Functional Assessment of the Mitochondrial Pathway of Caspase Activation in Patients with Acute Myeloid Leukemia (AML).

Chemotherapeutic agents used in the treatment of AML such as cytarabine and etoposide can induce caspase activation through the mitochondrial pathway. Blocks in this pathway can render cells resistant to these agents and may contribute to poor patient outcomes. To explore the frequency of defects in this pathway, the functional integrity of the downstream portion of the mitochondrial pathway of caspase activation was assessed in cytosolic lysates from patients with newly diagnosed (n=29) or relapsed (n=11) AML who had >80% blasts in their peripheral blood. Lysates were stimulated with cytochrome c and dATP (cy c/dATP) and hydrolysis of Ac-DEVD-AFC by effector caspases was measured over time. The specificity of the assay was tested by pretreating lysates with the BIR3 domain of XIAP (a caspase-9 inhibitor) or Crm A (a caspase-8 inhibitor). Pretreatment with BIR3, but not CrmA protein, inhibited cy c/dATP-induced effector caspase activation. In contrast, Crm A, but not BIR3 inhibited recombinant caspase-8-induced effector caspase activation, demonstrating the specificity of the assay. Cy c/dATP-induced effector caspase activation correlated with etoposide-induced apoptosis. Samples with the highest increases in effector caspase activation had the greatest sensitivity to etoposide (R²=0.69, p=0.006). To relate the functional status of the mitochondrial pathway to patient outcome, mitochondrial pathways were divided into intact and blocked based on the increase in effector caspase activity in normal mobilized peripheral blood mononuclear stem cells (PBSC) stimulated with cy c/dATP. AML samples with increases in effector caspases ≥ PBSC were deemed to have intact pathways, while AML samples with increases in effector caspases < PBSC were deemed to have blocked pathways. With this division, 8/11 (73%) samples from patients with relapsed disease had a blocked mitochondrial pathway vs 7/29 (24%) samples from patients with newly diagnosed disease (p = 0.009 by Fisher Exact test). In the 22 patients who completed induction chemotherapy, functional defects in caspase activation did not predict response, as 3/13 (23%) responders and 2/9 (22%) nonresponders had a blocked pathway (p = 1.0 by Fisher Exact test). Defects in the mitochondrial pathway of caspase activation may be due to blocks above or below the level of effector caspases. To distinguish between these possibilities, effector caspases were directly activated by stimulating lysates with recombinant active caspase-8. Data were dichotomized based on caspase-8-induced activation of effector caspases in lysates of PBSC. Active caspase-8 failed to activate effector caspases in 13/15 (87%) samples with a blocked mitochondrial pathway. In contrast, active caspase-8 failed to activate effector caspase in 3/27 (11%) samples with an intact mitochondrial pathway (p<0.0001 by Fisher Exact test). These results suggest that most samples with a blocked mitochondrial pathway had a distal defect in caspase activation. In conclusion, functional defects in the distal portion of the mitochondrial pathway of caspase activation were observed more frequently in patients with relapsed AML than newly diagnosed disease. Future studies with longitudinal samples are required to determine whether defects in relapsed patients were present at diagnosis or emerged at relapse. Nonetheless, for patients with blocked pathways, therapeutic strategies that overcome these distal blocks or induce cell death through caspase-independent mechanisms could be useful.