stem cells to tumorigenesis at pate in the effect of Bmi-1 on stem cell self-renewal. But the onset. Tumorigenesis is a process in which multiple whether p16 null has the same effect as p18 null does on genetic or epigenetic changes are accumulated, ulti- HSC self-renewal, if vigorously examined by the HSC mately resulting in the outgrowth of an abnormal tissue transplant models, is yet to be defined. More impor- over normal tissues. In theory, tumor development tantly, given the fact that p16 or Rb protein are much might initiate from stem cells because stem cells are the more frequently inactivated in cancer cells compared cell type that can self-renew for a lifetime in the body with other CKIs, it would be very interesting to with the most chances of being exposed to many chronic investigate what the physiological roles of these proteins oncogenic insults. In contrast, the majority of progeni- are under homeostasis and how differently these tor cells are short-lived with a high turnover rate. If a proteins may contribute to the formation of TSC under ‘normal’ stem cell does not directly transform into a oncogenic insults (more discussions below). TSC, then it may pass on a certain degree of tumorigenesis susceptibility into the progenitor cells depending on the history of oncogenic insults that the CKIs and TSCs stem cell has encountered. Given these assumptions, we may largely utilize the current knowledge of molecular Understanding the molecular mechanisms of cell cycle genetics or epigenetics in tumor biology. For example, it control in stem cells would not only guide us to design has been extensively studied that disruption of the two rational strategies for stem cell therapies in regenerative major tumor-suppressor pathways, namely the p16-Rb medicine but also offer insights into the molecular and ARF-p53 pathways, largely participates in the understanding of stem cell-involved tumorigenesis. As development of most cancer types (Sherr, 2001). It the cell surface markers for TSCs have just begun to be would be interesting to know how these critical events revealed in a few tissue types, data on the roles of CKIs may be manifested by TSCs upon causative oncogenic or other cell cycle regulators are not available in the insults. However, given that the TSCs are a very small strictly defined TSC populations. Discussions about cell population in a heterogeneous tumor tissue, like CKIs in TSCs here are based on analyses of normal stem normal stem cells, rarity of TSCs would certainly pose a cells, tumor cell lines, primary tumor samples and technical challenge to our research especially in solid phenotypes of mice deficient in the CKIs. Therefore, tumors. many points are speculative with indirect evidence. Relevance of CKIs to general tumor occurrence The concept of TSC In human cancers, Cip/Kip members (p21, p27 and p57) The existence of TSC was proposed several decades ago are rarely found to be mutated or silenced, but may be (Pardal et al., 2003), but phenotypic characterization of important mediators or targets of tumor suppressors or TSC has only gained experimental support in recent oncoproteins. For example, p21 is transcriptionally years. Following the work on human acute myeloid regulated by p53 in cell cycle arrest. p27 has been leukemia (AML) (Bonnet and Dick, 1997), it has been shown to be a target of several oncogenes such as E1A recently reported that TSC may also exist in human (Mal et al., 1996) and c-Myc (Steiner et al., 1995). The solid tumors such as breast cancer (Al-Hajj et al., 2003) lower expression of p27 protein has been reported to be and brain tumors. It is expected that similar findings will associated with the prognostic of many human cancer be reported in many other cancer types in the coming types (Bloom and Pagano, 2003). In contrast, the INK4 years (details reviewed by others in this issue). A key family member, p16 and its related protein p14ARF have element of the TSC concept is to apply the basic rules of been defined as tumor-suppressor genes, and are stem cell biology in the context of cancer development frequently inactivated in human cancers. For example, and therapies. By extension, a tumor is viewed as a based on a survey on 4700 primary cases of leukemia or heterogeneous and regenerative tissue that is able to out- lymphoma and some 320 continuous leukemia–lympho- compete its normal counterpart; TSC