As the major proteins of adult keratinocytes, keratins provide biochemical markers for exploring mouse epidermal embryogenesis. Here, we used a modified method of whole-mount in situ hybridization to track skin-specific expression of endogenous keratin mRNAs throughout embryogenesis. To monitor transcriptional regulation, we coupled this with beta-galactosidase expression of a human epidermal keratin promoter-driven transgene. These studies have radically changed our perception of how the program of gene expression becomes established during epidermal development. Specifically, we have discovered that (1) basal keratin (K5 and K14) genes are first detected at E9.5 in a highly regional fashion, and surprisingly as early as the single layered ectodermal stage; (2) the early patterns do not correlate with morphogenesis per se, but rather with regional variations in the embryonic origin of underlying mesenchyme, supporting morphogenetic criteria that early inductive cues are mesenchymal; (3) epidermal keratin genes are expressed in periderm, supporting the notion that this layer arises from ectodermal stratification, even though it is simple epithelial-like in morphology and is subsequently sloughed during development; (4) later embryonic patterns of K5 and K14 gene expression parallel proliferative capacity and not stratification; and (5) K1 and K10 mRNAs are first detected as early as E13.5, and their patterns correlate with differentiation and not stratification. These patterns of epidermal gene expression led us to explore whether potential transcriptional regulators of these genes are expressed similarly. We show that AP2 (but not Sp1) cRNAs hybridize in a pattern similar to, but preceding that of basal keratin cRNAs. Finally, using gene expression in cultured cells, we demonstrate that AP2 has a strong inductive effect on basal keratin expression in a cellular environment that does not normally possess AP2 activity.

Differentiation of skeletal muscle involves withdrawal of myoblasts from the cell cycle, fusion to form myotubes, and the coordinate expression of a variety of muscle-specific gene products. Fibroblast growth factor and type beta transforming growth factor specifically inhibit myogenesis; however, the transmembrane signaling pathways responsible for suppression of differentiation by these growth factors remain elusive. Because ras proteins have been implicated in the transduction of growth factor signals across the plasma membrane, we used DNA-mediated gene transfer to investigate the potential involvement of this family of regulatory proteins in the control of myogenesis. Transfection of the mouse skeletal muscle cell line C2 with the oncogenic forms of H-ras or N-ras completely suppressed both myoblast fusion and induction of the muscle-specific gene products nicotinic acetylcholine receptor and creatine kinase. Inhibition of differentiation by activated ras genes occurred at the level of muscle-specific mRNA accumulation. In contrast, proto-oncogenic forms of N-ras or H-ras had no apparent effects on the ability of C2 cells to differentiate. Myoblasts transfected with activated ras genes exhibited normal growth properties and ceased proliferating in the absence of mitogens, indicating that ras inhibited differentiation through a mechanism independent of cell proliferation. These results demonstrate that activated ras gene products mimic the inhibitory effects of fibroblast growth factor and type beta transforming growth factor on myogenic differentiation and suggest that each of these regulators of myogenesis may operate through a common intracellular pathway.

Retinoic acid (RA), a developmental morphogen, causes activation of a transcript of an endogenous retrovirus-related element in the human teratocarcinoma-derived cell line PA-1. This provirus is defective, and the provirus-related sequences exist as multicopy elements (more than 20 copies) in human DNA. This is the first human endogenous retroviral mRNA that is known to be transcriptionally activated by RA. The nucleotide sequence of the 3,357 bp of this viral cDNA was determined and shows a strong homology to the type C-related human endogenous retroviral proviruses ERV3 and 4-1. This cDNA contains 'R-U5-Apol-env-U3-R sequences of the provirus. Adjacent to the putative 5' long terminal repeat of this provirus there is an 18-bp sequence complementary to the 3' end of isoleucine tRNA. We named this RA-responsive virus RRHERV-I.

Microarrays are at the center of a revolution in biotechnology, allowing researchers to screen tens of thousands of genes simultaneously. Typically, they have been used in exploratory research to help formulate hypotheses. In most cases, this phase is followed by a more focused, hypothesis-driven stage in which certain specific biological processes and pathways are thought to be involved. Since a single biological process can still involve hundreds of genes, microarrays are still the preferred approach as proven by the availability of focused arrays from several manufacturers. Because focused arrays from different manufacturers use different sets of genes, each array will represent any given regulatory pathway to a different extent. We argue that a functional analysis of the arrays available should be the most important criterion used in the array selection. We developed Onto-Compare as a database that can provide this functionality, based on the Gene Ontology Consortium nomenclature. We used this tool to compare several arrays focused on apoptosis, oncogenes, and tumor suppressors. We considered arrays from BD Biosciences Clontech, PerkinElmer, Sigma-Genosys, and SuperArray. We showed that among the oncogene arrays, the PerkinElmer MICROMAX oncogene microarray has a better representation of oncogenesis, protein phosphorylation, and negative control of cell proliferation. The comparison of the apoptosis arrays showed that most apoptosis-related biological processes are equally well represented on the arrays considered. However, functional categories such as immune response, cell-cell signaling, cell-surface receptor linked signal transduction, and interleukins are better represented on the Sigma-Genosys Panorama human apoptosis array. At the same time, processes such as cell cycle control, oncogenesis, and negative control of cell proliferation are better represented on the BD Biosciences Clontech Atlas Select human apoptosis array.