Locus control region transcription plays an active role in long-range gene activation

Regulation of the human growth hormone gene family: possible role for Pit-1 in early stages of pituitary-specific expression and repression

The somatic cells of a multicellular organism contain an identical complement of genes that need to be expressed specifically and appropriately to allow the normal development and functions associated with an organism. In the eukaryotic cell nucleus, genes are packaged with nucleoprotein histones into chromatin. The human growth hormone (GH)/chorionic somatomammotropin (CS) gene family offers an excellent model to study the relationship between chromatin structure and transcription factor binding in terms of tissue-specific gene expression. The GH/CS gene family consists of five genes (GH-N, GH-V, CS-A, CS-B and CS-L), contained in a single locus on chromosome 17. Although they share approximately 94% sequence similarity, GH-N expression is restricted to pituitary somatotropes while the four placental GH/CS genes are expressed in the villus syncytiotrophoblast. Appropriate expression in vivo is dependent on remote sequences found 14-32 kb upstream of GH-N in the loci of adjacent genes, and these sequences are characterized by five (I-V) nuclease-hypersensitive sites (HS). Pituitary-specific factor Pit-1 binds at HS I/II and plays an essential role in chromatin remodeling and GH-N expression; however, the processes that lead to HS I/II accessibility are unknown. We discuss the possibility that Pit-1-driven remodeling at HS III may precede that at HS I/II in the pituitary. Also, in pituitary chromatin, all five GH/CS genes share similar nuclease sensitivity, suggesting that the conformation of the placental genes is not inhibitory to transcription. Given that the promoters of both GH-N and the placental GH/CS genes contain Pit-1-binding sites, possible mechanisms to restrict placenta GH/CS promoter activity in the pituitary are discussed, including active repression via P sequences located upstream of each of the placental GH/CS genes. Positively or negatively influencing those components known to be important for pituitary transcription may link epigenetic events to key transcription factors in the overall picture of tissue-specific control of gene expression.

Transcriptional landscape of the human and fly genomes: nonlinear and multifunctional modular model of transcriptomes

Regions of the genome not coding for proteins or not involved in cis-acting regulatory activities are frequently viewed as lacking in functional value. However, a number of recent large-scale studies have revealed significant regulated transcription of unannotated portions of a variety of plant and animal genomes, allowing a new appreciation of the widespread transcription of large portions of the genome. High-resolution mapping of the sites of transcription of the human and fly genomes has provided an alternative picture of the extent and organization of transcription and has offered insights for biological functions of some of the newly identified unannotated transcripts. Considerable portions of the unannotated transcription observed are developmental or cell-type-specific parts of protein-coding transcripts, often serving as novel, alternative 5' transcriptional start sites. These distal 5' portions are often situated at significant distances from the annotated gene and alternatively join with or ignore portions of other intervening genes to comprise novel unannotated protein-coding transcripts. These data support an interlaced model of the genome in which many regions serve multifunctional purposes and are highly modular in their utilization. This model illustrates the underappreciated organizational complexity of the genome and one of the functional roles of transcription from unannotated portions of the genome.