Large-scale Chromatin Structure and Chromonema Fibers

References:

A.S. Belmont, Large-scale Chromatin Structure, in "Genome Structure and Function", NATO Advanced Study Institute, Kluwer Acad. Pub., 261-278 (1997)

A. S. Belmont, K. Bruce, G. Li, Three-dimensional visualization of G1 chromosomes: a folded, twisted, supercoiled chromonema model of interphase chromatid structure, J. Cell Biol. 127: 287-302 (1994)

A. Belmont, M. Braunfeld, J. Sedat, D.A. Agard, Large-scale chromatin structural domains in mitotic and interphase chromosomes in vivo and in vitro, Chromosoma 98:129-143 (1989)

A. Belmont, J. Sedat, D.A. Agard, A three dimensional approach to mitotic chromosome structure: evidence for a hierarchical organization, J. Cell Biol. 105:77-92 (1987)

Background:

A fundamental question, how chromatin is packaged into interphase chromosomes, remains largely unanswered. We have a relatively detailed understanding of DNA folding into nucleosomes and the spacing of nucleosomes. Above this there is general agreement that the great majority of the genome is folded into 30 nm diameter chromatin fibers. Exactly how this folding of nucleosomes into 30 nm chromatin fibers remains an area of active research. Models varying from a highly regular, helical folding of nucleosomes to irregular folding of nucleosomes dictated by local geometry of DNA entering and exiting nucleosome core particles and statistical variations in linker DNA length have been proposed.

What is agreed upon is that together, the folding of DNA into nucleosomes and the folding of nucleosomes into 30 nm chromatin fibers results in a roughly 40 fold linear compaction. However, in situ measurements of interphase chromatin compaction by FISH in mammalian cell nuclei has yielded values ranging from hundreds to thousands for the compaction ratio, and individual chromosomes have been observed to fold within local chromosome "territories". Therefore it is clear that there remain additional levels of chromatin folding above the 30 nm chromatin fiber which we refer to as large-scale chromatin structure.

Several technical problems have severely limited progress in visualizing large-scale chromatin structure. These include the limited spatial resolution of light microscopy, the nonspecific staining of DNA by traditional EM stains, the difficulty of visualizing 3-D volumes, as opposed to thin sections of cells, by electron microscopy, and the sensitivity of chromatin conformation to even small changes in divalent or polycations.

Results:

We took a very simple approach to overcome some of these difficulties motivated by our previous work on isolated mitotic chromosomes. To overcome the DNA staining specificity problem for electron microsocopy we worked with permeabilized cells in which the soluble nucleoplasm is removed prior to fixation, leaving behind the chromatin and other nuclear structures. To minimize perturbation of large-scale chromatin structure by the permeabilization buffer, we applied a simple constaint in surveying potential buffers: specifically, we asked which isolation buffers preserved the light microscopy appearance, after optical section deconvolution, of mitotic chromosomes or interphase nuclei after permeablization relative to their appearance by light microscopy in living cells stained with cell permeable DNA dyes. As a second constraint we also asked which buffers preserved the appearance of mitotic chromosomes and interphase nuclei after permeabilization, fixation, and embedding for electron microscopy relative to what is seen in cells fixed directly with glutaraldehyde and embedded. Here we examined mitotic and interphase large-scale chromatin structure in embedded samples by both light microscopy, using DAPI staining of plastic sections, and electron microscopy.

Our approach revealed a DNA distribution within interphase nuclei from several mammalian cell lines suggestive of an underlying fibrillar large-scale chromatin organization. Electron microscopy revealed large-scale chromatin fibers, which we termed chromonema fibers, which appeared as short fiber segments or more continuous fiber lengths, roughly 100 nm in diameter (60-130 nm depending on cell cycle position and cell type). Serial section and tomography EM reconstructions demonstrated that these structures corresponded to actual fibers, which were coiled, supercoiled, and twisted within the interphase nuclei.

Left: Interphase DNA distribution within Hela cell observed live, using Hoescht vital staining, after deconvolution. A textured pattern suggestive of an underlying fibrillar organization is observed. (Similar images from live cells have now been obtained by other groups using a GFP-histone fusion protein).

Right: Transmission electron micrograph (shown as negative) of a semi-thick, 200 nm thick Epon section through a permeabilized cell nucleus. Electron dense regions appear white. An isolation buffer which preserves, as assayed by light microscopy, the large-scale chromatin organization visualized in living cells was used. Distinct large-scale chromatin domains highly suggestive of ~100 nm fibers are observed as short segments within this section. These can be seen looping off the nuclear envelope and the nucleolus (center).

Serial section reconstructions showing that these large-scale chromatin domains represent spatially discrete fibers. Thin sections were ~40 nm thick. Top panels in A & B represent projections through 12 consecutive sections, allowing visualization of entire chromonema fibers. Panels below these represent selected individual thin sections from these reconstructions demonstrating the fiber nature of the domains. Chromonema fiber in (A) can be followed as discete fiber for more than 2 um as it loops off the edge of the nucleolus (NU). Chromonema fiber in (B) is seen "corkscrewing" towards the nuclear envelope (NE).

Computational slice through EM tomography reconstruction of G1 interphase Hela cell nucleus. Larger fibers are ~100-130 nm chromonema fibers. Smaller fibers are 30 nm fibers which are folded within the chromonema fibers. To put this image in perspective, we point out a 30 nm fiber folding from one chromonema fiber edge to the other edge and back would represent ~20-25 kb of DNA. We are therefore observing the organized folding of thousands of kb of DNA within the chromonema fibers shown in this reconstruction. Note the occasional loop of 30 nm fiber protruding from these chromonema fibers, as well as the local regions of looser or more discontinuous folding of 30 nm fibers within these large-scale structures. It is tempting to speculate that these regions might correspond to large chromatin domains more sensitive to nuclease digestion, as described by various molecular biological assays. A key question is how reproducible these variations in large-scale chromatin folding are. This requires new methods for in situ visualization of specific chromosome regions. (Large-scale chromatin structure and chromonema fibers)

Conclusions:

Our results demonstrate that within these transformed, log phase cell lines the majority of the genome is folded within distinct large-scale chromatin fibers, which represent at least one additional level of folding beyond the 30 nm chromatin fiber. We are very interested in the physiological significance of this higher level of chromatin organization. Specifically what does it mean for chromatin to be folded within these structures with respect to transcriptional activity, DNA replication, and DNA recombination. In particular, how reproducible is this structure and does this chromatin folding limit accessibility of the large macromolecular protein assemblies required for most aspects of DNA function.