A negative regulatory element in the rabbit 3'IgH chromosomal region

Mouse and human IgH loci contain several 3'IgH enhancers. In rabbit, a single hs1,2 enhancer is located 3' of the distal germ line Calpha gene, Calpha13. We searched for additional regulatory elements in this region by using a luciferase reporter assay and nucleotide sequence analysis. Within 8 kb 3' of Calpha13, we identified a 1-kb fragment that negatively regulated the hs1,2 enhancement of the Ialpha promoter. This negative regulatory element, Calpha-NRE, contains a conserved 300-bp region that is associated with 8 of the 13 germ line Calpha genes. This conserved region contains an E box that, by electrophoretic mobility shift assay, binds an E47-like protein. At the 5' end, Calpha-NRE also includes a 270-bp region with 20-bp repeats nearly identical to those 3' of mouse and human Calpha genes, and these repeats bind unidentified nuclear protein(s). Calpha-NRE appears to be a novel regulatory element that may contribute to the regulation of IgH gene expression.     

Human and pig SRY 5' flanking sequences can direct reporter transgene expression to the genital ridge and to migrating neural crest cells.

Mechanisms for sex determination vary greatly between animal groups, and include chromosome dosage and haploid-diploid mechanisms as seen in insects, temperature and environmental cues as seen in fish and reptiles, and gene-based mechanisms as seen in birds and mammals. In eutherian mammals, sex determination is genetic, and SRY is the Y chromosome located gene representing the dominant testes determining factor. How SRY took over this function from ancestral mechanisms is not known, nor is it known what those ancestral mechanisms were. What is known is that SRY is haploid and thus poorly protected from mutations, and consequently is poorly conserved between mammalian species. To functionally compare SRY promoter sequences, we have generated transgenic mice with fluorescent reporter genes under the control of various lengths of human and pig SRY 5' flanking sequences. Human SRY 5' flanking sequences (5 Kb) supported reporter transgene expression within the genital ridge of male embryos at the time of sex determination and also supported expression within migrating truncal neural crest cells of both male and female embryos. The 4.6 Kb of pig SRY 5' flanking sequences supported reporter transgene expression within the male genital ridge but not within the neural crest; however, 2.6 Kb and 1.6 Kb of pig SRY 5' flanking sequences retained male genital ridge expression and now supported extensive expression within cells of the neural crest in embryos of both sexes. When 2 Kb of mouse SRY 5' flanking sequences (-3 to -1 Kb) were placed in front of the 1.6 Kb of pig SRY 5' flanking sequences and this transgene was introduced into mice, reporter transgene expression within the male genital ridge was retained but neural crest expression was lost. These observations suggest that SRY 5' flanking sequences from at least two mammalian species contain elements that can support transgene expression within cells of the migrating neural crest and that additional SRY 5' flanking sequences can extinguish this expression.     

Conserved RARE localization in amphioxus Hox clusters and implications for Hox code evolution in the vertebrate neural crest.

The Hox code in the neural crest cells plays an important role in the development of the complex craniofacial structures that are characteristic of vertebrates. Previously, 3' AmphiHox1 flanking region has been shown to drive gene expression in neural tubes and neural crest cells in a retinoic acid (RA)-dependent manner. In the present study, we found that the DR5-type RA response elements located at the 3' AmphiHox1 flanking region of Branchiostoma floridae are necessary and sufficient to express reporter genes in both the neural tube and neural crest cells of chick embryos, specifically at the post-otic level. The DR5 at the 3' flanking region of chick Hoxb1 is also capable of driving the same expression in chick embryos. We found that AmphiHox3 possesses a DR5-type RARE in its 5' flanking region, and this drives an expression pattern similar to the RARE element found in the 3' flanking region of AmphiHox1. Therefore, the location of these DR5-type RAREs is conserved in amphioxus and vertebrate Hox clusters. Our findings demonstrate that conserved RAREs mediate RA-dependent regulation of Hox genes in amphioxus and vertebrates, and in vertebrates this drives expression of Hox genes in both neural crest and neural tube. This suggests that Hox expression in vertebrate neural crest cells has evolved via the co-option of a pre-existing regulatory pathway that primitively regulated neural tube (and possibly epidermal) Hox expression.     

