Arsenic Trioxide Abrogate Mn1 Mediated Ra-Resistance In Acute Promyelocytic Leukemia

Introduction: Described as a well know marker of worse prognosis in acute myeloid leukemia (AML), MN1 overexpression has been associated with inv(16) or EVI1 overexpression (Heuser et al., Blood 2007). The promoter region of the MN1 gene has Retinoic Acid Response Elements (RAREs), and higher levels of MN1 expression have been associated with decreased response to retinoic acid (RA) in vitro. Nevertheless, in the context of acute promyelocytic leukemia, little is known about MN1 gene expression and functionality in vivo. Aims: Here, we investigated the effects of in vitro treatment with RA plus arsenic trioxide (ATO) in APL cell lines and primary blasts overexpressing MN1. Additionally, we quantified MN1 expression and correlated its levels with treatment outcome in a cohort of patients enrolled in the International Consortium on Acute Leukemia (ICAPL2006) study. Methods: Primary leukemic blasts from hCG-PMLRARα transgenic mice (TM; n=2) and APL patients (age, 36-45y; n=2) were transduced with MN1 or empty vector (EV, control) to evaluate cell proliferation, differentiation and apoptosis. Confirmatory assays were performed using transduced NB4 and NB4R2 (RA-resistant) cell lines. After synchronization using double thymidine block, transduced cells were submitted to proliferation and clonogenic (treated with RA and ATO, as well) assays. To evaluate the apoptotic rate, cells were treated with ATO (1 µM) alone or in combination with RA (1 µM each), for 24, 48 and 72 hours. The granulocytic differentiation in response to RA treatment alone (1 µM) or in combination with ATO (1 µM) was evaluated based on the CD11b and CD11c surface levels. In addition, 116 patients (age, 18-82y; 51% males) with newly diagnosed APL enrolled in the ICAPL2006 study were included. To validate our data, Bootstrap resampling procedure with 1000 repetitions from the original database was performed to assess the model bias. Results: Primary APL cells transduced with MN1 (from TM/APL patients) presented higher proliferation rates compared to controls (P<.05). Similarly, the overexpression of MN1 in APL cell lines was associated with increased proliferation (P=.001) and clonogenicity (P<.05). Furthermore, NB4-MN1 cells are able to form colonies in the presence of RA (1 µM) (P<.01) but not under ATO (1 µM) treatment (P>.05), while NB4R2-MN1 cells were able to form colonies in the presence of ATO (P<.05). To investigate whether MN1 promotes resistance to drug-induced apoptosis, we treated lentivirally transduced cells with RA plus ATO for 72 hours. No differences were founded between the MN1-transduced and the EV cells (P>.05). In accordance with our results using primary APL samples, ATO treatment (alone or in combination with RA) does not modulate the drug-induced apoptosis in a time-dependent manner in NB4 and NB4R2-MN1 cells (P<.05). For NB4 cells, the differentiation rate was lower in MN1-expressing cells under RA alone or in combination with ATO for 48 hours (P<.05), although these effects were abrogated after 72 hours of RA and ATO treatment. In contrast, NB4R2-MN1 cells exhibited decreased differentiation rate at 48 and 72 hours in presence of RA alone or in combination with ATO, in comparison with EV cells (P<.05). In the clinical setting, the retrospective analysis of patients enrolled in the ICAPL2006 study revealed that the baseline characteristics were similar between patients with low and high MN1 levels. According to PETHEMA/GIMEMA criteria for predicting relapse, 34% and 46% of patients assigned to the low- and high-MN1 groups were deemed high-risk patients, respectively (P=.045). High MN1 expression was associated with lower 5-y Disease-Free Survival rates (74%; 95% CI: 69-91%)(HR: 2.61, 95% CI: 0.86-8.36) and Cumulative Incidence of Relapse (25%; 95% CI: 13-34%) (HR: 2.27, 95% CI: 0.75-6. 8). Bootstrap analysis for internal validation resulted in an AUC (0.63, 95% CI: 0.57-0.809) very similar to the original described data. Conclusion: We show that MN1 is relevant for RA-induced differentiation both in vitro and in vivo and may be involved in RA-resistance. Additionally, treatment with ATO circumvented the differentiation blockage in MN1 cells.