Tumorigenesis by Meis1 overexpression is accompanied by a change of DNA target-sequence specificity which allows binding to the AP-1 element

Abstract

Meis1 overexpression induces tumorigenicity but its activity is inhibited by Prep1 tumor suppressor. Why does overexpression of Meis1 cause cancer and how does Prep1 inhibit? Tumor profiling and ChIP-sequencing data in a genetically-defined set of cell lines show that: 1) The number of Meis1 and Prep1 DNA binding sites increases linearly with their concentration resulting in a strong increase of “extra” target genes. 2) At high concentration, Meis1 DNA target specificity changes such that the most enriched consensus becomes that of the AP-1 regulatory element, whereas the specific OCTA consensus is not enriched because diluted within the many extra binding sites. 3) Prep1 inhibits Meis1 tumorigenesis preventing the binding to many of the “extra” genes containing AP-1 sites. 4) The overexpression of Prep1, but not of Meis1, changes the functional genomic distribution of the binding sites, increasing seven fold the number of its “enhancer” and decreasing its “promoter” targets. 5) A specific Meis1 “oncogenic” and Prep1 “tumor suppressing” signature has been identified selecting from the pool of genes bound by each protein those whose expression was modified uniquely by the “tumor-inducing” Meis1 or tumor-inhibiting Prep1 overexpression. In both signatures, the enriched gene categories are the same and are involved in signal transduction. However, Meis1 targets stimulatory genes while Prep1 targets genes that inhibit the tumorigenic signaling pathways.

New Insights into Cooperative Binding of Homeodomain Transcription Factors PREP1 and PBX1 to DNA

“…While PREP1 HD and PREP1 hd do not bind to the control probe, in the case of the PBX1 HD clear bands are formed, indicating little discrimination of PBX1 HD between sequences. This agrees with the results of the ChIP-seq analysis3031, which failed to identify a real consensus motif for DNA sites bound uniquely by PBX1. In the literature examples of PBX1 consensus sequences are reported3536.…”

“…The DNA sequence that was employed to measure binding to the HD, was based on ChIP-seq data on whole embryo trunk and several murine and human cell lines30313233. These studies have provided consensus sequences for DNA binding by PREP1-PBX1: the very frequent decameric (TGATTGACAG) and the less frequent octameric (TGATTGAT) oligonucleotides conform to the above consensus sequences.…”

“…MEIS-A and MEIS-B domains are PREP1 motifs required for heterodomerization with PBX1. Panel B: Consensus sequences for DNA binding by PREP1-PBX1: the very frequent decameric (PMH) and the less frequent octameric (PH) oligonucleotides303132, are both highlighted in red in the sequence. Control probes correspond to two sequences that do not contain any PBX1 orPREP1 binding site.…”

Deregulation of the HOXA9/MEIS1 axis in acute leukemia

“…This interaction can be targeted using GSK-3 inhibitors, leading to growth inhibition of cells transformed by either HOXA9/MEIS1 or MLL-fusion proteins (57–59). The oncogenic properties of MEIS1 are antagonistically regulated by PREP1, another TALE family protein, through direct competition for binding sites (68). PREP1 also competitively heterodimerizes and sequesters PBX proteins, thereby decreasing stability of MEIS1 and preventing the MEIS1-DDX3x/DDX5 interactions that are required for tumorigenesis (69, 70).…”

A tale of TALE, PREP1, PBX1, and MEIS1: Interconnections and competition in cancer

“…An in‐depth analysis of the DNA binding landscape (ChIP‐Seq analysis) of Prep1, Pbx1, and Meis1/2 was carried out in the whole trunk of the E11.5 mouse embryo and in embryonic stem (ES) cells, mouse embryonic fibroblasts (MEFs), pre‐adipocytes and has uncovered a set of general rules that can be applied to most TALE proteins.…”