is responsible for ma cell lines, p16 and p15 are frequently deleted or the initiation and maintenance of an abnormal tissue. silenced by the hypermethylation in their promoters; in This view is reasonable given that traditional che- contrast, very few cases had p18 or p19 deletion or motherapies or radiotherapies can shrink a tumor but hypermethylation (Drexler, 1998). often cannot completely eradiate it. This limitation Interestingly, these genetic or epigenetic data in might be partially attributed to the inability of these human cancers seem to be at least partially in agreement conventional therapies to effectively target TSCs em- with the occurrence of tumor development in the mice bedded in a given tumor mass (Reya et al., 2001). deficient in CKIs. All the knockout animals except p57- Studies in leukemia development indicate that TSC null mice (Zhang et al., 1997) exhibit normal organ may directly derive from either HSCs or HPCs, based on development, but sharply differ in tumor susceptibility. the experiments where a specific oncogenic fusion Therefore, CKIs can be divided into three categories protein was transduced into phenotypically defined based on the frequency of tumor formation in those

Oncogene Cell cycle inhibitors in stem cells T Cheng 7263 knockout mice. (1) Tumor suppressors (p16 and p15): 2003). However, whether the b-catenin other self- Mice lacking p16 have thymic hyperplasia with en- renewing pathways involve the same downstream cell hanced mitogenic responsiveness in T cells. Like p19ARF cycle inhibitors during the cell cycle is not clear. deficiency, p16 deficiency is associated with an increased In an attempt to further demonstrate the different incidence of spontaneous and carcinogen-induced can- involvements of CKIs in TSC formation, reconstitution cers (Sharpless et al., 2001). To a less degree, loss of p15 experiments with p18-null hematopoietic cells have been confers proliferative advantage to the cultured cells and followed up for more than 2 years (Yuan, et al., 2003). sensitizes these cells to H-ras induced transformation Enhanced self-renewal of p18null-HSC has been sus- and results in a low incidence of angiosarcomas in the tained without evidence of leukemogenesis at the stem mice (Latres et al., 2000). In addition, p15 deficiency cell level. At the later stage, a small percentage of the also enhances susceptibility to retrovirus-induced AML transplanted mice developed acute T-cell leukemia, but in mice. (2) Haplo-insufficient tumor suppressors (p27 the development of the leukemia involved the methyla- and p18): Disruption of both alleles of p27 or p18 gene tion of p15 or p16 promoter in addition to the absence in mice results in organomegaly with higher cellularity of p18. Interestingly, HSCs isolated from these leukemic and an increase in the incidence of spontaneous mice did not readily transfer the leukemic phenotype in pituitary tumor with advanced age or multiple tumor the secondary recipients. These results support a notion types in the presence of carcinogens (Fero et al., 1996; that increased stem cell self-renewal by the down- Nakayama et al., 1996; Franklin et al., 1998, 2000; Bai regulation of p18 is not sufficient to cause leukemia et al., 2003). In addition, p27 or p18 is able to sensitize and the disruption of p16 or p15 may set the stage of mice to gamma-irradiation- or chemical carcinogen- forming the leukemia stem cells to outcompete the induced tumorigenesis (Fero et al., 1998; Bai et al., normal hematopoiesis (Figure 2). These data, together 2003). (3) Nontumor suppressors (p21 and p19): Unlike with our studies about the physiological roles of p21 and other CKIs, p21-null or p19-null mice have a normal life p27 described above, pose an intriguing possibility that span and do not spontaneously develop tumors (Bru- normal and TSCs favor different downstream mediators garolas et al., 1995; Deng et al., 1995; Zindy et al., 2000). such as different CKIs. Targeting some less tumor-prone Among the above different groups, there are also CKI genes might be a worthwhile approach for stem cell specific collaborative effects between some of the CKIs expansion as indicated in previous studies (Antonchuk or their related proteins, such as p16 and p19ARF et al., 2002; Krosl et al., 2003; Stier et al., 2003). We (Sharpless et al., 2004) or p27 and p18 (Franklin et al., must, however, bear in