One reason for the difficulty in developing effective treatments for myelodysplastic syndrome is that there are no myelodysplastic syndrome cell lines which can be used to model or study the disease. Although numerous investigators have attempted to develop xenograft models for myelodysplastic syndrome, these attempts have met with little success. Given that the NUP98-HOXD13 (NHD13) mice develop a highly penetrant myelodysplastic syndrome that closely resembles the human disease, we have begun studies to determine if these mice are a useful pre-clinical model for myelodysplastic syndrome. Our initial studies have used the DNA-methyltransferase inhibitor 5-azacytidine. Our pilot study included 3 groups of mice [NHD13 mice injected with 5-azacytidine (n=6), NHD13 mice injected with saline (n=4), and WT mice injected with 5-azacytidine (n=4)]. Although the results from this pilot were encouraging, they showed no clear survival benefit. To determine whether we were achieving levels of 5-azacytidine adequate to cause cytosine demethylation, we examined global and gene-specific methylation status, in collaboration with Dr. J.P. Issa (M.D. Anderson);these experiments demonstrated hypermethylation in the NHD13 mice compared to controls, and partial reversal of the hypermethylation with 5-azacytidine treatment. Some of these results have recently been published (Genome Res. 4:580-91, 2014).However, since the experiments discussed above used transgenic mice, effective treatment with 5-azacytidine could not replace the MDS bone marrow with completely normal (ie, wildtype or WT) bone marrow, since all of the bone marrow was transgenic. Therefore, in order to distinguish improvement in peripheral blood cytopenia due to differentiation of the MDS clone from elimination of the MDS clone, we have repeated the experiments using chimeric mice, that have both WT and NHD13 bone marrow. These repeat experiments have been performed using Decitabine (DAC), a related DNA methyltransferase inhibitor. Mice treated with DAC showed hematologic improvement and a survival advantage compared with saline-treated control mice;this experiment has now been repeated three times. We have sorted BM cells from treated mice into WT and NHD13 populations, and sent DNA from these samples to our collaborator (Dr. J.P. Issa), who has shown clear differences in global cytosine methylation between the NHD13 and WT samples, and partial reversal of this hypermethylation with DAC treatment, using a deep sequencing technique. A portion of this data was published in abstract form, and a manuscript describing these findings is currently being prepared. Despite the recent FDA approval of three drugs for MDS, the only curative option for patients with MDS remains allogeneic hematopoietic stem cell transplantation (HSCT). Allogeneic HSCT has two essential components, high-dose cytotoxic chemo-radiotherapy, and an immune-mediated graft versus host (GVH), or graft versus leukemia (GVL) effect. However, the relative contributions of high dose cytotoxic therapy and GVL are not well established in MDS patients. We propose to use transplantation of the NHD13 mice as a means to investigate this question. Our initial experiments, using either 650 or 1000 CGy of radiation, indicated that 650 CGy was ineffective, but that 1000 CGy could induce a remission of 26-38 weeks, defined by normalization of peripheral blood counts and less than 2% circulating host cells. However, despite this period of prolonged remission, and prolonged survival compared to non-transplanted mice, all of the mice have ultimately relapsed, indicating that this myeloablative therapy was not curative. To address the question of a GVL effect, we crossed C57bl6 mice with C3h.SW mice. C3H.SW mice are identical to C57Bl6 mice at the major histocompatibility loci, but have numerous mismatches at minor histocompatibility loci108. For this reason, transplantation of C3H.SW donor cells into C57Bl6 host mice has been used to study GVH and GVL. We transplanted BM from C57Bl6/C3HSW mice into C57Bl6 NHD13 recipients. The mice developed little GVH or GVL under these conditions. Subsequent experiments using higher doses of BM and 5 x 10E06 peripheral T cells showed severe GVH, that was lethal to 3 of 5 mice. A third trial, using a higher dose of bone marrow cells (10E07)has led to survival 52 weeks post-transplant. These results demonstrate a survival benefit for bone marrow transplant;ongoing experiments are aimed at determining whether donor lymphocyte infusion or co-tranplantation of specific T cell subsets will lead to yet greater survival. A portion of this data was presented at the Dec. 2013 American Society of Hematology (ASH) Annual meeting. In addition to the experiments outlined above, we have transferred NHD13 mice to colleagues at several academic institutions, and have licensed NHD13 mice to at least two separate biotech companies for pre-clinical studies. These colleagues have plans to treat NHD13 mice with a variety of agents, including histone deacetylase inhibitors, apoptosis inhibitors, and angiogenesis inhibitors. One of these compounds (ACE-536) is now in phase II clinical trials for patients with MDS. As stated in the goals, we have generated leukemic mice and cell lines in which the leukemia is driven by the binding of a leukemic fusion protein (NP23) to H3K4Me3 residues. We have treated these NP23 cell lines with compounds that have been shown to inhibit binding of PHD domains (including that present in the NP23 fusion) to H3K4Me3 in solution. NP23 cell lines are completely killed by a 12 hr treatment with one of these compounds (disulfiram), whereas control cell lines are not killed by this compound. Furthermore, cell death is associated with a marked (more than 5-fold) decrease in binding of the NP23 fusion protein to selected H3K4Me3 residues. A manuscript describing these findings was recently published (Cancer Discov. 5:564-77, 2014), and an in vivo trial of disulfiram for NP23 leukemias has begun.
Showing the most recent 10 out of 13 publications
