End joining at Caenorhabditis elegans telomeres

Abstract

Critically shortened telomeres can be subjected to DNA repair events that generate end-to-end chromosome fusions. The resulting dicentric chromosomes can enter breakage-fusion-bridge cycles, thereby impeding elucidation of the structures of the initial fusion events and a mechanistic understanding of their genesis. Current models for the molecular basis of fusion of critically shortened, uncapped telomeres rely on PCR assays that typically capture fusion breakpoints created by direct ligation of chromosome ends. Here we use independent approaches that rely on distinctive features of Caenorhabditis elegans to study the frequency of direct end-to-end chromosome fusion in telomerase mutants: (1) holocentric chromosomes that allow for genetic isolation of stable end-to-end fusions and (2) unique subtelomeric sequences that allow for thorough PCR analysis of samples of genomic DNA harboring multiple end-to-end fusions. Surprisingly, only a minority of end-to-end fusion events resulted from direct end joining with no additional genome rearrangements. We also demonstrate that deficiency for the C. elegans Ku DNA repair heterodimer does not affect telomere length or cause synthetic effects in the absence of telomerase.

Figure 1.—

Telomere-length homeostasis in C. elegans NHEJ mutants. (A) Telomere length of cku-80 mutant strains. Genomic DNA was isolated from strains that were homozygous for their given mutations for eight generations. (B) Time to sterility for trt-1 single and trt-1; NHEJ double mutant strains. Arrows indicate median. (C) Telomere erosion in trt-1 single and trt-1; cku-80 double mutant strains.

Figure 2.—

Characterization of outcrossed end-to-end chromosome fusions. (A) Fusion breakpoints of 38 X-autosome end-to-end chromosome fusions, as determined by linkage analysis. Different chromosomes are depicted as rectangles in various colors with green squares representing telomeres at every chromosome end. Genetic names of each independent fusion are to the right of their fusion orientation. End-to-end fusions that were amenable to PCR are indicated in boldface type and underlined. Although the fusion breakpoint of ypT23 was spanned by PCR, it involved a small inversion event and therefore was not a direct fusion breakpoint. (Figure 3B) (B) Terminal deletion analysis to map fusion breakpoints molecularly. PCR results are shown for one chromosome end involved in a fusion. Arrows indicate primers. In this example, one fusion strain was tested for terminal deletion at XR with primers targeting 5 kb of subtelomeric DNA. N2, wild type control; f, X-autosome fusion strain. (C) Distribution of extent of telomere deletion at chromosome ends. (D) Frequency of fusion breakpoint configurations. Observed frequencies (%), expected frequencies (Exp), and P-values are shown.

Figure 3.—

Fusion breakpoint structures of two direct fusions. (A) The drawing depicts the fusion breakpoint for ypT44 where sequencing revealed several subtelomeric direct repeats (yellow) remain intact. Below, the structure of inverted repeats and direct repeats at wild-type IIR, as well as the position of the fusion breakpoint for ypT44, are shown. (B) An inversion at the fusion breakpoint of ypT23 is shown. A 3-bp deletion occurred at the inversion, denoted by a dots above the breakpoint sequence. Boldface type and underlining indicate microhomology at the breakpoints.

Figure 4.—

PCR amplification of end-to-end fusions from mid- to late-generation telomerase deficient strains. PCR products were analyzed by separation on an ethidium bromide-stained agarose gel. Three primer pairs were utilized on 16 strains, including the wild-type control, as indicated. M, 1-kb ladder (Invitrogen). (A) PCR targeting IIR and VL. All of the PCR reactions containing genomic DNA from telomerase-deficient strains displayed nonspecific, weak bands, or smears. (B) PCR targeting IIL and VL. Arrowheads indicate PCR products for one strain, trt-1(ok410).a, for all three primer pairs. While all three reactions contained the same primer targeting IIR, each reaction contained a different primer targeting VL at sites 1371 bp, 1651 bp, and 1768 bp from the start of the telomere repeats for primers VL3, VL29, and VL31, respectively.

Figure 5.—

The physical structure of two fusion breakpoints that were refractory to PCR, as investigated by Southern analysis. Wild-type and mutant DNA digests using the indicated enzymes are shown. ypT27 fusion breakpoint was probed with (A) XR and (B) VL subtelomeric probes. For wild type, the signals agreed with the predicted sizes and appearances: >1.4 kb, >1.3 kb, and >1.0 kb (XR) and 4.3 kb, 4.5 kb, and >7.2 kb (VL) for AvaII, HindIII, or PflMI, respectively. The smeary signals associated with terminal restriction fragments (containing telomeric DNA) are delimited by a bracket to the left of a given lane. (C) The restriction fragments predicted to be detected by each probe for wild type and ypT27 are shown. ypT21 was probed with (D) XR and (E) IVR subtelomeric probes. Signals for wild type agreed with predictions: >1.6 kb and >1.7 kb (XR) or >3.2 kb and >3.1 kb (IVR) for HpaII and PmlI, respectively. (F) The restriction fragments predicted to be detected by each probe for wild type and ypT21 are shown.