Regulation of the PU.1 Gene by Sense and Functional Antisense RNAs Generated through the Same Chromatin Architecture.

The transcription factor PU.1 is an important regulator of hematopoiesis and correct expression levels in specific lineages are critical for normal hematopoietic development. Specifically, PU.1 is maintained or upregulated in specific lineages, and failure to downregulate PU.1 in other lineages can lead to a block in development of that lineage and/or leukemia. In vivo expression of PU.1 is dependent on an upstream regulatory element called the URE. Disruption of the URE leads to downregulation of PU.1 and development of leukemia and lymphoma, but the other distal elements regulating PU.1 have not been defined. Here we show that other phylogenetically conserved elements participate in the initiation of antisense transcription, and that these antisense RNAs function as important modulators of proper dosages of PU.1. Specifically, antisense transcripts originate from specific conserved sites in introns 1 and 3, and that the intron 3 site contains binding sites for transcription factors such as AML1 and Ets factors. The conserved intron 3 element also possesses anti-sense promoter activity. These antisense transcripts are present at about 15% of PU.1 sense transcripts in PU.1 expressing cells. They negatively regulate PU.1 sense RNA, as introduction of siRNA molecules which specifically target antisense transcripts lead to 3–5 fold increases in PU.1 sense RNA and protein. Both sense and antisense PU.1 gene RNAs are dependent on the URE and are transcribed from the same chromatin architecture, in which the conserved elements, including URE and sense and antisense promoters are located in the same nuclear fraction and can be shown to exist in the same nucleoprotein complex by chromosome conformation capture (3C). We are currently testing the mechanisms involved in the formation of such complexes, and specifically whether the complexes are mediated by binding of AML1 to the URE. Since we do not observe significant differences in antisense transcript levels between PU.1 high-expressing and PU.1 low-expressing cells, we hypothesize that the function of these antisense transcripts is to modulate rather than absolutely control PU.1 levels: in PU.1 high-expressing cells, such as myeloid cells, the antisense transcripts trim PU.1 levels to prevent overexpression, while in cell types in which PU.1 is not expressed, such as T cells, the antisense transcripts prevent any expression of PU.1. We propose that such a mechanism will likely be important in fine-tuning the regulation of many genes and may be the reason for the large number of overlapping complementary transcripts with so far unknown function.