Mitotic crossing-over and segregation in man.

Abstract

Although clear genetic evidence of mitotic crossing-over is lacking in man, observations of mitotic chiasmata in normal cells (0.1–1 per 1000) and in Bloom's syndrome (BS) cells (5–150 per 1000) demonstrate its occurrence. That mitotic chiasmata are true exchanges is concluded from the occurrence of heteromorphic “bivalents” and the pattern of sister chromatid exchanges in mitotic “bivalents”. Several observations demonstrate that chiasmata are different in principle from chromatid translocations which simply happen to take place at homologous loci. For example, the ratio of adjacent exchanges to mitotic chiasmata is 1/20–1/60, whereas this ratio is approximately 1:1 for chromatid translocations. Furthermore, mitotic chiasmata make up a very high proportion of total quadriradials (QRs): 48% in normal untreated cells and 90% in BS cells. Close proximity of homologous chromosomes promotes mitotic crossing-over. Thus in normal diplochromosomes, the incidence is increased a hundred-fold as compared to diploid cells. However, closeness of homologues is not the only factor promoting crossing-over; the BS gene specifically promotes exchanges between homologous segments as shown by the roughly 15-fold increase of chiasmata in BS diplochromosomes as compared to normal diplochromosomes. Mitotic chiasmata are distributed extremely nonrandomly in different chromosomes and chromosome segments. The preferred sites are short Q-dark regions, 3p21, 6p21, 11q13, 12q13, 17q12, and 19p13 or q13 being veritable hot spots. Our preferred hypothesis is that the hot spots have higher gene densities than other regions. Consequently they are active and extended in interphase which would promote their pairing and chiasma formation. Segregation after mitotic corssing-over in satellite stalks can be demonstrated by means of distinct satellites. In a BS patient there were 31 different patterns for Q-bright satellites in 58 cells. Segregation after presumed crossing-over has also been seen in three dicentric chromosomes with one centromere inactivated. Recombination in satellite stalks in BS resulted in 12/58 cells homozygous for Q-bright satellites. In two of these cells, two chromosomes were homozygous for Q-bright satellites, and in one cell, three chromosomes were homozygous. This high degree of homozygosity which obviously applies to other chromosome regions too, may explain the high incidence of malignant disease in BS on the assumption that cancer is caused by recessive genes.