Overexpression of GATA1 Confers Chemotherapy Resistance in Pediatric Acute Megakaryocytic Leukemia.

Abstract 2039

Poster Board II-16

Acute megakaryocytic leukemia (AMkL; M7) is a biologically heterogeneous form of AML, representing ∼10% of pediatric and 1-2% of adult AML cases. AMkL is the most common AML subtype of children with Down syndrome (DS). DS children with AMkL have an excellent prognosis with EFS rates of 80-100% when treated with ara-C/anthracycline-based protocols, in contrast to the <30% EFS rates of non-DS children with AMkL. This also contrasts to the ∼50% EFS rates of non-DS children with AML overall, indicating that AMkL is an extremely poor risk group amongst non-DS children with AML despite the use of intensive chemotherapy-based protocols. These clinical data make a compelling argument that new therapies are essential to improve the treatment outcome of this aggressive disease. Acquired somatic mutations of the transcription factor gene, GATA1 (localized to Xp11.23), have been detected uniformly in nearly all DS AMkL cases, but not in non-DS AML and non-AMkL DS leukemia cases. The net effect of GATA1 mutations is an introduction of early stop codons and synthesis of a shorter GATA1 protein (designated GATA1s) that has altered transactivation activity, potentially contributing to the uncontrolled proliferation of immature megakaryocytes. It is conceivable that the altered GATA1 function between DS and non-DS AMkL may account for differential expression of GATA1 target genes in these two groups of patients. On the other hand, overexpression of GATA1 in megakaryoblasts from non-DS children with AMkL compared to myeloblasts from non-DS children with other subtypes of AML may contribute to differences in chemotherapy sensitivity via regulation of GATA1 target genes. We previously reported that GATA1 mutations in DS AMkL are associated with decreased expression of cytidine deaminase (encodes an enzyme which can convert ara-C to ara-U, the inactive form of the drug), thus contributing to the enhanced ara-C sensitivity of DS AMkL blasts. Further, when GATA1 was ectopically expressed in a DS AMkL cell line, CMK, it caused significantly increased resistance to ara-C. In the present study, we confirmed overexpression of GATA1 in non-DS AMkL blasts compared to non-DS AML blasts by real-time RT-PCR quantitation of GATA1 transcripts in our cohort of patient samples. shRNA knockdown of GATA1 in a non-DS AMkL cell line, Meg-01, resulted in significantly increased sensitivities to ara-C and daunorubicin, the two main drugs used for AML treatment, and significantly increased basal level apoptosis. This was accompanied by significantly decreased Bcl-xL transcript and protein levels in the GATA1 shRNA knockdown clones compared to a shRNA negative control. Binding of GATA1 to the two GATA elements in Bcl-x promoter and transactivation of Bcl-x promoter activity by GATA1 was demonstrated by ChIP assays and luciferase reporter assays, respectively, in Meg-01 cells. In our cohort of non-DS AMkL and AML patient samples, significant overexpression of Bcl-xL in non-DS AMkL compared to non-DS AML cases and a significant correlation between Bcl-xL and GATA1 transcripts were detected. Besides Bcl-xL, additional GATA1 targets (e.g. TNF) related to apoptosis were also identified by gene expression and ChIP-on-ChIP microarray analyses. Interestingly, our microarray data also suggest that GATA1 may have an impact on PI3-kinase/Akt pathway through modulating directly or indirectly a group of genes within the pathway. Western blotting revealed increased phosphorylation of Akt in the GATA1 knockdown clones compared to the negative control cells. Previous studies reported that histone deacetylase inhibitors (HDACIs) treatment causes hyperacetylation and subsequent degradation of GATA1, suggesting that these agents may be effective in targeting GATA1 in AMkL. Treatment of Meg-01 cells with an HDACI, valproic acid (VPA), resulted in decreased protein levels for GATA1 and Bcl-xL and increased phosphorylation of Akt. Co-treatment of Meg-01 cells with VPA and ara-C resulted in synergistic induction of apoptosis and activation of caspase-3. This drug synergy was amplified when a non-toxic dose of the PI3-kinase inhibitor LY294002 was added. Our results demonstrate that GATA1 causes resistance to chemotherapy in non-DS AMkL by promoting AMkL blast survival through regulating its target genes. Treatment of AMkL may be improved by integrating HDACI and PI3-kinase or Akt inhibitors into the chemotherapy of this disease.