Expression analysis of cloned chromosomal segments of Escherichia coli.

The novel transcription system of bacteriophage T7 was used to express Escherichia coli genes preferentially with a new low-copy-number plasmid vector, pFN476, to minimize toxic gene effects. Selected E. coli chromosomal fragments from an ordered genomic library (Y. Kohara, K. Ikiyama, and K. Isono, Cell 50:495-508, 1987) were recloned into this vector, and their genes were preferentially expressed in vivo utilizing its T7 promoter. The protein products were analyzed by two-dimensional gel electrophoresis. By using DNA sequence information, the gel migration was predicted for the protein products of open reading frames from these segments, and this information was used to identify gene products visualized as spots on two-dimensional gels. Even in the absence of DNA sequence information, this approach offers the opportunity to identify all gene products of E. coli and map their genes to within 10 kb on the E. coli genome; with sequence information, this approach can produce a definitive expression map of the E. coli genome.

Transgenic overexpression of transforming growth factor alpha bypasses the need for c-Ha-ras mutations in mouse skin tumorigenesis.

The induction of skin papillomas in mice can be divided into two different stages. Chemical initiation frequently elicits mutations in the Ha-ras gene, leading to the constitutive activation of ras. The second step, promotion, involves repetitive topical application of phorbol esters or wounding, leading to epidermal hyperproliferation and papilloma formation. We have found that overexpression of transforming growth factor alpha (TGF-alpha) in the basal epidermal layer of transgenic mice yielded papillomas directly upon wounding or 12-O-tetradecanoylphorbol-13-acetate treatment without the need for an initiator. Moreover, papillomas from TGF-alpha mice did not exhibit mutations in the Ha-ras gene. Interestingly, TGF-alpha acted synergistically with 12-O-tetradecanoylphorbol-13-acetate to enhance epidermal hyperproliferation. Our results demonstrate a central role for TGF-alpha overexpression in tumorigenesis and provide an important animal model for the study of skin tumorigenesis.

A function for keratins and a common thread among different types of epidermolysis bullosa simplex diseases.

Previously we demonstrated that transgenic mice expressing a mutant keratin in the basal layer of their stratified squamous epithelia exhibited a phenotype bearing resemblance to a subclass (Dowling Meara) of a heterogeneous group of human skin disorders known as epidermolysis bullosa simplex (EBS) (Vassar, R., P. A. Coulombe, L. Degenstein, K. Albers, E. Fuchs. 1991. Cell. 64:365-380.). The extent to which subtypes of EBS diseases might be genetically related is unknown, although they all exhibit skin blistering as a consequence of basal cell cytolysis. We have now examined transgenic mice expressing a range of keratin mutants which perturb keratin filament assembly to varying degrees. We have generated phenotypes which include most subtypes of EBS, demonstrating for the first time that at least in mice, these diseases can be generated by different mutations within a single gene. A strong correlation existed between the severity of the disease and the extent to which the keratin filament network was disrupted, implicating perturbations in keratin networks as an essential component of these diseases. Some keratin mutants elicited subtle perturbations, with no signs of the tonofilament clumping typical of Dowling-Meara EBS and our previous transgenic mice. Importantly, basal cell cytolysis still occurred, thereby uncoupling cytolysis from the generation of large, insoluble cytoplasmic protein aggregates. Moreover, cell rupture occurred in a narrowly defined subnuclear zone, and seemed to involve three factors: (a) filament perturbation, (b) the columnar shape of the basal cell, and (c) physical trauma. This work provides the best evidence to date for a structural function of a cytoplasmic intermediate filament network, namely to impart mechanical integrity to the cell in the context of its tissue.

Point mutations in human keratin 14 genes of epidermolysis bullosa simplex patients: genetic and functional analyses.

Previously we demonstrated that transgenic mice expressing mutant basal epidermal keratin genes exhibited a phenotype resembling a group of autosomal dominant human skin disorders known as epidermolysis bullosa simplex (EBS). EBS diseases affect approximately 1: 50,000 and are of unknown etiology, although all subtypes exhibit blistering arising from basal cell cytolysis. We now demonstrate that two patients with spontaneous cases of Dowling-Meara EBS have point mutations in a critical region in one (K14) of two basal keratin genes. To demonstrate function, we engineered one of these point mutations in a cloned human K14 cDNA, and showed that a K14 with an Arg-125----Cys mutation disrupted keratin network formation in transfected keratinocytes and perturbed filament assembly in vitro. Since we had previously shown that keratin network perturbation is an essential component of EBS diseases, these data suggest that the basis for the phenotype in this patient resides in this point mutation.

Gene-protein database of Escherichia coli K-12: edition 3.

The first two editions of the E. coli Gene-Protein Index were published to provide identifications of protein spots resolved by two-dimensional gel electrophoresis as the products of known genes. This third edition has been expanded to include information about genes and proteins gained directly from two-dimensional gel analysis--including information about protein spots not yet characterized genetically or biochemically--and is therefore more properly called a cellular protein database. An alpha-numeric designation has been uniquely assigned to each of the 616 polypeptide spots in the current database. To this, information is linked about the polypeptide's identification (protein name, gene name, Enzyme Commission--EC number), location on reference gels (x-y coordinates), genetics (Genbank code, DNA sequence reference), biochemistry (molecular weight, isoelectric point), and physiology (steady state level of the protein as a function of media and temperature, membership in various regulons and stimulons).

Genomically linked cellular protein databases derived from two-dimensional polyacrylamide gel electrophoresis.

In its most useful form a cellular protein database should be genomically based, because it is the genome which determines both the total number of proteins a cell can make and the particular ones that will be made under any given condition. Such a database should trace each protein back to its structural gene, and should account for every structural gene of a cell. Recent advances in molecular biology greatly facilitate the construction of such gene-protein databases. The mapping of genes of unidentified proteins resolved from total cell extracts on two-dimensional gels can now be accomplished by largely biochemical methods, without the necessity of isolating mutants or performing genetic crosses. Other techniques permit one to search gels for the product of any newly discovered gene (or open reading frame) suspected of encoding a protein. Consequently, gene-protein indices can be built independently and simultaneously from either direction--deducing the genetic map from the protein pattern, or finding the protein pattern from information encoded in the genome. A database of this sort is being constructed for the bacterium, Escherichia coli. Given the current pace of DNA nucleotide sequencing, the development of total gene-protein indices for a variety of cells can be anticipated in the near future.