Isolation and characterization of nuclear scaffolds

The control of gene expression in eukaryotic organisms is generally considered to be multi-leveled, and one of the levels is thought to involve the effect of the structure of chromatin fibers on the access of transcription factors to DNA [ 1 ]. It has become increasingly apparent in recent years that the organization of chromatin fibers within the nucleus also has a role to play in the control of gene expression. Evidence has accumulated that chromatin fibers are organized through association with a proteinaceous network, variously referred to as the nuclear scaffold or nuclear matrix, which runs throughout the nucleus [2]. Chromatin fibers are believed to associate with the scaffold at specific, AT-rich DNA sequences that have been termed SARs (for scaffold associated regions) or MARs (for matrix associated regions).2 The chromatin fibers between the attachment sites would then form ‘loop domains’ along the chromatin fibers. Some of the evidence supporting this view comes from electron micrographs of histone depleted chromosomes in which DNA segments can be seen to form large loops (on the order of 50 to 100 kb) attached at their bases to a proteinaceous scaffold [3]. In at least some cases, these SAR-bounded physical loop domains coincide with transcriptionally active chromatin domains as assayed by DNase I sensitivity (reviewed by Zlatanova and Van Holde [4]. This has led to the idea that SARs form the boundaries of topologically isolated chromatin domains that function as units of gene regulation [5, 6].