Our initial studies on p15Ink4b provided evidence that the gene is a tumor suppressor for myeloid leukemia in mice using p15Ink4b-deficient mice. Use of the mice in a retrovirus-induced leukemogenesis protocol showed that mice heterozygous for Ink4b deficiency had increased susceptibility to myeloid leukemia. Mice were inoculated with retrovirus MOL4070LTR, a unique myeloid disease-inducing virus, which was recently constructed and produced in our laboratory. The retrovirus provided the cooperating events through insertional mutagenesis. This allowed the demonstration of the susceptibility to leukemia provided by loss of one Ink4b allele. Mice deficient in both alleles were not more susceptible than those deficient in one allele, raising the possibility that there are opposing forces to the development of myeloid leukemia in these Ink4b null mice. This may be due to the deletion throughout multiple tissues of the mouse. To assist in our understanding of why Ink4b-/- mice are not more susceptible to leukemia than Ink4b-/+ mice we have developed a conditional knockout of Ink4b based on the Cre-loxP system. These mice, when crossed with LysMCre mice for specific deletion in myeloid lineage cells, are highly susceptible to retrovirus-induced leukemia. Mice homozygous for loss of the gene have a 50% incidence of myeloid disease.
The loss of Ink4b expression in a high proportion of myeloid based diseases and increased AML susceptibility in Ink4b-deficient mice implies an important function for this CDKI in the maintenance of normal myelopoiesis. Despite this, little is known about how loss of Ink4b contributes to myelopoietic diseases or of the potential role of Ink4b in myeloid cell formation. We have recently investigated the role of Ink4b in hematopoiesis using Ink4b knock-out mice (Ink4b-/-). Bone marrow (BM) from Ink4b-/- mice show no changes in the frequency of hematopoietic stem cells but contained greater numbers of bi-potent granulocyte-macrophage progenitors (GMP) and this was found to be intrinsic to the p15Ink4b cells. Interestingly, Ink4b-/- progenitors did not cycle more frequently than wild-type progenitors and showed no differences in apoptotic potential. However, loss of Ink4b was shown to affect differentiation of the common myeloid progenitors (CMP). The resulting imbalance favors the production of myeloid progenitor cells, GMPs at the expense of erythroid progenitors. This work demonstrates a novel cell cycle-independent role for Ink4b during cell fate at the erythroid/myeloid bifurcation.
Another area of research our laboratory has examined human AML subtypes for hypermethylation of the INK4b locus. Inhibitors of DNA methylation are promising options for the treatment of human AML and MDS. So far, it was not investigated whether or not INK4b is hypermethylated in all cytogenetic subtypes of AML. A comparison of levels of INK4b methylation in AML with the three most common cytogenetic alterations, inv(16), t(8;21) and t(15;17), revealed a strikingly low level of methylation in all leukemias with inv(16) compared to the other types. Surprisingly, the expression level of INK4b in inv(16)+ AML samples was low and comparable to that of the other subtypes. An investigation into an alternative mechanism of INK4b silencing determined that the loss of INK4b expression was caused by inv(16)-encoded CBFβ-SMMHC (β subunit of the core binding factor fused in frame to the smooth muscle myosin heavy chain). The silencing was manifested in an inability to activate the normal expression of INK4b RNA. CBFβ-SMMHC was shown to displace RUNX1 from a newly determined core binding factor (CBF) site in the promoter of INK4b. Importantly, this study 1) establishes that the gene encoding the tumor suppressor p15INK4b is a target of CBFβ-SMMHC, a finding relevant to the leukemogenesis process and 2) indicates that in patients with inv(16)-containing AML, re-expression from the INK4b locus in the leukemia would not be predicted to occur using hypomethylating drug