mind that ‘benign gene’ in 1998) or p15 and p18 (Latres et al., 2000), indicating the tumorigenesis is a relative term and many genes can crosstalk of different pathways. Together, these results contribute to cancer development depending on the indicate the differential roles of CKIs in cancer actual context. The long-term effect of modifying these development. p16 and p15 are the most relevant CKIs genes or gene products should be carefully examined to cancers based on the current data in both humans and and must be in a regulatable manner in order to develop mice. Give the broad expression of CKIs and similar a clinically acceptable protocol. substrates of each CKI subfamily in cell cycle, it is Future studies on TSC will focus on the following puzzling why these molecules exhibit strikingly different questions. How do different CKIs respond to the tumor susceptibilities with limited compensatory effects. upstream signaling pathways such as Hox proteins, Wnt-b-catenin, Notch-1, and Bmi-1? Is there a common Potential roles of CKIs in TSC and therapeutic cell cycle switch in different stem cell populations? How implications different is the switch in normal and TSCs? For those relatively ‘benign’ CKIs such as p21 or p18, to what The differences of tumor susceptibility in the CKI extent can we target or express the molecule for normal knockout animals suggest that these proteins may act in stem cell expansion without the occurrence of tumor- different differentiation stages and pathways, thereby igenesis? intersecting differently with other key pathways involved in tumorigenesis. The distinct roles of different CKIs in stem/progenitor cells offer unique insights into Closing remarks to the roles of CKIs in tumorigenesis. If there is a TSC, then the TSC needs to have an increased cell survival or The critical roles of cell cycle inhibitors, particularly a higher self-renewing potential rate than that of the different CKIs, represent an interface of complex normal stem cells in a given tissue in order to develop a extracellular signals and their transduction in stem cell dominant tumor mass. Thus, understanding the cell regulation. Different members of the CKI subfamilies cycle basis underlying self-renewal of TSCs versus play distinct roles in stem or progenitor cell populations. normal stem cells is a key subject in the field. Recent The function of the different CKIs appears to be highly studies showed that TSCs and normal stem cells share differentiation-stage-specific and confers an important the mechanisms for self-renewal, such as the Bmi-1 and level of regulation in stem or progenitor cells to b-catenin pathways (Lessard and Sauvageau, 2003; maintain homeostasis. The mouse models deficient in Reya et al., 2003; Park et al., 2004). p16 and p19ARF p21, p27 or p18 have offered unique tools for studying can partially mediate the role of Bmi-1 in the cell cycle stem cell quiescence, repopulation capacity or self- of the stem cells (Molofsky et al., 2003; Park et al., renewal. Cooperative effects between members of the

Oncogene Cell cycle inhibitors in stem cells T Cheng 7264 two CKI subfamilies are likely, given the evidence of creased cell death or a disrupted differentiation program interplay between RB and p53 pathways. How the CKIs in order to outcompete the normal counterpart. In exert distinct effects and the pathways converging on addition, there must be an appropriate local micro- these regulators are the subjects of ongoing studies and environment (possibly altered stromal cell elements or will potentially provide further insights for manipula- ‘niches’ for TSCs) and newly developed vascularization tion of stem and progenitor cells. While the distinct roles to cultivate the tumor growth. Finally, TSCs must also for these CKIs have been largely deﬁned in hemato- acquire an ability to elude the body’s immune system. poietic cells, several lines of evidence suggest that these Nevertheless, deregulated cell cycle control must be one distinctions may be preserved across stem and progeni- of the fundamentally intrinsic steps leading to TSC- tor pools from different adult tissues, thereby offering a derived tumorigenesis (Figure 2). A greater under- platform for future therapeutic efforts based on the standing of the coordination of cell cycle regulation molecular commonality of cell cycle control. It should with cell differentiation and