Self-renewal versus transformation: Fbxw7 deletion leads to stem cell activation and leukemogenesis

Abstract

Recent reports have demonstrated that specific tumor suppressors are important for both maintaining hematopoietic stem cell (HSC) quiescence and preventing leukemia development, suggesting a connection between these two activities. Matsuoka and colleagues (pp. 986–991) have further illustrated this theme by demonstrating that inactivation of the tumor suppressor Fbxw7 leads to HSC depletion by active cell cycling and the initiation of leukemia.

Adult hematopoietic stem cells (HSCs) are believed to be relatively quiescent, contributing to blood homeostasis throughout an organism’s life span. Aberrant, overactive cycling of HSCs can lead to exhaustion of the stem cell pool and blood cell depletion or leukemia development. Significantly, numerous tumor suppressors and proto-oncogenes have been implicated in the regulation of HSC self-renewal, including PTEN, Bmi-1, Notch, c-Myc, JunB, and MEF/ELF4 (Calvi et al. 2003; Park et al. 2003; Passegue et al. 2004; Wilson et al. 2004; Lacorazza et al. 2006; Yilmaz et al. 2006; Zhang et al. 2006; Hosen et al. 2007). In the previous issue of Genes & Development, Matsuoka et al. (2008) demonstrate a role for the tumor suppressor Fbxw7 in maintaining HSC quiescence and preventing leukemia formation.

First identified in Caenorhabditis elegans as a negative regulator of Notch, Fbxw7 is localized within the human 4q32 locus, which is frequently deleted in diverse cancers (Akhoondi et al. 2007). On average, 6% of tumors exhibit mutation in Fbxw7; however, higher frequencies are observed in certain types of malignancy. For instance, Fbxw7 is mutated in ∼30% of patients with T-cell acute lymphoblastic leukemia (T-ALL) but is rarely mutated in other types of leukemia. As a ubiquitin ligase, Fbxw7 plays a key role in protein turnover by targeting specific proteins for proteosomal degradation. Aberrant protein degradation can contribute to tumor formation, as tumor suppressors may be targeted for rapid degradation while oncogenes may resist degradation in certain cancers (Welcker and Clurman 2008). Fbxw7 may be of particular interest in this regard because Fbxw7’s substrates include several prominent proto-oncogenes including Notch, c-Myc, JunB, and cyclin E. As such, Fbxw7 has been identified as an important tumor suppressor with loss-of-function mutations leading to chromosomal instability (Perez-Losada et al. 2005; Nakayama and Nakayama 2006; Welcker and Clurman 2008).

By negatively regulating cyclin E, Fbxw7 plays a key role in cell cycle control. Cyclin E is involved in driving quiescent (G0) or G1 cells into S phase and, predictably, is frequently dysregulated in cancer (Hwang and Clurman 2005). Notch is also negatively regulated by Fbxw7 and can serve as an oncogene in T-ALL, where >50% of these malignancies exhibit activating mutations in Notch (Weng et al. 2004; Grabher et al. 2006). Notch is important for directing lymphoid lineage cell fate determination and has also been implicated in HSC self-renewal. In particular, Notch is expressed by HSCs while its ligand, Jagged, is expressed by the HSC niche, and increased Jagged/Notch activation results in increased HSC number and niche expansion (Calvi et al. 2003). Thus, Notch activity may increase both self-renewal capacity and T-cell lineage commitment, which may significantly contribute to T-ALL development. Interestingly, Fbxw7 seems to play a particularly prominent role in regulating the Notch pathway by targeting not only Notch itself but also several upstream and downstream components of the Notch pathway (Welcker and Clurman 2008). This includes the prominent proto-oncogene c-Myc, which is a direct Notch target and is often linked to Notch-associated leukemia (Weng et al. 2006).

These themes are illustrated in the current study by Matsuoka et al. (2008) Here, conditional deletion of Fbxw7 was examined in the hematopoietic system. Fbxw7 is expressed in most hematopoietic lineages including lineage marker-negative, Sca-1⁺, Kit⁺ (LSK) cells, which are highly enriched in HSCs. The highest expression is observed in lymphoid lineages, especially CD4⁺ CD8⁺ (double positive or DP) T lymphocytes. Although immature (lineage marker-negative) cells are reduced in Fbxw7-deficient mice, Fbxw7 appears to be dispensable for multilineage terminal differentiation. Regarding HSCs, the repopulating capability of mutants is severely impaired. A relatively quiescent HSC population, LSK CD34⁻ cells, exhibited reduced frequency of cells in G0 in Fbxw7 mutants, suggesting that Fbxw7 may have a role in maintaining HSC quiescence. Although actively cycling, HSCs are not reported to accumulate in Fbxw7 mutants. Instead, HSCs eventually decline. Matsuoka et al. (2008) examined several well-known substrates of Fbxw7 including JunB, c-Myc, and Notch. No changes in JunB protein levels were observed in LSK CD34⁻ cells; however, both Notch and c-Myc protein levels were substantially increased in Fbxw7-deficient LSK CD34⁻ cells. Differential disease manifestation was observed in Fbxw7 mutants, with ∼30% exhibiting severe reduction in white blood cells (leukopenia) at 12 wk post-induction due to p53-induced apoptosis. In contrast, Fbxw7-deficient mice that did not exhibit leukopenia have reduced levels of p53 and normal frequency of apoptotic cells. Correspondingly, most mice without leukopenia develop T-ALL with DP T lymphocytes predominating. In contrast, none with leukopenia develop leukemia. Given the 3- to 4-mo latency and differential manifestation of disease in Fbxw7 mutants, additional mutations appear to be necessary for transformation, and p53 is a prime candidate for one such additional mutation. To confirm p53’s potential role in suppressing leukemic transformation, p53-deficient/Fbxw7-deficient double mutant mice were examined. These double mutants exhibited both reduced leukopenia and much shorter disease latency. Thus, p53 appears to protect from oncogenic activity resulting from Fbxw7 deficiency. Leukemic transformation occurs only by selection of cells with suppressed p53.

