The tumour suppressor p53 transcriptionally regulates a range of target genes that control cell growth and survival. Mutations of p53 have been implicated in the development of ∼50% of human cancers, including those instigated by exposure to mutagens. Although numerically rare, cancers can arise as a consequence of inherited mutations, such as in the Li–Fraumeni syndrome, which is caused by mutation of one p53 allele. Gene-targeted mice deficient for p53 have been generated to study this familial cancer syndrome. On a C57BL/6 background, p53 -deficient mice develop primarily thymic lymphoma and more rarely sarcoma. Evasion of apoptosis is considered to be essential for neoplastic transformation. As proteins of the Bcl-2 family are the critical regulators of apoptosis, we investigated the role of the pro-survival members Bcl-2, Bcl-x L and Bcl-w in cancer development in p53 +/− and p53 −/− mice by testing whether ABT-737, a pharmacological inhibitor of these proteins, could prevent or delay tumourigenesis. Our studies showed that ABT-737 prophylaxis only caused a minor delay and reduction in γ -radiation-induced thymic lymphoma development in p53 −/− mice, but this was accompanied by a concomitant increase in sarcoma. These data show that, collectively, Bcl-2, Bcl-x L and Bcl-w have only minor roles in thymic lymphoma development elicited by defects in p53, and this may indicate that Mcl-1 and/or A1 may feature more prominently in this process.

Evasion of apoptosis is critical for tumorigenesis, and sustained survival of nascent neoplastic cells may depend upon the endogenous levels of pro-survival BCL-2 family members. Indeed, previous studies using gene-targeted mice revealed that BCL-XL, but surprisingly not BCL-2, is critical for the development of c-MYC-induced pre-B/B lymphomas. However, it remains unclear whether another pro-survival BCL-2 relative contributes to their development. MCL-1 is an intriguing candidate, because it is required for cell survival during early B-lymphocyte differentiation. It is expressed abnormally high in several types of human B-cell lymphomas and is implicated in their resistance to chemotherapy. To test the B-cell intrinsic requirement for endogenous MCL-1 in lymphoma development, we conditionally deleted Mcl-1 in B-lymphoid cells of Eμ-Myc transgenic mice. We found that MCL-1 loss in early B-lymphoid progenitors delayed MYC-driven lymphomagenesis. Moreover, the lymphomas that arose when MCL-1 levels were diminished appeared to have been selected for reduced levels of BIM and/or increased levels of BCL-XL. These results underscore the importance of MCL-1 in lymphoma development and show that alterations in the levels of other cell death regulators can compensate for deficiencies in MCL-1 expression.

Impaired apoptosis is considered one of the prerequisites for the development of most, if not all, cancers, but the mechanisms that guarantee the sustained survival of most cancer cells remain unknown. Members of the Bcl-2 family are key regulators of apoptosis and include proteins essential for cell survival and those required to initiate cell death. Studies with transgenic mice have shown that over-expression of Bcl-2 or related pro-survival family proteins, such as Bcl-xL or Mcl-1, can promote tumorigenesis, particularly in conjunction with mutations that deregulate cell cycle control, such as deregulated c-myc expression. However, chromosomal translocations or other mutations that promote Bcl-2 over-expression only account for a small proportion of cancers, such as follicular center B lymphoma. As most tumors do not harbor such cytogenetic abnormalities, it is likely that expression of endogenous Bcl-2-like anti-apoptotic proteins, driven by pathways activated by oncogenic lesions, is required to maintain the survival of cells undergoing neoplastic transformation. Using Eμ-myc transgenic mice, a well-characterized model of human Burkitt's lymphoma, and Bcl-2-deficient mice we investigated the role of endogenous Bcl-2 in lymphoma development. Bcl-2 was found to be largely dispensable for the development and survival of pro-B, pre-B and immature B cells expressing the transgene but reduced the numbers of mature Eμ-myc B lymphocytes by ∼10-fold. Despite the resulting marked deficit in pre-malignant B cells, the absence of bcl-2 neither delayed the onset nor reduced the incidence of Eμ-myc B lymphoma. In contrast, loss of Bcl-xL, which also reduced the numbers of pre-leukaemic Eμ-myc B lymphoid cells by ∼10-fold, completely abrogated Eμ-myc induced lymphoma development. These results demonstrate that Bcl-xL but not Bcl-2 is critical to sustain the survival of cells undergoing neoplastic transformation in the Eμ-myc mouse lymphoma model. We are currently investigating whether Bcl-xL is also critical for the development of other types of cancers and whether it is essential for sustaining the growth of already established tumours.

