Reproducible but Dynamic Positioning of Specific Sequences in Mitotic Chromosomes

Reference:

Dietzel, S. and Belmont, A.S. (2001) "Reproducible but dynamic positioning of DNA within chromosomes during mitosis," Nature Cell Biology In press.

Background:

How DNA is folded into chromosomes is unknown. Mitotic chromosome banding demonstrates reproducibility in longitudinal compaction at several megabasepairs (Mbp) resolution. Less clear is whether DNA sequences are targeted laterally to specific locations. In vitro chromosome assembly of prokaryotic DNA suggests lack of sequence requirements for chromosome condensation, implying an absence of DNA targeting. However, protein extraction experiments suggest binding of specific sequences to a chromosome scaffold. Chromosome banding patterns, using dyes with differential sequence specificity, were interpreted as resulting from alignment of AT rich sequences within a partially helically folded chromosome scaffold. Fluorescence in situ hybridization (FISH) experiments, though, at best have shown only slight deviation from a random, lateral sequence distribution 5, perhaps due to technical limitations.

Results:

Using gene amplified chromosome regions, we have shown highly reproducible targeting of specific chromosome segments to the metaphase chromatid axis, but localization of these segments to the prophase and telophase chromosome periphery. Unfolding intermediates during anaphase and telophase suggest sequence repositioning occurs through the global uncoiling of an underlying chromatid structure.

Mitotic chromosomes stained with the DNA dye, DAPI, are in red. Lac operator tagged sequences binding GFP-lac repressor are in green. Left panel shows labeled sequences localizing to the axis of the early anaphase chromatids. Middle panel shows late anaphase plate in which GFP signal assumes more of a zig-zag pattern. Right panel shows telophase nuclei where axial location is lost and GFP spots frquently localize to chromosome periphery.

Details:

Previously we described development of an in vivo labeling method for localizing specific DNA sequences using a GFP - lac repressor fusion protein bound to a lac operator direct repeat. We combined this approach with gene amplification to investigate mitotic condensation and decondensation of labeled sequences in structurally well preserved chromosomes. The simultaneous detection of multiple GFP signals within the same chromosome arm allowed recognition of highly reproducible targeting of specific sequences within mitotic chromosomes in one of the cell clones generated.

The EP1-4 cell line, a derivative of CHO cells containing a single vector copy insertion of a plasmid containing a lac op direct repeat, was subjected to gene amplification. This was followed by transfection and selection of stable clones expressing GFP-lac repressor. Two clones were characterized.

Both clones showed localized clusters of labeled dots within interphase nuclei. Transgenes in both clones consistently located at the border of chromatin domains, but otherwise no pattern was obvious. Surprisingly, metaphase chromosomes in Ah clone cells showed a strikingly non-random alignment of labeled dots. Specifically, the GFP label was found lying along the axis of the chromatids with the diameter of the GFP signal close to the resolution limit of the light microscope and roughly half the chromatid diameter. None of the metaphase cells from the Cd clone showed this alignment. Instead, the GFP spots were distributed apparently randomly within metaphase chromosomes.

In early prophase and late anaphase/telophase spots in the Ah clone frequently localized to the chromosome periphery. In anaphase GFP staining showed changes from a linear to more curvilinear or zig-zag patterns. At later stages of decondensation in late anaphase / early telophase examples of discontinuous patches of GFP staining which nevertheless connect into a roughly co-linear or zig-zag pattern were observed; these patches located at the edges of the underlying, condensed chromatid substructure but in areas which are centrally located relative to the larger envelope of DAPI staining.

 

Conclusions:

We have described a highly reproducible targeting of sequences to the metaphase chromatid axis. Key to recognition of this targeting was our experimental approach in which a structurally nonperturbing, in vivo labeling method allowed simultaneous visualization of many gene amplified copies of a marked transgene to the chromosome axis. Independent of the underlying mechanism of the axial positioning our results demonstrate a high level of accuracy and reproducibility in the architecture of mitotic chromosomes. Moreover, we observed a dynamic redistribution of location towards the chromosome exterior in prophase and anaphase. This contradicts predictions of the original radial loop model in which sequences forming the bases of the DNA loops would be located at the central axis throughout mitosis. Rather, this redistribution demonstrates that a major structural transition occurs at a late stage of chromosome condensation between prophase and metaphase. This structural transition is consistent with all chromosome models involving coiling/uncoiling of an underlying chromatid, including coiled-radial loop and coiled-chromonema models.