Full text

Turn on search term navigation

ARTICLE

Received 7 Jul 2014 | Accepted 15 Oct 2014 | Published 27 Nov 2014

Most long non-coding RNAs (lncRNAs) encoded by eukaryotic genomes remain uncharacterized. Here we focus on a set of intergenic lncRNAs in ssion yeast. Deleting one of these lncRNAs exhibited a clear phenotype: drug sensitivity. Detailed analyses of the affected locus revealed that transcription of the nc-tgp1 lncRNA regulates drug tolerance by repressing the adjacent phosphate-responsive permease gene transporter for glycerophosphodiester 1 (tgp1 ). We demonstrate that the act of transcribing nc-tgp1 over the tgp1 promoter increases nucleosome density, prevents transcription factor access and thus represses tgp1 without the need for RNA interference or heterochromatin components. We therefore conclude that tgp1 is regulated by transcriptional interference. Accordingly, decreased nc-tgp1 transcription permits tgp1 expression upon phosphate starvation.

Furthermore, nc-tgp1 loss induces tgp1 even in repressive conditions. Notably, drug sensitivity results directly from tgp1 expression in the absence of the nc-tgp1 RNA. Thus, transcription of an lncRNA governs drug tolerance in ssion yeast.

DOI: 10.1038/ncomms6576 OPEN

Long non-coding RNA-mediated transcriptional interference of a permease gene confers drug tolerance in ssion yeast

Ryan Ard1, Pin Tong1 & Robin C. Allshire1

1 Wellcome Trust Centre for Cell Biology and Institute of Cell Biology, School of Biological Sciences, The University of Edinburgh, Max Born Crescent, Edinburgh EH9 3BF, Scotland, UK. Correspondence and requests for materials should be addressed to R.C.A. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 5:5576 | DOI: 10.1038/ncomms6576 | http://www.nature.com/naturecommunications

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6576

Eukaryotic genomes are pervasively transcribed. Frequently this transcription generates long non-coding RNAs (lncRNAs), which may be transcribed antisense to pro

tein-coding genes, from within introns, or from intergenic regions of the genome. RNA polymerase II (RNAPII) is responsible for generating both messenger RNAs (mRNAs) and lncRNAs1. As with mRNAs, many lncRNAs are processed (that is, capped, spliced, polyadenylated), however, in contrast to protein-coding mRNAs, lncRNAs are predominantly nuclear and many are rapidly degraded by the exosome2, the major cellular 30-50 RNA degradation machinery3. Consequently, the majority of lncRNAs exhibit low steady-state levels compared with mRNAs. This instability coupled with their general lack of primary sequence conservation has lead to the suggestion that many lncRNAs might simply result from spurious, inconsequential transcriptional noise4. Nonetheless, accumulating evidence indicates that an increasing number of lncRNAs act to regulate gene expression2,5.

The mere act of lncRNA transcription, including accompanying chromatin modications and resulting changes in nucleosome density6, can have a profound impact on neighbouring gene expression. In the simplest scenario, lncRNA expression can provide an environment that is either suitable or unsuitable for transcription factor binding. For example, cascading lncRNA transcription upstream of the ssion yeast Schizosaccharomyces pombe fbp1 gene is required to induce fpb1 expression following glucose starvation7. In addition, in a process termed transcriptional interference, serine-mediated repression of the budding yeast Saccharomyces cerevisiae SER3 gene is brought about by lncRNA transcription into the gene promoter, which increases nucleosome density and prevents transcription factor access810. These examples illustrate the positive and negative inuence that lncRNA transcription can exert on gene regulation in response to environmental changes.lncRNAs can also be processed into smaller regulatory RNAs (for example, short interfering RNA)11. In S. pombe, lncRNAs transcribed from centromeric outer repeats are processed by Dicer (Dcr1) into short interfering RNAs, which target the Clr4 H3K9 methyltransferase via Ago1 (within the RNA-induced transcriptional silencing complex) to establish repressive heterochromatin through the methylation of lysine 9 on histone H3 (refs 1215). In addition, lncRNAs may directly associate with and recruit factors that alter chromatin status, in cis or in trans, silencing genes or behaving as enhancers16,17. For example, lncRNAs aid the response of S. cerevisiae cells to specic changes in nutrient availability by recruiting chromatin-modifying complexes (for example, histone deacetylases) to dynamically regulate several genes1820. Related mechanisms have since been reported in multicellular eukaryotes21,22. Recent analyses also suggest that patches of transient heterochromatin can form under particular conditions at specic euchromatic loci inS. pombe2325. This mechanism involves the RNA-binding protein Mmi1, which recruits the RNA-surveillance machinery to specic determinant of selective removal (DSR) motifs in target transcripts, leading to their exosome-mediated degradation26. Mmi1 and its associated factor Red1 are reported to also recruit chromatin-modifying activities via nascent mRNA and lncRNA targets to deposit H3K9 methylation (H3K9me2) at these locations23,25,27,28. It is therefore evident that lncRNAs employ a variety of mechanisms to regulate gene expression.

Despite rapid advances in lncRNA identication, only a small number have been characterized in detail. A clear challenge in assigning function has been a lack of lncRNA sequence conservation between even the most closely related species29. However, the order of genes anking the transcription units that encode lncRNAs can be preserved through evolution30 (that is,

synteny) and provides another criterion by which we can identify potential functionally conserved lncRNAs whose primary sequences might have diverged too much so as not to retain detectable homology.

Only a few of the B500 annotated intergenic lncRNAs inS. pombe are conserved at the sequence level in three divergent Schizosaccharomyces species, although many retain synteny with anking genes in at least one other species31. We identied eight discrete intergenic lncRNAs that exhibit synteny in at least three of the four Schizosaccharomyces species. Deletion of one of these loci (SPNCRNA.1343 or ncRNA.1343 for short) exhibited a drug-sensitivity phenotype. We demonstrate that ncRNA.1343 encodes a bidirectional lncRNA promoter and that its deletion causes loss of expression of the divergent unstable transcript nc-tgp1. Our analyses reveal that nc-tgp1 is targeted for Mmi1-directed exosome degradation and is required to repress a downstream phosphate-responsive gene (SPBC1271.09 designated transporter for glycerophosphodiester 1 (tgp1 )). However, rather than involving transient heterochromatin formation as a result of targeted RNA degradation, the regulation of tgp1 by the nc-tgp1

RNA appears to be mediated by transcriptional interference. Most importantly, tolerance of S. pombe to a broad spectrum of compounds relies on the regulation of tgp1 by nc-tgp1.

ResultsDeletion of SPNCRNA.1343 causes drug hypersensitivity. TheS. pombe genome is predicted to encode B500 intergenic lncRNAs32. Although few of these lncRNAs exhibit detectable sequence conservation, B100 are conserved in synteny with putative lncRNA orthologues in at least one of the three other known Schizosaccharomyces species31. For example, the functionally characterized telomerase RNA (ter1 /SPNCRNA.214) is syntenic despite its lack of sequence conservation (see Supplementary

Fig. 1a).

