Search

Heterochromatin in Drosophila has unusual genetic, cytological and molecular properties. Highly repeated DNA sequences (satellites) are the principal component of heterochromatin. Using probes from cloned satellites, we have constructed a chromosome map of 10 highly repeated, simple DNA sequences in heterochromatin of mitotic chromosomes of Drosophila melanogaster. Despite extensive sequence homology among some satellites, chromosomal locations could be distinguished by stringent in situ hybridizations for each satellite. Only two of the localizations previously determined using gradient-purified bulk satellite probes are correct. Eight new satellite localizations are presented, providing a megabase-level chromosome map of one-quarter of the genome. Five major satellites each exhibit a multi-chromosome distribution, and five minor satellites hybridize to single sites on the Y chromosome. Satellites closely related in sequence are often located near one another on the same chromosome. About 80% of Y chromosome DNA is composed of nine simple repeated sequences, in particular (AAGAC)n (8 Mb), (AAGAG)n (7 Mb) and (AATAT)n (6 Mb). Similarly, more than 70% of the DNA in chromosome 2 heterochromatin is composed of five simple repeated sequences. We have also generated a high resolution map of satellites in chromosome 2 heterochromatin, using a series of translocation chromosomes whose breakpoints in heterochromatin were ordered by N-banding. Finally, staining and banding patterns of heterochromatic regions are correlated with the locations of specific repeated DNA sequences. The basis for the cytochemical heterogeneity in banding appears to depend exclusively on the different satellite DNAs present in heterochromatin.

The X and Y chromosomes of Drosophila melanogaster each contain a cluster of several hundred ribosomal RNA genes (rDNA). A nontranscribed spacer region separates adjacent rRNA genes and contains tandem copies of 240 bp repeats that include the initiation site for RNA polymerase I transcription. We show here that Drosophila simulans, a sibling species of D. melanogaster, contains few, if any, rRNA genes on its Y chromosome but carries instead a large block (3,000 kb or 12,500 copies) of 240 bp nontranscribed spacer repeats. The repeats are located at the tip of the long arm of the simulans Y chromosome, in contrast to their location among rRNA genes on the short arm of the Y chromosome of D. melanogaster. The bobbed mutation in homozygous females of D. melanogaster shortens and thins the bristles, owing to a partial deletion of rRNA genes on the X chromosome. The bristles of bobbed/Y males are normal owing to the presence of a full complement of rRNA genes on the Y chromosome. Peculiarly, in bobbed/Y males of D. simulans the short bristle phenotype does not return to normal but is enhanced by the presence of the Y chromosome. We propose that the 12,500 nontranscribed spacer repeats on the Y chromosome are responsible for this biological effect by competition for a protein factor(s) essential for normal levels of rDNA transcription at the X-linked locus.

The high variability of chromomeric patterns in near-terminal regions of polytene chromosome arms has been explored in a number of races, strains and hybrids of Drosophila melanogaster. Traditional explanations for tip differences between strains (differential compaction of chromatin, somatic or germinal deletion) are examined and, in the light of the reported observations, rejected. The range of polytene tip variability and rates of change in wild races are greater than has been supposed: strains formerly considered to be terminally deleted appear to gain terminal bands; others, formerly considered normal, appear to have lost them. Strains with high cell-to-cell tip variability are also described. Cell-to-cell variations, as well as much of the observed rapid changes in tip appearance, are probably due to heritable differences in the location of an abrupt transition zone between polytene and nonpolytene chromatin. A quantitative relationship between the amount of certain subterminal bands present and the frequency of tip association of nonhomologous chromosomes is shown and its possible significance for chromosome is shown and its possible for chromosome pairing discussed.

Most of some 33 X-ray-induced duplications recovered as Suppressors of Minute loci proved to be direct tandem duplications. When heterozygous, most duplications were crossover suppressors, and duplications of short to moderate size did not reduce the fitness of their bearers. Crossover suppression by tandem duplication may be attributed to intrastrand foldbacks of the type regularly seen in somatic polytene chromosomes. As a consequence, linkage disequilibrium between duplicated elements and normal chromosomes should be more profound than has been supposed. Tandem duplications appear to be predisposed by reason of frequency of generation, crossover suppression and fitness effects to serve as the primary source of new genes.

The frequency of recovered X-ray-induced (4000R) rearrangements that, in all probability mimic terminal deletions of the X chromosome was only one of, roughly, 10-5 X chromosomes screened for tip deficiencies. Although the single exception looks terminally deleted, it is probably capped by a very short or nonpolytene telomeric segment. It is apparent from these data that the probability of "healing" or stabilization of a terminally deleted X in the zygotic nucleus or developing embryo of Drosophila melanogaster is vanishingly small. The telomeric caps in two obviously interstitial deficiencies that were recovered represent, roughly, 1/500 of the length of a mitotic chromosome. These findings give some indication of the extreme difficulty of detecting short telomeric segments capping either deleted polytene chromosomes or deleted metaphase chromosomes of, for example, humans.

Overlapping duplications recovered as suppressors of Minute loci have been used to localize M(2)z and M(3)w(124) to specific polytene bands 25A1(2) and 95A1(2). The surprising efficiency of M localization by duplication may result from the tendency of M suppressors to be at least a visible fraction of a polytene band in length.

The behavior of T(3;4)10 provides evidence that the centromere of 4 is not in the doublet 101D3,4, and that most of what has been called the left arm of 4 should be called basal 4R. Basal 4R usually does not replicate as much as the tip of 4R when 4 is in its normal position in the chromocenter. Translocated to the tip of 3R, however, basal 4R attains a width equal to that of the rest of the arm-a possible position effect on DNA replication.