survival in stem cells under be noted that mouse ES stem cells have not shown the physiological cues or oncogenic insults is a central same cell cycle characteristics as adult stem cell question in TSC biology as well as tissue regeneration. populations (Aladjem et al., 1998; Burdon et al., 2002) The answer to this question will maximize our ability to and so the content discussed in this article would not expand stem cells for normal tissue regeneration, and apply to ES cells. In addition, the ultimate success of meanwhile prevent the normal stem cells from becoming therapeutic manipulations for stem cells in regenerative cancerous or selectively eradicate malignant stem cells. medicine will also require a greater understanding of the Does this sound too good to be true? Once we know the interaction between stem cell and microenvironment molecular boundary between normal and TSCs and (‘niche’) through the signaling circuitry and cell cycle have the ability to manipulate the key circuitry, it will progression involved in achieving self-renewal (Calvi become a reality! et al., 2003). Uncovering the mechanisms for cell cycle control Acknowledgements underlying stem cell self-renewal will eventually lead to My research work has been supported by NIH grants greater insights into stem cell biology, thereby providing DK02761, HL70561. I was a recipient of the Faculty Scholar new strategies for normal tissue regeneration, tumor- Award from the American Society of Hematology (2003). Due igenesis and possibly tumor therapies. However, the cell to the limited space and the targeted topic, many related cycle control is only one of the important aspects in TSC publications cannot be cited in this review, but I very much appreciate the important contributions from many other biology. The self-renewing divisions of stem cells must investigators in the ﬁeld. I thank Paulina Huang for the be balanced physiologically with cell death or differ- artwork, Jon Kilner for editorial assistance and Jian Yufor entiation. If a tumor originates from the transformation critical reading of this manuscript. I specially thank David of a normal cell, then the TSC must require a Scadden for his valuable input, and great inspiration to my competitive self-renewal potential coupled with de- research on cell cycle regulation of stem cells.

References

Abkowitz JL, Catlin SN and Guttorp P. (1996). Nat. Med., 2, Bonnet D and Dick JE. (1997). Nat. Med., 3, 730–737. 190–197. Bradford GB, Williams B, Rossi R and Bertoncello GB. Aladjem MI, Spike BT, Rodewald LW, Hope TJ, Klemm M, (1997). Exp. Hematol., 25, 445–453. Jaenisch R and Wahl GM. (1998). Curr. Biol., 8, 145–155. Braun SE, Mantel C, Rosenthal M, Cooper S, Liu L, Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ and Robertson KA, Hromas R and Broxmeyer HE. (1998). Clarke MF. (2003). Proc. Natl. Acad. Sci. USA, 100, 3983– Blood Cells Mol. Dis., 24, 138–148. 3988. Brugarolas J, Chandrasekaran C, Gordon JI, Beach D, Jacks Antonchuk J, Sauvageau G and Humphries RK. (2002). Cell, T and Hannon GJ. (1995). Nature, 377, 552–557. 109, 39–45. Burdon T, Smith A and Savatier P. (2002). Trends Cell Biol., Austin TW, Solar GP, Ziegler FC, Liem L and Matthews W. 12, 432–438. (1997). Blood, 89, 3624–3635. Burton PB, Raff MC, Kerr P, Yacoub MH and Barton PJ. Bai F, Pei XH, Godfrey VL and Xiong Y. (2003). Mol. Cell. (1999). Dev. Biol., 216, 659–670. Biol., 23, 1269–1277. Calvi LM, Adams GB, Weibrecht KW, Weber JM, Olson DP, Bartek J and Lukas J. (2001). FEBS Lett., 490, 117–122. Knight MC, Martin RP, Schipani E, Divieti P, Bringhurst Becker PS, Nilsson SK, Li Z, Berrios VM, Dooner MS, FR, Milner LA, Kronenberg HM and Scadden DT. (2003). Cooper CL, Hsieh CC and Quesenberry PJ. (1999). Exp. Nature, 425, 841–846. Hematol., 27, 533–541. Cardoso AA, Li ML, Batard P, Hatzfeld A, Brown EL, Berardi AC, Wang A, Levine JD, Lopez P and Scadden DT. Levesque JP, Sookdeo H, Panterne B, Sansilvestri P, Clark (1995). Science, 267, 104–108. SC and Hatzfeld J. (1993). Proc. Natl. Acad. Sci. USA, 90, Bloom J and Pagano M. (2003). Semin. Cancer Biol., 13, 41– 8707–8711. 47. Carrano AC, Eytan E, Hershko A and Pagano M. (1999). Nat. Bonfanti L, G A, Galli R and Vescovi AL. (2001). Stem Cells Cell Biol., 1, 193–199. and CNS Development Rao MS (ed). Humana Press: Cashman JD, Clark-Lewis I, Eaves AC and Eaves CJ. (1999). Totowa, NJ, pp. 31–48. Blood, 94, 3722–3729.

