RAR beta 2-mediated growth inhibition in HeLa cells.

Retinoic acid inhibits the growth of a variety of normal and transformed cells in vitro and in vivo. How retinoic acid inhibits cell growth is poorly understood but involves interactions between the ligand and a series of nuclear and cytoplasmic receptors. The nuclear receptors for retinoic acid are of two types, the RARs and the RXRs. Each can function as a ligand-inducible transcription enhancing factor. In previous studies, we have demonstrated that an isoform of one RAR, RAR beta 2, is transcriptionally up-regulated in senescent human dermal fibroblasts and senescent human mammary epithelial cells. Moreover, we have also shown that RAR beta 2 can inhibit oncogene-induced focus formation, in primary rat embryo fibroblasts, as effectively as the tumor suppressor gene p53. Here, we extend our studies of retinoid-regulated signal transduction pathways that inhibit cell proliferation by demonstrating that HeLa cells expressing an RAR beta 2 construct are growth inhibited by greater than 50% when compared to the parent cell lines. The RAR beta 2-expressing cell lines are inhibited further by the addition of exogenous all-trans-retinoic acid. Finally, soft agar assays show that the RAR beta 2-expressing cell lines also demonstrate an inhibition of growth in soft agar, when compared to the parent growth cell lines, and are inhibited further in the presence of added all-trans-retinoic acid. These data definitively show that RAR beta 2 can inhibit cell proliferation in an established tumor cell line and provide more strength to the notion that this isoform is an effective growth inhibitor in vitro and, most likely, in vivo.

Effect of cellular senescence and retinoic acid on the expression of cellular retinoic acid binding proteins in skin fibroblasts.

Retinoic acid (RA) has significant effects on a variety of cellular processes, including growth and differentiation. Retinoic acid has also been implicated as a major morphogen during embryogenesis. In the skin, both the epidermis and the dermis are extremely responsive to the effects of retinoids; however, the molecular mechanisms through which retinoids act in this tissue remain poorly understood. Two classes of proteins play roles in mediating the biological effects of retinoic acid. The nuclear receptors for retinoic acid are of two types, the initially described RARs, and the RXRs. Each of these nuclear receptor families has multiple isoforms and can function as a ligand-inducible transcription enhancing factor. A second class of receptor proteins exists for retinoic acid that are found in the cytoplasm, the cellular retinoic acid binding proteins (CRABPs). These proteins play a role in the binding, transport, and metabolism of retinoic acid. In previous studies, we have demonstrated that retinoic acid induces the expression of RAR beta and RAR gamma in human dermal fibroblasts. Moreover, we have also shown the selective transcriptional up-regulation of the RAR beta 2 isoform in senescent dermal fibroblasts and senescent human mammary epithelial cells. In order to further define molecules important in regulating the response of senescent dermal fibroblasts to retinoids, we demonstrate here that retinoic acid induces CRABP-II messenger RNA in human dermal fibroblasts in a dose-dependent manner. Moreover, we show that the induction by RA can be inhibited by actinomycin D, suggesting that the up-regulation may be mediated by a transcriptional mechanism. We further demonstrate that cycloheximide also has an effect on the up-regulation, suggesting a role for protein synthesis in the regulation of CRABP-II gene expression. We show that CRABP-II is a very stable messenger RNA species, in contrast to the mRNAs for RAR alpha, RAR beta, and RAR gamma. Of interest, we demonstrate no significant difference in the expression of CRABP-II between presenescent and senescent fibroblasts. Taken together, these data suggest that retinoic acid plays a central role in the regulation of CRABP-II gene expression in the dermal fibroblast and that this molecule is the major mediator of the cytoplasmic effects of retinoids in dermal fibroblasts. However, in contrast to RAR beta 2, there is no apparent change in the regulation of CRABP-II in senescent dermal fibroblasts.

Cellular aging and transformation suppression: a role for retinoic acid receptor beta 2.

Cellular senescence is characterized by a finite proliferative capacity in vitro. Moreover, the proliferative capacity of dermal fibroblasts harvested from humans is inversely proportional to the age of the donor, suggesting that senescence in culture is a manifestation, at the cellular level, of processes that occur during in vivo human aging. As cellular senescence is a program that ultimately decreases cell proliferation, it has been hypothesized that the genetic mechanisms responsible for the negative growth regulation of senescence may also be involved in the suppression of neoplastic transformation. Retinoic acid (RA) and its derivatives are effective negative growth regulators and are known to inhibit tumor growth, in vitro and in vivo. As a first step in examining a role for retinoic acid in the regulation of cellular aging in human fibroblasts, we examined the expression of the nuclear receptors for RA (RAR alpha, RAR beta, and RAR gamma) in human donors of different ages. These studies demonstrate a selective up-regulation of RAR beta, in response to RA, in fibroblasts that manifest a decreased proliferative capacity. We extend these observations to show that this finding is independent of the age of the donor and correlates with the proliferative capacity of the culture as a whole. Nuclear run-on studies show that the increase in RAR beta mRNA accumulation is mediated by a striking increase in the transcription of the RAR beta 2 isoform. Senescent fibroblasts manifesting the transcriptional increase of the RAR beta 2 isoform also demonstrate transcriptional repression of the protooncogene, c-fos. Functional studies demonstrate that RAR beta 2, like the tumor suppressor gene p53, can inhibit oncogene-induced focus formation. These data provide further support for the contention that genetic events important in cellular senescence may also play a significant role in tumor suppression in humans. Moreover, these observations suggest that RA, through transcriptional regulation of RAR beta 2, may mediate aspects of the negative growth control that characterizes both states.

