The Novel Thiosemicarbazone Derivative Coti-2 Induces Mitochondrial Apoptosis in Acute Myeloid Leukemia Cells Via a p53-Independent Mechanism

IntroductionTP53 mutations are rare (< 5%) in de novo acute myeloid leukemia (AML) but if present, they are associated with a very poor prognosis (< 1% overall survival at 3 years). p53 mutations have been frequently detected in poor-prognosis patients with complex karyotype or therapy-related AML. COTI-2 is a third generation thiosemicarbazone derivative that was identified as a small molecule candidate against a diverse group of human cancer cell lines using a proprietary in silico drug screening method. COTI-2 causes cancer cell death via apoptosis. COTI-2 has been proposed to restore non-functional mutant p53 conformation to functional wild-type conformation in solid cancer cells, and it has recently entered a clinical trial in patients with advanced or recurrent gynecologic malignancies (NCT02433626). The molecular mechanisms of COTI-2 to induce apoptosis, however, remain largely unknown. In this study, we investigated activity and possible modes of action of COTI-2 against AML cells, especially focusing on its p53-independent properties.

METHODS AND RESULTS A total of 10 AML cell lines were exposed to COTI-2 for 72 hours and its anti-leukemia effects were determined by viable cell counts and annexin V staining. OCI-AML2, OCI-AML3, MOLM-13, MOLM-14 and MV4;11 express wild-type (WT) p53; THP-1and Kasumi-1 express mutant (MUT) p53; KG-1, U937 and HL-60 does not express p53 protein (p53 NULL). COTI-2 inhibited leukemia cell growth and increased the percentage of annexin V-positive cells, irrespective of cellular p53 status. The IC50 values (concentration at which cell growth is inhibited by 50%) for were not statistically different between p53 WT and MUT/NULL cells (10.3 ± 4.5 nM vs 20.2 ± 11.5 nM, P = 0.44). So did the ED50 values (effective concentration inducing 50% cell killing as measured by Annexin V positivity) (115.0 ± 50.9 nM vs 237.8 ± 109.9 nM, P = 0.34). We next examined the apoptotic effect on primary AML cells from patients with AML. A total of 14 samples [11 p53 WT cases and 3 p53 MUT (C135W, R248Q and C242Y) cases] were examined. COTI-2 induced apoptosis both in p53 WT (40.3 ± 6.4% compound-specific annexin V induction) and MUT samples (48.0 ± 17.7%) (P = 0.62). To elucidate the p53-independent activity of COTI-2, mRNA expression levels of TP53 (p53), CDKN1A (p21) and BBC3 (PUMA) were determined in four AML cell lines (OCI-AML3, MV4;11, HL-60 and Kasumi-1) after exposure to COTI-2. COTI-2 did not induce TP53 or its transcriptional targets CDKN1A or BBC3 in any of the cell lines. To investigate if COTI-2 activates an intrinsic pathway to induce apoptosis in AML, we used p53 NULL HL-60 cells overexpressing BCL-2 (HL-60/BCL-2), BCL-XL (HL-60/BCL-XL) and their controls (HL-60/neo). HL-60/BCL-2 and HL-60/BCL-XL respectively expressed BCL-2 and BCL-XL at greater than 4 times higher levels than HL-60/neo. Interestingly, overexpression of either BCL-2 or BCL-XL almost completely abrogated COTI-2-induced apoptosis. Similar results were obtained in OCI-AML3/BCL-2. MCL-1 overexpression only modestly inhibited COTI-2-induced apoptosis. Involvement of the extrinsic pathway was modest as JurkatI9.2 (a Jurkat clone deficient in caspase-8) and its control showed similar sensitivity to COTI-2. COTI-2 appeared to reduce MCL-1 expression levels through mTORC1 inhibition. Finally, COTI-2 strongly synergized with the BCL-2 inhibitor ABT-199 to induce apoptosis in AML cells.

Conclusion COTI-2 induces mitochondrial apoptosis in AML cells, irrespective of their p53 mutational status. COTI-2-induced apoptosis depends on BCL-2 and BCL-XL expression but not on MCL1 expression or p53 activation. COTI-2 has clinical potential in AML that often expresses high levels of MCL-1, especially in combination with BCL-2 inhibition.