Oncogene Cell cycle inhibitors in stem cells T Cheng 7265 Cheng M, Olivier P, Diehl JA, Fero M, Roussel MF, Roberts Harrison DE and Lerner CP. (1991). Blood, 78, JM and Sherr CJ. (1999). EMBO J., 18, 1571–1583. 1237–1240. Cheng T, Rodrigues N, Dombkowski D, Stier S and Scadden Harrison DE, Astle CM and Lerner C. (1988). Proc. Natl. D. (2000a). Nat. Med., 6, 1235–1240. Acad. Sci. USA, 85, 822–826. Cheng T, Rodrigues N, Shen H, Yang Y, Dombkowski D, Hartwell LH and Weinert TA. (1989). Science, 246, Sykes M and Scadden DT. (2000b). Science, 287, 1804–1808. 629–634. Cheng T, Shen H, Rodrigues N, Stier S and Scadden DT. Hatzfeld A, Batard P, Panterne B, Taieb F and Hatzfeld J. (2001). Blood, 98, 3643–3649. (1996). Hum. Gene Ther., 7, 207–213. Cheshier SH, Morrison SJ, Liao X and Weissman IL. (1999). Hatzfeld J, Li ML, Brown EL, Sookdeo H, Levesque JP, T Proc. Natl. Acad. Sci. USA, 96, 3120–3125. OT, Gurney C, Clark SC and Hatzfeld A. (1991). J. Exp. Clarke AR, Maandag ER, van Roon M, van der Lugt NM, Med., 174, 925–929. van der Valk M, Hooper ML, Berns A and te Riele H. Hirabayashi Y, Matsuda M, Aizawa S, Kodama Y, Kanno J (1992). Nature, 359, 328–330. and Inoue T. (2002). Exp. Biol. Med. (Maywood), 227, Classon M and Harlow E. (2002). Nat. Rev. Cancer, 2, 474–479. 910–917. Horner PJ, Power AE, Kempermann G, Kuhn HG, Palmer Coats S, Flanagan WM, Nourse J and Roberts JM. (1996). TD, Winkler J, Thal LJ and Gage FH. (2000). J. Neurosci., Science, 272, 877–880. 20, 2218–2228. Conlon I and Raff M. (1999). Cell, 96, 235–244. Jacks T, Fazeli A, Schmitt EM, Bronson RT, Goodell MA and Cotsarelis G, Cheng SZ, Dong G, Sun TT and Lavker RM. Weinberg RA. (1992). Nature, 359, 295–300. (1989). Cell, 57, 201–209. Karnezis AN, Dorokhov M, Grompe M and ZhuL. (2001). Cunningham JJ and Roussel MF. (2001). Cell Growth Differ., J. Clin. Invest., 108, 383–390. 12, 387–396. Kay HE. (1965). Lancet, 15, 418–419. Dao MA, Taylor N and Nolta JA. (1998). Proc. Natl. Acad. Krosl J, Austin P, Beslu N, Kroon E, Humphries RK and Sci. USA, 95, 13006–13011. Sauvageau G. (2003). Nat. Med., 9, 428–432. de Nooij JC and Hariharan IK. (1995). Science, 270, 983–985. LaBaer J, Garrett MD, Stevenson LF, Slingerland JM, de Nooij JC, Letendre MA and Hariharan IK. (1996). Cell, 87, SandhuC, ChouHS, Fattaey A and Harlow E. (1997). 1237–1247. Genes Dev., 11, 847–862. Deng C, Zhang P, Harper JW, Elledge SJ and Leder P. (1995). Latres E, Malumbres M, Sotillo R, Martin J, Ortega S, Cell, 82, 675–684. Martin-Caballero J, Flores JM, Cordon-Cardo C and Doetsch F, Caille I, Lim DA, Garcia-Verdugo JM and Barbacid M. (2000). EMBO J., 19, 3496–3506. Alvarez-Buylla A. (1999). Cell, 97, 703–716. Lee EY, Chang CY, HuN, Wang YC, Lai CC, HerrupK, Lee Doetsch F, Verdugo JM, Caille I, Alvarez-Buylla A, Chao MV WH and Bradley A. (1992). Nature, 359, 288–294. and Casaccia-Bonneﬁl P. (2002). J. Neurosci., 22, 2255–2264. Legrier ME, Ducray A, Propper A, Chao M and Kastner A. Drexler HG. (1998). Leukemia, 12, 845–859. (2001). Cell Growth Differ., 12, 591–601. Ducos K, Panterne B, Fortunel N, Hatzfeld A, Monier MN Lemischka IR, Raulet DH and Mulligan RC. (1986). Cell, 45, and Hatzfeld J. (2000). J. Cell Physiol., 184, 80–85. 