To visualize binding of the glucocorticoid receptor to its DNA target in living cells, we are using a specially constructed cell line, 3617 cells, characterized by two key features. First, 3617 cells are stably transformed with a GFP-tagged glucocorticoid receptor (GFP-GR), and second the 3617 genome contains an ~200-fold tandem array of the mouse mammary tumor virus (MMTV) promoter. Since GR binds to the MMTV, we should be able to use fluorescence light microscopy to visualize GFP-GR binding to the MMTV tandem array in the 3617 cells. We have now demonstrated that this can be done. When 3617 cells are stimulated with hormone, GFP-GR moves into the nucleus. In addition to a punctate staining pattern seen in control cells lacking an array, a significant percentage of 3617 cells also exhibit a GFP-GR staining pattern indicative of the array. The array appears as either a large, amorphous spot or a more elongated linear structure. Such structures are in fact the array, as demonstrated by fluorescence in-situ hybridization (FISH). This technique reveals RNA transcript accumulation. In 3617 cells, FISH demonstrates that the RNA transcripts known to be regulated by the MMTV promoter colocalize uniquely with the array structures visualized by GFP-GR. The ability to visualize GR binding in living cells now paves the way for an in vivo analysis of steroid receptor function by a variety of light microscopy methods. Two aspects of the GFP-GR array are being studied. First, we are using FRAP microscopy (fluorescence recovery after photobleaching) to study exchange rates of molecules at the MMTV template. We have shown that GFP-GR exchanges rapidly at this site in live cells, and present studies seek to understand the mechanism of this rapid exchange. This is being done by studying exchange rates for GR in the presence of receptor antagonists and for GR mutants. Second, we are examining structural changes in the array over time. We find by time lapse microscopy that hormone induces decondensation followed by recondensation of the array. We have also shown using RNA FISH that decondensed arrays produce more transcript. Thus, the observed decondensation and recondensation of the array likely reflect the activation and subsequent down regulation of the MMTV known to occur from previous run on transcription studies. We are beginning to investigate whether the deondensed states observed for the array are characteristics of other regions of chromatin where a set of contiguous genes are simultaneously activated.
Showing the most recent 10 out of 15 publications
