Search

Recent evidence has demonstrated the existence of a small subset of the tumour mass that is wholly responsible for the sustained growth and propagation of the tumour. This cancer stem cell (CSC) compartment is also likely to be responsible both for disease relapse and the resistance to therapy that often accompanies relapse. The evidence for CSCs in various malignancies is presented. The failure of existing therapeutics to eradicate CSCs suggests that they are relatively resistant to present cancer treatments. This resistance may reflect the preservation of normal stem cell protective mechanisms, such as an increased expression of drug efflux pumps or alterations in apoptotic, cell cycle and DNA repair mechanisms. Targeting these mechanisms, and taking advantage of potential differences in the biology of normal stem cells and CSCs, such as differences in surface phenotype, self renewal/quiescence and stem cell–niche interactions are discussed and preliminary preclinical or clinical data are presented. Finally, the authors give their opinion of the direction in which one must travel to successfully target the CSC and improve treatment outcomes in malignant disease.

Self-renewal is a prerequisite for cancer initiation and maintenance, however the origins and molecular basis of self-renewal in malignant cells remain elusive. Recently, we and others have shown that certain leukemia-associated fusion oncogenes (such as MOZ-TIF2 and MLL fusions) can alter self-renewal in, and generate acute myeloid leukemia (AML) from, committed murine myeloid progenitor cells. AML is a highly heterogeneous disease, both genetically and biologically, and is associated with many differing combinations of mutations and a variable clinical course. However, it is not currently known whether an alteration of self-renewal is a more widespread finding in AML. Moreover, it is also not known whether alterations of self-renewal are mediated by common genetic programs downstream of multiple individual mutations or through the engagement of unique programs downstream of individual mutations. This distinction is important, as the demonstration of common pathways may identify common critical molecular targets for the treatment of AML. We now demonstrate, using retroviral expression of the AML1-ETO (RUNX1-RUNX1T1) and NUP98-HOXA9 fusion proteins in committed myeloid progenitors, that the ability to alter self-renewal is a more generalized property of leukemia-associated transcription factor fusion oncogenes. Furthermore, to investigate the molecular basis for the reestablishment of self-renewal we compare expression differences between granulocyte monocyte progenitors (GMP) directly transduced with MOZ-TIF2, AML1-ETO and NUP98-HOXA9 retroviruses and GMP transduced with empty vector retrovirus, to identify common candidate genes and molecular pathways involved in self-renewal downstream of these fusion oncogenes. Thereafter, we validate the expression of these early genetic programs in functionally defined murine leukemia stem cells and bulk human leukemias and perform functional experiments to validate the effects of individual candidate genes within the programs on self-renewal and transformation. Our data demonstrate AML1-ETO and NUP98-HOXA9 alter self-renewal in and NUP98-HOXA9 generates AML from committed myeloid progenitors. Disparate leukemia-associated oncogenes (MOZ-TIF2, AML1-ETO and NUP98-HOXA9) initiate an early, common self-renewal signature of 167 genes (FDR level p<0.05). Significant elements of this self-renewal signature can be detected in established, functionally defined leukemia stem cells from a murine model of AML. Significant elements of this early signature can also be detected across a large cohort of 253 patients with differing AML subtypes, where the signature predicts for disease biology and patient survival. Single genes from the signature, including Sox4 and Bmi1 can themselves phenocopy the leukemia-associated oncogenes and alter self-renewal in committed murine progenitors. Taken together, our findings have important implications for the biology of AML and, furthermore, may suggest common targets for therapeutic intervention.

Caudal-related (Cdx) genes are transcription factors that function as master homeobox regulators of HOX genes. Three Cdx genes (Cdx1, Cdx2 and Cdx4) have been identified in mammals, and function in part as mediators anterior-posterior patterning during development. However, accumulating data suggests a central role for Cdx genes in vertebrate hematopoiesis. This link was first established in the zebrafish model whereby Cdx4-deficient fish exhibited severe anaemia within the first day of development and fail to specify blood progenitors (Davidson et al, Nature 2003). Given the high degree of conservation of Cdx genes between species, we anticipated that the loss of Cdx4 in the mouse would result in a phenotype similar to the lethal bloodless phenotype observed in zebrafish. This initial hypothesis was supported by reports that overexpression of Cdx4 enhances hematopoietic potential from murine embryonic stem cells or embryoid bodies (McKinney-Freeman et al, Blood 2008). To establish the function of the Cdx4 gene in mammalian hematopoiesis, we analyzed the hematopoietic system during development and in the adult hematopoietic compartment, using mice in which Cdx4 was either deleted in the germline, or conditionally deleted in the adult hematopoietic system using Mx-Cre mediated excision. Germline deletion resulted in fertile and viable animals born with expected Mendelian rations. Surprisingly, there were no apparent gross phenotypic manifestations of Cdx4 deletion in either context despite complete excision of the Cdx4 gene in the hematopoietic compartment. Peripheral blood counts were normal, as was flow cytometric analysis for the myeloid and lymphoid compartments as assessed using Mac1/Gr1 and various B- & T-cell markers. Using multi-parameter flow analysis and methylcellulose plating assays, we observed that loss of Cdx4 in both mouse models only caused subtle hematopoietic perturbations, such as an increase in thymus weight and a modest decrease in pre-B colony formation ability in vitro. However, there were no other apparent defects in the hematopoietic stem and progenitor compartments including GMP, CMP & MEP. There were also no marked differences in the expression of Hox genes in the hematopoietic compartment, even though striking dysregulation of various Hox genes was observed in Cdx4-deficient zebrafish model. We also assessed the effect of Cdx4 deletion on the development of leukemia mediated by MLL fusion genes. Using in vivo retroviral transduction/bone marrow transplantations and in vitro serial-replating assays, we observed that Cdx4 null mice remain susceptible to leukemic transformation by MLL-AF9, indicating that Cdx4 is not necessary for MLL-AF9 mediated leukemias. The only subtle differences observed were that Cdx4−/− progenitors transduced with MLLAF9 generated more differentiated transformed blast colony-forming units in vitro, and induced leukemias with longer disease latencies compared to their wild-type littermate controls. Thus, in contrast with the zebrafish system, we found that Cdx4 is dispensable for normal mammalian hematopoiesis, as well as for MLL-AF9 mediated leukemia. The striking differences we observed between the studies of the same gene in the two different vertebrate systems emphasizes the importance of not solely relying on a single model system to assess the role of transcription factors in development, and that the deficiency of even a highly conserved gene might perturb function to variable extents in different animal models.