917–927. Dunnwald M, Chinnathambi S, Alexandrunas D and Bick- Lessard J and Sauvageau G. (2003). Nature, 423, 255–260. enbach JR. (2003). J. Cell Physiol., 195, 194–201. LiuY, Martindale JL, Gorospe M and Holbrook NJ. (1996). D’Urso G and Datta S. (2001). Stem Cell Biology Daniel R Cancer Res., 56, 31–35. Marshak, Gardner RL and Gottlieb D (eds). Cold Spring Mahmud N, Devine SM, Weller KP, Parmar S, Sturgeon C, Harbor Laboratory Press: New York, pp. 61–94. Nelson MC, Hewett T and Hoffman R. (2001). Blood, 97, el-Deiry WS. (1998). Curr. Top Microbiol. Immunol., 227, 3061–3068. 121–137. Mal A, Poon RY, Howe PH, Toyoshima H, Hunter T and el-Deiry WS, Kern SE, Pietenpol JA, Kinzler KW and Harter ML. (1996). Nature, 380, 262–265. Vogelstein B. (1992). Nat. Genet., 1, 45–49. Mantel C, Luo Z, Canﬁeld J, Braun S, Deng C and Broxmeyer Ellisen LW, Carlesso N, Cheng T, Scadden DT and Haber HE. (1996). Blood, 88, 3710–3719. DA. (2001). EMBO J., 20, 1897–1909. Marone M, Pierelli L, Mozzetti S, Masciullo V, Bonanno G, Ema H, Takano H, Sudo K and Nakauchi H. (2000). J. Exp. Morosetti R, Rutella S, Battaglia A, Rumi C, Mancuso S, Med., 192, 1281–1288. Leone G, Giordano A and Scambia G. (2000). Br. J. Fero ML, Randel E, Gurley KE, Roberts JM and Kemp CJ. Haematol., 110, 654–662. (1998). Nature, 396, 177–180. Missero C, Di Cunto F, Kiyokawa H, Koff A and Dotto GP. Fero ML, Rivkin M, Tasch M, Porter P, Carow CE, Firpo E, (1996). Genes Dev., 10, 3065–3075. Polyak K, Tsai LH, Broudy V, Perlmutter RM, Kaushansky Miyazawa K, Himi T, Garcia V, Yamagishi H, Sato S and K and Roberts JM. (1996). Cell, 85, 733–744. Ishizaki Y. (2000). J. Neurosci., 20, 5756–5763. Fortunel N, Batard P, Hatzfeld A, Monier MN, Panterne B, Molofsky AV, Pardal R, Iwashita T, Park IK, Clarke MF and Lebkowski J and Hatzfeld J. (1998). J. Cell Sci., 111, Morrison SJ. (2003). Nature, 425, 962–967. 1867–1875. Morgan DO. (1995). Nature, 374, 131–134. Fortunel NO, Hatzfeld A and Hatzfeld JA. (2000). Blood, 96, Morris RJ, LiuY, Marles L, Yang Z, TrempusC, Li S, Lin JS, 2022–2036. Sawicki JA and Cotsarelis G. (2004). Nat. Biotechnol., 22, Franklin DS, Godfrey VL, Lee H, Kovalev GI, Schoonhoven 411–417. R, Chen-Kiang S, SuL and Xiong Y. (1998). Genes Dev., 12, Morshead CM, Reynolds BA, Craig CG, McBurney MW, 2899–2911. Staines WA, Morassutti D, Weiss S and van der Kooy D. Franklin DS, Godfrey VL, O’Brien DA, Deng C and Xiong Y. (1994). Neuron, 13, 1071–1082. (2000). Mol. Cell. Biol., 20, 6147–6158. Nakanishi M, Adami GR, Robetorye RS, Noda A, Venable Gao FB, Durand B and Raff M. (1997). Curr. Biol., 7, SF, Dimitrov D, Pereira-Smith OM and Smith JR. (1995). 152–155. Proc. Natl. Acad. Sci. USA, 92, 4352–4356.

