HeT‐A elements are a new family of transposable elements in Drosophila that are found exclusively in telomeric regions and in the pericentric heterochromatin. Transposition of these elements onto broken chromosome ends has been implicated in chromosome healing. To monitor the fate of HeT‐A elements that had attached to broken ends of the X chromosome, we examined individual X chromosomes from a defined population over a period of 17 generations. The ends of the X chromosomes with new HeT‐A additions receded at the same rate as the broken ends before the HeT‐A elements attached. In addition, some chromosomes, approximately 1% per generation, had acquired new HeT‐A sequences of an average of 6 kb at their ends with oligo(A) tails at the junctions. Thus, the rate of addition of new material per generation matches the observed rate of terminal loss (70–75 bp) caused by incomplete replication at the end of the DNA molecule. One such recently transposed HeT‐A element which is at least 12 kb in length has been examined in detail. It contains a single open reading frame of 2.8 kb which codes for a gag‐like protein.

HeT-A elements are Drosophila melanogaster LINE-like retroposons that transpose to broken chromosome ends by attaching themselves with an oligo(A) tail. Since this family of elements is believed to be involved in the vital function of telomere elongation in Drosophila, it is important to understand their transposition mechanism and the molecular aspects of activity. By comparison of several elements we have defined here the unit length of HeT-A elements to be approximately 6 kb. Also, we studied an active HeT-A element that had transposed very recently to the end of a terminally deleted X chromosome. The 12 kb of newly transposed DNA consisted of a tandem array of three different HeT-A elements joined by oligo(A) tails to each other and to the chromosome end broken in the yellow gene. Such an array may have transposed as a single unit or resulted from rapid successive transpositions of individual HeT-A elements. By sequence comparison with another recently transposed HeT-A element, conserved domains in the single open reading frame (ORF), encoding a gag-like polypeptide, of these elements were defined. We conclude that for transposition an intact ORF is required in cis, while the reverse transcriptase is not encoded on the HeT-A element but is provided in trans. This would make HeT-A elements dependent on an external reverse transcriptase for transposition and establish control of the genome over the activity of HeT-A elements. This distinguishes the Drosophila HeT-A element, which has been implicated in Drosophila telomere elongation, from the other, 'selfish' LINE-like elements.

The telomeric regions in Drosophila cause transcriptional silencing of integrated transgenes. A complex satellite has recently been identified in the subterminal region of the left arm of chromosome 2 that is a good candidate for the source of the observed telomeric silencing, because genetically marked transposable elements that have inserted into this subtelomeric array show repression and variegation of the reporter gene. We asked whether this satellite can also cause transcriptional repression in ectopic chromosomal positions by placing it upstream of a mini-white reporter gene in P element constructs used for germ line transformation. The transgenes are shielded from external influences at the integration site using SU(HW) binding sites at either end. It was found that the satellite represses transcription of the reporter gene in an orientation dependent and an array length dependent manner. The satellite does not, however, induce variegation under the conditions used. The repressed transgenes do not respond to typical modifiers of centromeric position effect variegation, such as Su(var)205 5 , Su(var)2±1 1 , Su(var)3±1 1 , and Su(var)3±6 1 , or to the addition of a Y chromosome. However, as with the original variegating telomeric insertion, suppression in the transgenes is relieved by Su(z)2 5 , suggesting that suppression induced by the subtelomeric satellite retains aspects of telomeric silencing in ectopic positions.

In most eukaryotes telomeres are extended by telomerase. Drosophila melanogaster, however, lacks telomerase, and telomere-specific non-LTR retrotransposons, HeT-A and TART, transpose specifically to chromosome ends. A Drosophila strain, Gaiano, that has long telomeres has been identified. We extracted the major Gaiano chromosomes into an Oregon-R genetic background and examined the resulting stocks after 60 generations. In situ hybridization using HeT-A and TART sequences showed that, in stocks carrying either the X or the second chromosome from Gaiano, only the Gaiano-derived chromosomes display long telomeres. However, in stocks carrying the Gaiano third chromosome, all telomeres are substantially elongated, indicating that the Gaiano chromosome 3 carries a factor that increases HeT-A and TART addition to the telomeres. We show that this factor, termed Telomere elongation (Tel), is dominant and localizes as a single unit to 69 on the genetic map. The long telomeres tend to associate with each other in both polytene and mitotic cells. These associations depend on telomere length rather than the presence of Tel. Associations between metaphase chromosomes are resolved during anaphase, suggesting that they are mediated by either proteinaceous links or DNA hydrogen bonding, rather than covalent DNA-DNA bonds.

A mutator, mu2, in Drosophila melanogaster has been identified recently that potentiates the recovery of terminal deficiencies. The deleted chromosomes behave as if they had been capped; that is, they are protected from degradation and from fusion with other chromosome fragments. The mutator maps near the telomere on the left arm of chromosome 3. Using the selectable marker Aprt, 150 deficiencies for region 62 of the cytological map have been recovered. These deficiencies identify the map position of mu2 as 62B11-C1. A yeast artificial chromosome spanning this region has been subcloned into lambda phage, and the positions of deficiency breakpoints on either side of the mu2 gene have been identified within the subclones. These positions limit the location of the left end of the gene to a 23 kb region. In the course of these experiments, three additional, presumptive mutant alleles were identified, suggesting that other mutator alleles remain undiscovered in many standard laboratory stocks.