To identify other potential functionally conserved lncRNAs, we selected eight lncRNAs, including ter1 as a control, where surrounding gene order is retained in S. pombe and at least two other Schizosaccharomyces species. Each lncRNA gene was deleted by replacement with a loxP-anked ura4 marker (Supplementary Fig. 1b). Apart from ter1D, the selected lncRNAs were not essential for normal cell growth (Supplementary Figs 1c and 2). However, since many characterized lncRNAs regulate gene expression in response to environmental changes and stress33, we tested the growth of these lncRNA deletion strains in response to the following stresses: temperature, the microtubule destabilizing drug thiabendazole (TBZ), DNA synthesis inhibitor hydroxyurea (HU), ultraviolet-induced DNA damage, H2O2-induced oxidative stress and caffeine, an inhibitor of cyclic AMP phosphodiesterase. Cells lacking SPNCRNA.1343 (ncRNA.1343 for short) displayed a phenotype: hypersensitivity to TBZ, HU and caffeine but not to temperature extremities, ultravioletirradiation or oxidative stress (Supplementary Fig. 1c and Supplementary Fig. 2).

Drug sensitivity of 1343D cells is caused by tgp1 induction. lncRNAs can act in cis to regulate the expression of nearby genes2. To determine the cause of drug sensitivity in 1343D cells we examined the expression of genes anking the locus by real-time quantitative reverse transcriptase-PCR (RTqPCR) in wild-type cells, cells with ncRNA.1343 replaced by loxP-anked ura4 marker (1343D::ura4 ) and cells with the ura4 marker subsequently removed (1343D; Fig. 1a). SPBC1271.09 transcript levels increased 450-fold in both 1343D::ura4 and 1343D cells (Fig. 1b), while the expression of other neighbouring genes was unaltered. SPBC1271.09 encodes a conserved

2 NATURE COMMUNICATIONS | 5:5576 | DOI: 10.1038/ncomms6576 | http://www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6576 ARTICLE

340 kb

350 kb 352 kb 354 kb 356 kb 358 kb 360 kb

Ch. II

sme2 +

sme2+

1271.10c 1271.09

(tgp1 +)

ncRNA.1343 1271.08c 1271.07c mug96

RTqPCR

1343

100

70 40 10

Wild-type

3 kb2.5 kb

EtBr

Wild-type

Relative expression

1343 ::ura4 +

tgp1 + mRNA

1343

1271.10c

1271.09

(tgp1+)

1271.08c

mug96

YES TBZ HU Caffeine

Wild-type

1343 ::ura4 +

1343

tgp1 + 1343

Figure 1 | Drug sensitivity following ncRNA.1343 deletion is due to increased tgp1 expression. (a) Schematic representation of genes anking ncRNA.1343. (b) RTqPCR experiments measured transcript levels for nearby genes in wild-type cells and following replacement of ncRNA.1343 with ura4 (1343D::ura4 ) or deletion (1343D). Error bars represent s.e.m. resulting from at least three independent replicates.

(c) Northern analysis of tgp1 transcript levels in wild-type and 1343D cells grown in the presence of phosphate. (d) Serial dilutions of wild-type, 1343D::ura4 , 1343D and tgp1D1343D double mutant spotted on non-selective YES medium or in the presence of TBZ (20 mg ml 1), HU (10 mM)

or caffeine (15 mM), respectively.

glycerophosphodiester membrane transporter (designated as tgp1) orthologous to the S. cerevisiae permease GIT1. As with

S. cerevisiae GIT1, the tgp1 gene is repressed when cells are grown in the presence of phosphate and induced upon phosphate starvation34,35. Northern analysis conrmed that tgp1 was indeed highly expressed in 1343D cells but not wild-type cells, both grown in the presence of phosphate (repressed condition; Fig. 1c).

To determine whether the drug sensitivity of 1343D cells is a direct result of increased tgp1 expression, the tgp1 gene was deleted from 1343D cells (tgp1D1343D). This manipulation restored TBZ, HU and caffeine tolerance to levels comparable with wild-type cells (Fig. 1d). We conclude that increased tgp1 expression is directly responsible for the drug-sensitivity phenotype of cells lacking ncRNA.1343.

Bidirectional lncRNA promoter upstream of tgp1 . Previous RNA-seq analysis indicates that an lncRNA is transcribed in the sense orientation upstream of tgp1 (refs 27,31). We identied two divergent transcriptional start sites arising within ncRNA.1343: one lncRNA transcribed towards the tgp1 gene (nc-tgp1) and the other in the opposite orientation (nc-1343; Fig. 2a;

Supplementary Fig. 3). lacZ reporter assays demonstrate that the bidirectional promoter drives greater levels of transcription in the nc-tgp1 direction (Supplementary Fig. 3). This nding is consistent with Rpb1 Chromatin Immunoprecipitation (ChIP) analysis showing that RNAPII is enriched over the nc-tgp1 transcription unit, while much lower RNAPII levels are detected on nc-1343 (Fig. 2b).

We next examined the regulation of the nc-1343 and nc-tgp1 transcripts produced from this bidirectional promoter. A B0.9 kb transcript for nc-1343 was readily detected in wild-type cells. The size and levels of the nc-1343 transcript increased in exosome defective (rrp6D) cells, but not cells lacking Mmi1 or Red1 (Fig. 2c,d; Supplementary Fig. 4). The lncRNA corresponding to nc-tgp1 was previously detected in rrp6D and red1D cells27.

We identied a consensus DSR motif for Mmi1 binding at position 820 nt within the nc-tgp1 transcript and RNA

IP (RIP) experiments conrmed a direct interaction between Mmi1 and the nc-tgp1 RNA (Supplementary Fig. 5). Northern analysis identied that an B1.9 kb nc-tgp1 RNA accumulates in rrp6D, mmi1D and red1D, but not in wild-type cells (Fig. 2e,f; Supplementary Fig. 4). Interestingly, a recent study found that the repressive lncRNA transcribed upstream of the phosphate-responsive pho1 gene in S. pombe also contains a

DSR motif and is targeted by Mmi1 for exosome-mediated degradation28, indicating that a similar regulatory mechanism might control expression of tgp1 and pho1 . In sum, both nc-1343 and nc-tgp1 transcripts are processed by the exosome, but only nc-tgp1 is regulated by Mmi1-mediated recruitment of the nuclear exosome.