Oncogene Cell cycle inhibitors in stem cells T Cheng 7266 Nakayama K, Ishida N, Shirane M, Inomata A, Inoue T, Sherr CJ and Roberts JM. (1995). Genes Dev., 9, 1149–1163. Shishido N, Horii I, Loh DY and Nakayama K. (1996). Cell, Sherr CJ and Roberts JM. (1999). Genes Dev., 13, 85, 707–720. 1501–1512. Oostendorp RA, Audet J and Eaves CJ. (2000). Blood, 95, Sherr CJ and Weber JD. (2000). Curr. Opin. Genet. Dev., 10, 855–862. 94–99. Palmer TD, Schwartz PH, Taupin P, Kaspar B, Stein SA and Steiner P, Philipp A, Lukas J, Godden-Kent D, Pagano M, Gage FH. (2001). Nature, 411, 42–43. Mittnacht S, Bartek J and Eilers M. (1995). EMBO J., 14, Pardal R, Clarke MF and Morrison SJ. (2003). Nat. Rev. 4814–4826. Cancer, 3, 895–902. Steinman RA, Hoffman B, Iro A, Guillouf C, Liebermann DA Pardee AB. (1989). Science, 246, 603–608. and el-Houseini ME. (1994). Oncogene, 9, 3389–3396. Park IK, Morrison SJ and Clarke MF. (2004). J. Clin. Invest., Steinman RA, Huang J, Yaroslavskiy B, Goff JP, Ball ED and 113, 175–179. Nguyen A. (1998). Blood, 91, 4531–4542. Park IK, Qian D, Kiel M, Becker MW, Pihalja M, Steinman RM and Nussenzweig MC. (2002). Proc. Natl. Acad. Weissman IL, Morrison SJ and Clarke MF. (2003). Nature, Sci. USA, 99, 351–358. 423, 302–305. Stier S, Cheng T, Dombkowski D, Carlesso N and Scadden Passegue E, Jamieson CH, Ailles LE and Weissman IL. (2003). DT. (2002). Blood, 99, 2369–2378. Proc. Natl. Acad. Sci. USA, 100 (Suppl 1), 11842–11849. Stier S, Cheng T, Forkert R, Lutz C, Dombkowski DM, Pawliuk R, Eaves C and Humphries RK. (1996). Blood, 88, Zhang JL and Scadden DT. (2003). Blood, 102, 1260–1266. 2852–2858. Taniguchi T, Endo H, Chikatsu N, Uchimaru K, Asano S, Phelps DE, Hsiao KM, Li Y, HuN, Franklin DS, Westphal E, Fujita T, Nakahata T and Motokura T. (1999). Blood, 93, Lee EY and Xiong Y. (1998). Mol. Cell. Biol., 18, 2334– 4167–4178. 2343. Tong X and Srour EF. (1998). Exp. Hematol., 26, 684. Phelps DE and Xiong Y. (1998). Cell Growth Differ., 9, 595– Tschan MP, Peters UR, Cajot JF, Betticher DC, Fey MF and 610. Tobler A. (1999). Br. J. Haematol., 106, 644–651. Polyak K, Lee MH, Erdjument-Bromage H, Koff A, Roberts Uchida N, Dykstra B, Lyons KJ, Leung FY and Eaves CJ. JM, Tempst P and Massague J. (1994). Cell, 78, 59–66. (2003). Exp. Hematol., 31, 1338–1347. Potten MLAC (ed). (1997). Stem Cells and Cellular Pedigrees – Uchida N, Friera AM, He D, Reitsma MJ, Tsukamoto AS and A Conceptual Introduction. Academic Press: London. Weissman