To determine the basis for the extraordinarily selective tissue-specific expression of PLAC1, we have now confirmed that the gene is expressed from two promoters, P1 and P2, spaced 105 Kilobases apart. Transcription from P1 leads to expression of 3 additional non-coding exons and the gene structure was revised to include these newly defined exons and the gene now spans 200 Kb. Further, there are at least five splice isoforms. We cloned both promoters from mouse and human and fused them to a Luciferase reporter gene and have defined the minimal promoter regions. In silico analysis of the minimal promoter region suggested the presence of binding sites for nuclear receptors Retinoic Acid X Receptor alpha (RXR-alpha), and Steroidogenic factor 1 (SF1)/Estrogen related receptor beta (ERR-beta) at promoter sites. We have now shown that RXR-alpha can directly interact with the binding sites along with Liver X receptor-beta (LXR-beta) as a heterodimer. In the presence of RXR-alpha, LXR-beta and SF-1/ERR-beta and their respective agonists, transcription is stimulated >10 fold. Various mutations in the binding sites lead to impaired binding of nuclear receptors and loss of stimulation. Transcription from the P2 promoter dominates in placenta compared to P1 promoter. However, in several cancer cell-lines we tested, either P1 or P2 is predominantly used. This may reflect the levels of auxiliary factors in cells. This work is in press in Placenta. We have also assessed the proteomics profile of placenta and compared it to its transcription profile. Using pre-fractionated total proteins from placenta by SDS gel-electrophoresis and subjecting the recovered size-fractionated proteins to trypsin digestion and two-dimensional micro-high pressure liquid chromatography coupled to tandem mass spectrometric (MS/MS) analysis, we identified 21,781 peptide signatures, 13,409 of which were unique and were assigned to 6,415 proteins. Using our computing resources, we curated the NCBI mouse protein NR database to collate and unify multiple protein IDs represented in the Genbank database. The recovered proteins represent all known intracellular compartments and a full range of isoelectric charge;thus the fractionation method showed no apparent bias. For 2,809 proteins matching ESTs from placenta were found in a set of 8,387 ESTs in the NCBI EST database. Mass spectrometric results provide direct evidence for expression of the remaining 3,606 proteins. Of particular interest, 1299 proteins that had been predicted solely on the basis of sequence analysis have now been substantiated as true products of translation from transcribed genes. In comparative ongoing work we have analyzed the proteomics data for mouse R1-9 ES cells and identified 9,370 expressed proteins, and for total kidney, 6,000 proteins. Analysis of the metabolic clustering of inferred proteins show that complex tissue proteomics can be analyzed in terms of metabolic and signaling pathways as well to specifying the components of functional organelles and structures. Concerning the function of the gene, we have now derived a mouse strain in which Plac1 is knocked out. Currently we are characterizing the effects of lack of Plac1 expression in placenta by a) cytological examination and fate of the embryo;b) transcriptional expression from homozygous and heterozygous mice;and c) effects of paternal versus maternal inheritance of the knock-out allele. For Foxl2, we have isolated sequences as much as 200 kb upstream of the transcription start site that contribute to its regulation, and are testing to identify which sequences and factors account for the tissue-specific expression. To identify Foxl2 binding targets and the genes that contain the binding sites, we carried out chromatin immuno-precipitation with anti-flag antibody from mouse embryonic stem cells in which Foxl2 tagged with Flag is over expressed. Parallel immuno-precipitation from mouse ovarian tissue (and a hypophysis cell-line where the gene is secondarily expressed) was also done using anti-Foxl2 antibody, and the binding site motifs were compared. The results show a correspondence of binding sites and expression profiling results for positive dependence of the expression of a cohort of genes.
