Characterization of the mouse gene, human promoter and human cDNA of TSCOT reveals strong interspecies homology.

The regulation of gene expression in thymic epithelial cells is critical for T cell development. The mouse thymic epithelial gene Tscot encodes a protein with weak homology to bacterial 12 transmembrane co-transporters. Using competitive reverse transcription-polymerase chain reaction (RT-PCR), we show that low level Tscot expression is detectable in several other tissues. Tscot was mapped to chromosome 4 and was also detected in other mammalian species by Southern blotting. The human cDNA clone showed 77% amino acid identity with the mouse sequence. The highest conservation was in the TM regions and in a small segment of the central cytoplasmic loop. Genomic clones spanning 17164 bases of the Tscot gene revealed four exons with nine of the TM domains encoded in the first exon. The major transcriptional start site in mouse was identified by a primer extension analysis and confirmed by RT-PCR. Comparison of 1.7 kb of the human and mouse promoters identified six conserved possible regulatory elements, one containing a potential binding site for an interferon alpha inducible factor. Finally, as a functional test, 3 kb of the murine promoter was used to create a transgenic mouse that expresses enhanced green fluorescent protein message strongly in the thymus, weakly in the kidney and undetectably in the spleen, liver and heart.

A putative 12 transmembrane domain cotransporter expressed in thymic cortical epithelial cells.

We have isolated a full-length cDNA clone (thymic stromal origin (TSO)-1C12) from a SCID thymus library using a probe from a PCR-based subtractive library enriched for sequences from fetal thymic stromal cells. TSO-1C12 mRNA is expressed mainly in the thymic cortex and is highly enriched in SCID thymus. Expression per cell is highest during fetal thymus development and decreases after day 16. Antipeptide Abs immunoprecipitated a hydrophobic, plasma membrane glycoprotein (thymic stromal cotransporter, TSCOT) whose translated sequence has weak homology to bacterial antiporters and mammalian cation cotransporters with 12 transmembrane domains. TSCOT represents a new member of this superfamily that is highly expressed in thymic cortical epithelial cells.

Spatial, a gene expressed in thymic stromal cells, depends on three-dimensional thymus organization for its expression.

Although the importance of thymic stroma in thymopoiesis has been recognized, the underlying molecular details regarding stromal cell biology remain obscure. To study this area, we have cloned genes expressed in thymic stromal cells. Spatial is alternatively spliced to generate two mRNAs in thymus and lymph node (LN) but it is not expressed in the spleen. In mouse embryos, the short form begins expression at day 10 while the long form is not detected until day 12. Both mRNAs encode proline rich proteins and their closest homology is to homeobox and POU domain transcription factors. Spatial is not expressed in thymocytes, but it is expressed in 2-deoxyguanosine-treated day 14 fetal thymic organ culture (FTOC) and in reaggregated FTOC. These data suggest that a normal three-dimensional organization of stromal cells is required for Spatial expression. An antiserum raised against a C-terminal peptide detected proteins of 38 and 32 kDa in Western blots of total thymus proteins. In frozen thymus sections, subcapsular epithelial cells were stained with the anti-Spatial antiserum. Paracortical subcapsular cells of unknown function were also stained in the LN. Both forms of Spatial fused to the green fluorescent protein (GFP) localize to the nucleus in transfected cells.

A subtractive PCR-based cDNA library made from fetal thymic stromal cells.

We describe our initial approach to clone and characterize genes expressed preferentially in thymic stromal cells, in an attempt to generate molecular reagents to study the role of these cells in thymopoiesis and thymic function. Thymic stromal cells were prepared from fetal thymic organ cultures by treating them with 2-deoxyguanosine and depleting the remaining hematopoietic cells with anti-CD45 antibody. A cDNA library was then prepared after subtraction and amplification by PCR. The cloned inserts were sequenced and compared for homology with known genes in the data base. Unidentified cDNAs were then examined for expression in normal and SCID thymus and in a set of SV40-transformed thymic epithelial cell lines, by Northern blotting and a dot blot assay. In this report we describe the development of the library and present a general description of the genes identified from the initial 249 cDNAs sequenced. Among these, a relatively high percentage (55%) do not show any homology to previously identified genes. Several genes with a limited expression pattern were selected for further study.

