Non-coding telomeric and subtelomeric transcripts are differentially regulated by telomeric and heterochromatin assembly factors in fission yeast

Abstract

While telomere repeat-containing non-coding RNA has been identified in a variety of eukaryotes, its biological role is not yet clear. We have identified telomeric transcripts in fission yeast, a model system that combines precise genetic manipulability with telomeres remarkably similar to those of human. Like human and budding yeast, fission yeast harbours a population of telomeric RNA molecules containing G-rich telomeric repeats transcribed from the subtelomere to the telomere. In addition, we detect substantial levels of C-rich telomeric RNA whose appearance is independent of the RNA-dependent RNA polymerase, suggesting that the telomere repeats themselves serve as promoter sites; multiple distinct subtelomeric RNAs are also present. The regulation of these transcripts depends on the telomere-associated proteins Taz1 and Rap1, as deletion of taz1(+) or rap1(+) leads to increased levels of both telomere repeat-containing and subtelomeric transcripts. In contrast, loss of the heterochromatin proteins Swi6 or Clr4 or the telomerase regulator Rif1 results in elevated subtelomeric RNA levels while telomere-repeat containing transcript levels remain repressed. Coupled with the large body of knowledge surrounding the functions of telomeric and heterochromatin factors in fission yeast, these in vivo analyses suggest testable models for the roles of TERRA in telomere function.

Figure 1.

Telomeric RNA is detectable in wt and taz1Δ cells. (A) Schematic of the three fission yeast chromosomes with telomeres indicated by black rectangles versus the circular chromosomes of Otrt1Δ strains that lack telomeres entirely. (B) Northern blot analysis was performed on 40 µg total cellular RNA from each strain, with or without RNase A. The ethidium bromide (EtBr)-stained gel prior to transfer is shown on the left. The blot was hybridized with Telo1, which contains 250 bp of telomere repeats plus 32 bp of STE sequence (see text); hybridization to ribosomal RNA is also observed. The smear that appears above the telomeric RNA in wt and taz1Δ lanes is not RNase-sensitive, and therefore likely to represent residual DNA in the sample. (C) Southern blot of ApaI-digested genomic DNA hybridized with Telo. The blot was stripped and re-probed with an internal region to control for loading (below). (D) Telomeric RNA in taz1Δ cells is predominantly C-rich. Northern blot analysis of 20 µg total cellular RNA hybridized with strand-specific telomere probes. dsTelo: a dsDNA telomere-repeat containing fragment that allows comparison of hybridization intensity between G- and C-rich probes. Samples were run on the same gel and transferred, then the membrane was cut and probed with either a C-rich probe (left panel) to detect G-rich RNA, or a G-rich (right panel) probe to detect C-rich RNA. The G-rich probe detects telomeric RNA in taz1Δ samples, whereas the C-rich probe only detects bands that are present in Otrt1Δ samples as well and are therefore not telomere-specific; background hybridization to rRNA provides a loading control.

Figure 2.

Rdp1 is required neither to generate TERRA nor C-rich telomeric RNA. (A) RT-PCR reveals telomeric and subtelomeric RNA transcribed from the subtelomere outwards. Top: Schematic of telomeric region. Oligonucleotides used are designated as arrows pointing in the 5′-to-3′ direction. First strand synthesis with or without reverse transcriptase (RT+/−) using the primers indicated (first strand) was performed on RNA from the indicated strains. cDNA was then amplified with the oligonucleotides indicated (PCR). Genomic DNA (DNA ctrl) used as template was included as a control. o3 primes cDNA from RNA transcribed in a telomere-to-subtelomere direction, whereas oC primes cDNA from RNA of the opposite polarity. oG likely primes C-rich telomeric cDNA, but subtelomeric primers would not amplify this cDNA. This serves as an important negative control, as under standard denaturation conditions, endogenous priming of telomeric RNA can occur; such endogenous priming was removed under the conditions used here (see ‘Materials and Methods’ section). The act1 control demonstrates the presence of RNA in all samples. (B) Transcription of the telomeric C-strand is independent of Rdp1. Strand-specific northern blot analysis of total cellular RNA as in Figure 1D. Telomere-specific signals are absent from the blot probed with C-rich RNA, as all signals are present in the Otrt1Δ sample. In contrast, the G-rich probe detects telomere-specific C-strand signal in taz1Δ and taz1Δrdp1Δ samples. C-rich and G-rich probes hybridize approximately equally to ds telomeric DNA (data not shown).

Figure 3.

Detection of subtelomeric RNA. (A) Schematic representation of the ends of Chromosomes I and II. Telomeres (Telo) are ∼270 bp in length and the subtelomeric elements are ∼1.8 kb (for STE1) and ∼5 kb (STE2) in length. Probes are indicated by black solid lines. (B) Northern blot analysis was performed as in Figure 1B, except that a subtelomeric sequence (STE1) probe was used. RNase-sensitive bands of ∼0.4, 0.6, 0.8 and 1.2 kb are detected in wt and taz1Δ but not Otrt1Δ samples.

Figure 4.

Telomeric and heterochromatin proteins regulate telomeric and subtelomeric RNA. (A) Telomeric and subtelomeric RNAs are differentially regulated by telomeric and heterochromatin proteins. Total cellular RNA of 20 µg was loaded in each lane, blotted and hybridized with Tel80, which contains only telomere sequence. The blot was then stripped and hybridized with the STE2 probe. (B) Northern analysis was carried out as in A on total cellular RNA from the indicated strains.