As many oncogenic changes, such as Myc overexpression, promote apoptosis, the survival of emerging neoplastic clones may often initially depend upon endogenous levels of particular pro-survival members of the Bcl-2 protein family. Pertinently, we recently showed that in lymphoma-prone Eμ-myc transgenic mice, which overexpress Myc in all B-lymphoid cells, endogenous Bcl-x(L) is critical for the survival, as well as the expansion of preneoplastic B-lymphoid cells and the development of malignant disease. This discovery raised the possibility that pharmacological blockade of Bcl-x(L) might impede Myc-driven lymphoma development. Indeed, we report here that treatment of preleukaemic Eμ-myc transgenic mice with the Bcl-2 homology (BH)3 mimetic drug ABT-737, which inhibits Bcl-x(L), as well as Bcl-2 and Bcl-w, augmented apoptosis of preneoplastic B-lymphoid cells, reduced their numbers and greatly prolonged lymphoma-free survival. These findings reveal that BH3 mimetic drugs may provide a prophylactic strategy to prevent the development of certain tumours, particularly those driven by deregulated Myc expression. Moreover, such treatment may help in the management of patients with hereditary cancer syndromes and perhaps also in the prevention of tumour relapses.

Tumour development requires a combination of defects that allow nascent neoplastic cells to become self-sufficient for cell proliferation and insensitive to signals that normally restrain cell growth. Among the latter, evasion of programmed cell death (apoptosis) has proven to be critical for the development and sustained growth of many, perhaps all, cancers. Apoptotic cell death is regulated by complex interactions between pro-survival members and two subgroups of pro-apoptotic members of the B-cell lymphoma-2 (Bcl-2) protein family. In this invited review article, we reminisce on the discovery of Bcl-2, the first regulator of cell death identified, we discuss the mechanisms that control apoptotic cell death, focussing on how defects in this process promote the development and sustained growth of tumours and also affect their responses to anticancer therapeutics and, finally, we describe how current knowledge of the regulatory networks of apoptosis is exploited to develop novel approaches for cancer therapy.

BH3-only proteins, such as Bim and Bad, contribute to tissue homeostasis by initiating apoptosis in a cell type- and stimulus-specific manner. Loss of Bim provokes lymphocyte accumulation in vivo and renders lymphocytes more resistant to diverse apoptotic stimuli and Bad has been implicated in the apoptosis of haematopoietic cells upon cytokine deprivation. To investigate whether their biological roles in apoptosis overlap, we generated mice lacking both Bim and Bad and compared their haematopoietic phenotype with that of the single-knockout and wild-type (wt) animals. Unexpectedly, bad−/− mice had excess platelets due to prolonged platelet life-span. The bim−/−bad−/− mice were anatomically normal and fertile. Their haematopoietic phenotype resembled that of bim−/− mice but lymphocytes were slightly more elevated in their lymph nodes. Although resting B and T lymphocytes from bim−/−bad−/− and bim−/− animals displayed similar resistance to diverse apoptotic stimuli, mitogen activated bim−/−bad−/− B cells were more refractory to cytokine deprivation. Moreover, combined loss of Bim and Bad-enhanced survival of thymocytes after DNA damage and accelerated development of γ-irradiation-induced thymic lymphoma. Unexpectedly, their cooperation in the thymus depended upon thymocyte–stromal interaction. Collectively, these results show that Bim and Bad can cooperate in the apoptosis of thymocytes and activated B lymphocytes and in the suppression of thymic lymphoma development.