A moderate increase in tgp1 transcript levels has previously been reported in cells lacking Mmi1 (ref. 23). In agreement with this, we detected a similar increase (approximated four-fold) in tgp1 transcript levels in mmi1D or exosome (rrp6D or dis3-54)

mutant cells by RTqPCR, however, this increase is signicantly less than the 450-fold upregulation of tgp1 observed in 1343D cells (Fig. 2g,h; Supplementary Fig. 4). Moreover, we failed to detect the tgp1 transcript in rrp6D or mmi1D cells by Northern analysis, indicating that tgp1 is not induced in the absence of these factors. Thus, Mmi1-mediated exosome degradation is not the predominant mechanism involved in tgp1 regulation.

tgp1 is repressed by the nc-tgp1 lncRNA. The presence of the unstable nc-tgp1 RNA upstream of tgp1 suggests that either nctgp1, nc-1343 or both regulate tgp1 expression. To test the involvement of these lncRNAs in tgp1 regulation, a series of strategic genetic manipulations were performed (Fig. 3a). Truncations of nc-1343 (that is, AD and BD) that retain its 50 end did not result in the drug-sensitivity phenotype presented by 1343D cells (Fig. 3b) and, similarly, did not induce tgp1 expression (Fig. 3c). This indicates that full-length nc-1343 is not required for tgp1 repression. We next tested if nc-tgp1 is involved in repressing tgp1 . Our analyses show that transcription of nc-tgp1 starts within the encoded ncRNA.1343 transcription unit (Supplementary Fig. 3). Thus, deletion of the entire locus (1343D)

removes the nc-tgp1 promoter, and the 50 end of its transcript, resulting in the observed loss of nc-tgp1 expression (Figs 2f and 3c). The AD and BD truncations of nc-1343, which retain the nc-tgp1 promoter, do not affect nc-tgp1 transcription or relieve repression of tgp1 . In contrast, interruption of the nc-tgp1 transcription unit by insertion of the ura4 marker gene (nc-tgp1:ura4 ) prevented nc-tgp1 transcription, induced tgp1 expression to levels observed in 1343D levels and increased sensitivity of these cells to TBZ, HU and caffeine (Fig. 3b,c). These analyses demonstrate that it is nc-tgp1, not nc-1343, that is critical for repressing tgp1 in the presence of phosphate.

Phosphate starvation induces tgp1 by repressing nc-tgp1. Upon phosphate starvation of ssion yeast, several genes involved in the phosphate response are induced (including tgp1 and pho1 ) (ref. 35). To determine how the transcription of nc-tgp1 is altered in response to phosphate and how it might inuence tgp1 expression we assessed expression in phosphate-rich

NATURE COMMUNICATIONS | 5:5576 | DOI: 10.1038/ncomms6576 | http://www.nature.com/naturecommunications

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6576

352 kb 353 kb 354 kb

355 kb

tgp1+

ncRNA. 1343

nc-1343

nc-tgp1

nc-tgp1 nc-1343

%IP

ChIP: Rpb1

543210 1 2 3 4 5 6

RPKM

500

300

Primer pairs:

Northern probes:

Wild-type

tgp1+

2 3 4 5 6

rrp6

mmi1

Wild-type

rrp6

1343

mmi1

Wild-type

rrp6

1343

mmi1

1.5 kb 2.5 kb

1.5 kb

EtBr

2 kb

1 kb

EtBr

nc-1343

IncRNA

3kb

2 kb

EtBr

2.5 kb

nc-tgp1 IncRNA

tgp1+

mRNA

RT-qPCR

Wild-type

rrp6 1343

0 nc-1343 nc-tgp1 tgp1+

mmi1

RT-qPCR

4 15

100 80 60 40 20

RT-qPCR

Relative

expression

Relative

expression

Relative

expression

mmi1

Wild-type

rrp6 1343

mmi1

Wild-type

rrp6 1343

Figure 2 | Two distinct lncRNAs are transcribed from a bidirectional promoter upstream of tgp1 . (a) Previously published strand-specic RNA-Seq analysis (Rhind et al.,31) upstream of SPBC1271.09/tgp1 , represented as reads per kilobase per million (RPKM). Location of qPCR primer pairs and probes for Northern analysis are shown below. (b) Rbp1 ChIPqPCR experiments performed in wild-type cells. (c,e,g) Northern analysis of nc-1343, nc-tgp1 and tgp1 transcript levels in wild-type, rrp6D, mmi1D and 1343D, respectively. (d,f,h) RTqPCR experiments measured nc-1343, nc-tgp1 and tgp1 transcript levels in wild-type, rrp6D, mmi1D and 1343D, respectively. Error bars represent s.e.m. resulting from at least three independent replicates.

( PO4) and phosphate-deprived ( PO4) conditions. As

expected, the levels of tgp1 and the pho1 control increased upon phosphate starvation (Fig. 4a,b). In contrast, the levels of both nc-tgp1 and nc-1343 RNAs decreased signicantly in the absence of phosphate (Fig. 4a; Supplementary Fig. 6). The observed reduction in nc-tgp1 levels is consistent with a situation whereby loss or reduction of nc-tgp1 transcription permits tgp1 induction. In agreement with this, signicantly less RNAPII associates with the nc-tgp1 transcription unit in both phosphate-starved wild-type cells and phosphate-replete 1343D cells, which do not transcribe nc-tgp1 (Fig. 4c). Therefore, preventing nc-tgp1 transcription, even in phosphate-rich medium, recapitulates the changes in RNAPII occupancy that normally accompany tgp1 induction upon phosphate deprivation.

RNAi-directed heterochromatin does not repress tgp1 . Cells with defective exosome function (rrp6D) accumulate non-coding RNAs, some of which have been reported to attract Mmi1-dependent RNA elimination factors, along with RNA interference (RNAi) components and the Clr4 H3K9 methyltransferase, leading to the formation of transiently regulated HOODs (heterochromatin domains)25. The regions containing the tgp1 and pho1 genes are included in HOOD-17 and HOOD-24, respectively, and both form a region of Mmi1-directed transient heterochromatin in rrp6D cells24,27. The nc-tgp1 transcript is clearly regulated by Mmi1-directed exosome degradation

(Fig. 2e,f), however, we do not detect methylated H3K9 (H3K9me2) over the tgp1 , nc-tgp1 or nc-1343 genes within

HOOD-17 in wild-type cells (Fig. 5a). Likewise, only very low levels of H3K9me2, slightly above background in cells lacking the H3K9 methyltransferase (clr4D), could be detected on the pho1 gene and the upstream Mmi1-targeted lncRNA (nc-pho1) within

HOOD-24. Moreover, this low level of H3K9me2 did not drop appreciably upon induction of tgp1 and pho1 ( PO4;

Fig. 5a). Equivalent background levels of H3K9me2 were detectable on another Mmi1-targeted lncRNA gene (sme2 )

and the highly expressed actin gene (act1 ). In contrast, H3K9me2 was B100-fold enriched over the centromeric outer repeats (dg) in wild-type cells, but reduced to background in clr4D cells, indicating that H3K9-methylated chromatin had been efciently immunoprecipitated. In addition, the transcript levels of tgp1, nc-tgp1, nc-1343, pho1 and nc-pho1 were unaffected by loss of RNAi (for example, ago1D or dcr1D) or heterochromatin components (for example, clr4D or swi6D) (Fig. 5b; Supplementary Fig. 7a). Nor were the kinetics of tgp1 or pho1 induction following phosphate starvation altered in cells lacking heterochromatin (Supplementary Fig. 7b,c). In contrast, nc-tgp1, nc-pho1 and sme2 RNA levels were clearly elevated in cells lacking Mmi1-mediated exosome degradation (mmi1D and rrp6D). Thus, although H3K9me2 accumulates at particular regions in rrp6D cells (for example, HOOD-17: tgp1 and HOOD-24: pho1 ), we conclude that RNAi and heterochromatin play no appreciable role in regulating these

4 NATURE COMMUNICATIONS | 5:5576 | DOI: 10.1038/ncomms6576 | http://www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6576 ARTICLE

