Where does transcription start? 5'-RACE adapted to next-generation sequencing.

The variability and complexity of the transcription initiation process was examined by adapting RNA ligase-mediated rapid amplification of 5' cDNA ends (5'-RACE) to Next-Generation Sequencing (NGS). We oligo-labelled 5'-m(7)G-capped mRNA from two genes, the simple mono-exonic Beta-2-Adrenoceptor (ADRB2R)and the complex multi-exonic Glucocorticoid Receptor (GR, NR3C1), and detected a variability in TSS location that has received little attention up to now. Transcription was not initiated at a fixed TSS, but from loci of 4 to 10 adjacent nucleotides. Individual TSSs had frequencies from <0.001% to 38.5% of the total gene-specific 5' m(7)G-capped transcripts. ADRB2R used a single locus consisting of 4 adjacent TSSs. Unstimulated, the GR used a total of 358 TSSs distributed throughout 38 loci, that were principally in the 5' UTRs and were spliced using established donor and acceptor sites. Complete demethylation of the epigenetically sensitive GR promoter with 5-azacytidine induced one new locus and 127 TSSs, 12 of which were unique. We induced GR transcription with dexamethasone and Interferon-γ, adding one new locus and 185 additional TSSs distributed throughout the promoter region. In-vitro the TSS microvariability regulated mRNA translation efficiency and the relative abundance of the different GRN-terminal protein isoform levels.

Role of the 5'-untranslated regions in post-transcriptional regulation of the human glucocorticoid receptor.

GR transcripts display a remarkable heterogeneity in their 5' untranslated regions (5'UTRs). These variable 5'UTRs are encoded by a series of alternative 1st exons, and together with their associated promoters they maintain tissue-specific GR expression levels. In this study we over-expressed GR transcripts containing individual 1st exons, and assessed their effect on RNA stability, 3'-splicing, translation initiation and protein isoform production. We showed that these alternative 5'UTRs influence the predicted mRNA structure and free energy, and were associated with differential levels of functional spliced mRNA. However, the 5'UTR had little influence on the relative levels of the two principal 3' splice transcripts, GR-α and -ß. The overall mRNA length, the free energy of the transcript and the translational efficiency directly influenced total GR levels. However, individual N-terminal protein isoform levels appeared to depend upon elements within the 5'UTR. Membrane-GR specific labelling suggested that the mGR originates from transcripts containing exon 1D and possibly 1H, although the specific trafficking sequences or structures within these transcripts remain unidentified. The role of the alternative first exons and their associated 5'UTRs has now been expanded to translational control, influencing total GR levels, individual constituent isoform levels, as well as trafficking to the cell surface.

Membrane glucocorticoid receptor activation induces proteomic changes aligning with classical glucocorticoid effects.

Glucocorticoids exert rapid nongenomic effects by several mechanisms including the activation of a membrane-bound glucocorticoid receptor (mGR). Here, we report the first proteomic study on the effects of mGR activation by BSA-conjugated cortisol (Cort-BSA). A subset of target proteins in the proteomic data set was validated by Western blot and we found them responding to mGR activation by BSA-conjugated cortisol in three additional cell lines, indicating a conserved effect in cells originating from different tissues. Changes in the proteome of BSA-conjugated cortisol treated CCRF-CEM leukemia cells were associated with early and rapid pro-apoptotic, immune-modulatory and metabolic effects aligning with and possibly "priming" classical activities of the cytosolic glucocorticoid receptor (cGR). PCR arrays investigating target genes of the major signaling pathways indicated that the mGR does not exert its effects through the transcriptional activity of any of the most common kinases in these leukemic cells, but RhoA signaling emerged from our pathway analysis. All cell lines tested displayed very low levels of mGR on their surface. Highly sensitive and specific in situ proximity ligation assay visualized low numbers of mGR even in cells previously thought to be mGR negative. We obtained similar results when using three distinct anti-GR monoclonal antibodies directed against the N-terminal half of the cGR. This strongly suggests that the mGR and the cGR have a high sequence homology and most probably originate from the same gene. Furthermore, the mGR appears to reside in caveolae and its association with caveolin-1 (Cav-1) was clearly detected in two of the four cell lines investigated using double recognition proximity ligation assay. Our results indicate however that Cav-1 is not necessary for membrane localization of the GR since CCRF-CEM and Jurkat cells have a functional mGR, but did not express this caveolar protein. However, if expressed, this membrane protein dimerizes with the mGR modulating its function.

Proteomic profiling of rapid non-genomic and concomitant genomic effects of acute restraint stress on rat thymocytes.

In order to investigate rapid non-genomic effects of acute stress, rats were restrained for 15 min which was sufficient to activate the hypothalamus-pituitary-adrenal (HPA) axis but too short to induce massive genomic effects of cortisol. Subcellular fractions of thymocytes (cytosol, nucleus, membrane) were investigated using quantitative 2D DIGE with MALDI-TOF/TOF mass spectrometry. In total, 108 proteins with differential subcellular localizations were identified. The specificity of the changes induced by psychological stress was reflected by the prominent modulation of proteins involved in the HPA and sympathoadrenal medullar (SAM) axis such as HMGB1 and NHERF1. Intracellular trafficking was characterized by a dominant protein exodus from the cytosol. Real translocation was observed for 9 proteins with 6 that shuttled from the cytosol to the nucleus (HYOU1, HNRPF, HNRPC, STRAP, PSA1, PPA1) and 3 from the nucleus to the cytosol (HMGB1, NHERF1, PSMA1). Proteins showing subcellular reshuffling were largely involved in transcription and translation processes (39 of 108) with a significant enrichment of RNA splicing factors. Bioinformatics analysis revealed significant enrichment for protein kinase A and 14-3-3 signaling, probably reflecting real non-genomic effects. This is the first study investigating rapid effects of stress-induced HPA activation in vivo at the proteome level.

Transcriptional control of the glucocorticoid receptor: CpG islands, epigenetics and more.

The unique variability in the 5' region of the GR gene, with 9 alternative first exons and 13 splice variants plays a critical role in transcriptional control maintaining homeostasis of the glucocorticoid receptor (GR). This 5'm RNA heterogeneity, common to all species investigated, remains untranslated since the alternative first exons are spliced to exon 2 immediately upstream of the translation initiation codon. These alternative first exons are located either immediately upstream of the coding exons in the CpG island (exons B-H and J), or further upstream (exons 1A and 1I). The mechanisms regulating the differential usage of these first exons in different tissues and individuals, and the role of the 5' UTR in the splicing of the coding exons are still poorly understood. Here we review some of the mechanisms that have so far been identified. Data from our laboratory and others have shown that the multiple first exons represent only a first layer of complexity orchestrated probably by tissue-specific transcription factors. Modulation of alternative first exon activity by epigenetic methylation of their promoters represents a second layer of complexity at least partially controlled by perinatal programming. The alternative promoter usage also appears to affect the 3' splicing generating the different GR coding variants, GRα, GRß, and GR-P. Aberrant GR levels are associated with stress-related disorders such as depression, and affect social behaviour, mood, learning and memory. Dissecting how tissue-specific GR levels are regulated, in particular in the brain, is a first step to understand the significance of aberrant GR levels in disease and behaviour.