Genome Wide Screening Reveals Distinct Signaling Pathways of the ETS Transcription Factor ERG In Primary Acute Leukemia Blasts

The Ets related gene, ERG, encodes a transcription factor with a crucial role in hematopoiesis. ERG rearrangements are associated with subtypes of myeloid leukemia, Ewing sarcoma and prostate carcinoma. Overexpression of ERG predicted adverse outcome in patients with acute T-lymphoblastic leukemia (T-ALL) and acute myeloid leukemia (AML). Recently, it was shown that a single ERG missense mutation in mice produced severe defects in embryonic hematopoiesis and adult stem cell maintenance. The significance of ERG signaling contributing to this disadvantageous phenotype in mice as well as in patients with acute leukemia is unknown. Thus, a genome wide screen to identify targets involved in ERG signaling was conducted.

Chromatin immunoprecipitation on chip (ChIP-chip) was carried out in 5 primary AML samples (new diagnosed AML A – E), one primary T-ALL sample and one normal bone marrow sample. ChIP-chip was performed on all samples using an ERG-specific antibody and unspecific IgG antibody. In addition, ChIP-chip was conducted in the HL60 cell line (that lack ERG product) as a negative control. Duplicate ChIP-chips were performed for each immunoprecipitation. Subsequently, the ERG enriched DNA was co-hybridized to total chromatin on a human promoter array. The array represents ∼35.000 human gene loci representing 11.000 independent promoters. Following normalization, a class comparison was carried out between the potential ERG targets against non-targets identified with IgG ChIP-chip.

Significantly enriched genes with a P-value ≤ 0.05 included 164 genes from AML A, 1053 genes from AML B, 829 genes from AML C, 4112 genes from AML D, 895 genes from AML E, 1266 genes from T-ALL, and 1941 genes from normal bone marrow ChIP-chips. Functional clustering by gene ontology (DAVID) of significantly ERG enriched gene sets revealed overrepresentation of pathways involved in oncogenesis, proliferation and cell signaling including: the MAPK signaling pathway (KEGG pathway analysis) was enriched in two AML ChIP-chips (AML B, P= 0.007; AML D, P= 0.003) and in the normal bone marrow ChIP-chip (P= 0.02). Moreover, the Ras GTPase protein domain (INTERPRO) was enriched in three of the five AML samples (AML B, P=0.04; AML C, P= 0.02; AML D, P= 0.04). Both, MAPK pathway and Ras GTPase protein domains were not enriched in the HL60 negative control. Furthermore, proliferation assays were performed in an ERG tet-on inducible leukemia cell line system to explore functional consequences of ERG overexpression. In stably transfected cells with potent ERG overexpression a significantly higher proliferation rate was observed compared to non-induced cells.

By genome wide ChIP-chip in primary leukemic samples, potential ERG downstream signaling pathways were revealed. Furthermore these data suggest that the ERG induced proliferation induction may be mediated by cellular programs including the MAPK pathway. Further validation of these findings will define the role of ERG involvement in the cellular signaling cascades and provide insights for molecular based treatment approaches to improve outcome for high risk leukemia characterized by high ERG expression.

No relevant conflicts of interest to declare.