Wild-type

YES TBZ

RT-qPCR

nc-1343

nc-tgp1 1343

tgp1+

RT-qPCR

65 45 25

+PO4 PO4

ura4+

nc-tgp1:ura4+

Relative expression

%IP

3 kb

Wild-type

1343

+ +

PO4

2.5 kb

EtBr

tgp1+ mRNA

1271.10c tgp1+ nc-tgp1

Primer pairs: 1

HU Caffeine

pho1+

tgp1+

nc-tgp1

nc-1343

1343

nc-1343

tgp1+

nc-tgp1

nc-tgp1:ura4+

2 3

ChIP: Rpb1 Wild-type (+PO4)

Wild-type (PO4)

4 5 6

100

Relative expression

Wild-type

1343

1343 (+PO4)

nc-tgp1:ura4+

1 0

1 2 3 4 5 6

tgp1+ nc-tgp1 nc-1343

Figure 3 | nc-tgp1, not nc-1343, represses tgp1 to confer drug tolerance. (a) Schematic diagram indicating strategic manipulations of lncRNAs upstream of tgp1 including 1343D, shorter deletions of ncRNA.1343(AD and BD) and ura4 integration within the nc-tgp1 lncRNA locus

(nc-tgp1:ura4) in wild-type background. (b) Serial dilutions of wild-type, 1343D, AD, BD and nc-tgp1:ura4 were spotted on non-selective YES medium or in the presence of TBZ (20 mg ml 1), HU (10 mM) or caffeine (15 mM), respectively. (c) RTqPCR experiments measured tgp1, nc-tgp1 and nc-1343 transcript levels in wild-type, 1343D, AD, BD and nc-tgp1:ura4 cells. Error bars represent s.e.m. resulting from three independent replicates.

Figure 4 | Phosphate starvation induces tgp1 and reduces lncRNA transcription. (a) RTqPCR experiments measured tgp1 , nc-tgp1 and nc-1343 transcript levels in wild-type cells grown in phosphate-rich medium ( PO4) or in the absence of phosphate ( PO4). pho1 is a positive

control for phosphate starvation. (b) Northern analysis of tgp1 in wild-type cells grown in the presence or absence of phosphate and 1343D grown in the presence of phosphate. (c) Rbp1 ChIPqPCR experiments performed in wild-type cells grown in the presence or absence of phosphate and 1343D grown in the presence of phosphate. Error bars represent s.e.m. resulting from three independent replicates.

genes under normal physiologically repressive conditions or during their induction.

nc-tgp1 prevents Pho7 transcription factor binding. The above analyses indicate that nc-tgp1 is transcribed into the tgp1 promoter and suggest that production of this upstream lncRNA represses tgp1 expression. We therefore investigated if transcription of nc-tgp1 interferes with the induction mechanism of tgp1 in response to phosphate starvation. The Pho7 transcription factor has previously been shown to engage phosphate-response gene promoters in phosphate-starved cells35,36. Our ChIP analyses conrmed that Pho7green uorescent protein (Pho7GFP) accumulates on the pho1 promoter in phosphate-depleted cells (Supplementary Fig. 8). In addition, Pho7GFP levels accumulate over the region upstream of tgp1 when activated (Fig. 6a). However, in cells unable to transcribe nc-tgp1 (1343D), higher levels of Pho7GFP associate with the region upstream of tgp1 even in repressive conditions (that is, PO4).

We conclude that loss of nc-tgp1 expression due to phosphate starvation or by preventing production of this lncRNA (for example, 1343D) allows Pho7 binding and subsequent tgp1 induction.

Active RNAPII promoters display reduced nucleosome density37. lncRNA transcription over promoters can increase nucleosome density and prevent gene induction8,10,20. We

found that histone H3 levels were greater over the tgp1 gene and upstream region when it is repressed ( PO4) compared with

when it is expressed ( PO4; Fig. 6b). In contrast, H3 levels over

control loci (act1, sme2 and dg repeats) were unaffected by phosphate availability. Thus, upstream transcription appears to alter nucleosome density over the tgp1 promoter and thereby occlude Pho7 binding. Likewise, a considerable drop in H3 levels was observed on the pho1 gene and nc-pho1 lncRNA region upstream in phosphate-poor conditions, implying a similar mechanism may also operate to regulate the expression of pho1 . We conclude that transcription of the upstream lncRNA inhibits expression of tgp1 by a transcriptional interference mechanism that alters the chromatin landscape, preventing access to the key phosphate-responsive transcription factor Pho7.

To directly test if transcriptional interference of tgp1 by nctgp1 is responsible for tgp1 repression, we replaced the nc-tgp1 promoter with the strong, thiamine-regulated nmt1 promoter (nmt1-nc-tgp1) (Fig. 7a). Transcription of nc-tgp1 from the nmt1 promoter is rendered unresponsive to phosphate. Instead, nc-tgp1 is repressed or derepressed in the presence or absence of thiamine, respectively. When nc-tgp1 was transcribed from the nmt1 promoter, tgp1 remained repressed regardless of phosphate availability (Fig. 7b). In contrast, repression of nmt1-driven nc-tgp1 by thiamine resulted in the induction of tgp1 expression in phosphate-rich media and consequently caused drug sensitivity (Fig. 7b,c). In addition, H3 levels over the region upstream of tgp1 were high when nc-tgp1 was transcribed and

NATURE COMMUNICATIONS | 5:5576 | DOI: 10.1038/ncomms6576 | http://www.nature.com/naturecommunications

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6576

tgp1+ nc-tgp1

nc-1343

Primer pairs: 1

2 3 4 5 6

ChIP: H3K9me2 RT-qPCR2

1.5

0.5

Wild-type (+PO4)

Wild-type

Relative expression

9 5

clr4 swi6 dcr1 ago1

%IP

0.6

Wild-type (PO4) clr4

0.4

0.2

nc-pho1

sme2+

0 1 2 3 4 5 6

act1+

pho1+

tgp1+

nc-tgp1

nc-1343

Figure 5 | tgp1 is not regulated by RNAi/heterochromatin. (a) H3K9me2 ChIPqPCR experiments performed in the presence or absence of phosphate. clr4D was used as a negative control. The euchromatic actin gene (act1 ) and centromeric dg repeats (dg) are positive and negative controls for heterochromatin. pho1 is a phosphate-regulated gene repressed by nc-pho1, a lncRNA target of Mmi1. sme2 is another lncRNA target of Mmi1.

H3K9me2 to bulk H3 ratio has not been presented due to background methyl H3K9 levels detected at these loci. (b) RTqPCR experiments measured tgp1 , nc-tgp1 and nc-1343 transcript levels in wild-type cells and cells lacking factors involved in heterochromatin formation and stability, respectively. Error bars represent s.e.m. resulting from at least three independent replicates.

0.16

ChlP: Pho7-GFP tgp1+ promoter

+PO4

PO4

Primerpairs: 2 3 4 5 6

ChIP: Histone H3

4 3 2 1 0

1 2 3 4 5 6

act1+

tgp1+ nc-tgp1

nc-1343

0.120.080.04 0

%IP

+PO4

PO4

%IP

Untagged

Wild-type

1343

pho1

nc-pho1

sme2+

Pho7-GFP

Figure 6 | nc-tgp1 transcription prevents stable Pho7 binding and increases nucleosome density upstream of tgp1. (a) GFP ChIPqPCR experiments were performed in the presence or absence of phosphate in cells with C-terminally GFP-tagged Pho7. An untagged strain was used as a negative control.

