Telomeres: Do the ends justify the means?

Abstract

Elizabeth H. Blackburn Department of Molecular Biology University of California Berkeley, California 94720 For decades it has been known that the ends of linear eucaryotic chromosomes are specialized structures (re- viewed in Blackburn and Szostak, Ann. Rev. Biochem. 53, 163-194, 1984). This fact was most evident when chro- mosomes were broken by mechanical rupture or X-irradia- tion Unlike true telomeres, the ends of broken chromo- somes are usually unstable; they fuse chromosomes end- to-end and manifest other abnormal reactions, One likely function of a telomere is to allow completion of the replication of the end of the linear chromosomal DNA molecule. All known template-dependent DNA polym- erases require a primer (RNA or DNA) to initiate 5’*3’ strand synthesis, which after chromatid replication would leave some unreplicated bases at each of its ends. Studies on linear viral DNAs show that viruses exploit various types of structures at their DNA termini to solve this problem, but the question of whether any of these viral models apply to cellular, chromosomal DNAs in eucaryotic nuclei re- mains Molecular Structure of Telomeres in Lower Eucaryotes Direct analysis of chromosomal ends has not been prac- ticable until recently, because telomeric sequences repre- sent a minute proportion of the enormously long chromo- somal DNAs. However, the nuclei of certain lower eucary otes contain relatively short linear DNAs (<lo0 kb), which can be purified in sufficient quantity for direct molecular analysis of their ends. Examples include linear ribosomal RNA gene molecules (rDNA) and the subchromosomal linear DNAs that together comprise the somatic macronu- clear genomes of several ciliated protozoa. Their terminal sequences and structures are all remarkably similar in spite of the evolutionary diversity of the different species (see table). First, the sequence closest to the end consists of short, tandemly repeated DNA sequences. Second, the rDNA of the ciliated protozoan Tetrahymena contains a specific array of nonligatable, single-strand breaks, limited to the distal portion of the tandem repeats (Blackburn and Gall, JMB 120, 33-53, 1978). Similar nonligatable breaks are at the ends of the rDNA of Physarum (Johnson, Cell 22, 875-886, 1980). Third, as described below, the ex- treme ends of such linear rDNAs appear to be blocked in some way (Blackburn and Gall,