Chromatin structures: dissecting their mixed patterns in nuclease digests.

Four separate features could be distinguished in Fe-DNAase-1 digestions of human lymphoblast nuclei: a di-nucleosomal (2N) repeat, a mono-nucleosomal (1N) repeat, a component of "random" DNA, and triple splitting of major peaks. The random component is major, is unlikely to be completely artifactual, and is what would be expected from the face to face layering model of Subirana et. al., (1). The 2N pattern appeared to be associated with compact, metaphase-type chromatin, whereas the 1N pattern was associated with more exposed chromatin. These two modes are explained in terms of orderly back-to-back folding of zig-zag nucleofilaments, and face-to-face folding respectively. Hybridization studies indicated that the centromeric classes of repetitive DNA had the same digestion spectra as the major interspersed classes of repetitive DNA, and DNA enriched in transcriptionally active sequences. It is suggested that current coil models are all inadequate explanations of higher order chromatin packing.