Faculty Bibliography

Epidermal injury results in activation of keratinocytes which produce and respond to growth factors and cytokines and become migratory. Activated keratinocytes express a specific pair of keratin proteins, K6 and K16, distinct from the keratins in the healthy epidermis. Keratinocytes can be activated, for example, by binding of the appropriate ligands to the epidermal growth factor receptor (EGFR). We have analyzed the effects of EGFR activation on keratin gene transcription by transfecting DNAs containing keratin promoters linked to a reporter gene into primary cultures of human epidermal keratinocytes in the presence or absence of EGF or transforming growth factor alpha (TGF alpha), two growth factors that activate EGFR. The activation of EGFR had no effect on the promoters of simple epithelial, basal-layer-specific, or differentiation-specific keratins. In contrast, the expression of K6 and K16 was strongly and specifically induced. A 20-bp DNA segment of the K16 gene promoter conveyed the EGF regulation, functioned in a heterologous construct, and therefore constituted an EGF-responsive element. A nuclear protein specifically bound to this element and to the analogous sequence of the K6 promoter. Thus, EGF specifically induces K6 and K16, markers of activated keratinocytes, via nuclear proteins that bind to EGF-responsive elements in the promoters of these keratin genes

The normal pattern of keratin expression in epidermis is altered in carcinomas as well as in nonmalignant diseases such as psoriasis and wound healing. Under these circumstances, the transcription of differentiation-specific keratins K1 and K10 is suppressed, whereas the activation- and hyperproliferation-associated keratins K6 and K16 are induced. Very little is known regarding transcriptional regulators involved in this switch. To investigate the nuclear factors that participate in regulation of expression of the K6 gene, we have characterized the binding sites for nuclear proteins on the promoter DNA of the K6 gene by gel retardation assays and site-specific deletion mutagenesis. We found four nuclear protein binding sites in the K6 gene promoter. Two are near the TATA box, but their ability to bind HeLa or keratinocyte nuclear extracts is independent of the TATA box-binding protein complex. The third binding site is a large palindrome. The sequences of these three sites do not correspond to any described target sequences for characterized transcriptional factors. The fourth is an AP-1 site, the target sequence for the proto-oncoproteins fos and jun. All four sites are independent of the previously characterized epidermal growth factor-responsive element, EGF-RE. These findings suggest that there may be two parallel pathways of induction of K6 transcription. One proceeds through the EGF-RE, which may be involved in nonmalignant hyperproliferation processes; the other, through the AP-1 site and the fos-jun proto-oncoproteins, may be related to induction in malignant processes

A common feature of all epithelial cells is the presence of keratin proteins that assemble into an intermediate filament cytoskeletal network. Whereas other cell types often use a specific master transcription factor to coordinate cell type-specific transcription, analysis of transcriptional regulation of keratin genes suggests that specific groupings of widely expressed transcription factors, acting on clusters of recognition elements in the promoter regions, confer epithelia-specific transcription. We define such a cluster of three sites that binds five transcription factors in the human K5 keratin gene. Within this cluster, an unusual Sp1 site binds the Sp1 transcription factor and two additional proteins. Flanking the Sp1 site are an AP2 site and another sequence, Site A; each binds a transcription factor. Similar clusters of recognition sites for the same five transcription factors have been also identified in other keratin genes. Such clusters may play a role in epithelia-specific expression of keratins

Transcription factor AP2 plays an important role in transcription of keratin genes, and it has been suggested that AP2 confers epithelial specificity. Promoters of keratin genes contain AP2 sites, usually within tight clusters of binding sites for other nuclear transcription factors. The role of AP2 was examined by in vitro gel shift analysis, AP2 binding site mutagenesis, and stable and transient transfection experiments. Nonepithelial cells, such as GM10 fibroblasts and melanocytes, neither express keratin nor become phenotypically epithelial when transfected with an AP2-expressing vector. However, in 3T3 and HeLa cells, co-transfection of an AP2-expressing vector increases the level of transcription from keratin gene promoters. This increase requires an intact AP2 binding site. Thus, the role of AP2 in keratin gene expression is quantitative rather than qualitative. AP2 interacts with other transcription factors and may convey extracellular regulatory signals to the transcription complex in the promoters of keratin genes

The promoter of human K14 keratin gene, specific for the basal layer of stratified epithelia, is regulated by nuclear receptors for retinoic acid and thyroid hormone. However, the DNA sequences responsible for this regulation have not yet been identified. To identify the retinoic acid-responsive site, we have devised a simple site-specific mutagenesis method and introduced mutations into the K14 keratin gene promoter. These mutations identify the retinoic acid-responsive site. The site consists of a cluster of consensus palindrome half-sites in various orientations. As shown previously, retinoic acid and thyroid hormone receptors can recognize and bind common sequences in regulated genes. Here, we describe mutations that abolish regulation by both receptors. Interestingly, the hormone-dependent and -independent regulatory sites of the thyroid hormone nuclear receptor can be separated. Clusters of half-sites that share structural organization with the K14 regulatory site were found in the K5 and K10 keratin gene promoters. Similar clusters may be responsible for retinoic acid-mediated transcription regulation in epidermis

