Spontaneous replication fork collapse regulates telomere length homeostasis in wild-type yeast

In most eukaryotes, the enzyme telomerase maintains the termini of linear chromosomes through the addition of repetitive telomeric sequences. It is widely assumed that the primary site of action for telomerase is the single-stranded G-rich overhang at the ends of linear chromosomes. However, increasing evidence indicates that a second telomerase substrate is generated in response to replication stress as the replisome traverses duplex telomeric DNA. We show here that this second substrate, created by spontaneous replication fork collapse in wild-type budding yeast, is elongated by telomerase at a much higher frequency than fully replicated chromosome termini. Furthermore, as much as ~200 nucleotides can be added in a single cell division to newly collapsed forks, indicating that replication fork collapse and the subsequent response by telomerase is a major determinant of telomere length homeostasis. This challenges a long-standing model which posits a length-sensing mechanism that assesses individual telomeres to determine whether yeast chromosome ends are in "telomerase-extendible" versus "telomerase-non-extendible" states. We propose that these two states are instead structurally and temporally distinct substrates for telomerase, generated by two different processes (fork collapse versus completion of DNA replication). We also show that fork collapse at telomeres is kept in check by a telomere-dedicated Cdc13/Stn1/Ten1 complex in collaboration with the canonical RPA complex, indicating that these two complexes stabilize single-stranded DNA exposed at the replication fork. Although failures during DNA replication are often genotoxic events, this represents an opposing example in which fork collapse has been co-opted to promote genome stability.