Comparative analysis of human and mouse 3' Igh regulatory regions identifies distinctive structural features

Immunoglobulin heavy chain (Igh) locus rearrangements are controlled in part by an approximately 30 b complex 3' regulatory region located 3' of C alpha: this region contains several enhancers. We report here the comparison of the genomic sequences of the 3' regulatory region and further downstream sequences from mouse, rat, human and chimpanzee. Only short segments of homology were detected in the 3' regulatory region, and these were located in the vicinity of the known 3' enhancers. The nearest highly conserved segment is the nearest non-Igh gene, hole, which is located approximately 62 kb downstream of mouse C alpha. Analysis of murine 3' Igh sequences by single nucleotide polymorphism (SNP) and restriction fragment length polymorphism (RFLP) detected a transition region (high to low SNP or RFLP density) approximately 120 kb downstream of mouse C alpha. Although there is only limited sequence identity between rodent and primate 3' Igh regulatory regions, all of these regulatory regions contain a palindrome and locally repetitive elements. Locally repetitive elements in primates comprise blocks of "switch-like" sequences that differ from the families of inverted and tandem repeats that are present in rodents. We propose that together with enhancers, these "conserved" structural features are essential for the activity of the 3' Igh regulatory region in vivo.     

Th2-specific chromatin remodeling and enhancer activity in the Th2 cytokine locus control region

We recently identified a 3' region of the rad50 gene possessing strong enhancer activity as well as activity consistent with function as a locus control region (LCR) for the flanking Th2 cytokine genes. In this study, we identify several functional elements within this region by examining chromatin changes as well as activity in transgenic mice. We find within this region four DNase I hypersensitive clusters, three of which are highly conserved and predominantly expressed in Th2 cells. Histone acetylation of this region is elevated in Th2 cells. Further, one of the hypersensitive sites (RHS7) is rapidly demethylated in Th2, but not Th1, cells. In transgenic mice, these hypersensitive sites impart strong, Th2-specific enhancer activity as well as copy number-dependent expression of the reporter gene, recapitulating LCR function. We postulate that these sites function alone or in combination with other regulatory elements to coordinate gene expression in the Th2 cytokine locus.     

Long-range conserved non-coding SHOX sequences regulate expression in developing chicken limb and are associated with short stature phenotypes in human patients

Defects in long-range regulatory elements have recently emerged as previously underestimated factors in the genesis of human congenital disorders. Léri-Weill dyschondrosteosis is a dominant skeletal malformation syndrome caused by mutations in the short stature homeobox gene SHOX. We have analysed four families with Léri-Weill dyschondrosteosis with deletions in the pseudoautosomal region but still with an intact SHOX coding region. Using fluorescence in situ hybridization and single nucleotide polymorphism studies, we identified an interval of approximately 200 kb that was deleted in all tested affected family members but retained in the unaffected members and in 100 control individuals. Comparative genomic analysis of this interval revealed eight highly conserved non-genic elements between 48 and 215 kb downstream of the SHOX gene. As mice do not have a Shox gene, we analysed the enhancer potential in chicken embryos using a green fluorescent protein reporter construct driven by the beta-globin promoter, by in ovo electroporation of the limb bud. We observed cis-regulatory activity in three of the eight non-genic elements in the developing limbs arguing for an extensive control region of this gene. These findings are consistent with the idea that the deleted region in the affected families contains several distinct elements that regulate Shox expression in the developing limb. Furthermore, the deletion of these elements in humans generates a phenotype apparently undistinguishable to those patients identified with mutations in the SHOX coding region and, for the first time, demonstrates the potential of an in vivo assay in chicken to monitor putative enhancer activity in relation to human disease.     

Evidence for a complex regulatory array in the first intron of the human adenosine deaminase gene

Adenosine deaminase (ADA) is expressed ubiquitously by diverse mammalian cells and tissues but at levels that vary according to tissue and species. In humans, the thymus exhibits levels of the enzyme up to 100-fold higher than most other tissues. Using transgenic mice, we identified human ADA gene regulatory domains. Up to 3.7 kb of 5'-flanking and first exon DNA from the human ADA gene failed to promote the expression of a chloramphenicol acetyl transferase (CAT) reporter gene in an efficient, reproducible, or tissue-appropriate manner in transgenic mice. However, when 12.8 kb of DNA from the first intron of the human ADA gene was placed 3' of CAT-coding and -processing sequences, transgenic mice reproducibly expressed CAT activity in most tissues, with profoundly high levels in the thymus. DNase I hypersensitivity studies demonstrated that among transgenic mouse tissues, human thymus, and a variety of human cell lines, a region of the intron 4-10 kb downstream of the first exon exhibited an array of hypersensitive sites that varied according to tissue and cell type. Deletion of this region from the gene construction eliminated high-level expression in transgenic mice. In transfection-transient expression assays, the 12.8-kb intron fragment exhibited enhancer activity in several cell types. A 1.3-kb fragment encompassing two of the hypersensitive sites exhibited some of these activities. The results of these studies suggest that the diverse pattern of human ADA gene expression is determined, in part, by a cluster of cis-regulatory elements contained within its large first intron.