Bone marrow niche of acute myeloid leukemia (AML) regulates leukemic progression by favoring leukemic cell quiescence and chemoresistance. Niche components, especially mesenchymal stromal cells (MSCs) protect leukemic cells from chemotherapy-induced apoptosis. Direct intercellular communications between hematopoietic stem cells and MSCs involve connexins, key components of gap junctions, which could be interresting targets to reverse this phenomenom.
We postulated that blocking the assembly of connexins could modify cell-cell interactions in the leukemic niche and consequently reverse the chemoresistance. We evaluated the effects of carbenoxolone (CBX), a glycyrrhetinic acid derivative known to block gap junctions.
Primary bone marrow (BM) AML blasts, 8 AML cell lines and primary BM CD34+ cells were exposed to CBX to assess cell viability and proliferation (MTT assay). The combined drug effects of CBX and cytarabine (Ara-C) were studied by isobolograms. The metabolic effects of CBX on AML cells were investigated with Seahorse® analyzer and high-throughput metabolic profiling by OmniLog® technology. Direct cell-cell interactions were studied in a co-culture system of AML cells and primary BM MSCs (48 h, 37°C, 5% CO2)exposed or not to CBX (150 μM) and Ara-C (1 μM). BM MSC-adherent AML cells were collected and the two populations were discriminated by flow cytometry studies to analyses the cell cycle (Ki67/7AAD assay) and to quantify apoptosis/necrosis (annexinV/7AAD assay). Gap junction inhibition by CBX was checked by calcein transfer from labeled MSCs to AML cells. CBX in vivo effects were evaluated in NSG mice transplanted with the human OCI-AML-3 cells and treated or not with Ara-C. Statistical analyses were performed with Mann-Whitney-Wilcoxon test.
CBX reduced cell growth and viability of AML cell lines in a time- and dose-dependent manner (IC50≈150 μM). The exposure to CBX (48 h, 150 μM) of primary AML blast induced a 2-fold decrease in the number of viable cells. CBX did not have any deleterious effect on normal BM-CD34+ cells. Interestingly, the isobolograms of the 8 AML cells lines identified 3 different profiles of resistance to Ara-C and a synergistic effect between CBX and Ara-C. Regarding the energy metabolism, CBX induced a major decrease in oxidative phosphorylation and glycolysis of leukemic cells regardless their chemoresistance level to Ara-C. In the co-culture experiments, contact with MSCs promoted quiescence and resistance to Ara-C of leukemic cells (4- and 6-fold decrease in apoptosis/necrosis rate of KG1a cells and primary blasts) and co-treatment of leukemic cells with CBX reduced this quiescence and reversed the MSC-induced chemoresistance (5- and 2-fold increased in apoptosis/necrosis rate of KG1a cells and primary blasts). CBX induced a 48%>decrease in calcein transfer between leukemic cells and MSCs, highlighting its role as gap junction inhibitor. Finally, in vivo survival analyses showed that untreated mice died after 32–33 days, while CBX treatment and even more CBX and Cyt infusions significantly improved mice survival (up to 85 days). Moreover, CBX-treated mice showed a limited splenomegaly, hepatomegaly and leukemic infiltration in the liver.
Niche-induced chemoresistance is associated with AML relapse after initial well-conducted chemotherapy. Combined to chemotherapy, CBX, a drug already clinically evaluated in the treatment of gastric/duodenal ulcer, could be of interest to reduce the deleterious effects of leukemic niche by targeting gap junctions.