     |
|
|
|
||
| Research |
Gordon L. Hager, Ph.D.
|
 |
|
|
|
|
|
|
Laboratory of Receptor Biology and Gene Expression | |||||||||||||||||
|
||||||||||||||||||||
 |
|
|
||||||||||||||||||
|
||||||||||||||||||||
| Dr. Hager received his Ph.D. in genetics at the University of Washington in the lab of Ben Hall. He pursued postdoctoral studies with Dick Epstein at the Institut de Biologie Moleculaire in Geneva and with Dr. William Rutter at the University of California-San Francisco. He carried out the first molecular cloning of retroviruses at the NIH and also reported the first identification of steroid responsive regulatory elements. Dr. Hager is currently chief of the Laboratory of Receptor Biology and Gene Expression, where his program interests include the role of chromatin structure in gene regulation, the mechanism of steroid receptor function, and the architecture of active genes in the interphase nucleus. | ||||||||||||||||||||
|
||||||||||||||||||||
|
Research
|
||||||||||||||||||||
| Steroid Receptors, Chromatin, and Nuclear Structure
Mechanism of Nucleoprotein Remodeling by Nuclear Receptors on Reconstituted Chromatin We have used the mouse mammary tumor virus promoter (MMTV) for several years as a model system to study the impact of chromatin structure on gene activation, particularly by steroid receptors. In the current chromatin program, we are attempting to characterize the wide variety of activities that are recruited to a gene target in response to steroid receptor activation and to understand in molecular detail the nature of the chromatin remodeling events. We originally proposed that the nucleoprotein transition in MMTV involved the reorganization of one nucleosome, Nucl B, in the MMTV phased array of nucleosomes. Two subsequent findings indicated that this view was oversimplified. First, each translational setting in the nucleosome array results not from the presence of a single, uniquely positioned core, but rather from a “family” of closely grouped nucleosomes. Second, the chromatin remodeling event maps at higher resolution not to the B family, but to a more extended region of transition including all of the B family and the 3’ half of the C family. These findings indicated that the receptor-induced transition must be understood in terms of a more complex model, perhaps invoking higher order chromatin fiber structure. We have recently succeeded in reconstructing the receptor-induced chromatin remodeling event with MMTV nucleosome arrays assembled in vitro. Surprisingly, the receptor-dependent nucleoprotein transition can be precisely duplicated with these assembled arrays using highly purified glucocorticoid receptor (GR) and a HeLa extract as a source of remodeling activity. The chromatin remodeling event in vitro is dependent on ATP and maps precisely to the B and 3’ half of C nucleosomes, as is found in vivo. These results indicate that the in vitro system accurately recapitulates the in vivo transition and places us in position to study the molecular details of the transition with the much more powerful tools that are available in reconstituted systems. Steroid/Nuclear Receptor Trafficking in Living Cells In 1996 we demonstrated that intracellular distribution and trafficking by the glucocorticoid receptor could be effectively studied in living cells with fusions to the green fluorescent protein. We have now tagged essentially all of the classic steroid/nuclear receptors, including GR, PR, ER, TR, AhR, and the PPAR family, and find that the technology is generally applicable to all of the receptor groups. These advances have led to several unexpected and important findings for members of the nuclear receptor superfamily. In particular, we find that behavior with regard to subcellular localization and ligand dependent redistribution for several members of the family is quite different from the classic view. Our results indicate that the currently accepted description of steroid/nuclear receptor trafficking is seriously deficient. Trafficking by the progesterone receptor. The progesterone receptor occurs naturally in two forms, A and B. Potential roles for these forms have been the subject of much investigation, but little is understood concerning the functional significance of alternate forms. Using GFP labeled derivatives, we discovered that the A and B forms traffic differentially in living cells despite identical nuclear localization signals. These findings indicate that as yet undescribed interactions between the two receptor forms and either nuclear or cytoplasmic interacting proteins (or both) are determinative in compartmentalization, and that these interactions play a large role in distribution of the forms. This view contrasts with the view that nuclear localization signals (NLS) are the primary determinant of distribution. Trafficking and distribution by the estrogen receptor, alpha form. The estrogen receptor also occurs naturally in two forms, alpha, and the recently described beta form. With GFP labeled variants, we have shown that ER alpha is constitutively located in the nucleus in living cells, but moves within intranuclear compartments in response to ligand activation. The receptor is also differentially distributed in the nuclei of ER negative and ER positive human breast cancer cells, indicating that states of intranuclear organization change in the progression from ER(-) to ER(+) stages of the disease. These findings indicate that states of subcellular architectural organization may be important in development of hormone-independent disease. Trafficking and distribution by the thyroid receptor. The thyroid receptor has been described as a “nuclear receptor” since its original discovery and characterization; that is, it is found constitutively present in the nucleus in the presence or absence of ligand. Using the GFP technology, we examined TR distribution in living cells for the first time, and discovered the classic dogma to be incorrect. A significant portion of the receptor is found cytoplasmically localized in living cells, and moves to the nucleus with ligand treatment. Following on this observation, we have characterized in very recent work mutations of TR that disrupt interactions with corepressors and find that these mutant receptors are located completely in the cytoplasm in untreated cells, but translocate to the nucleus in a ligand-dependent fashion. These are quite unexpected results, and indicate that the classic literature with regard to TR trafficking is incorrect. Mechanism of Transcriptional Repression by the Glucocorticoid Receptor The glucocorticoid receptor, as well as other members of the nuclear receptor superfamily, will both activate and repress gene expression. Using the GnRH promoter as a model, we discovered a novel mechanism for GR function in transcriptional repression. GR does not bind to a DNA recognition site in this promoter, but rather binds indirectly to the promoter via ligand-dependent protein-protein interactions with Oct 1. Oct 1 has a noncanonical binding site in the promoter, and GR repression is dependent on this site. We showed by in vitro DNA binding experiments that purified GR will interact directly with Oct 1 when it is bound to the GnRH *Oct 1* site, but will not bind to Oct 1 on a wild-type Oct 1 binding site. Furthermore, we showed that the Oct 1 protein is in a different conformation when bound to the *Oct 1* site; that is, the DNA binding site for Oct 1 serves as an allosteric effector for the Oct 1 interaction with GR. This provides direct evidence for a tethering mechanism for GR function at this promoter and is the first demonstration in vitro of differential GR protein-protein interactions determined by the architecture of an interacting protein DNA binding site. |
||||||||||||||||||||
| This page was last updated on 5/21/2004. | ||||||||||||||||||||
|
||
|
||
|
 |
|