In all organisms, regulated gene expression is used to specify complex patterns of development. The vertebrate beta-globin family of genes provides a rich opportunity to study fundamental properties of developmentally regulated gene expression. The human a-globin genes are organized within a 70 kb region in which the 5? to 3? order of the genes, epsilon, G-gamma, A-gamma, delta and beta, mirrors their temporal induction during development. The beta-globin locus control region (LCR) functions as a strong enhancer for all the genes. The LCR includes four erythroid specific DNase I hypersensitive sites, HS1-HS4, 10 to 60 Kb upstream of the globin genes that are the sites of interaction of a cohort of erythroid specific and ubiquitous transcription activators. We use biochemistry and cell and molecular biology approaches to understand the regulation of this human gene locus. Our experimental systems include (1) a manipulable in vivo model episomal system, (2) endogenous chromosomes in human and mouse erythroid cell lines and mouse embryonic stem cells (ES cells), and (3) human a-globin YAC transgenic mice and mice carrying endogenous globin locus alterations introduced by homologous recombination in ES cells. Over the last few years we have contributed to an understanding that the co-activator complexes recruited to the beta-globin LCR access the distant promoters both by spreading across the intervening DNA to create a domain of histone modifications and by physically interacting with the target genes, referred to as looping.? ? Recently, we have investigated insulator-enhancer antagonism. Insulators are DNA elements that are proposed to limit LCR activity by blocking any effect on non-target genes. One of our studies using replicating episomes in vivo showed that a chromatin insulator, 5?HS4 from the chicken beta globin locus, blocked formation of a domain of histone hyperacetylation when placed between the LCR HS2 globin enhancer and a globin gene. We discovered that enhancer blocking correlated with a significant depletion of nucleosomes in the core region of the insulator. Thus, one means by which insulators function is to creation an altered nucleosome structure incompatible with spreading of histone modifications across chromatin. We tested insulator function of a site analogous to chicken 5?HS4, human 5?HS5. An ectopic HS5 inserted between the LCR and the globin genes on a transgene in mice had no affect on histone acetylation in the LCR, but blocked histone acetylation, and transcription, of the downstream genes consistent with our studies in episomes. These studies support the idea that HS5 in the human globin locus has a portable enhancer blocking activity. ? ? Flanking the human globin LCR and genes are more distant DNase I hypersensitive sites that loop together and are proposed to be the boundaries of the globin transcriptional domain. We studied epigenetic modification and chromatin structure at high resolution across 400 Kb of chromosome 11 including the globin locus and upstream and downstream flanking regions. We observed distinctive signatures of histone H3K4, K36 and K9 methylation leading us to conclude that mono-methylation is a widely distributed mark on permissive chromatin and that hyper-methylated forms appear at highly transcribed genes at the expense of the mono-methyl mark, consistent with a transcription coupled mechanism of enzymatic conversion. However, H3K36me3 was strongly detected in transcribed coding as well as non-coding regions of the locus leading us to propose that in mammals H3K36me3 is a stable mark on sequences transcribed at any level. The human globin locus flanking hypersensitive sites did not correspond to transitions in histone methylation (or nuclease sensitivity) calling into question the role of the potential boundaries. At this juncture, it seems that new models of enhancer-insulator antagonism are needed in this evolving field.? ? We continue to use biochemical and genetic approaches to gain insight into how genes are activated by remote enhancers and how enhancer activity is restricted to the appropriate gene. It is increasingly clear that mutations affecting long range gene activation, chromatin remodeling and histone modifying proteins play a role in mis-regulation of genes that contribute to genetic diseases and cancers. Thus, understanding how chromatin is regulated during development and differentiation may offer insights into the treatment of numerous diseases. Although we focus on understanding regulation of the human globin locus at the molecular level, the underlying mechanisms of long range gene regulation have widespread relevance to mammalian development in health and disease.
Showing the most recent 10 out of 13 publications
