Search

Background Recently, several transcription factors were found to possess large-scale chromatin unfolding activity; these include the VP16 acidic activation domain, BRCA1, E2F1, p53, and the glucocorticoid and estrogen steroid receptors. In these studies, proteins were fluorescently labeled and targeted to a multimerized array of DNA sequences in mammalian cultured cells, and changes in the appearance and/or size of the array were observed. This type of experiment is impeded by the low throughput of traditional microscopy. Methods We report the application of automated microscopy to provide unattended, rapid, quantitative measurements of fluorescently labeled chromosome regions. Results The automated image collection routine produced results comparable to results previously obtained by manual methods and was significantly faster. Using this approach, we identified two subdomains within the E domain of estrogen receptor α capable of inducing large-scale chromatin decondensation. Conclusions This work confirms that, like BRCA1, the activation function 2 region of the estrogen receptor has more than one distinct chromatin unfolding domain. In addition, we demonstrate the feasibility of using automated microscopy as a high-throughput screen for identifying modulators of large-scale chromatin folding. The Supplementary Material referred to in this article can be found at the CYTO Part A website (http://www.interscience.wiley.com/jpages/0196-4763/suppmat/v58A.html) © 2004 Wiley-Liss, Inc.

DNA, the hereditary material of the cell, is folded into protein-containing fibers in the nucleus of eukaryotes. The structure and location of these fibers is thought to be important in determining whether molecular machines can access the DNA for important cellular processes such as gene activation. Keywords: chromosome; chromatin; nucleus; transcription; structure; topology; organization; territory

Theoretical predictions within the Born approximation of the expected differential light scattering of circularly polarized light (CIDS) were made for the 300-A chromatin fiber, modeled as a helical array of dielectric ellipsoids. Computed CIDS values were strongly dependent on the exact geometry of the solenoid model, depending particularly on parameters relaed to the chiral nature of the fiber and the orientation of the nucleosomes within the helix, in contrast to the values of the total light scattering, which mainly probed size and shape. In particular, both a superbead model and a strict linear 110-A nucleosome filament would be predicted as giving rise to zero CIDS (in disagreement with the finite values observed). At the same time, helical models in which the normal vectors to the nucleosome faces were exactly parallel to the helical axis also yielded zero CIDS. Confirming earlier expectations, CIDS values were significantly less dependent on helical length than total light scattering. Finally, comparison of these calculated results from those extrapolated from available experimental data indicates that predicted CIDS values, based on currently accepted models of solenoid structure, are within an order of magnitude of those experimentally observed. Together, these results indicate the potential of differential light scattering measurements as a probe of chromatin higher-order structure, complementary to existing scattering measurements.