Bioelectric Reprogramming of the Bone Marrow Niche in Acute Myeloid Leukemia

Background. A large body of evidence indicates that cancer is not comprised solely of malignant cells, but that it represents part of a complex network of transformed and non-transformed cells, including mesenchymal stromal cells (MSCs), immune, nervous, endothelial, stem cells and fibroblasts that, together with soluble mediators, make up the tumor microenvironment. More recently, molecular mechanisms underlying a novel system of developmental control have been described to define the role of bioelectric gradients. Differences in endogenous resting membrane potential (V _mem) of non-neural cells have importance in cell-cell interactions and regulation or dysregulation in embryogenesis, regeneration patterns, and cell behavior, especially proliferation and differentiation. Recently, the same issues have been suggested to be related to cancer where abnormal depolarization of resting V _mem is associated to neoplastic cells and the role of ion channels conductivity in cancer initiation and progression deserves great attention in tumor microenvironment remodeling and therapy response.

Methods and results. We characterized MSCs derived from bone marrow (BM) of pediatric acute myeloid leukemia patients (AML-MSCs) and identified significant changes in ion channel expression profile, including the Voltage-gated L-type calcium channel CaV1.2. We investigated the role of CaV1.2 and documented a reduced expression of CaV1.2 channel in 14 primary AML-MSCs (<53% by flow cytometry, p<0.001) mediating more frequent spontaneous calcium oscillations with higher amplitude than those found in MSCs derived from healthy donors (h-MSCs) (21% and 7%, respectively, p<0.05). Measurement of V _mem by patch clamp and a fluorescent potential-sensitive probe DiBAC revealed that AML-MSCs showed a more depolarized V _mem (-14.7 mV) with respect to h-MSCs (-30.5 mV, p<0.01). We hypothesized that depolarization may play a role in the AML-MSCs gene expression and functional reprogramming. Hence, we exposed h-MSCs to AML blasts demonstrating that MSCs V _mem depolarized (-11.8 mV, n=10), resembling the AML-MSCs V _mem, with a concomitant decrease of CaV1.2 channel expression level (<44% by flow cytometry, p<0.01). We treated h-MSCs with depolarizing pharmacological agents (ouabain and potassium gluconate) finding that the pharmacological-induced V _mem depolarization promotes a pro-leukemia cell phenotype like the AML-MSCs. Briefly, MSCs exhibited higher proliferation level (ouabain, 2.5-fold; Kgluc, 1.9-fold, p<0.05), higher release of IL-6 (1.7-fold average, p<0.05), lower immunomodulatory potential and reduced the CaV1.2 levels (<35% by flow cytometry, p<0.05) with respect to untreated h-MSCs. On the contrary, we treated AML-MSCs with hyperpolarizing agents (lubiprostone and ivermectin), which reverted their phenotype to that of h-MSCs, including lower proliferation level (ivermectin -1.9-fold, p<0.05), higher immunomodulatory potential, and higher CaV1.2 levels (1.8-fold by flow cytometry, p<0.05) with respect to untreated AML-MSCs, suggesting that depolarization is reversible and that by modeling MSCs V _mem we can restore an healthy stroma within the BM niche. We transduced AML-MSCs with a CaV1.2 expressing lentiviral vector (namely re-AML-MSCs, n=7) inducing a stable hyperpolarized V _mem and documented that their gene expression profile reverted toward the h-MSCs profile. Finally, we observed that re-AML-MSCs propagated their acquired bioelectrical state to the contiguous AML-MSCs establishing a healthy bio-electric field.

Conclusions. We demonstrated an unprecedented AML niche bioelectrical code that governs the state of MSCs, demonstrating that the contact with AML blasts is the driving force inducing the h-MSCs V _mem depolarization shift and the acquisition of pro-leukemia functions. In addition, we unveiled CaV1.2 channels as a key player in this process and postulated that rewriting the bioelectric code of the patient' AML-MSCs, through the restoration of voltage-dependent calcium channel CaV1.2 activity, could be a strategy to restore their healthy phenotype. These findings provide the biological rationale to evaluate the potential of novel MSCs gene or pharmacological therapies that could both disadvantage AML progression and restore a permissive BM niche.