Structural basis for Rap1-Rif1-Rif2 assembly : insights into budding yeast telomere architecture and functions

Telomeres form the ends of eukaryotic linear chromosomes and are composed of specialized nucleoprotein complexes. They have been the subject of intense investigation over several decades, as telomere dysfunction has been associated with genome instability and the development of cancer. Yeast (Saccharomyces cerevisiae) telomeric DNA is comprised of irregular TG1-3 repeats, bound in a sequence-specific manner by multiple copies of Rap1, forming Rap1 arrays. Together with its telomere binding proteins Rif1 and Rif2, arrays of Rap1, Rif1 and Rif2 form a protective proteinaceous cap that regulates telomere length, modulates Sir-mediated transcriptional silencing, and prevents unwanted DNA-repair events. As a general transcription regulator in budding yeast, about 90% of cellular Rap1 is found in promoters or silencers, whereas Rif1 and Rif2 can only be detected at telomeres.
The following questions remain unresolved for the major telomere capping proteins Rap1, Rif1, and Rif2. What determines the exclusive telomeric localization of Rif1 and Rif2, and their absence elsewhere in the genome, given that the association of Rif1 and Rif2 at telomeres is solely dependent on the Rap1 C-terminal domain (Rap1RCT)? What is the molecular basis behind the competition between Rif and Sir proteins at telomeres? How do the telomere-associated proteins Rap1, Rif1, and Rif2 influence telomerase activities? And, of central importance to genome stability, what are the roles of Rif1 and Rif2 in damping DNA repair at telomeres?
To address these questions, I determined the structures of Rif2, Rif2-Rap1RCT, Rif1-Rap1RCT, and the outermost Rif1 C-terminal domain, Rif1CTD, using x-ray crystallography. Structural studies, combined with in vitro reconstitution and cellular assays, demonstrated that Rif1 and Rif2 are the long-sought elements that interlink Rap1 units cooperatively. The long- and short-range protein interactions from Rif1 and Rif2, the multimerization module present in Rif1CTD, and the trans interaction between Rap1 and Rif2 provide a network of Rap1-Rif1-Rif2 complexes. This protein network allows the formation of higher-order structures at telomeres. The organizing principle that controls Rap1-Rif1-Rif2 assembly relies on the presence of arrays of Rap1-binding sites, which are exclusively found in telomeric regions. This explains why Rif1 and Rif2 are restricted to telomeric regions, and are not localized to the other ~300 single/double Rap1-binding sites at promoters or silencers within the S. cerevisiae genome.
In addition, I was able to provide the molecular basis for Rif1- and Rif2-modulated silencing at telomeres and HMR. Structural studies, combined with in vivo analysis, allowed me to identify the interaction domains within Rif1 (Rif1RBM) and Rif2 (Rif2RBM), which block the transcriptional repressor Sir3 from accessing the common binding cleft on Rap1. Thus, Rif1 and Rif2 directly compete with Sir3 for the RBM binding groove on Rap1. This protein-binding groove therefore enables Rap1 to integrate opposing cues coming from the Sir3 and Rif1/Rif2 RBMs into a composite silencing response. The partially redundant assembly of Rif1 and Rif2 on Rap1 also elucidates the reported synergistic function of Rif1 and Rif2 in modulating transcriptional silencing at telomeres and HMR.
In this study, I could demonstrate that the Rif2-mediated anti-checkpoint function is dependent on its telomeric localization through the protein interaction with Rap1. I further identified a novel function of Rif1 as a direct DNA-binding protein for protecting resected telomeres from being accidentally recognized as DNA double-strand breaks. Both in vivo and in vitro studies illustrate the remarkable ability of Rif1 to directly outcompete the yeast RPA complex from single-stranded DNA next to single-/double-stranded DNA junctions. The architecture of Rap1-Rif1-Rif2 assemblies favors Rif1 binding the resected telomeric DNA, once the telomere capping function is compromised. Notably, the structure-function studies of Rif1CTD and the Rif1 N-terminus (Rif1NTD) provide strong evidence for applying the principle of inhibiting checkpoint activation from yeast to human.
The work presented here details how the yeast shelterin complex Rap1-Rif1-Rif2 directly influences transcriptional silencing, telomere length regulation, and telomere protection against inadvertent DNA-damage checkpoint activation.