page 237, Cornforth et al. address this question, which has been debated vigorously for nearly a century, and conclude that the distribution of chromosomes within the nucleus of human cells is mostly, but not entirely, random. The work provides the most comprehensive analysis to date of chromosome–chromosome spatial associations in human interphase nuclei, and helps to explain why earlier studies sometimes reached conflicting conclusions.
Previous chromosome painting studies using fluorescence in-situ hybridization (FISH) demonstrated that each chromosome generally occupies its own space during interphase, but this did not clarify whether or not particular pairs of chromosomes tended to remain close together. Studies that examined just a few chromosomes at a time sometimes gave seemingly conflicting answers. In the new work, the authors exposed human cells to ionizing radiation to produce chromosome breaks, and used 24-color whole-chromosome painting to examine all possible interchanges between heterologous chromosomes after the breaks rejoined. The frequency of interchanges between two chromosomes should indicate whether or not they were in close proximity.
The large number of chromosomal interchanges in the system strengthened the statistical analysis, enabling the authors to identify very small deviations from spatial randomness. Among all 22 autosomes, most of the deviation from randomness is explained by a single cluster of five chromosomes, with the other chromosomes distributed randomly in the interphase nucleus. The results raise the possibility that this five-chromosome cluster, which has been observed previously, might be functionally significant.
In addition to this cluster, the data show some evidence of other spatial associations that have been suggested by earlier work, but that fail to reach statistical significance in the new work. One possibility is that specific parts of chromosomes might interact without requiring the entire chromosomes to be in close proximity. ▪