Evolutionary-conserved telomere-linked helicase genes of fission yeast are repressed by silencing factors, RNAi components and the telomere-binding protein Taz1

Abstract

In Schizosaccharomyces pombe the RNAi machinery and proteins mediating heterochromatin formation regulate the transcription of non-coding centromeric repeats. These repeats share a high sequence similarity with telomere-linked helicase (tlh) genes, implying an ancestral relationship between the two types of elements and suggesting that transcription of the tlh genes might be regulated by the same factors as centromeric repeats. Indeed, we found that mutants lacking the histone methyltransferase Clr4, the Pcu4 cullin, Clr7 or Clr8, accumulate high levels of tlh forward and reverse transcripts. Mutations and conditions perturbing histone acetylation had similar effects further demonstrating that the tlh genes are normally repressed by heterochromatin. In contrast, mutations in the RNAi factors Dcr1, Ago1 or Rdp1 led only to a modest derepression of the tlh genes indicating an alternate pathway recruits heterochromatin components to telomeres. The telomere-binding protein Taz1 might be part of such a redundant pathway, tlh transcripts being present at low levels in Deltataz1 mutants and at higher levels in Deltataz1 Deltadcr1 double mutants. Surprisingly, the chromodomain protein Chp1, a component of the Ago1-containing RITS complex, contributes more to tlh repression than Ago1, indicating the repressive effects of Chp1 are partially independent of RITS. The tlh genes are found in the subtelomeric regions of several other fungi raising the intriguing possibility of conserved regulation and function.

Figure 1

Regions of sequence homology in S.pombe centromeres and subtelomeres. (A) Schematic representation of S.pombe centromeres. dh and dg repeats are indicated in red and blue, respectively. Regions of homology (>60% sequence identity) between tlh genes and centromeric repeats are indicated by open arrows. These regions map within 3921 bp in the 5662 bp tlh1 ORF, 1102 bp (seven segments; <2% gap) of tlh1 displaying an identity to centromeric sequences >75%. Centromeric repeats that have not been sequenced are indicated by double bars. dh and dg repeats were placed according to published data (8) and our own sequence comparisons. The green arrowhead indicates a previously unidentified fragment from a tf2-retrotransposable element. Available cosmid sequences (8) are represented by bars below each centromere. (B) Schematic representation of S.pombe subtelomeric regions. The subtelomeric cosmids SPAC212, SPBCPT2R1 and SPAC750 are represented. Dashed lines between SPAC212, SPBCPT2R1 and SPAC750 point to regions of synteny. The tlh homology to centromeric repeats is represented in red (dh) or blue (dg). Green arrows indicate the presence of LTRs. Open arrows denote ORFs. Bracketed grey lines in the left portion of tel1R represent a hypothetical tlh gene that is not part of the SPAC750 sequence. Its presence and position are suggested by hybridization experiments (35) and by the syntenic relationship between the three represented telomeres (see text).

Figure 2

Conservation of tlh genes in fungi. The displayed subtree was created from a full alignment including the product of 58 S.pombe helicase genes and 42 RecQ helicases from other species (see text). Black squares mark the products of telomeric ORFs (according to annotations); grey squares mark ORFs that are first or second in contigs >500 kb; open circles mark ORFs that are more than two ORFs away from the end of a DNA segment; and open squares mark single ORFs not assigned to any segment, or bacterial ORFs.

Figure 3

Transcriptional regulation of the S.pombe tlh genes. (A) Schematic representation of the tlh1 gene and of a portion of centromere 3 indicating the location of primers used in (B and C) and fragments amplified with these primers. (B) Abundance of forward and reverse tlh transcripts in heterochromatin, RNAi and Taz1 mutants. Strand specific RT–PCR was performed to estimate the levels of tlh RNA by amplifying a region of tlh with high sequence similarity to the centromeric dh repeats (tlh dh). Reverse transcription was primed with OKR41 to detect the tlh forward strand (For), or with OKR40 to detect the tlh reverse strand (Rev). Centromeric forward (cen For) and reverse (cen Rev) RNAs were detected as described previously (19). The bands observed above the centromeric forward and reverse PCR products in the Δclr8 panels originate from the mating-type region in h⁹⁰ strains. The actin transcript (act) was amplified to estimate the amount of total RNA in the samples. −RT was performed with actin primers. wt:SPK10; Δclr4:SPK20; clr6-1Δclr3:SPK27; Δchp1:PG2870; Δtaz1:J11; Δago1:TV292; Δdcr1:TV293; Δrdp1:TV296; Δclr7:SPA15; Δclr8:PG3389; Δpcu4:PG3435. (C) Variations in the abundance of reverse strand along the tlh ORF. RT–PCR was performed as in (A). In addition, a region of tlh RNA encoding the DEAH helicase domain (tlh DEAH) was amplified, the reverse transcription being primed with OKR45 to detect the forward strand (For) or with OKR44 to detect the reverse strand (Rev).

Figure 4

Recovery from TSA treatment. (A) Centromeric and subtelomeric RNAs in cells recovering from TSA treatment. RT–PCR was performed on wild-type cells (SPK10) not treated with TSA (−TSA), or treated with TSA and maintained in culture for the indicated times following TSA removal (+TSA). (B) Normal cell morphology (−TSA) is perturbed by TSA treatment (6 h recovery) and re-established within a few generations following TSA removal (24 h recovery). White arrowheads point to cases of pseudohyphal growth and white arrows to multiple septa in a single cell.

Figure 5

tlh RNA levels in Δtaz1, Δdcr single and double mutants. MNE values were obtained for tlh transcripts by real-time RT–PCR analysis, using actin transcripts for the normalization as described in Materials and Methods. (A) wt:SPK10; Δclr4:SPK20; Δtaz1Δdcr1:SPK61; Δtaz1:J11; Δdcr1:TV293. (B) Tetrad analysis of a cross between J11 and PG3039: Tetrad 1 (parental ditype) [Δdcr1:SPK50; Δtaz1:SPK51; Δdcr1:SPK52; Δtaz1:SPK53], Tetrad 2 (non-parental ditype) [wt:SPK54; Δtaz1Δdcr1:SPK55; wt:SPK56; Δtaz1Δdcr1:SPK57].

Figure 6

The tlh genes affect the expression of neighboring genes. (A) Schematic representation of constructs integrated at the ura4 locus and phenotypes of corresponding strains. Ten-fold serial dilutions of cell suspensions were spotted on the indicated media, allowed to grow for 4 days at 33°C, and photographed. (B) Three individual PG3273 colonies in the red (PG3273R) or white (PG3273W) epigenetic state were spotted on the indicated media. (C) A FOA-resistant isolate of PG3273 (PG3273F) was plated along with a red and white isolate of the same strain and two tester strains, PTI3 (Ade⁻ Ura⁻) and PTI14 (Ade⁺ Ura⁺).

Figure 7

Sensitivity of an ectopic tlh2-ade6⁺ reporter to histone acetylation. PG3263 cells were treated with TSA (+TSA) or mock treated (−TSA), plated on the indicated selective or indicator media, incubated for 4 days at 33°C, and plates were scored and photographed.

Figure 8

Model for the generation of a centromere by telomere fusion. Partial tlh ORFs are found in subtelomeric regions in addition to the full-length genes, suggesting that tlh genes were once present in several copies at these sites (dotted arrows). A dashed line shows how a telomeric end to end fusion of the PT2R1 (Tel IIR) and c212 (Tel IL) regions would create a symmetrical structure like those at centromeres.