MN1 Inhibits Myeloid Differentiation by Transcriptional Repression of EGR2

Abstract 229

Overexpression of MN1 (meningioma 1) is a negative prognostic factor in acute myeloid leukemia (AML) patients with normal cytogenetics, and induces a rapidly lethal AML in mice. We have shown previously that MN1, a transcription cofactor of retinoic acid receptor alpha (RARA), increases resistance to all-trans retinoic acid (ATRA) by greater than 3000-fold in an in-vitro differentiation model. We investigated the molecular mechanisms involved in the MN1-induced myeloid differentiation block by fusing potent transcriptional activation or repression domains to MN1, conducting a structure-function analysis of MN1, gene expression profiling, ChIP-on chip experiments, and functional validation of MN1 target genes. We found that (1) MN1 inhibits myeloid differentiation through transcriptional repression; (2) the C-terminal domain of MN1 is critical for induction of resistance to ATRA; (3) EGR2 is a putative direct target of MN1 and RARA that is repressed in MN1 leukemias; and (4) that constitutive upregulation of EGR2 in MN1 leukemias permits differentiation and prevents engraftment of transplanted cells. To investigate whether MN1 impacts on myeloid differentiation through transcriptional activation or repression we fused a strong transcriptional activation domain (VP16) or repression domain (M33) to MN1. MN1VP16 immortalized murine bone marrow cells, however, these cells could differentiate to mature granulocytes, and succumbed to cell cycle arrest upon treatment with ATRA. Mice receiving transplants of MN1VP16 cells had a median survival of 143 days (n=16) compared to 35 days in mice receiving MN1-transduced cells (n=18; p<.001). Morphologic analysis of bone marrow mostly showed mature granulocytes with less than 20 percent immature forms consistent with a diagnosis of myeloproliferative-like disease. Conversely, mice receiving transplants with cells transduced with the fusion of MN1 to the transcriptional repression domain of M33 (n=7) developed leukemia with a similar latency and phenotype as mice receiving transplants from MN1-transduced cells (survival, P=.6). These data suggest that MN1 inhibits myeloid differentiation by transcriptional repression rather than activation of its target genes. A structure-function analysis was performed to identify the domain(s) of MN1 required to inhibit myeloid differentiation. Consecutive stretches of 200 amino acids of MN1 were interrogated The deletion constructs were subsequently transduced into bone marrow cells immortalized by NUP98-HOXD13 (ND13). ND13 cells are very sensitive to ATRA-induced differentiation and cell cycle arrest with an IC50 of 0.1 μ M, whereas overexpression of MN1 increases resistance greater than 3000-fold. Interestingly, deletion of the 200 C-terminal amino acids of MN1 restored ATRA sensitivity of ND13 cells compared to full-length MN1, suggesting that the C-terminus of MN1 is required for inhibition of myeloid differentiation. To identify MN1-regulated genes important for the myeloid differentiation block we performed gene expression profiling of MN1- and MN1VP16-transduced bone marrow cells. To further identify genes that might be directly regulated by MN1 we performed ChIP-on-chip using anti-MN1 and anti-RARA antibodies. EGR2, CCL5, CMAH, among others, were identified as targets of both MN1 and RARA whose gene expression was low in MN1 but high in MN1VP16 cells. Overexpression of these genes in MN1-transduced leukemic cells was used to validate their function. Blast percentage of in vitro cultured bone marrow cells was 93, 58, 83, and 41 percent in MN1+CTL cells, MN1+EGR2, MN1+CCL5, and MN1+CMAH cells, respectively. MN1+EGR2 cell engraftment in peripheral blood of mice declined from 2.2 percent at 4 weeks to undetectable levels at 8 weeks (n=4), whereas MN1+CCL5 and MN1+CMAH cell engraftment was 23 (n=4) and 26 (n=4) percent at 4 weeks, and 14 and 30 percent at 8 weeks, respectively. At time of death, EGR2 was not detectable in mice whereas leukemias of mice receiving MN1+CCL5 or MN1+CMAH- transduced cells were positive for CCL5 or CMAH, respectively.

In conclusion, our data suggest that MN1 inhibits myeloid differentiation by transcriptional repression of a subset of its target genes, and that re-expression of EGR2, a zinc-finger transcription factor, may prevent outgrowth of MN1 leukemias in mice. Pharmacologic activation of EGR2 may become a novel antileukemic strategy.