Apoptosis is a genetically programmed process for killing unwanted or dangerous cells. It plays a critical role in the immune system, reflected by the observation that defects in this process can cause autoimmune disease or lymphoid malignancies. Vertebrates have two distinct but ultimately converging apoptotic pathways. One is initiated by death receptors, members of the TNF-R family with an intra-cellular 'death domain' and requires the cysteine protease, caspase-8, and its adaptor FADD. The other pathway is triggered by developmentally programmed cues, growth factor deprivation and cytotoxic stimuli and is regulated by the interplay of pro- and anti-apoptotic members of the Bcl-2 protein family. We will present data from our analysis of multiple gene-targeted and transgenic mice on the identification of the functions of Bcl-2 family members and death receptors in the control of apoptosis of cells of the immune system, with emphasis on their roles in autoimmune disease and lymphoma/leukaemia.

ment and maintenance. The first concerns the nature of the cells that perpetuate a tumor. It is nowwell established that normal tissues that turn over, such as the blood or gut, are sustained by rare tissue stem cells (see Weissman; Visvader et al.; both this volume). That principle stimulated the notion that the relentless growth of a tumor might be sustained not by most of its cells, but instead exclusively by a rare subpopulation, commonly termed the “cancer stem cells” (Wang and Dick 2005; Clarke et al. 2006). As reviewed recently (Clarke et al. 2006; Campbell and Polyak 2007; Adams and Strasser 2008; Vermeulen et al. 2008), this issue is attracting enormous interest, both because of its fundamental importance for tumor biology and its implications for therapy. We have investigated this issue for mouse hematopoietic tumors (Kelly et al. 2007a). The second issue we address concerns the role of endogenous prosurvival members of the Bcl-2 family in tumorigenesis. Abatement of apoptosis is a key step in tumor development (Hanahan and Weinberg 2000; Cory and Adams 2002), and these proteins are the principal guardians against apoptosis (Adams and Cory 2007; Youle and Strasser 2008). Because overexpression of Bcl-2 can promote lymphoma development (Strasser et al. 1990), we have explored whether endogenous Bcl-2 is required for lymphomagenesis (Kelly et al. 2007b).

One critical issue for cancer biology is the nature of the cells that drive the inexorable growth of malignant tumors. Reports that only rare cell populations within human leukemias seeded leukemia in mice stimulated the now widely embraced hypothesis that only such "cancer stem cells" maintain all tumor growth. However, the mouse microenvironment might instead fail to support the dominant human tumor cell populations. Indeed, on syngeneic transplantation of mouse lymphomas and leukemias, we and other investigators have found that a substantial proportion (>10%) of their cells drive tumor growth. Thus, dominant clones rather than rare cancer stem cells appear to sustain many tumors. Another issue is the role of cell survival in tumorigenesis. Because tumor development can be promoted by the overexpression of prosurvival genes such as bcl-2, we are exploring the role of endogenous Bcl-2-like proteins in lymphomagenesis. The absence of endogenous Bcl-2 in mice expressing an Emu-myc transgene reduced mature B-cell numbers and enhanced their apoptosis, but unexpectedly, lymphoma development was undiminished or even delayed. This suggests that these tumors originate in an earlier cell type, such as the pro-B or pre-B cell, and that the nascent neoplastic clones do not require Bcl-2 but may instead be protected by a Bcl-2 relative.

Numerous p53 target genes have been implicated in DNA damage-induced apoptosis signaling, but proapoptotic Bcl-2 (B-cell leukemia 2) family members of the BH3 (Bcl-2 homolog region [BH] 3)-only subgroup appear to play the critical initiating role. In various types of cultured cells, 3 BH3-only proteins, namely Puma (p53 up-regulated modulator of apoptosis), Noxa, and Bim (Bcl-2 interacting mediator of cell death), have been shown to initiate p53-dependent as well as p53-independent apoptosis in response to DNA damage and treatment with anticancer drugs or glucocorticoids. In particular, the absence of Puma or Bim renders thymocytes and mature lymphocytes refractory to varying degrees to death induced in vitro by growth factor withdrawal, DNA damage, or glucocorticoids. To assess the in vivo relevance of these findings, we subjected mice lacking Puma, Noxa, or Bim to whole-body gamma-radiation or the glucocorticoid dexamethasone and compared lymphocyte survival with that in wild-type and BCL2-transgenic mice. Absence of Puma or Bcl-2 overexpression efficiently protected diverse types of lymphocytes from the effects of gamma-radiation in vivo, and loss of Bim provided lower but significant protection in most lymphocytes, whereas Noxa deficiency had no impact. Furthermore, both Puma and Bim were found to contribute significantly to glucocorticoid-induced killing. Our results thus establish that Puma and Bim are key initiators of gamma-radiation- and glucocorticoid-induced apoptosis in lymphoid cells in vivo.