Primer pair #3 was used to detect Pho7 binding at the tgp1 promoter. (b) Nucleosome density was measured by histone H3 ChIPqPCR experiments in wild-type cells grown in the presence or absence of phosphate. Error bars represent s.e.m. resulting from three independent replicates.

nmt1-nc-tgp1

PMG

25 RT-qPCR

Thiamine/ +PO4 Thiamine/ PO4 +Thiamine/ +PO4

Relative

expression

20 15 10

5 0

nmt1

Primer pairs:

tgp1+ nc-tgp1

2 3 4 5

tgp1+

nc-tgp1

TBZ HU CAF

ChIP: Histone H3

Thiamine

+ Thiamine

1343 nmt1-nc-tgp1

Thiamine + Thiamine

%IP

1343 nmt1-nc-tgp1

0 1 2 3 4 5

Figure 7 | nmt1 controlled nc-tgp1 alters drug tolerance in response to thiamine. (a) Schematic diagram of nc-tgp1 under the control of the strong, thiamine-repressible nmt1 promoter. (b) RTqPCR experiments measured tgp1 and nc-tgp1 levels in response to thiamine and phosphate availability using nmt1-nc-tgp1 cells. (c) Serial dilutions of wild-type, 1343D and nmt1-nc-tgp1 cells were spotted on non-selective PMG medium or in the presence of TBZ, HU or caffeine, respectively, with or without thiamine as indicated. (d) H3 ChIPqPCR experiments in nmt1-nc-tgp1 cells grown in the presence or absence of thiamine. Error bars represent s.e.m. resulting from three independent replicates.

reduced when nc-tgp1 was repressed by thiamine (Fig. 7d). Lastly, exogenous expression of full-length nc-tgp1 from a plasmid failed to repress tgp1 , ruling out the possibility that nc-tgp1 operates in trans (Supplementary Fig. 9). Collectively, these ndings

conrm that it is the transcription of nc-tgp1 over the tgp1 promoter that alters nucleosome density to regulate tgp1 induction (see Fig. 8) and, as a consequence, drug tolerance of ssion yeast cells.

6 NATURE COMMUNICATIONS | 5:5576 | DOI: 10.1038/ncomms6576 | http://www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6576 ARTICLE

+PO4

(Repressed)

PO4 or

Preventing nc-tgp1 transcription (induced)

tgp1+

Pol II

tgp1+

nc-tgp1 nc-tgp1

Pol II

Pho7

Figure 8 | Model for transcriptional interference at tgp1 . The presence of phosphate induces transcription of an unstable lncRNA (nc-tgp1). lncRNA transcription increases nucleosome density, occludes Pho7 transcription factor binding and thus represses tgp1 expression. nc-tgp1 transcription is reduced following phosphate starvation, decreasing nucleosome density, allowing Pho7 to stably engage the tgp1 promoter and induce tgp1 expression.

DiscussionAn increasing number of lncRNAs have been shown to tightly regulate eukaryotic gene expression following intra-/extra-cellular environment changes that require rapid, integrated responses at the level of transcription2. In S. pombe, for example, the balance of antisense lncRNAs and sense transcription controls various stress-response pathways33,38. However, little is known about the majority of S. pombe intergenic lncRNAs. Here we selected and deleted eight stable, discrete lncRNAs in S. pombe that show conserved synteny in at least two of the three other known Schizosaccharomyces species. Excluding the ter1 control, only deletion of ncRNA.1343 exhibited a denitive phenotype: sensitivity to various compounds due to induction of a nearby phosphate-responsive permease gene (tgp1 ). Closer inspection revealed that the ncRNA.1343 promoter is bidirectional.

Furthermore, transcription from this bidirectional promoter favours the production of a previously unannotated and unstable lncRNA (nc-tgp1) towards the tgp1 gene under repressive conditions.

Recent studies in ssion yeast have implicated lncRNAs in directing repression of specic genes by a mechanism involving transient RNAi-dependent heterochromatin formation27. For example, the Mmi1-targeted lncRNA upstream of pho1 has recently been reported to recruit RNAi-directed heterochromatin to repress pho1 in response to phosphate availability28.

However, these ndings differ from genome-wide H3K9me2 mapping which show that tgp1 and pho1 , both of which are regulated by upstream lncRNAs that are targeted for exosome-mediated degradation by Mmi1 (Fig. 2; ref 28), only accumulate RNAi-directed H3K9me2 in mutants with defective RNA processing/degradation (for example, rrp6D) and not in wild-type cells grown under repressive phosphate-rich conditions24.

The signicance of rrp6D-dependent heterochromatin at the tgp1 and pho1 genes is therefore unclear. Cells lacking Rrp6 accumulate aberrant RNAs and exhibit disrupted heterochromatin globally, including signicantly decreased H3K9me2 over centromeric repeats39. Therefore caution must be exercised when interpreting the analyses of mutants with such severe defects in RNA processing/degradation. Importantly, we do not detect signicant levels of H3K9me2 enrichment on the tgp1 and pho1 promoters/genes in wild-type cells under repressive (phosphate-rich) conditions. We cannot exclude the possibility that distinct assay conditions in a previous report allowed detection of low H3K9me2 levels on the pho1 promoter when repressed28, however, the consequence of such H3K9me2 remains uncertain given that our analyses show that the expression of pho1 or tgp1 is unaffected by loss of RNAi/

heterochromatin. We note that our ndings are in agreement with previous expression proling analyses, showing unaltered

tgp1 and pho1 levels in S. pombe cells lacking RNAi/ heterochromatin40. In contrast, transcripts arising from bone de heterochromatin in centromeric outer repeats are clearly elevated when RNAi/heterochromatin is defective. Thus, our analyses indicate that the repression of both tgp1 and pho1 is unlikely to involve regulated heterochromatin in wild-type cells. Instead, we favour a model whereby tgp1 and pho1 are repressed by a transcriptional interference mechanism.

Transcriptional interference is well-established in many systems. In the bacterium Escherichia coli, the gene encoding the clr transcriptional activator is repressed in response to nitrogen starvation by the act of lncRNA transcription from an alternate upstream promoter41. In the single celled eukaryoteS. cerevisiae, which lacks RNAi and heterochromatin, transcription of the SRG1 lncRNA into the SER3 promoter, or heterologous promoters, was found to alter nucleosome density and interfere with transcription factor binding810. Similarly, in S. cerevisiae, non-coding transcription over the IME1 (ref. 20), GAL7 (ref. 42) and FLO11 (ref. 43) promoters prevent gene induction. Analogous mechanisms have also been reported in multicellular eukaryotes. For example, the Drosophila Ubx gene44, the human dihydrofolate reductase gene45 and the imprinted Igf2r gene in mammals46 are repressed independent of RNAi or transient heterochromatin formation by non-coding transcription into their respective promoters. These examples illustrate that transcriptional interference is a simple, conserved mechanism for modulating specic genes without requiring additional trans-acting regulatory factors. Our results are consistent with both nc-tgp1 and nc-pho1 mediating repression of downstream genes (tgp1 and pho1 , respectively) by transcriptional interference, not by the formation of transient heterochromatin.