Keratin K5 is expressed in the basal layer of stratified epithelia in mammals and its synthesis is regulated by hormones and vitamins such as retinoic acid. The molecular mechanisms that regulate K5 expression are not known. To initiate analysis of the protein factors that interact with the human K5 keratin gene upstream region, we have used gel-retardation and DNA-mediated cell-transfection assays. We found five DNA sites that specifically bind nuclear proteins. DNA-protein interactions at two of the sites apparently increase transcription levels, at one decrease it. The importance of the remaining two sites is, at present, unclear. In addition, the location of the retinoic acid and thyroid hormone nuclear receptor action site has been determined, and we suggest that it involves a cluster of five sites similar to the consensus recognition elements. The complex constellation of protein binding sites upstream from the K5 gene probably reflects the complex regulatory circuits that govern the expression of the K5 keratin in mammalian tissues

Among extrinsic modulators of keratinization are certain hormones and vitamins, which makes them potentially important pharmacological tools for treatment of keratinization disorders. Vitamin D3 and vitamin A, and their metabolites, promote and inhibit keratinization, respectively. We have shown that retinoic acid, via its nuclear receptor, directly suppresses the expression of the keratin genes which are markers of keratinocyte differentiation. Here we present evidence that 1,25(OH)2 vitamin D3 and its nuclear receptor do not directly regulate keratin gene expression. Co-transfection of a vector expressing the nuclear receptor for vitamin D3 with responder DNA constructs containing keratin gene promoters had no effect on the level of activity of keratin gene promoters either in the presence or in the absence of vitamin D3. We conclude that vitamin D3, unlike retinoic acid, modifies keratin synthesis indirectly, by changing the differentiation phenotype of the keratinocyte

Several methods for DNA-mediated cell transfection were tested to determine the optimal conditions for transfection of human epidermal keratinocytes. The following methods were compared: electroporation, lipofection, Ca3(PO4)2 co-precipitation, DEAE-dextran, and polybrene-mediated transfection. The transfected DNA included human keratinocyte-specific promoter for keratin K14 as well as SV40 and RSV viral promoters. Enzyme assays and in situ staining were used to evaluate both quantitative and qualitative aspects of transfection, and both subconfluent and post-confluent, stratifying keratinocytes were examined. Lipofection, Ca3(PO4)2 co-precipitation, and polybrene methods transfect very efficiently, but lipofection is expensive and Ca++ in the co-precipitation procedure induces keratinocytes to differentiate. We have found that polybrene-mediated transfection followed by a 27% DMSO shock is optimal for introducing DNA into human epidermal keratinocytes

The promoters of epidermal keratin genes, K5, K6, and K10 were cloned and their functions compared with that of the previously described promoter of the K14 keratin gene in non-epithelial and transformed epithelial cell lines, as well as in primary cultures of cells derived from simple and stratified epithelia. The four promoters were functional only in epithelial cells. Although the promoter for the basal cell-specific, acidic-type K14 gene was active in all epithelial cells tested, its basic-type partner, K5, and the promoter for the hyper-proliferation-associated K6 were active only in primary cultures of stratified epithelia. The promoter for the epidermal differentiation-specific K10 keratin gene was active at a low level in primary cultures of stratified epithelial cells on non-epidermal origin. Thus, the K14 gene promoter is functional in all epithelial cells, but the upstream regions of the K5 and K6 keratin genes restrict their expression to stratified epithelia, whereas the epidermal determinants of the K10 gene are not in the proximal upstream sequences

In the epidermis, retinoids regulate the expression of keratins, the intermediate filament proteins of epithelial cells. We have cloned the 5' regulatory regions of four human epidermal keratin genes, K#5, K#6, K#10, and K#14, and engineered constructs in which these regions drive the expression of the CAT reporter gene. By co-transfecting the constructs into epithelial cells along with the vectors expressing nuclear receptors for retinoic acid (RA) and thyroid hormone, we have demonstrated that the receptors can suppress the promoters of keratin genes. The suppression is ligand dependent; it is evident both in established cell lines and in primary cultures of epithelial cells. The three RA receptors have similar effects on keratin gene transcription. Our data indicate that the nuclear receptors for RA and thyroid hormone regulate keratin synthesis by binding to negative recognition elements in the upstream DNA sequences of the keratin genes. RA thus has a twofold effect on epidermal keratin expression: qualitatively, it regulates the regulators that effect the switch from basal cell-specific keratins to differentiation-specific ones; and quantitatively, it determines the level of keratin synthesis within the cell by direct interaction of its receptors with the keratin gene promoters