Although acute myeloid leukemia (AML) is a molecularly diverse disease, all AML patients have one thing in common, and that is the threat of disease recurrence. Relapses occur even after myeloablative chemotherapy and stem cell transplant. The biggest obstacle to improving the outlook for these patients, therefore, is our limited understanding of chemotherapy resistance. The ultimate goal of the Redell lab is to identify the mechanisms by which AML cells withstand chemotherapy, and by disabling those mechanisms, to achieve a meaningful advance in AML treatment. Toward that goal, the objective of this application is to determine the genomic, functional, and clinical differences between AML cases with intact environment- induced Stat3 signaling and cases with defective signaling. In contrast to prior evaluations of Stat3 in cancer, this project focuses not on the levl of Stat3 activity, but on the capacity for inducible signaling, as a predictor of chemotherapy response. Published data show that in pediatric AML, a normal Stat3 response to G-CSF and IL-6 stimulation is associated with an excellent outcome; unfortunately, most samples do not respond normally, despite abundant receptor expression, and these outcomes are poor. New preliminary data show that samples that fail to activate Stat3 in response to these ligands also fail to activate other pathways, and fail to respond to multiple stroma-derived factors. In contras, samples with intact Stat3 activation upon single ligand stimulation also activate other pathways, and they respond very similarly to stroma-derived factors. This result suggests that, in some poor-risk cases at least, Stat3 resistance is due to a generalized signaling defect. The central hypothesis of this application is that insensitivity of the Stat3 pathway to ligand stimulation is due to generalized signaling dysfunction, and is a characteristic of cells with enhanced DNA damage responses and self-renewal capacities. The four Specific Aims of this proposal will use a large cohort of pediatric AML samples. Specific Aim 1 is to demonstrate that a ligand-resistant signaling pattern is associated with poor outcome and more effective DNA damage repair. Multiple intracellular parameters will be measured by FACS, and the ability of signaling patterns to identify groups with good v. poor outcomes will be confirmed. Specific Aim 2 is to compare gene expression profiles between ligand-sensitive and ligand-resistant AML samples, and to identify which differentially expressed genes are functionally related to signaling and apoptosis. Specific Aim 3 will test the hypothesis that lipid raft dysfunction is a unifying mechanism underlying signaling failure and resistance to chemotherapy-induced apoptosis. Lipid raft clustering will be evaluated by microscopy, and the effects of lipid raft disruption on signaling and apoptosis will be investigated. Specific Aim 4 will compare the self-renewal capacities of ligand-sensitive v. ligand-resistant AML samples, using in vitro and in vivo methods. The benefits of this research to cancer patients are extensive. Our results will lead to a better understanding of chemoresistance in AML, and ultimately to new strategies that reduce resistance or augment chemosensitivity.