A cell-specific and selective effect on transactivation by the androgen receptor.

The androgen (AR) and glucocorticoid receptors (GR) are related ligand-activated transcriptional regulators which bind the same cis-acting element and are coexpressed in a variety of cell types. Despite a shared DNA binding site, these receptors mediate diverse cellular responses. To explain this paradox, the existence of cell-specific factors that interact with, and modulate the function of, distinct receptors has been proposed. Prostate epithelial cell growth is sensitive to androgens, but is not affected by glucocorticoids, even though both AR and GR are expressed in these cells. We have recently isolated a unique panel of prostate epithelial cell lines from normal rats and have used these cell lines to examine cell-specific steroid responses. In this study, we compared the abilities of AR and GR to enhance transcription of several different reporter genes regulated by simple (i.e., noncompsite) hormone response elements (HREs) in prostate and nonprostate cell lines. The cell-specific effect occurred independently of the AR hormone binding domain and could be observed with a GAL4 fusion protein containing only the AR N-terminal regulatory domain. Gel shift analyses showed that the relative DNA binding affinity of AR for a probe containing a simple HRE was similar in prostate and nonprostate cell extracts. Presently, the only factors known to mediate steroid receptor-specific gene regulation are cJun and cFos, but there were no cell-specific differences in the functional levels of these proteins which could account for a preferential effect on AR-dependent transcription. Taken together, these results suggest that cell-specific activities exist which can preferentially modulate transcriptional transactivation by AR.

Recessive mutations in a common pathway block thymocyte apoptosis induced by multiple signals.

The glucocorticoid receptor (GR) is a ligand-regulated transcription factor that controls genes necessary to initiate glucocorticoid-induced thymocyte apoptosis. We have performed a genetic analysis of thymocyte cell death by isolating and characterizing a panel of GR+ dexamethasone-resistant mutants of the murine WEHI7.2 thymocyte cell line. These apoptosis-defective (Apt-) mutants were used to identify previously unknown early steps in the apoptotic pathway. The Apt- mutants contain nonglucocorticoid receptor, recessive mutations in genes that represent multiple complementation groups. These mutations block apoptosis induced by dexamethasone, gamma irradiation, and c-AMP treatment before the point where Bcl-2 exerts its protective effect. We propose that different signals share a common apoptotic pathway, and that the induction of apoptosis involves multiple precommitment steps that can be blocked by recessive mutations.

Elevated glutathione S-transferase gene expression is an early event during steroid-induced lymphocyte apoptosis.

Based on the finding that glutathione S-transferase Yb1 (GST) gene expression is elevated in the regressing prostate of androgen-ablated rats, we analyzed GST transcript levels during steroid-induced lymphocyte cell death. It was found that GST gene expression was induced in steroid-sensitive cells within 4 hr of dexamethasone treatment, required functional glucocorticoid receptor, and was dose-dependent with regard to hormone. GST expression was not induced in an apoptosis-defective variant that contained normal levels of functional receptor, indicating that GST up-regulation was the result of secondary events that occur during steroid-mediated apoptosis. Using the calcium ionophore A23817 to induce lymphocyte cell death, GST RNA levels were increased in both steroid-sensitive and steroid-resistant cell lines, supporting the conclusion that elevated GST expression was the result of cellular processes associated with apoptosis, rather than a direct consequence of steroid-mediated transcriptional control. The cells were also treated with dibutyryl cAMP to cause cell death; however, this mode of killing did not result in GST up-regulation. Taken together, these results suggest that GST induction in dexamethasone-treated T-lymphocytes occurs early in the steroid-regulated apoptotic pathway and that this may be a marker of calcium-stimulated cell death. Based on the known function of GST as an antioxidant defense enzyme and its transcriptional regulation by reactive oxygen intermediates, we propose that the gene product of a primary GR target gene(s) directly or indirectly effects the redox state of the cell. Thus activation of GST gene expression in apoptotic lymphocytes is likely a indicator of oxidative stress, rather than a required step in the pathway.