We base this conclusion on our ndings that: (i) tgp1 and pho1 expression is unaffected by loss of RNAi/heterochromatin;

(ii) H3K9me2 is not associated with tgp1 or pho1 loci in wild-type cells; (iii) nc-tgp1 transcription declines when tgp1 is induced ( PO4); (iv) loss of the nc-tgp1 transcript allows

induction of tgp1 under normally repressive ( PO4)

conditions (similarly, loss of lncRNA transcription upstream induces pho1 in repressive medium27,28); (v) transcription of nc-tgp1 by a thiamine-repressible promoter brings tgp1 under the control of thiamine, rather than phosphate; (vi) RNAPII and nucleosome density is increased over the tgp1 promoter region when the repressive nc-tgp1 RNA is transcribed and (vii) the

Pho7 activator binds the tgp1 promoter region when nc-tgp1 transcription is lost.

Genome-wide RNA sequencing has allowed the detection of a large number of lncRNAs in a variety of species. However, it remains unclear how many of these lncRNA are functional transcripts that act to inuence gene expression and/or chromatin landscapes. Examples such as Xist RNA in mammals and roX RNAs in Drosophila represent functional transcripts that are critical for mediating dosage compensation by altering chromatin status and expression levels from sex chromosomes47. However, enthusiasm for lncRNA function has been somewhat dampened by reports showing that the ablation in animal models of some of the best-characterized lncRNAs (for example, HOTAIR, MALAT1, Kcnq1ot1, NEAT1) exhibited less dramatic or undetectable phenotypes4853. Of the discrete stable lncRNAs that we deleted in ssion yeast, only one (ncRNA.1343) had an obvious phenotype in the growth conditions tested. Detailed analysis was required to reveal that deletion of ncRNA.1343 actually affected expression of a divergent unstable lncRNA (nc-tgp1) transcribed in the opposite orientation as the annotated locus. Only after further manipulation and analyses could we conclude that the expression of nc-tgp1 interferes with the expression of tgp1 downstream. The fact that the unstable

NATURE COMMUNICATIONS | 5:5576 | DOI: 10.1038/ncomms6576 | http://www.nature.com/naturecommunications

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6576

nc-tgp1 transcript is the functional partner of the apparently nonfunctional stable nc-1343 RNA transcribed from the same bidirectional promoter demonstrates the importance of comprehensive analyses of ncRNAs and the consequences of their deletion. Based on our analyses. we surmise that the low level expression of nc-1343 represents transcriptional noise, resulting as a byproduct of ample nc-tgp1 transcription. The syntenic conservation of ncRNA.1343 within the Schizosaccharomyces genus31 hints at the possibility of a conserved regulatory mechanism that involves lncRNA transcription into the promoter region of tgp1 in related species. Thus, although genome-wide approaches can rapidly catalogue the presence and response of various lncRNAs to different conditions, much more detailed locus-specic analyses is required to pinpoint the function of each individual lncRNA with respect to cis regulation of nearby genes or trans regulation of genes at distal loci.

Methods

Yeast strains, plasmids and standard techniques. S. pombe strains used in this study are listed in Supplementary Table 1. Standard methods were used for ssion yeast growth, genetics and manipulations54. All strains were grown in Yeast extract plus supplement medium (YES), unless otherwise indicated. For phosphate starvation experiments, cells were grown to mid-log phase in YES medium, washed twice in dH2O, and then grown for indicated times in Pombe minimal glutamate (PMG) synthetic medium without Na2HPO4 ( PO4). Genetic deletions and

protein tagging were carried out by lithium acetate transformation. All genetic modications were conrmed by colony PCR. Plasmids were transformed by electroporation. Selections were performed on PMG/agar plates with according auxotrophy or on YES/agar plates with appropriate antibiotic(s) and grown at32 C. Serial (1:4) dilutions of equal number of cells were spotted onto YES/agar and grown at 32 C, unless indicated otherwise. For drug-sensitivity experiments, cells were spotted onto YES/agar or PMG/agar with DMSO or TBZ (20 mg ml 1),

HU (10 mM), caffeine (15 mM) and H2O2 (1 mM). For ultraviolet-sensitivity experiments, spotted cells were ultraviolet-irradiated at 80 J m 2 with a

Stratalinker UV Crosslinker and grown in the dark at 25 C. The plasmids containing lacZ under the control of the nc-tgp1 and nc-1343 bidirectional promoter were cloned as follows. The non-coding promoter was amplied from S. pombe genomic DNA in both orientations (using lacZ_1_F/lacZ_1_R and lacZ_2_F/lacZ_2_R primer pairs; see Supplementary Table 2) and ligated into pREP vector containing lacZ using PstI/SalI restriction sites. To test if nc-tgp1 can repress tgp1 in trans, the nc-tgp1 transcription unit was amplied from S. pombe genomic DNA (using nc-tgp1_SalI_F and nc-tgp1_XmaI_R primer pairs, see

Supplementary Table 2) and ligated into pREP3x using SalI/XmaI restriction sites.

Liquid assay for b-galactosidase activity. Assays for b-galactosidase activity were performed as described55. Briey, yeast containing vectors expressing lacZ under the control of various promoters were grown to log phase (OD595 of B0.5) in

selective media. Cells were permeabolized by SDS/chloroform. Cell extracts were equilibrated at 30 C for 5 min before the addition of ortho-Nitrophenyl-bgalactoside (ONPG). The reaction was stopped with Na2CO3 once the solution turned yellow and elapsed time was recorded. Cell debris was spun and the OD420

was measured. Units were calculated as follows: Units/OD 1000 (OD420/

Volume Time OD595).

Chromatin and RIP. Cells were grown to mid-log phase at 32 C in YES. For phosphate starvation experiments, cells in mid-log phase were washed twice in dH2O before being grown in PMG ( PO4) for 4 h. ChIP was performed

essentially as described12. Briey, cells were xed with 1% paraformaldehyde for 15 min at room temperature. Cells were lysed by bead beating (Biospec Prodcutes) and sonicated using a Bioruptor (Diagenode) sonicator at 5 C on high for a total of 20 min (30 s ON/OFF cycles). Five microlitres of Rpb1 antibody (#2629; Cell Signaling), 2 ml GFP antibody (G10362; Life Technologies), 2 ml H3 antibody (ab1791;

Abcam) and 1 ml of H3K9me2 antibody (m5.1.1; ref. 56) were used for IPs. RIP experiments were performed essentially as described13. Hisx6-TEV-Protein A-tagged Mmi1 was captured from cell lysate with IgG Dynabeads (Life Technologies). Mmi1-bound RNA was isolated by phenol-chloroform extraction, DNase treated and reverse transcribed. Quantitative analysis was performed by qPCR.

RNA analysis. RNA was isolated from S. pombe using RNeasy Mini- or Midi-Kits as per manufacturers instructions (Qiagen). For RTqPCR experiments, rst strand complementary DNA synthesis was performed on Turbo DNase (Life Technologies) treated RNA using random hexamers and Superscript III (Invitrogen) as per manufacturers instructions. Negative controls lacking RT were performed alongside all RTqPCR experiments. Northern analysis of long non-coding

transcripts was performed using UTP-[a32P]-labelled RNA probes as described57. Transcription start sites were mapped using the SMARTer RACE complementary DNA Amplication Kit as per manufacturers instructions (Clontech).

