The multifunctional C terminus of the hematopoietic AML1 transcription factor interacts with coregulatory proteins, supports the convergence and integration of physiological signals, and contains the nuclear matrix targeting signal, the protein motif that is necessary and sufficient to target AML1 to subnuclear sites. The (8;21) chromosomal translocation, which replaces the C terminus of AML1 with the ETO protein, modifies subnuclear targeting of AML1 in acute myeloid leukemia (AML) and results in defective myelopoiesis. We therefore addressed the relevance of AML1 subnuclear targeting and associated functions that reside in the C terminus to myeloid differentiation. A single amino acid substitution that abrogates intranuclear localization was introduced in the AML1 subnuclear targeting signal. Expression of the mutant AML1 protein blocks differentiation of myeloid progenitors to granulocytes in the presence of endogenous AML1 protein, as also occurs in the (8;21) chromosomal translocation, where only one allele of the AML1 gene is affected. The cells expressing the mutant AML1 protein continue to proliferate, maintain an immature blast-like morphology, and exhibit transformed properties that are hallmarks of leukemogenesis. These findings functionally link AML1 subnuclear targeting with competency for myeloid differentiation and expression of the transformed͞leukemia phenotype.acute myeloid leukemia ͉ AML1-ETO fusion protein ͉ nuclear matrix ͉ subnuclear organization C hanges in nuclear morphology are diagnostic hallmarks of leukemic cells and have been traditionally linked to structural and biochemical properties of hematopoietic lineage cells (1-3). In addition, the mammalian nucleus is functionally organized into subnuclear domains that contain regulatory machinery for transcription, replication, and repair (4-6). Alterations in the composition and distribution of these regulatory domains are observed in leukemic phenotypes (5, 7-9). Thus, modifications in nuclear structure and subnuclear organization are associated with normal myeloid cell growth͞differentiation and leukemic transformation.Acute myeloid leukemia (AML) is a prevalent hematopoietic malignancy characterized by abnormal proliferation and abrogated differentiation of myeloid progenitor cells (10, 11). Numerous cytogenetic abnormalities associated with AML involve the gene encoding acute myelogenous leukemia transcription factor 1 (AML1; also designated as Runt-related transcription factor 1, Runx1) (12, 13). AML1 is a nuclear matrix-associated transcription factor essential for hematopoiesis, thereby providing means to investigate nuclear structure-function interrelationships that are linked to biological control and are altered during disease (5, 14). For example, the (8;21) translocation, a predominant mutation identified in Ϸ15% of adult AML patients, produces a chimeric protein (AML1-ETO) in which the C terminus of AML1 is replaced by the ETO protein (13). The AML1-ETO translocation-fusion protein is a dominantnegative inhibitor of the endog...

Many mammalian SWI/SNF complexes use Brahma-related gene 1 (Brg1) as a catalytic subunit to remodel nucleosomes for transcription regulation. In several mesenchymal cells and tissues, expression of a defective Brg1 protein negates the normal activity of the SWI/SNF complex and delays or blocks differentiation. To investigate the role of SWI/SNF complexes during myelopoiesis, we stably expressed a dominant negative (dn) Brg1 mutant in the myeloid lineage. Forced expression of dnBrg1 in IL-3-dependent murine 32Dcl3 myeloid progenitor cells results in a profound delay in the granulocyte-colony stimulating factor (G-CSF) induced granulocytic maturation. These cells also exhibit a significant decrease in the expression of both CD11b and Gr-1 surface receptors, which are normally upregulated during granulopoiesis, and show sustained expression of myeloperoxidase, which is synthesized primarily during the promyelocytic (blast) stage of myeloid development. Thus, dnBrg1 expression causes a developmental block at the promyelocytic/metamyelocytic stage of myeloid differentiation. Our findings indicate that the normal chromatin remodeling function of Brg1 is necessary for the G-CSF dependent differentiation of myeloid cells towards the granulocytic lineage. This dependency on Brg1 may reflect a stringent requirement for chromatin remodeling at a critical stage of hematopoietic cell maturation.

Many types of acute myelogenous leukemia involve chromosomal translocations that target the C-terminus of Runx1/AML1 transcription factor, a master regulator of hematopoiesis. The C-terminus of Runx1/AML1 that includes the nuclear matrix targeting signal (NMTS) is essential for embryonic development, hematopoiesis, and target gene regulation. During the onset and normal progression of hematopoiesis, several lineage-specific factors such as C/EBPalpha and PU.1 interact with Runx1 to regulate transcription combinatorially. Here we addressed the functional interplay between subnuclear targeting of Runx1 and gene activation during hematopoiesis. Point mutations were generated in the NMTS of the human Runx1 protein and tested for their effect on transcriptional cooperativity with C/EBPalpha and PU.1 at myeloid-specific promoters. We characterized five mutants that do not alter nuclear import, DNA binding or C/EBPalpha-dependent synergistic activation of the target gene promoters. However a critical tyrosine in the NMTS is required for subnuclear targeting and activation of the granulocyte-macrophage colony stimulating factor (GM-CSF) promoter. Furthermore, this point mutation is defective for transcriptional synergism with PU.1 on the macrophage colony stimulating factor (MCSF) receptor c-FMS promoter. Our results indicate that the NMTS region of Runx1 is required for functional interactions with PU.1. Taken together, our findings establish that subnuclear targeting of Runx1 is a critical component of myeloid-specific transcriptional control.

Runx1 is a key hematopoietic transcription factor required for definitive hematopoiesis and is a frequent target of leukemia-related chromosomal translocations. The resulting fusion proteins, while retaining DNA binding activity, display loss of subnuclear targeting and associated transactivation functions encoded by the C-terminus of the protein. To define the precise contribution of the Runx1 C-terminus in development and leukemia, we created a knock-in mouse with a C-terminal truncation by introducing a single nucleic acid substitution in the native Runx1 locus. This mutation (Runx1(Q307X)) models genetic lesions observed in patients with leukemia and myeloproliferative disorders. The Runx1(Q307X) homozygous mouse exhibits embryonic lethality at E12.5 due to central nervous system hemorrhages and a complete lack of hematopoietic stem cell function. While able to bind DNA, Runx1(Q307X) is unable to activate target genes, resulting in deregulation of various hematopoietic markers. Thus, we demonstrate that the subnuclear targeting and transcriptional regulatory activities of the Runx1 C-terminus are critical for hematopoietic development. We propose that compromising the C-terminal functions of Runx1 is a common mechanism for the pathological consequences of a variety of somatic mutations and Runx1-related leukemic fusion proteins observed in human patients.

No abstract