RNAi and Ino80 complex control rate limiting translocation step that moves rDNA to eroding telomeres

Abstract

The discovery of HAATIrDNA, a telomerase-negative survival mode in which canonical telomeres are replaced with ribosomal DNA (rDNA) repeats that acquire chromosome end-protection capability, raised crucial questions as to how rDNA tracts 'jump' to eroding chromosome ends. Here, we show that HAATIrDNA formation is initiated and limited by a single translocation that juxtaposes rDNA from Chromosome (Chr) III onto subtelomeric elements (STE) on Chr I or II; this rare reaction requires RNAi and the Ino80 nucleosome remodeling complex (Ino80C), thus defining an unforeseen relationship between these two machineries. The unique STE-rDNA junction created by this initial translocation is efficiently copied to the remaining STE chromosome ends, independently of RNAi or Ino80C. Intriguingly, both RNAi and Ino80C machineries contain a component that plays dual roles in HAATI subtype choice. Dcr1 of the RNAi pathway and Iec1 of Ino80C both promote HAATIrDNA formation as part of their respective canonical machineries, but both also inhibit formation of the exceedingly rare HAATISTE (where STE sequences mobilize throughout the genome and assume chromosome end protection capacity) in non-canonical, pathway-independent manners. This work provides a glimpse into a previously unrecognized crosstalk between RNAi and Ino80C in controlling unusual translocation reactions that establish telomere-free linear chromosome ends.

Figure 1.

Whole genome sequencing suggests two-step model for HAATI^rDNA translocation. (A) Comparative analysis of whole genome Illumina sequencing of a single HAATI^rDNA clone. Regardless of the genome coverage (15X vs 300X), only a single STE-rDNA junction is identified. (B) Proposed two-step model for rDNA translocation. Linear chromosomes of S. pombe genomes are represented with telomeres (purple) flanked by STE (STE, red) on Chr I and II, and sub-terminally positioned rDNA on Chr III (green). Upon telomerase loss, HAATI^rDNA is proposed to arise via two steps: First, a unique STE–rDNA junction (black solid line) arises wherein a single illegitimate translocation places rDNA from Chr III on a STE tracts of Chr I or II. Second, the STE–rDNA junction is copied to the remaining STE-containing chromosomal ends, presumably via BIR.

Figure 2.

Generation of S.pombe genome with a pre-existing STE-rDNA junction on Chr I. (A) When plasmid-borne Trt1 is introduced into HAATI^rDNA cells, telomeres are added and the rearranged genomes (with rDNA at all subtermini) are stabilized; these HAATI^rDNA+Trt1 genomes are ‘pre-rearranged’. (B) Generation of cells harboring a single HAATI^rDNA+Trt1 chromosome. wt cells containing LacO repeats integrated at the A67: 5479451 and Sod2: 3185470 loci on Chr I were mated with HAATI^rDNA+Trt1 cells. Offspring lacking or containing LacO repeats were selected based on inheritance or absence of nearby selectable antibiotic resistance markers. Offspring lacking the LacO repeats (and lacking antibiotic resistance) have inherited Chr I from the HAATI^rDNA+Trt1 parent (with STE-rDNA junctions near either chromosome end), and were further screened for the absence of rDNA on Chr II (C). (C) Whole chromosome PFGE reveals which chromosomes have been inherited from HAATI^rDNA+Trt1 parent. Several trt1Δ clones (1–5) that lack or contain LacO repeats on Chr I are shown. Ethidium bromide staining shows the presence of all three chromosomes (left); Southern blot analysis with an rDNA probe (right) shows that in wt genomes (WT lane), rDNA is restricted to Chr III. Based on rDNA signal on Chr I or II, trt1Δ^junction (carrying Chr I from HAATI+Trt1 parent) and trt1Δ^nojunction (carrying Chr I without STE-rDNA junction) were selected for subsequent experiments.

Figure 3.

A pre-existing STE-rDNA junction guarantees HAATI^rDNA formation irrespective of growth conditions. (A) Left: schematic of trt1Δ^nojunction genome (top) or the trt1Δ^junction genome (bottom). Right: five-fold serial dilutions spotted on rich media with or without MMS; five of 10 representative replicates analyzed are shown for each growth condition. trt1Δ^nojunction populations form HAATI^rDNA under competitive conditions and circular survivors (‘O’) under noncompetitive conditions. trt1Δ^junction survivors show MMS resistance indicative of HAATI formation when raised under either condition. (B) Validation of genome arrangements of ten trt1Δ^junction survivors raised by non-competitive growth by PFGE. NotI digestion of wt chromosomes releases four terminal fragments (L, M, I, and C) from Chr I and II; in circulars, these are replaced by fusion fragments L+I and C+M, while in HAATI cells, the vast majority of NotI fragments are retained in the well. All trt1Δ^junction survivors are HAATI. The absence of STE hybridization (right panel) further validates HAATI^rDNA formation; a HAATI^STE genome is included as a reference. (C) Table summarizing the effect of a pre-existing junction on HAATI^rDNA formation. Number of replicates indicates number of independent cultures with identical genotypes that were propagated to raise survivors upon trt1+ deletion. A representative experiment (of three performed) is summarized; percentage of HAATI^rDNA formation was scored via MMS resistance, PFGE pattern and the absence of STE amplification. Identical results were obtained in additional replicates performed for data shown in Figures 4 and 7.

Figure 4.