Self-renewal is a prerequisite for cancer initiation and maintenance, however the origins and molecular basis of self-renewal in malignant cells remain elusive. Recently, we and others have shown that certain leukemia-associated fusion oncogenes (such as MOZ-TIF2 and MLL fusions) can alter self-renewal in, and generate acute myeloid leukemia (AML) from, committed murine myeloid progenitor cells. AML is a highly heterogeneous disease, both genetically and biologically, and is associated with many differing combinations of mutations and a variable clinical course. However, it is not currently known whether an alteration of self-renewal is a more widespread finding in AML. Moreover, it is also not known whether alterations of self-renewal are mediated by common genetic programs downstream of multiple individual mutations or through the engagement of unique programs downstream of individual mutations. This distinction is important, as the demonstration of common pathways may identify common critical molecular targets for the treatment of AML. We now demonstrate, using retroviral expression of the AML1-ETO (RUNX1-RUNX1T1) and NUP98-HOXA9 fusion proteins in committed myeloid progenitors, that the ability to alter self-renewal is a more generalized property of leukemia-associated transcription factor fusion oncogenes. Furthermore, to investigate the molecular basis for the reestablishment of self-renewal we compare expression differences between granulocyte monocyte progenitors (GMP) directly transduced with MOZ-TIF2, AML1-ETO and NUP98-HOXA9 retroviruses and GMP transduced with empty vector retrovirus, to identify common candidate genes and molecular pathways involved in self-renewal downstream of these fusion oncogenes. Thereafter, we validate the expression of these early genetic programs in functionally defined murine leukemia stem cells and bulk human leukemias and perform functional experiments to validate the effects of individual candidate genes within the programs on self-renewal and transformation. Our data demonstrate

Caudal-related (Cdx) genes are transcription factors that function as master homeobox regulators of HOXgenes. Three Cdx genes (Cdx1, Cdx2and Cdx4) have been identified in mammals, and function in part as mediators anterior-posterior patterning during development. However, accumulating data suggests a central role for Cdxgenes in vertebrate hematopoiesis. This link was first established in the zebrafish model whereby Cdx4-deficient fish exhibited severe anaemia within the first day of development and fail to specify blood progenitors (Davidson et al, Nature 2003). Given the high degree of conservation of Cdxgenes between species, we anticipated that the loss of Cdx4in the mouse would result in a phenotype similar to the lethal bloodless phenotype observed in zebrafish. This initial hypothesis was supported by reports that overexpression of Cdx4 enhances hematopoietic potential from murine embryonic stem cells or embryoid bodies (McKinney-Freeman et al, Blood 2008). To establish the function of the Cdx4gene in mammalian hematopoiesis, we analyzed the hematopoietic system during development and in the adult hematopoietic compartment, using mice in which Cdx4was either deleted in the germline, or conditionally deleted in the adult hematopoietic system using Mx-Cre mediated excision. Germline deletion resulted in fertile and viable animals born with expected Mendelian rations. Surprisingly, there were no apparent gross phenotypic manifestations of Cdx4deletion in either context despite complete excision of the Cdx4gene in the hematopoietic compartment. Peripheral blood counts were normal, as was flow cytometric analysis for the myeloid and lymphoid compartments as assessed using Mac1/Gr1 and various B- & T-cell markers. Using multi-parameter flow analysis and methylcellulose plating assays, we observed that loss of Cdx4 in both mouse models only caused subtle hematopoietic perturbations, such as an increase in thymus weight and a modest decrease in pre-B colony formation ability in vitro.However, there were no other apparent defects in the hematopoietic stem and progenitor compartments including GMP, CMP & MEP. There were also no marked differences in the expression of Hoxgenes in the hematopoietic compartment, even though striking dysregulation of various Hoxgenes was observed in Cdx4-deficient zebrafish model. We also assessed the effect of Cdx4deletion on the development of leukemia mediated by MLL fusion genes. Using in vivoretroviral transduction/bone marrow transplantations and in vitroserial-replating assays, we observed that Cdx4null mice remain susceptible to leukemic transformation by MLL-AF9, indicating that Cdx4is not necessary for MLL-AF9 mediated leukemias. The only subtle differences observed were that Cdx4−/− progenitors transduced with MLLAF9 generated more differentiated transformed blast colony-forming units in vitro, and induced leukemias with longer disease latencies compared to their wild-type littermate controls. Thus, in contrast with the zebrafish system, we found that Cdx4is dispensable for normal mammalian hematopoiesis, as well as for MLL-AF9 mediated leukemia. The striking differences we observed between the studies of the same gene in the two different vertebrate systems emphasizes the importance of not solely relying on a single model system to assess the role of transcription factors in development, and that the deficiency of even a highly conserved gene might perturb function to variable extents in different animal models.