Lithium Treatment Potentiates Both in Vitro and in Vivo Retinoic Acid Efficacy in APL.

Abstract 2614

We previously demonstrated that although retinoic acid (RA) has targeted efficacy in Acute Promyelocytic Leukemia (APL), heterogeneity exists leading to the appearance of un-targeted clones at the time of relapse. Characterization of these clones is not yet fully unraveled though we and others have previously highlighted the roles of RARα mutations, pharmacogenomics or APL miRNome. We recently identified that the ERK1/2 pathway synergized with RA to restore the transcriptional activity of RA in resistant APL cells, thus restoring RA induced differentiation (Cassinat et al. Mol Cell Biol 2011). These results suggest that targeting interconnected signaling pathways could optimize differentiation therapy efficacy. To this effect, we studied known signaling pathway activators or inhibitors that could potentiate with RA and identified Lithium chloride (LiCl).

Treatment of the ATRA sensitive-APL NB4 cell line with LiCl (25mM) decreases proliferation and increases apoptosis (25% and 40% of Annexin V-positive cells at day 1 and 2 respectively) with evidence of caspase 3 cleavage at day 2. Because NB4 cells fully differentiated with RA alone we were unable to observe any synergy when combined with LiCl. Treatment of the RA-resistant APL UF-1 cell line with RA or LiCl alone does not induce differentiation. Combination of RA+LiCl restores differentiation after 3 days of culture (65% CD11b positive and 55% NBT test positive cells). Similar results were obtained with different GSK3 inhibitors, suggesting that the LiCL effects were in part linked to its well characterized GSK3 inhibitory activity. Interestingly, we noted that LiCl treatment induces rapid phosphorylation of ERK1/2 and pretreatment with the MEK/ERK1/2 inhibitor UO126 fully abolished the differentiation induced by the RA+LiCl combination. The combination restores in UF-1 the expression of RA target genes (such as RARα2) to the same levels obtained in NB4 cells treated by RA alone. The level of luciferase activity of an RA responsive element reporter gene was increased with the RA+LiCl combination compared to RA alone. Both target gene expression and luciferase activiy were abolished after inhibition of the MEK/ERK1/2 pathway. Thus, increase in differentiation of UF-1 cells by RA+LiCl is linked to increased RA transcriptional activation. Similar studies in fresh APL patient cells confirmed both the increase in differentiation and level of RA target gene expression and their inhibition by UO126. Finally, to translate these findings in vivo, we used the APL-transplantable mouse model. Plasma lithium levels in treated mice were measured between 0.6 and 1.05 mmol/l, levels reached in humans. When LiCl was combined with RA we repeatedly observed a pronounced survival advantage compared to mice treated by RA alone as evaluated by Kaplan Meier analysis.

In this work we demonstrate that LiCl, a well tolerated agent in humans, has the potential, when combined with RA, to restore RA induced transcriptional activation and differentiation in RA resistant APL cells. Furthermore, this combination also increases RA efficacy in an in vivo APL mouse model.