The MMTV LTR adopts a specific nucleoprotein organization when introduced into cells. The 1300 base pairs of the LTR are wrapped on a series of six positioned nucleosomes, providing a unique and powerful model to explore the mechanisms involved in the establishment of highly ordered chromatin structures. Experiments demonstrated that the cores are positioned at very high resolution, potentially at the nucleotide level, suggesting two general classes of models. (1) Cores could be restricted to certain positions by sequence dependent resistance to binding; this is referred to as rotational positioning. (2) Alternatively, boundary conditions could be imposed by sequence specific binding of a """"""""phasing"""""""" factor, thereby excluding cores from a certain site and setting up a phased set on either side of the boundary factor; this is generally referred to as statistical positioning. A critical test of these models was provided by establishing the LTR as a functional promoter in Saccharomyces cerevisiae, which would not contain a phasing factor unique to mouse cells. Cores were found to adopt the same positions in yeast, arguing strongly that DNA sequence is the primary determinant in positioning. A separate issue concerns the transition in nucleosome structure induced by hormone activation of transcription. A new technique was developed involving high resolution positioning of formaldehyde induced DNA-histone crosslinks by liner Taq polymerase amplification. This method permitted a base pair resolution analysis of the structural alteration in nucleosome B that occurs upon steroid receptor binding. This analysis indicated that the B nucleosome was not displaced by receptor binding, but rather adopts a new conformation that is compatible with secondary loading of other transcription factors (NF1/CTF) responsible for recruitment of the basal transcription preinitiation complex.