Quantitative real-time PCR. Primers used in this study are listed in Supplementary Table 2. qPCR was performed using SYBR Green on a Roche Lightcycler. Data was analyzed with LightCycler 480 Software 1.5.0.39. RTqPCR levels were calculated by normalizing product of interest to an internal reference gene (act1 ). Expression levels were expressed relative to levels detected in wild-type cells. ChIP enrichments were calculated as the ratio of product of interest from IP sample normalized to the corresponding input sample and expressed as %IP. Error bars represent s.e.m., resulting from at least three independent replicates.

References

1. Guttman et al. Chromatin signature reveals over a thousand highly conserved large non-coding RNAs in mammals. Nature 458, 223227 (2009).

2. Ponting, C. P., Oliver, P. L. & Reik, W. Evolution and functions of long noncoding RNAs. Cell 136, 629641 (2009).

3. Mitchell, P., Petfalski, E., Shevchenko, A., Mann, M. & Tollervey, D. The exosome: a conserved eukaryotic RNA processing complex containing multiple 30-50 exoribonucleases. Cell 91, 457466 (1997).

4. Struhl, K. Transcriptional noise and the delity of initiation by RNA polymerase II. Nat. Struct. Mol. Biol. 14, 103105 (2007).

5. Wilusz, J. E., Sunwoo, H. & Spector, D. L. Long noncoding RNAs: functional surprises from the RNA world. Genes Dev. 23, 14941504 (2009).

6. Li, B., Carey, M. & Workman, J. L. The role of chromatin during transcription. Cell 128, 707719 (2007).

7. Hirota, K. et al. Stepwise chromatin remodeling by a cascade of transcription intiation of non-coding RNAs. Nature 456, 130134 (2008).

8. Hainer, S. J., Pruneski, J. A., Michell, R. D., Monteverde, R. M. & Martens, J. A. Intergenic transcription causes repression by directing nucleosome assembly. Genes Dev. 25, 2940 (2011).

9. Martens, J. A., Laprade, L. & Winston, F. Inter Thebault genic transcription is required to repress Saccharomyces cerevisiae SER3 gene. Nature 429, 571574 (2004).

10. Thebault, P. et al. Transcription regulation by the noncoding RNA SRG1 requires Spt2-dependent chromatin deposition in the wake of RNA polymeraseII. Mol. Cell Biol. 31, 12881300 (2011).11. Fejes-Toth, K. et al. Post-transcriptional processing generates a diversity of 50-modied long and short RNAs. Nature 457, 10281032 (2009).

12. Bayne, E. H. et al. Stc1: a critical link between RNAi and chromatin modication required for heterochromatin integrity. Cell 140, 666677 (2010).

13. Motamedi, M. R. et al. Two RNAi complexes, RITS and RDRC, physically interact and localize to noncoding centromeric RNAs. Cell 119, 789802 (2004).

14. Verdel, A. et al. RNAi-mediated targeting of heterochromatin by the RITS comples. Science 303, 672676 (2004).

15. Volpe, T. A. et al. Regulation of heterochromatin silencing and histone H3 lysine-9 methylation by RNAi. Science 297, 18331837 (2002).

16. Rinn, J. L. et al. Functional demarcation of active and silent chromatin domains in human HOX loci by non-coding RNAs. Cell 129, 13111323 (2007).

17. rom, U. A. et al. Long noncoding RNAs with enhancer-like function in human cells. Cell 143, 4658 (2010).

18. Camblong, J., Iglesias, N., Fickentscher, C., Dieppois, G. & Stutz, F. Antisense RNA stabilization induces transcriptional gene silencing via histone deacetylation in S. cerevisiae. Cell 131, 706717 (2007).

19. Houseley, J., Rubbi, L., Grunstein, M., Tollervey, D. & Vogelauer, M. A ncRNA modulates histone modication and mRNA induction in the yeast GAL gene cluster. Mol. Cell 32, 685695 (2008).

20. van Werven, F. J. et al. Transcription of two long non-coding RNAs mediates mating-type control of gametogenesis in budding yeast. Cell 150, 11701181 (2012).

21. Heo, J. B. & Sung, S. Vernalization-mediated epigenetic silencing by a long intronic noncoding RNA. Science 331, 7679 (2011).

22. Wang, K. C. et al. A long noncoding RNA maintains active chromatin to coordinate homeotic gene expression. Nature 472, 120124 (2011).

23. Hiriart, E. et al. Mmi1 RNA surveillance machinery directs RNAi complex RITS to specic meiotic genes in ssion yeast. EMBO J. 31, 22962308 (2012).

24. Yamanaka, S. et al. RNAi triggered by specialized machinery silences developmental genes and retrotransposons. Nature 493, 557560 (2013).

25. Zofall, M. et al. RNA elimination machinery targeting meiotic mRNAs promotes facultative heterochromatin formation. Science 335, 96100 (2012).

26. Harigaya, Y. et al. Selective elimination of messenger RNA prevents an incidence of untimely meiosis. Nature 442, 4550 (2006).

27. Lee, N. N. et al. Mtr4-like protein coordinates nuclear RNA processing for heterochromatin assembly and for telomere maintenance. Cell 155, 10611074 (2013).

8 NATURE COMMUNICATIONS | 5:5576 | DOI: 10.1038/ncomms6576 | http://www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms6576 ARTICLE

28. Shah, S., Wittmann, S., Kilchert, C. & Vasilieva, L. lncRNA recruits RNAi and the exosome to dynamically regulate pho1 expression in response to phosphate levels in ssion yeast. Genes Dev. 28, 213244 (2014).

29. Pang, K. C., Frith, M. C. & Mattick, J. S. Rapid evolution of noncoding RNAs: lack of conservation does not mean lack of function. Trends Genet. 22, 15 (2006).

30. Ulitsky, I., Shkumatava, A., Jan, C. H., Sive, H. & Bartel, D. P. Conserved function of lincRNAs in vertebrate embryonic development despite rapid sequence evolution. Cell 147, 15371550 (2011).

31. Rhind, N. et al. Comparative functional genomics of the ssion yeasts. Science 332, 930936 (2011).

32. Wilhelm, B. T. et al. Dynamic repertoire of a eukaryotic transcriptome surveyed at single-nucleotide resolution. Nature 453, 12391243 (2008).

33. Leong, H. S. et al. A global non-coding RNA system modulates ssion yeast protein levels in response to stress. Nat. Commun. 5, 3947 (2014).34. Almaguer, C., Mantella, D., Perez, E. & Patton-Vogt, J. P. Inositol and phosphate regulate GIT1 transcription and glycerophophoinositol incorporation in Saccharomyces cerevisiae. Eukaryot. Cell 2, 729736 (2003).

35. Carter-OConnell, I., Peel, M. T., Wykoff, D. D. & OShea, E. Genome-wide characterization of the phosphate starvation response in Schizosaccharomyces pombe. BMC Genomics 13, 697 (2012).

36. Henry, T. C. et al. Systematic screen of Schizosaccharomyces pombe deletion collection uncovers parallel evolution of the phosphate signal transduction pathway in yeasts. Eukaryot. Cell 10, 198206 (2011).

