A Comprehensive Genomic Approach Using Gain of Function Cell Models, Patient Specimens and ChIP-on-CHIP Technologies Identifies PLZF-RARα Target Genes with Potential Roles in t(11;17) Associated APL.

The t(11;17)(q23;q21) translocation involves the production of reciprocal fusion proteins PLZF-RARα and RARα-PLZF, which mediate malignant transformation by binding to and dysregulating RARα/RXR and PLZF target genes, respectively. In order to investigate the molecular basis for PLZF-RARα induced leukemogenesis, we used a gain of function model in which PLZF-RARα was ectopically expressed in U937 leukemia cells. After demonstrating in our system that PLZF-RARα is capable of inducing a G1 cell cycle arrest and inhibiting cell growth and myeloid differentiation, we sought to identify genes directly bound and transcriptionally regulated by PLZF-RARα. Chromatin from U937PLZF-RARα expressing cells (+10nM RA) was immunoprecipitated using PLZF antibodies, amplified by ligation-mediated PCR and biological triplicates were hybridized to NimbleGen 2.7kB promoter arrays, which represent 24,275 human promoters. We identified 1797 genes that are directly bound by PLZF-RARα in at least 2 out of 3 arrays, and the majority of these genes (89%) are also bound in the absence of exogenously added RA. Quantitative real time PCR using primary ChIP samples was used to validate ChIP-on-CHIP results and all genes tested to date (n=11) were confirmed as direct targets of PLZF-RARα. Ontological analyses of genes identified by ChIP-on-CHIP revealed enrichment for genes involved in myeloid cell functions including immune, inflammatory and defense responses, in addition to genes involved in apoptosis and signal transduction pathways. Furthermore, genes encoding nuclear proteins were also highly enriched and these included previously identified RARα/RXR target genes (ie. CEBPε, RARβ2, PRAM1, NFE-2), which are likely targeted by the PLZF-RARα oncoprotein, as well as novel PLZF-RARα targets, many of which have roles in blood cell development and have been implicated in leukemia (ie. RUNX1, MLL2, MCL1, PIM1, FANCB). Of these 1797 genes, a significant percentage (22%) are also transcriptionally regulated by PLZF-RARα (>1.5 fold, p<0.05). To identify genes specific to the PLZF-RARα fusion generated in t(11;17) APL, we compared gene expression profiles of 26 PML-RARα and 4 PLZF-RARα expressing APL patient blasts. A comparison of differentially expressed genes in the patient specimens with those both directly bound and regulated by PLZF-RARα in U937 cells, identified a small subset of genes including RUNX1, KLF10, a transcriptional regulator and inhibitor of cell growth, as well as ID1 and ID2, whose expression level has been shown to correlate with myeloid differentiation. Although the expression of these genes was variable in PML-RARα blasts, expression was consistently lower in PLZF-RARα APL blasts (>2 fold, p<0.03). In U937 cells, PLZF-RARα repressed RUNX1, KLF10 and ID1 in the absence of exogenous RA. Intriguingly, RUNX1, KLF10 and ID2 were also identified as direct target genes of PLZF in the KG1a cell line and were transcriptionally regulated by PLZF in U937 cells, suggesting that PLZF and PLZF-RARα may co-regulate a subset of target genes. Given the roles of RUNX1, KLF10, ID1 and ID2 in myeloid differentiation and growth inhibition, these genes may represent PLZF-RARα specific targets that potentially contribute to the pathogenesis of t(11;17) APL.