A beta 2RARE-LacZ transgene identifies retinoic acid-mediated transcriptional activation in distinct cutaneous sites.

Retinoic acid and its derivatives (retinoids) exert profound influences on epithelial growth differentiation in a variety of tissues, including the skin. How retinoic acid mediates these effects is not fully understood. The recent cloning of a series of nuclear receptors for retinoic acid (RARs) has demonstrated that these proteins can function as ligand-inducible transcriptional enhancing factors. Moreover, all receptors are members of the steroid/thyroid hormone multigene family. In vitro studies have demonstrated the expression of RAR alpha, RAR beta, and RAR gamma in various cell types found in the skin. While multiple isoforms exist for each of the three RARs, it is unclear where each of these receptors functions in vivo to mediate the tissue-specific effects of retinoic acid. As a first step in determining sites of retinoic acid-mediated transcriptional activation in the skin and its appendages, we developed a transgenic model in which the retinoic acid response element (RARE) of the RAR beta 2 isoform is linked to a beta-galactosidase reporter gene. Our observations consistently demonstrate that retinoic acid transcriptionally activates the beta 2RARE in distinct areas of the skin. Of interest, certain of these areas are known to contain stem cells. These data clearly demonstrate that this type of transgenic "reporter" model can be used to further define retinoic acid-regulated signal transduction pathways in the skin, as well as other complex tissues. Furthermore, these observations raise the possibility that transcriptional activation of RAR beta 2 may regulate the growth and differentiation programs of selected populations of stem cells in the skin and its appendages.

Regulation of retinoic acid receptor expression in dermal fibroblasts.

Retinoic acid (RA) is known to exert profound effects on growth and differentiation in a variety of cell types in the skin. In vitro studies have also shown that RA modulates gene expression in both fibroblasts and keratinocytes. Recently, three nuclear receptors specific for retinoic acid (RAR alpha, RAR beta, and RAR gamma) have been cloned and all are members of a large multigene family of ligand-inducible transcription enhancer factors. As a first step in defining the role of each receptor in the retinoid response of the skin, we examined the regulation of RAR alpha, RAR beta, and RAR gamma gene expression in human dermal fibroblasts by all-trans-retinoic acid. We demonstrate that human dermal fibroblasts express modest basal levels of RAR alpha and RAR gamma, but not RAR beta. When treated with 1 microM RA, the messenger RNAs for both RAR beta and RAR gamma are induced. In contrast, RAR alpha remains unchanged. The induction of RAR gamma is attenuated by the protein synthesis inhibitor, cycloheximide, while the induction of RAR beta increases slightly. Studies with actinomycin D and cycloheximide show that all three receptors have different half-lives, with RAR gamma having the longest half-life at 8 h. Gel shift analysis of known retinoic acid response elements (RAREs) in the RAR beta and RAR gamma genes demonstrates that the upregulation of these by genes by RA involves increased binding of complexes to the individual RAREs. In summary, these data demonstrate that fibroblasts express all three receptor types. Moreover, striking differences exist in the regulation of RAR gene expression in skin-derived fibroblasts and suggest that each receptor may well have a separate and discrete function in the retinoid response of the skin.

Cultured human melanocytes express the intermediate filament vimentin.

Human melanocytes are neural crest-derived cells that synthesize the pigment melanin. These cells migrate from a central location to the dermal-epidermal junction early in gestation and situate themselves between keratinocytes of the basal layer of the epidermis, extending thin dendritic processes upwards into the epidermis. In vitro, neonatal melanocytes can assume a variety of morphologies, depending on the culture conditions. Using standard immunofluorescent, immunoblotting, and Northern blotting techniques, we investigated the expression of intermediate filament proteins and demonstrate here that cultured human melanocytes express vimentin gene and protein under a variety of culture conditions. Vimentin is a 57-kD intermediate filament protein synthesized primarily by cells of mesenchymal origin. It is transcribed as a single-messenger RNA species of 2.0 kb and the human gene is located on chromosome 10. As a member of the intermediate filament family of proteins, we suggest that vimentin is an important component of the cytoskeleton of neonatal, human melanocytes.

Clonal origin of dermatofibrosarcoma protuberans.

Dermatofibrosarcoma protuberans (DFSP) is a malignant tumor originating in the dermis. Although it is known to be locally aggressive, it only rarely metastasizes and will recur unless completely excised. The exact cell responsible for the development of a DFSP has been a matter of controversy for several decades; however, most histochemical and electron microscopic studies support a fibroblastic origin, with the tumor cells staining uniformly for vimentin and containing active endoplasmic reticulum synthesizing collagen. Cytogenetic analysis of some of these tumors has demonstrated at least two specific chromosomal abnormalities in DFSP and suggested that this tumor may be polyclonal in origin. To further address the clonal origin of this locally invasive, mesenchymal tumor, we analyzed DNA from two female patients by restriction fragment length polymorphisms and methylation analysis. Our data strongly support the concept that DFSP is monoclonal in origin and that this tumor mass reflects the clonal expansion of a single cell.