A cis-acting control region is required exclusively for the tissue-specific imprinting of Gnas.

Genomic imprinting brings about allele-specific silencing according to parental origin. Silencing is controlled by cis-acting regulatory regions that are differentially marked during gametogenesis and can act over hundreds of kilobases to silence many genes. Two candidate imprinting control regions (ICRs) have been identified at the compact imprinted Gnas cluster on distal mouse chromosome 2, one at exon 1A upstream of Gnas itself and one covering the promoters for Gnasxl and the antisense Nespas (ref. 8). This imprinted cluster is complex, containing biallelic, maternally and paternally expressed transcripts that share exons. Gnas itself is mainly biallelically expressed but is weakly paternally repressed in specific tissues. Here we show that a paternally derived targeted deletion of the germline differentially methylated region at exon 1A abolishes tissue-specific imprinting of Gnas. This rescues the abnormal phenotype of mice with a maternally derived Gnas mutation. Imprinting of alternative transcripts, Nesp, Gnasxl and Nespas (ref. 13), in the cluster is unaffected. The results establish that the differentially methylated region at exon 1A contains an imprinting control element that specifically regulates Gnas and comprises a characterized ICR for a gene that is only weakly imprinted in a minority of tissues. There must be a second ICR regulating the alternative transcripts.

Analysis of genes from inner ear developmental-stage cDNA subtraction reveals molecular regionalization of the otic capsule.

Although the gross embryology of inner ear development has been documented for several different vertebrate species at a descriptive level, our understanding of the molecular mechanisms involved remains rudimentary. Therefore, we have used cDNA subtraction and normalization procedures to define genes upregulated in the 13.5dpc mouse inner ear, a developmental stage where inner ear morphogenesis and tissue remodeling is active and differentiation of future hair cells is being initiated. We recovered 33 different genes from this subtraction and using gene-specific primers have confirmed the transcriptional upregulation of 26 of these in the 13.5dpc inner ear. Northern analyses were used to investigate splicing differences between the inner ear and the whole embryo at 13.5dpc. Spatial localization of expression was determined through whole-ear in situ hybridization analysis, and selected genes were analyzed in more detail through in situ hybridization of tissue sections. These data illustrate that the genes isolated in this study are expressed in the developing otic capsule and/or neuroepithelium. Furthermore, the expression patterns also reveal molecular heterogeneity in the developing capsule and indicate that for some genes, the chondrogenic otic capsule is composed of distinct domains of gene expression.

Identification and analysis of genes from the mouse otic vesicle and their association with developmental subprocesses through in situ hybridization.

The otic vesicle (otocyst) occupies a pivotal position in inner ear development, bridging the gap between otic placode determination, and morphogenesis of vestibular and auditory compartments. The molecular mechanisms underlying the progressive subdivision of the developing inner ear into different compartments, and the molecular control and execution of the different developmental processes involved, are largely unknown. Since relatively few genes have been implicated in these processes, we have undertaken this study to identify genes involved in these early embryonic stages. We have used cDNA subtractions of mouse otic vesicle against adult liver cDNA, and describe a set of 280 candidate genes. We have also performed otic vesicle RNA hybridizations against DNA chips to not only confirm the efficacy of the library approach, but also to investigate the utility of DNA array alternatives. To begin to dissect potential developmental roles, we investigated the spatial pattern of gene expression for a selected set of 80 genes in developing mouse embryos at mid-gestation by whole-mount in situ hybridization. These data illustrate the compartmentalisation of gene expression in the otic vesicle for the majority of genes tested, and furthermore, implicate many of the genes tested with distinct developmental subprocesses.

Determination of downstream targets of FGF signalling using gene trap and cDNA subtractive approaches.

Signalling through the fibroblast growth factor family (FGF) of ligands is essential for normal mammalian embryonic development. At a cellular level, many details of the molecular basis of the signal transduction process have been uncovered, but our knowledge of the identity of the downstream effectors of the FGF signal in the developing embryo remains limited. We have used two independent approaches to begin to identify downstream targets of FGF signalling in the embryo: (1). a gene trap approach and (2). cDNA subtraction, using mouse embryonic stem (ES) cells as a cellular system representative of an early window on the developing embryo. Both approaches led to the identification of a number of targets of FGF signalling, and we provide data to show that the chaperone Mrj, the tumour antigen Tum, collapsin mediator response protein Crmp, a novel transcriptional repressor Nac1 and ribophorin are all differentially regulated following FGF signalling. Independent gene trapping of Mrj previously indicated a role for the gene in embryogenesis [Development 126 (1999) 1247], and we present transcript data implicating a number of the newly isolated FGF target genes in different embryonic processes.

Sonic hedgehog signaling from the urethral epithelium controls external genital development.

External genital development begins with formation of paired genital swellings, which develop into the genital tubercle. Proximodistal outgrowth and axial patterning of the genital tubercle are coordinated to give rise to the penis or clitoris. The genital tubercle consists of lateral plate mesoderm, surface ectoderm, and endodermal urethral epithelium derived from the urogenital sinus. We have investigated the molecular control of external genital development in the mouse embryo. Previous work has shown that the genital tubercle has polarizing activity, but the precise location of this activity within the tubercle is unknown. We reasoned that if the tubercle itself is patterned by a specialized signaling region, then polarizing activity may be restricted to a subset of cells. Transplantation of urethral epithelium, but not genital mesenchyme, to chick limbs results in mirror-image duplication of the digits. Moreover, when grafted to chick limbs, the urethral plate orchestrates morphogenetic movements normally associated with external genital development. Signaling activity is therefore restricted to urethral plate cells. Before and during normal genital tubercle outgrowth, urethral plate epithelium expresses Sonic hedgehog (Shh). In mice with a targeted deletion of Shh, external genitalia are absent. Genital swellings are initiated, but outgrowth is not maintained. In the absence of Shh signaling, Fgf8, Bmp2, Bmp4, Fgf10, and Wnt5a are downregulated, and apoptosis is enhanced in the genitalia. These results identify the urethral epithelium as a signaling center of the genital tubercle, and demonstrate that Shh from the urethral epithelium is required for outgrowth, patterning, and cell survival in the developing external genitalia.

Regulatory analysis of the mouse Fgf3 gene: control of embryonic expression patterns and dependence upon sonic hedgehog (Shh) signalling.

Fgf3 displays a dynamic and complex expression pattern during mouse embryogenesis. To address the molecular mechanisms underlying Fgf3 expression, we used a transgenic approach to assay genomic regions from the mouse Fgf3 gene for regulatory activity. We identified an enhancer that mediates major components of embryonic expression, governing expression in the midbrain, hindbrain, surface ectoderm, dorsal roots and dorsal root ganglia (DRG), proximal sensory ganglia, and the developing central nervous system (CNS). Deletional analysis of the enhancer further delimited this regulatory activity to a 5.7-kb fragment. We have also revealed sonic hedgehog (Shh) -dependent and Shh-independent aspects of Fgf3 expression through breeding the Fgf3 reporter transgene into Shh mutants. In the absence of Shh signalling, Fgf3 reporter expression is lost in the ventral CNS, DRG, and superior cervical nerves, whereas activation of reporter expression in cranial ganglion cells is Shh independent. Moreover, detailed re-examination of the Shh phenotype revealed that Shh signalling is required for the correct development/maturation of the DRG.