Molecular structure of hydroxycinnamic acid derivatives is critical for the ability to synergize with the polyphenol carnosic acid against acute myeloid leukemia cells

Introduction: Acute myeloid leukemia (AML) is a devastating blood malignancy characterized by unrestrained proliferation of immature myeloid blasts. Despite initial responses to chemotherapy prognosis remains poor, especially for patients who are unfit for aggressive treatment. We have previously shown that the combination of the plant polyphenols curcumin (CUR) and carnosic acid (CA) applied at non-cytotoxic concentrations of each agent synergistically induces massive apoptosis selectively in human AML cells both in vitro and in vivo, without affecting normal hematopoietic cells. This effect was specifically mediated by sustained cytosolic Ca²⁺ overload. In the present study, we synthesized a series of hydroxycinnamic acid derivatives, such as methyl hydroxycinnamates and hydroxybenzylideneacetones, and screened these compounds for the ability to cooperate with either CUR or CA in producing antileukemic effects in vitro. Material and method: Proliferation, viability and apoptosis were assessed in KG-1a and U937 human AML cells by standard assays. Intracellular levels of reactive oxygen species (ROS) and Ca²⁺ were measured using fluorescent probes. Results and discussion: All the eight hydroxycinnamic acid derivatives tested were capable of reducing the growth and viability of AML cells in a dose-dependent manner and to a varying extent. Evaluation of the structure-activity relationship of these compounds in combination with CUR or CA revealed that none was capable of cooperating with CUR and that only methyl 4-hydroxycinnamate (KS-3) and methyl 3-methoxy-4-hydroxycinnamate (KS-6) demonstrated the ability to synergize with CA at non-cytotoxic concentrations to induce a strong and rapid (4-8 h) apoptotic effect, along with a dramatic reduction in cell numbers and viability following 72 h. In addition, we observed striking similarities between the antileukemic features of KS-3+CA and CUR+CA in that the apoptotic effect of both combinations was caused by increased cytosolic Ca²⁺ levels and was not accompanied by elevated ROS levels. Further, similar to CUR+CA, the KS-3+CA combination was not cytotoxic to freshly isolated normal peripheral blood mononuclear cells. Conclusion: Our results indicate that, similar to CUR, KS-3 and KS-6 are capable of cooperating with CA to specifically kill AML cells and that this effect strongly depends on both the position of the hydroxyl group on the aromatic ring and the modification of the carbonyl group in a hydroxycinnamic acid derivative. The synergistically acting combinations of phenolic compounds studied here may provide a prototype of novel efficacious and safe therapeutic tools for AML treatment, particularly, in elderly or unfit patients. (Supported by the Israel Science Foundation grant 226/16).