IL. (1997). Blood, 90, 4354–4362. Prost S, Bellamy CO, Clarke AR, Wyllie AH and Harrison DJ. Varnum-Finney B, Xu L, Brashem-Stein C, Nourigat C, (1998). FEBS Lett., 425, 499–504. Flowers D, Bakkour S, Pear WS and Bernstein ID. (2000). QiuJ, T Y, Harada J, RodriguesN, Moskowitz MA, Scadden Nat. Med., 6, 1278–1281. DT and Cheng T. (2004). J. Exp. Med., 199, 937–945. Verfaillie CM. (2002). Nat. Immunol., 3, 314–317. Quesenberry PJ, Colvin GA and Lambert JF. (2002). Blood, Vogelstein B and Kinzler KW. (1992). Cell, 70, 523–526. 100, 4266–4271. Wang J and Walsh K. (1996). Science, 273, 359–361. Reya T, Duncan AW, Ailles L, Domen J, Scherer DC, Willert Willert K, Brown JD, Danenberg E, Duncan AW, Weissman K, Hintz L, Nusse R and Weissman IL. (2003). Nature, 423, IL, Reya T, Yates III JR and Nusse R. (2003). Nature, 423, 409–414. 448–452. Reya T, Morrison SJ, Clarke MF and Weissman IL. (2001). Wlodarski P, Wasik M, Ratajczak MZ, Sevignani C, Hoser G, Nature, 414, 105–111. Kawiak J, Gewirtz AM, Calabretta B and Skorski T. (1998). Rivard N, G LA, Bartek J and Pouyssegur J. (1996). J. Biol. Blood, 91, 2998–3006. Chem., 271, 18337–18341. Yaroslavskiy B, Watkins S, Donnenberg AD, Patton TJ and Roy V and Verfaillie CM. (1999). Exp. Hematol., 27, 302–312. Steinman RA. (1999). Blood, 93, 2907–2917. Sharpless NE, Bardeesy N, Lee KH, Carrasco D, Castrillon Yuan Y, Boyer M, Shen H, Cao S, Yu H and Cheng T. (2003). DH, Aguirre AJ, Wu EA, Horner JW and DePinho RA. Blood, 102, 3177 (Abstract). (2001). Nature, 413, 86–91. Yuan Y, Shen H, Franklin DS, Scadden DT and Cheng T. Sharpless NE, Ramsey MR, Balasubramanian P, Castrillon (2004). Nat. Cell Biol., 6, 436–442. DH and DePinho RA. (2004). Oncogene, 23, 379–385. Zhang J, NiuC, Ye L, HuangH, He X, Tong WG, Ross J, Shen H, Cheng T, Preffer FI, Dombkowski D, Tomasson MH, Haug J, Johnson T, Feng JQ, Harris S, Wiedemann LM, Golan DE, Yang O, Hofmann W, Sodroski JG, Luster AD Mishina Y and Li L. (2003). Nature, 425, 836–841. and Scadden DT. (1999). J. Virol., 73, 728–737. Zhang P, Liegeois NJ, Wong C, Finegold M, HouH, Sherr CJ. (1994). Cell, 79, 551–555. Thompson JC, Silverman A, Harper JW, DePinho RA and Sherr CJ. (1996). Science, 274, 1672–1677. Elledge SJ. (1997). Nature, 387, 151–158. Sherr CJ. (2000). Cancer Res., 60, 3689–3695. Zindy F, van Deursen J, Grosveld G, Sherr CJ and Roussel Sherr CJ. (2001). Nat. Rev. Mol. Cell. Biol., 2, 731–737. MF. (2000). Mol. Cell. Biol., 20, 372–378.

Oncogene