Multiple pathways inhibit NHEJ at telomeres

Abstract

The nonhomologous end-joining (NHEJ) repair pathway is inhibited at telomeres, preventing chromosome fusion. In budding yeast Saccharomyces cerevisiae, the Rap1 protein directly binds the telomere sequences and is required for NHEJ inhibition. Here we show that the Rap1 C-terminal domain establishes two parallel inhibitory pathways through the proteins Rif2 and Sir4. In addition, the central domain of Rap1 inhibits NHEJ independently of Rif2 and Sir4. Thus, Rap1 establishes several independent pathways to prevent telomere fusions. We discuss a possible mechanism that would explain Rif2 multifunctionality at telomeres and the recent evolutionary origin of Rif2 from an origin recognition complex (ORC) subunit.

Figure 1.

NHEJ inhibition at telomeres requires the Rap1 C-terminal domain, Rif2, and Sir4. (A) Schematic representation of S. cerevisiae Rap1. (B) Schematic representation of X and Y′ telomeres in S. cerevisiae and relative positions of the primers used for PCR amplification. (C) Fusions amplified after 26 cycles. The mean lengths of the X and Y′ telomeres and of the amplified fusions are indicated in base pairs below each lane. (D) Fusions amplified after 33 cycles.

Figure 2.

Amplification of telomere–telomere fusions in TEL1 cells. For each strain, PCRs from two independent samples are displayed. DNA was diluted with DNA from the wild-type strain.

Figure 3.

NHEJ inhibition by Rap1 in the absence of Rif2 and Sir4. For each strain, PCRs from two independent samples are displayed. The amount of amplified product is indicated in nanograms below each visible smear. DNA were diluted with DNA from the wild-type strain.

Figure 4.

NHEJ inhibition at double-strand breaks. (A) Schematic representation of the NHEJ assay. (B) NHEJ inhibition by internal telomeric repeats. (C) NHEJ inhibition by ectopic targeting of Rap1 C-terminal domain. The mean relative survival on galactose versus glucose were observed in three experiments.

Figure 5.

Alignment of the Rif2 and Orc4 sequences. (A) The first domain (domain I) of the AAA⁺ module. The positions of the observed secondary structures of Pyrobaculum aerophilum Cdc6 (Liu et al. 2000) are shown at the bottom, together with those of AAA⁺ motifs. (WA) Walker A; (WB) Walker B; (S1) sensor 1. Identities are shown white on a black background, and similarities are shaded gray (dark gray for hydrophobic residues, light gray for other similarities). Positions of the Walker A lysine, Walker B aspartate, and S. cerevisiae Orc4 Arg 267 are indicated by red arrowheads. (B) The second domain (domain II) of the AAA⁺ module. Only the sequences of Rif2 and Orc4 are aligned, as the alignment cannot be extended in an accurate way to Cdc6. However, secondary structure prediction made on the basis of this alignment using PHD (Rost et al. 2004) indicated an all-α fold, with four main helices, reminiscent of the domain II structure observed in Cdc6. Note that in Cdc6, the domain II is discontinuous, with a short N-terminal fragment located upstream of domain I. Such an N-terminal extension might be found in the Rif2 and Orc4 sequences.