37. Yuan, G. C. et al. Genome-scale identication of nucleosome positions in S. cerevisiae. Science 309, 626630 (2005).

38. Bitton, D. A. et al. Programmed uctuations in sense/antisense transcript ratios drive sexual differentiation in S. pombe. Mol. Syst. Biol. 7, 559 (2011).

39. Reyes-Turcu, F. E., Zhang, K., Zofall, M., Chen, E. & Grewal, S. I. Defects in RNA quality control factors reveal RNAi-independent nucleation of heterochromatin. Nat. Struct. Mol. Biol. 18, 11321138 (2011).

40. Hansen, K. R. et al. Global effects on gene expression in ssion yeast by silencing and RNA interference machineries. Mol. Cell Biol. 25, 590601 (2005).

41. Zafar, M. A., Carabetta, V. J., Mandel, M. J. & Silhavy, T. J. Transcriptional occlusion caused by overlapping promoters. Proc. Natl Acad. Sci. USA 111, 15571561 (2014).

42. Greger, I. H., Aranda, A. & Proudfoot, N. Balancing transcription interference and initation on the GAL7 promoter of Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA 97, 84158420 (2000).

43. Bumgarner, S. L. et al. Toggle involving cis-interfering noncoding RNAs controls variegated gene expression in yeast. Proc. Natl Acad. Sci. USA 106, 1832118326 (2009).

44. Petruk, S. et al. Transcriptional elongations of non-coding bxd RNAs promoted by the Trithorax TAC1 complex represses Ubx by a transcriptional interference mechanism. Cell 127, 12091221 (2006).

45. Martianov, I., Ramadass, A., Barros, A. S., Chow, N. & Akoulitchev, A. Repression of the human dihydrofolate reductase gene by a non-coding interfering transcript. Nature 445, 666670 (2007).

46. Latos, P. A. et al. Airn transcriptional overlap, but not its lncRNA products, induces imprinted Igf2r silencing. Science 338, 14691472 (2012).

47. Lee, J. T. & Bartolomei, M. S. X-activation, imprinting, and long non-coding RNAs in health and disease. Cell 152, 13081323 (2013).

48. Eimann, M. et al. Loss of the abundant nuclear non-coding RNA MALAT1 is compatible with life and development. RNA Biol. 9, 10761087 (2012).

49. Korostowski, L., Sedlak, N. & Engel, N. The Kcnq1ot1 long non-coding RNA affects chromatin and expression of Kcnq1, but does not regulate its imprinting in the developing heart. PLOS Genet. 8, e1002956 (2012).

50. Nakagawa, S. et al. Malat1 is not an essential component of nuclear speckles in mice. RNA 18, 14871499 (2012).

51. Schorderet, P. & Duboule, D. Structural and functional differences in the long non-coding RNA hotair in mouse and human. PLOS Genet. 7, e1002071 (2011).

52. Zhang, B. et al. The lncRNA Malat1 is dispensable for mouse development but its transcription plays a cis-regulatory role in the adult. Cell Rep. 2, 111123 (2012).

53. Nakagawa, S., Naganuma, T., Shioi, G. & Hirose, T. Paraspeckles are subpopulation-specic nuclear bodies that are not essential in mice. J. Cell Biol. 193, 3139 (2011).

54. Moreno, S., Klar, A. & Nurse, P. Molecular genetic analysis of ssion yeast Schizosaccharomyces Pombe. Methods Enzymol. 194, 795823 (1991).

55. Guarente, L. Yeast promoters and lacZ fusions designed to study expression of cloned genes in yeast. Methods Enzymol. 101, 181191 (1983).

56. Nakagawachi, T. et al. Silencing effect of CpG island hypermethylation and histone modications on O6-methylguanine-DNA methyltransferase (MGMT) gene expression in human caner. Oncogene 22, 88358844 (2003).

57. Buscaino, A. et al. Raf1 is a DCAF for the Rik1 DDB1-like protein and has separable roles in siRNA generation and chromatin modication. PLOS Genet. 8, e1002499 (2012).

Acknowledgements

We would like to thank Sandra Catania, Alison Pidoux, Manu Shukla and Sharon White for their technical expertise and input. We are grateful to Takeshi Urano for the H3K9me2 (5.1.1) antibody, Tomoyasu Sugiyama and Lidia Vasilieva for strains, and Steven West and Sander Granneman for critically evaluating the manuscript. R.A. is supported by the Darwin Trust of Edinburgh. The Centre for Cell Biology is supported by core funding from the Wellcome Trust (092076/Z/10/Z). P.T. is supported by European Commission Network of Excellence EpiGeneSys (HEALTH-F4-2010-257082) to R.C.A. R.C.A. is a Wellcome Trust Principal Research Fellow and this research was supported by the Wellcome Trust (095021/Z/10/Z).

Authors contributions

R.A. and R.C.A. conceived and designed the experiments. R.A. performed the experiments. P.T. performed bioinformatics analysis. R.A., P.T. and R.C.A. analyzed the data. R.A. and R.C.A. wrote the paper.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Ard, R. et al. Long non-coding RNA-mediated transcriptional interference of a permease gene confers drug tolerance in ssion yeast. Nat. Commun. 5:5576 doi: 10.1038/ncomms6576 (2014).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

NATURE COMMUNICATIONS | 5:5576 | DOI: 10.1038/ncomms6576 | http://www.nature.com/naturecommunications

Word count: 8008

Show less

Abstract

Translate

Most long non-coding RNAs (lncRNAs) encoded by eukaryotic genomes remain uncharacterized. Here we focus on a set of intergenic lncRNAs in fission yeast. Deleting one of these lncRNAs exhibited a clear phenotype: drug sensitivity. Detailed analyses of the affected locus revealed that transcription of the nc-tgp1 lncRNA regulates drug tolerance by repressing the adjacent phosphate-responsive permease gene transporter for glycerophosphodiester 1 (tgp1⁺ ). We demonstrate that the act of transcribing nc-tgp1 over the tgp1⁺ promoter increases nucleosome density, prevents transcription factor access and thus represses tgp1⁺ without the need for RNA interference or heterochromatin components. We therefore conclude that tgp1⁺ is regulated by transcriptional interference. Accordingly, decreased nc-tgp1 transcription permits tgp1⁺ expression upon phosphate starvation. Furthermore, nc-tgp1 loss induces tgp1⁺ even in repressive conditions. Notably, drug sensitivity results directly from tgp1⁺ expression in the absence of the nc-tgp1 RNA. Thus, transcription of an lncRNA governs drug tolerance in fission yeast.

Details

Title

Long non-coding RNA-mediated transcriptional interference of a permease gene confers drug tolerance in fission yeast

Author

Ard, Ryan; Tong, Pin; Allshire, Robin C

Pages

5576

Publication year

2014

Publication date

Nov 2014

Publisher

Nature Publishing Group

e-ISSN

20411723

Source type

Scholarly Journal

Language of publication

English

DOI

https://doi.org/10.1038/ncomms6576

ProQuest document ID

1628237817

Long non-coding RNA-mediated transcriptional interference of a permease gene confers drug tolerance in fission yeast

Jump to:

Full text

Abstract

Details

Suggested sources