Over Expression of MN1 Confers Resistance to Chemotherapy In Vitro and In Vivo

Abstract 969

Acute myeloid leukemia (AML) is an accumulation of immature myeloid precursors that leads to progressive marrow failure and death. This disease affects approximately 12,000 people per year in the United States, causing 9,000 deaths. Despite decades of active research the overall 5 year survival remains a dismal 30–40%. The backbone of initial therapy for the last 30 years is combination chemotherapy containing cytarabine (Ara-C) and an anthracycline. Resistance to these therapies is a major problem and most patients diagnosed with AML will ultimately die from resistant disease.

AML is characterized by heterogeneous genetic alterations that can be used to delineate prognosis. Using standard karyotyping techniques patients can be divided into good, intermediate and poor prognostic categories. There is a clear link between these chromosomal aberrations and response to chemotherapy as complete remission rates are significantly different between groups. Patients with no detectable cytogenetic abnormality fall into an intermediate prognostic group with a very heterogeneous outcome. Recent work has begun to uncover submicroscopic genetic alterations that effect prognosis for these patients. These alterations can be mutations, over or under expression of a particular gene.

The MN1 gene encodes a transcription co-factor first identified by its involvement in a balanced translocation in a patient with a meningioma. Since its initial description it has been found over-expressed in multiple AML patient samples. There are several reports that over-expression of MN1 confers a worse prognosis in AML. High MN1 expressers were less likely to achieve a remission and had a lower 3 year survival rate. Additionally, over expression of MN1 in murine bone marrow leads to AML in transplanted recipients and predicts for resistance to ATRA in elderly AML patients. However, the effect of MN1 over expression on response to standard chemotherapy is currently unknown.

To answer this question we used a murine model of AML driven by MLL-ENL. AML blasts were infected with retroviral vectors that contained MN1 and a GFP reporter. Partially infected blast populations were then exposed to various concentrations of either Ara-C or doxorubicin and the ratio of GFP positive and negative cells was compared to untreated controls. When blasts were exposed to 150 nM Ara-C the GFP+ percentage went from 21.10 (+/− 0.5302) in the control samples to 35.68 (+/−1.230) in the treated samples. This result was even more profound when cells were treated with 15 ng/ml doxorubicin where the percentage went from 21.10 (+/− 0.5302) to 80.27 (+/−1.615). Both results were highly statistically significant by two tailed student's t test with p values of 0.004 and <0.0001 respectively. Consistent results were obtained in multiple different infections and with separately derived MLL-ENL lines. These data demonstrate that blasts expressing MN1 had an advantage when exposed to either Ara-C or doxorubicin although the effect was far more pronounced with doxorubicin exposure. MN1 expressing blasts were also resistant to the combination of Ara-C and doxorubicin.

In order to determine if MN1 conferred resistance to Ara-C and doxorubicin in vivo we injected sublethally irradiated, Ly5.1+ C57Bl6 recipients with a partially infected population of blasts. Ly5.1+ animals do not express the Ly5.2 allele; thus, staining cells for Ly5.2 allows differentiation of leukemic cells from endogenous marrow. Eight days after injection of blasts animals were treated with 100 mg/kg Ara-C plus 3 mg/kg doxorubicin daily for 5 days or observed. On day 6 animals were sacrificed and bone marrow from bilateral femurs was harvested, stained for Ly5.2 and analyzed by flow cytometery. Animals treated with Ara-C plus doxorubicin had 90.58% (+/−0.6638) Ly5.2+, GFP+ blasts compared to 55.38% (+/−5.245) in control animals. This result was highly statistically significant with a p value of <0.0001 by two tailed student's t test. This observation was reproducible in a separately derived MLL-ENL driven cell line.

These data suggest that over expression of MN1 in this murine AML model confers resistance to both Ara-C and doxorubicin in vitro and in vivo and provides a biological explanation for the clinical observation that it confers a worse prognosis. The mechanisms involved in this resistance are currently under study.