These data correspond with clinical observations that Fbxw7 deficiency is frequently found in conjunction with loss of p53. Notably, Fbxw7 has been described as a p53-dependent tumor suppressor, having multiple putative p53-binding sites within its regulatory regions. Thus, Fbxw7 may be a p53-inducible gene (Mao et al. 2004). Interestingly, most natural Fbxw7 mutations in cancers are heterozygous but also exhibit p53 mutation. Initially, p53 may suppress the consequences of Fbxw7 haploinsufficency; however, a second “hit” in p53 in cooperation with Fbxw7 deficiency may then lead to tumorigenesis. Indeed, p53 and Fbxw7 have been shown to cooperatively restrain cyclin E-induced chromosomal instability (Minella et al. 2007).

With rare exception, stem cells are unique in having self-renewal capability, a potentially dangerous property if hijacked by tumor-initiating or cancer stem cells (Reya et al. 2001). Other reports have also recently illustrated a connection between tumor suppressor deficiency resulting in HSC activation coupled with malignant transformation, notably in PTEN-deficient mice (Yilmaz et al. 2006; Zhang et al. 2006). As with PTEN mutants, Fbxw7 mutant mice exhibit increased cell cycle activation in HSCs leading to HSC exhaustion. This highlights the importance of maintaining relative quiescence of HSCs, which is believed to be sustained by the HSC osteoblastic niche (Calvi et al. 2003; Zhang et al. 2003; Arai et al. 2004). PTEN deletion results in HSC mobilization to the spleen with transient, but unsustainable, expansion. However, Fbxw7-deficient HSCs were not reported to be increased even transiently. This is likely the result of p53-mediated apoptosis. In the case of PTEN deficiency, p53 may fail to prevent an increase in HSCs because PTEN is reported to regulate p53 stability; thus, PTEN loss may lead to reduction in p53 (Stambolic et al. 2001; Freeman et al. 2003; Trotman and Pandolfi 2003). In the Fbxw7-deficient model, loss of p53 and its apoptotic activity appears to be necessary for malignant transformation, with increased Notch leading to uncontrolled T-cell proliferation. Interestingly, c-Myc may play an additional role in HSC activation in Fbxw7 mutants. c-Myc has been reported to regulate the interaction between HSCs and the osteoblastic niche. This is based on studies showing that conditional deletion of c-Myc results in the accumulation of HSCs with up-regulation of niche-specific adhesion molecules such as N-cadherin. Likewise, forced expression of c-Myc results in decreased N-cadherin and release of HSCs from their niche (Wilson et al. 2004). Thus, increased c-Myc in Fbxw7-deficient mice may lead to the release of HSCs from their quiescence-promoting osteoblastic niche, which may promote aberrant activation. However, further studies are necessary to test this.

Tumor suppressors appear to play a key role in regulating HSCs by strictly limiting their expansion and maintaining them as a rare population (Fig. 1). This normally prevents cancer stem cell development. Repression of potential oncogenic activity of HSCs by tumor suppressors such as Fbxw7 and PTEN may function by maintaining relative HSC quiescence and thus preventing HSC depletion. The role of the niche in this process merits further study. Loss of tumor suppressor activity results in HSC activation, which can be controlled by p53-mediated apoptosis or senescence. However, this aberrant activation leads to selection for p53-deficient cells, resulting in leukemogenesis. Similarly, proto-oncogenes such as c-Myc, Notch, and JunB have a role in governing HSC self-renewal and must be strictly limited to prevent leukemic transformation. Aberrant activation of proto-oncogenes, potentially via the loss of negative regulators such as Fbxw7, also leads to HSC activation, which sets the stage for leukemia development.

Figure 1.

Tumor suppessors and proto-oncogenes coordinately regulate HSC activation versus quiescence. HSCs are maintained in a relatively quiescent state within their niche, maintaining a delicate balance of specific tumor suppressor and proto-oncogene activity. Deficiency in specific tumor suppressors or increased activity of specific proto-oncogenes leads to HSC activation and increased cell cycle entry. HSC expansion may be inhibited by p53-mediated apoptosis or senescence. Loss of p53 activity— either by secondary mutation or the loss of p53 stabilizers such as PTEN—results in uncontrolled HSC expansion. Having self-renewal capacity coupled with mutations that result in chromosomal instability, this population initiates leukemia development.