RNAi pathway is essential only for the initial translocation step. Figure 4A–E represent five representative survivors of 10 raised for each genotype under each condition. (A) dcr1Δ trt1Δ^nojunction survivors propagated under competitive growth conditions and spotted in 5-fold serial dilutions yield MMS sensitive (upper three rows) and resistant (lower two rows) clones. The MMS resistant clones are presumably HAATI^STE. (B) dcr1Δ trt1Δ^nojunction survivors when propagated under non-competitive growth conditions showed extreme MMS sensitivity, indicating circular survivors. (C) dcr1Δ trt1Δ^junction survivors propagated under competitive growth show MMS resistance indicative of HAATI formation. (D) dcr1 trt1Δ^junction survivors propagated under non-competitive growth show MMS resistance indicative of HAATI formation. (E) Determination of genome arrangements in dcr1Δ trt1Δ^junction survivors raised under competitive (left) and noncompetitive (right) conditions by PFGE of Not1 digested chromosomes. Retention of hybridization signal in the well for all survivors indicates formation of HAATI (top, LMIC probe). The absence of STE hybridization (bottom, STE1 probe) confirms formation of HAATI^rDNA. A HAATI^STE sample is included as a reference. (F) Table summarizing the dispensibility of RNAi factors for HAATI^rDNA formation when a pre-existing STE-rDNA junction is present. Number of replicates indicates number of populations of identical genotype propagated to form survivors. One representative experiment (of two performed) is summarized. Percentage of HAATI^rDNA formation is based on MMS resistance, retention of PFGE signal and absence of STE amplification. Putative HAATI^STE is inferred from MMS resistance and previous results showing that dcr1Δtrt1Δ survivors are never HAATI^rDNA but can be HAATI^STE(11).

Figure 5.

STE are preferred ‘acceptor sequences’ for HAATI^rDNA translocation. (A) wt S. pombe genomes contain knob regions (yellow), whose chromatin marks and repressive properties depend on Sgo2 recruitment, just proximal to STE regions. STEΔ strains lack virtually all STE tracts, placing the knob regions in close proximity to the canonical telomeric repeats. (B) Representative trt1Δ survivors (4 of 10 analyzed) obtained under competitive growth conditions and spotted in 5-fold serial dilution show MMS resistance, indicating HAATI formation. Representative trt1Δ STEΔsurvivors (4 of 10 analyzed) obtained under the same conditions show extreme MMS sensitivity, indicating chromosome circularization. (C) Genome arrangements of survivors. (Left) Retention of hybridization signal in the well for representative survivors (5 of 10 analyzed) indicates HAATI formation. (Right) trt1Δ STEΔ survivors form inter- and intra-chromosomal fusions indicated by the sizes circularization products, sometimes accompanied with chromosome length changes and/or dichromosome circles. (D) PFGE of NotI digested chromosomes of trt1Δ STEΔ sgo2Δ survivors. 2 out of 10 tested survivors (indicated by red asterisks) show retention of PFGE signal in the well, suggesting HAATI formation. (E) Table summarizing the role of STE sequences and the Sgo2 bound knob on HAATI^rDNA formation. Survivors derived from three independent ‘starting point’ clones are tallied. Percentage of HAATI^rDNA formation is based on MMS resistance, retention of PFGE signal and absence of STE amplification.

Figure 6.

Ino80C is essential for HAATI^rDNA formation while Iec1, an Ino80C member, phenocopies Dcr1 in its role in HAATI subtype choice. (A) Progeny of heterozygous trt1Δ/trt1+ diploids carrying or lacking arp8⁺ were cultured under competitive conditions. Five representative survivors (of 10 analyzed) are shown for each genotype. HAATI survivors are absent upon loss of Arp8; instead, linear survivors form as indicated by PFGE pattern. (B) Heterozygous trt1Δ/trt1+ diploids carrying or lacking iec1⁺ were sporulated, and the indicated progeny cultured under competitive conditions and spotted in 5-fold serial dilution. Five representative survivors (of 10 analyzed) for each genotype are depicted. All iec1Δ trt1Δsurvivors show MMS resistance indicating HAATI formation. (C) (Left) Schematic shows location of STE1 probe (black line) in STE region (red). (Right) Representative PFGE of NotI-digested chromosomes from iec1Δ trt1Δ survivors. Most iec1Δ trt1Δ survivors are HAATI as indicated by retention of signal in the well probed for LMIC. The membrane was stripped and re-probed for STE1, which hybridizes strongly with all NotI restriction fragments, indicating HAATI^STE formation. (D) Tables summarizing the essential role for Arp8 and Iec1 in HAATI^rDNA formation as well as Iec1’s role in blocking HAATI^STE formation independent of Ino80C, phenocopying Dcr1. Analysis of survivor formation is shown under competitive conditions (top) and noncompetitive conditions (bottom); asterisk marks published results from (11) shown here for comparison.

Figure 7.

Ino80 complex is essential only for the first step of translocation. Five representative iec1Δ trt1Δ^junction or arp8Δ trt1Δ^junction survivors (of 10 analyzed for each genotype) are described in Figure 7A and B. (A) All iec1Δ trt1Δ^junction survivors raised under competitive growth (left) or noncompetitive (right) conditions show MMS resistance suggestive of HAATI formation. (B) All arp8Δ trt1Δ^junction survivors propagated under either competitive (left) or non-competitive conditions (right) show MMS resistance indicative of HAATI formation. (C) PFGE of Not1 digested chromosomes of 10 arp8Δ trt1Δ^junction survivors. Retention of hybridization signal in the well for all survivors indicates HAATI formation (LMIC probe, top). The absence of STE hybridization (STE probe, bottom) indicates formation of HAATI^rDNA. (D) Table highlighting the dispensibility of Ino80C for HAATI^rDNA formation in the presence of a pre-existing STE-rDNA junction. Regardless of growth conditions, all iec1Δ trt1Δ^junction or arp8Δ trt1Δ^junction survivors are HAATI. For comparison, trt1Δ^junction survivor formation data from Figures 3D and 6D is shown.