Transcriptional analyses of steroid-regulated gene networks.

It has been proposed that cell-specific responses to steroid action are the result of coordinate expression of steroid gene networks. Using three different cell systems, we have performed transcriptional analyses to determine if the observed hormone-induced alterations in gene expression are consistent with a limited number of potential target genes in any one cell type. Our results indicate that greater than 95% of the transcripts in dexamethasone-treated rat hepatoma (HTC), or mouse lymphoma (WEH17) cells, are similar to the mRNAs in untreated cells based on subtraction hybridization. In addition, we find that although the castration-induced expression of androgen-regulated transcripts in the rat ventral prostate (RVP) is significantly different between normal and castrated rats (19%), the majority of these mRNAs are accounted for by the over abundance of sulfated glycoprotein-2 sequences. Specifically, analysis of an RVP subtracted cDNA library revealed that sulfated glycoprotein-2 mRNA masked the presence of less abundant differentially expressed sequences, confirming that the actual size of the RVP androgen gene network is small. We conclude that steroid-mediated changes in transcription accurately reflect the expression of a few cell-specific target genes, and thus support the model of steroid gene networks. The potential to characterize key elements which determine both the time course and magnitude of cell-specific hormone responses is discussed.

Cadmium produces a delayed mitogenic response and modulates the EGF response in quiescent NRK cells.

Recent studies have shown that cadmium, at subtoxic levels, may induce a response characteristic of that elicited by a type of growth factor that supports the anchorage independent growth of cells that are not fully transformed. That is, Cd++ was found to replace transforming growth factor beta in supporting soft agar growth of NRK-49F cells. To test the extent to which Cd++ further mimics transforming growth factor beta in its effects and to establish response patterns that suggest possible molecular mechanisms of action, we have determined the effects of Cd++ and/or epidermal growth factor (EGF) on DNA synthesis in quiescent NRK-49F cells. We found that subtoxic doses of Cd++ modulate EGF-induced DNA synthesis in a dose-dependent fashion. Although Cd++ effects on early (16-24 hr) EGF-induced DNA synthesis are primarily inhibitory, later effects involve stimulation as well. Subtoxic doses of Cd++ did not stimulate DNA synthesis in quiescent cells within 24 hr of addition. At later times (40 or 64 hr), however, an increase in DNA synthesis of up to threefold was induced by 0.25 microM Cd++. This pattern of mitogenic response, involving inhibition of early growth-factor induced DNA synthesis and stimulation of late DNA synthesis, is consistent with that reported to be effected in some instances by transforming growth factor beta. Because a defined pattern of gene expression also is associated with the mitogenic responses to transforming growth factor beta, future studies at the molecular level can definitively test the degree to which Cd++ and transforming growth factor beta effects are common.

Expression of the art/trs protein of HIV and study of its role in viral envelope synthesis.

The art/trs transactivator protein of human immunodeficiency virus (HIV) was expressed in mammalian cells as a 19-kilodalton protein that was immunoreactive with sera from HIV-infected patients. Separate plasmids encoding the art/trs protein, the tat protein, or the envelope glycoprotein gp120 were used to demonstrate that both art/trs and tat are absolutely required for the synthesis of gp120 from its cognate messenger RNA. In addition, both the tat and art/trs proteins influence the level of envelope RNA. The results suggest that art/trs and tat may be ideal targets for potential anti-HIV agents in AIDS therapy.