Imatinib mesylate (imatinib) exerts the anti-Philadelphia-positive (Ph1+) leukemia activity both by the inhibition of cell proliferation and by the induction of apoptosis. Recent studies demonstrate that the induction of cell death is essential for eradication of Ph1+ leukemic clones in imatinib treatment; however, the molecular mechanisms have not yet been clearly described. By examining the effect of imatinib on parental K562 and subclones overexpressing either Bcl-2, Bcl-XL or a dominant interfering mutant of FADD/MORT1, which blocks death receptor apoptosis signalling, we found that imatinib triggers apoptosis exclusively via the Bcl-2 family-regulated intrinsic apoptotic pathway. We investigated the involvement of BH3-only proteins as apoptotic initiators in imatinib-induced cell death, because the cell life-or-death decision is arbitrated by the balance between pro-apoptotic BH3 only-proteins and anti-apoptotic Bcl-2 proteins. We found that imatinib treatment upregulated Bim in Ph1+ leukemic cell lines and bcr-c-abl transformed murine fetal liver cells (FLCs)-derived cell lines both by transcriptional and post-translational mechanisms. Imatinib also activated Bad through dephosphorylation and upregulated Bmf transcriptionally. To examine the role of Bim in imatinib-induced apoptosis, we examined the cell killing activity of imatinib in subclones of K562 and BV173 Ph1+ cells expressing abnormally reduced levels of Bim using stable RNA interference system. This revealed that the cell killing activity of imatinib largely dependent on Bim expression levels in these cell lines, although significant apoptosis was still evident. To further define the role of Bim, Bad and Bmf in imatinib-induced cell death, we examined the effect of imatinib on retrovirally bcr-c-abl transformed cell lines derived from FLCs from wild type C57BL/6, Bim-/-, Bad-/-, Bim-/-Bad-/- double KO and Bcl-2 transgenic fetuses. The bim-/-bcr-c-abl+ FLCs were shown to be more resistant to imatinib-induced cell death than wt.bcr-c-abl+ FLCs, however, bim-/-bcr-c-abl+ FLCs were eventually induced into cell death, indicating that Bim is not the only initiator of apoptosis. The bad-/-bcr-c-abl+ FLCs were also partially resistant to imatinib-induced cell death. Intriguingly, like in vav.bcl-2.bcr-c-abl+ FLCs, the cell death induction by imatinib (~5.0μM) was largely abrogated in bim-/-bad-/-bcr-c-abl+ FLCs, indicating that Bim collaborates with Bad for the apoptotic induction by imatinib. Importantly, we found that Bim was inducible by ex vivo imatinib treatment in primary Ph1+ leukemic cells only from clinically good responders but not from patients refractory to imatinib treatment. Collectively, these results demonstrate that Bim is the critical but not the only initiator required for imatinib-induced apoptosis of Bcr/Abl-positive hematopoietic cells; Bad and Bmf may be the ancillary BH3-only proteins in this process. Our results provide evidence for the therapeutic significance of regulation of BH3-only proteins, particularly Bim, for the eradication of Ph1+ leukemic cells.

The Bcl-2 protein family, which largely determines commitment to apoptosis, has central roles in tumorigenesis and chemoresistance. Its three factions of interacting proteins include the BH3-only proteins (e.g., Bim, Puma, Bad, Noxa), which transduce diverse cytotoxic signals to the mammalian pro-survival proteins (Bcl-2, Bcl-x(L), Bcl-w, Mcl-1, A-1), whereas Bax and Bak, when freed from pro-survival constraint, provoke the mitochondrial permeabilization that triggers apoptosis. We have discovered unexpected specificity in their interactions. Only Bim and Puma, which mediate multiple cytotoxic signals, engage all the pro-survival proteins. Noxa and Bad instead bind subsets and cooperate in killing, indicating that apoptosis requires neutralization of different pro-survival subsets. Furthermore, Mcl-1 and Bcl-x(L), but not Bcl-2, directly sequester Bak in healthy cells, and Bak is freed only when BH3-only proteins neutralize both its guards. BH3-only proteins such as Bim are tumor suppressors and mediate many of the cytotoxic signals from anticancer agents. Hence, compounds mimicking them may prove valuable for therapy. Indeed, the recently described ABT-737 is a promising "BH3 mimetic" of Bad. We find that, like Bad, ABT-737 kills cells efficiently only if Mcl-1 is absent or down-regulated. Thus, manipulation of apoptosis by targeting the Bcl-2 family has exciting potential for cancer treatment.