Genome exposure and regulation in mammalian cells.

Abstract

A method of measurement of exposed DNA (i.e. hypersensitive to DNase I hydrolysis) as opposed to sequestered (hydrolysis resistant) DNA in isolated nuclei of mammalian cells is described. While cell cultures exhibit some differences in behavior from day to day, the general pattern of exposed and sequestered DNA is satisfactorily reproducible and agrees with results previously obtained by other methods. The general pattern of DNA hydrolysis exhibited by all cells tested consists of a curve which at first rises sharply with increasing DNase I, and then becomes almost horizontal, indicating that roughly about half of the nuclear DNA is highly sequestered. In 4 cases where transformed cells (Raszip6, CHO, HL60 and PC12) were compared, each with its more normal homolog (3T3, and the reverse transformed versions of CHO, HL60 and PC12, achieved by dibutyryl cyclic AMP [DBcAMP], retinoic acid, and nerve growth factor [NGF] respectively), the transformed form displayed less genome exposure than the nontransformed form at every DNase I dose tested. When Ca++ was excluded from the hydrolysis medium in both the Raszip6-3T3 and the CHO-DBcAMP systems, the normal cell forms lost their increased exposure reverting to that of the transformed forms. Therefore Ca++ appears necessary for maintenance of the DNA in the more highly exposed state characteristic of the nontransformed phenotype. LiCl increases the DNA exposure of all transformed cells tested. Dextran sulfate and heparin each can increase the DNA exposure of several different cancers. Colcemid prevents the increase of exposure of CHO by DBcAMP but it must be administered before or simultaneously with the latter compound. Measurements on mouse biopsies reveal large differences in exposure in different normal tissues. Thus, the exposure from adult liver cells was greater than that of adult brain, but both fetal liver and fetal brain had significantly greater exposure than their adult counterparts. Exposure in normal human fibroblasts as revealed by in situ nick translation reveals a nuclear distribution pattern around the periphery, around the nucleoli and in punctate positions in the nuclear interior in parts of both S and G1 phases of the cell cycle. The same exposure pattern is duplicated by the pattern of DNA synthesis in S cells. It would appear that these nuclear regions represent positions of special activity. The previously proposed theory of genome regulation in mammalian cells is supported by these findings. The theory proposes that: a) gene activity requires exposure of the given locus followed by action of transcription factors on the exposed genes; b) the fiber system of the cell (cytoskeleton, nuclear fibers, and extracellular fibers) are required for normal exposure; c) active sites for gene expression and replication consist of the nuclear periphery where differentiation genes particularly are exposed; the nucleoli where at least some housekeeping genes are exposed; and possibly also punctate regions in the interior; d) noncoding sequences play a critical role in genome regulation, possibly including the transport of loci to be activated to appropriate exposure transcriptional and replicating locations. Cancer cells have lost specific differentiation gene activities, at least sometimes because of mutation of appropriate exposure genes; at least some protooncogenes and tumor suppressor genes are responsible for exposure and transport of specific differentiation gene loci to their appropriate exposure sites in the nucleus and for inducing exposure.