Control of Cdc28 CDK1 by a Stress-Induced lncRNAet al., 2010). Thus, Hog1 plays a key role in the regulation of mRNA biogenesis (de Nadal et al., 2011; de Nadal and Posas, 2010; Martı´nez-Montan˜e´s

Please cite this article in press as: Nadal-Ribelles et al., Control of Cdc28 CDK1 by a Stress-Induced lncRNA, Molecular Cell (2014), http://dx.doi.org/10.1016/j.molcel.2014.01.006

Molecular Cell

Article

Control of Cdc28 CDK1 by a Stress-Induced lncRNAMariona Nadal-Ribelles,1,3 Carme Sole,1,3 Zhenyu Xu,2 Lars M. Steinmetz,2 Eulalia de Nadal,1,* and Francesc Posas1,*1Cell Signaling Unit, Departament de Ciencies Experimentals i de la Salut, Universitat Pompeu Fabra (UPF), 08003 Barcelona, Spain2EMBL Heidelberg, Meyerhofstrasse 1, 69117 Heidelberg, Germany3These authors contributed equally to this work*Correspondence: [email protected] (E.d.N.), [email protected] (F.P.)

http://dx.doi.org/10.1016/j.molcel.2014.01.006

SUMMARY

Genomic analysis has revealed the existence of alarge number of long noncoding RNAs (lncRNAs)with different functions in a variety of organisms,including yeast. Cells display dramatic changes ofgene expression upon environmental changes. Uponosmostress, hundreds of stress-responsive genesare induced by the stress-activated protein kinase(SAPK) p38/Hog1. Using whole-genome tiling arrays,we found that Hog1 induces a set of lncRNAs uponstress. One of the genes expressing a Hog1-depen-dent lncRNA in antisense orientation is CDC28, thecyclin-dependent kinase 1 (CDK1) that controls thecell cycle in yeast. Cdc28 lncRNAmediates the estab-lishment of gene looping and the relocalization ofHog1 and RSC from the 30 UTR to the +1 nucleosometo induce CDC28 expression. The increase in thelevels of Cdc28 results in cells able to reenter the cellcycle more efficiently after stress. This may representa general mechanism to prime expression of genesneeded after stresses are alleviated.

INTRODUCTION

The existence of long noncoding RNAs (lncRNAs) is widespread

in eukaryotes from yeast to mammals (Guttman and Rinn, 2012;

Jacquier, 2009). Long noncoding transcripts in yeast influence

gene expression, revealing a new layer of transcriptional regula-

tion (Wei et al., 2011; Wu et al., 2012). LncRNAs might regulate

transcription at multiple levels. Sense-oriented lncRNAs of

IMD2 and URA2 alter expression by transcriptional interference

and transcription start site (TSS) selection (Kuehner and Brow,

2008; Thiebaut et al., 2008). Expression of lncRNAs can also

trigger changes in chromatin epigenetic state or nucleosome

occupancy (Houseley et al., 2008; Kim et al., 2012; Margaritis

et al., 2012; Pinskaya et al., 2009; vanWerven et al., 2012; Hainer

et al., 2011; Uhler et al., 2007). Although in a few cases expres-

sion of specific lncRNAs alters normal mRNA biogenesis, the

general biological relevance and functionality of lncRNAs re-

mains elusive. Remarkably, changes in nutrient availability result

in changes in lncRNA expression (Xu et al., 2009, 2011), indi-

cating that environmental insults and signal transduction path-

ways might affect lncRNA transcription.

Exposure of cells to stress requires immediate and specific

cellular responses for proper adaptation (Hohmann et al.,

2007). Thus, environmental insults require adaptive responses

for maximal cell survival (de Nadal et al., 2011). Stress-activated

protein kinases (SAPKs) serve to respond and adapt to extracel-

lular changes. Exposure of yeast to high osmolarity results in

activation of the p38-related Hog1 SAPK (Saito and Posas,

2012), which is essential to control cell cycle (Clotet and Posas,

2007; Duch et al., 2013) and gene expression (de Nadal and

Posas, 2010).

The Hog1 SAPK acts in multiple stages of the cell cycle by tar-

geting several core components of the cell cycle machinery. For

instance, Hog1 controls G1/S transition by the downregulation of

cyclin expression and the stabilization of the Sic1 cyclin-depen-

dent kinase inhibitor (CDKi) (Adrover et al., 2011; Escote et al.,

2004). Hog1 also modulates other phases of the cell cycle,

such as S phase (Duch et al., 2013). Cells unable to delay cell

cycle progression upon osmostress display reduced viability

upon those conditions, suggesting the need to delay cell cycle

for proper adaptation.

The Hog1 SAPK is a key element for reprogramming gene

expression in response to osmostress by acting on hundreds

of stress-responsive genes. Hog1 is recruited to chromatin to re-

cruit RNA polymerase II (Alepuz et al., 2003; Nadal-Ribelles et al.,

2012) and associated factors (De Nadal et al., 2004; Sole et al.,

2011; Zapater et al., 2007). Hog1 is present also at the open

reading frames (ORFs) of stress-responsive genes (Cook and

O’Shea, 2012; Nadal-Ribelles et al., 2012; Pokholok et al.,

2006; Proft et al., 2006), where it stimulates strong chromatin re-

modeling by the interplay of the INO80 and the RSC complexes

(Klopf et al., 2009; Mas et al., 2009). Chromatin dynamics set a

threshold for gene induction upon Hog1 activation (Pelet et al.,

2011). In addition to gene induction, Hog1 controls mRNA stabil-

ity (Miller et al., 2011; Molin et al., 2009; Romero-Santacreu et al.,

2009), export (Regot et al., 2013), and translation (Warringer

et al., 2010). Thus, Hog1 plays a key role in the regulation of

mRNA biogenesis (de Nadal et al., 2011; de Nadal and Posas,

2010; Martınez-Montanes et al., 2010; Weake and Workman,

2010).

Here, we show that the Hog1 SAPK also associates with and

controls the induction of a set of lncRNAs in response to osmo-

stress. One of the genes expressing a stress-induced lncRNA in

antisense orientation is CDC28, the cyclin-dependent kinase 1

(CDK1) that controls the cell cycle in yeast. Induction of the

CDC28 lncRNA permits the increase in the levels of Cdc28,

allowing cells more efficient reentry into the cell cycle after

stress. Therefore, Hog1 directly coordinates the regulation of

Molecular Cell 53, 1–13, February 20, 2014 ª2014 Elsevier Inc. 1

mailto:[email protected]

mailto:[email protected]


Molecular Cell

CDC28 lncRNA Regulates Cell Cycle upon Stress


transcription and cell cycle progression by controlling expres-

sion of a stress-induced lncRNA in CDC28.

RESULTS

Hog1 Mediates the Expression of a Set of Stress-Inducible lncRNAsMost transcriptome studies performed to define stress genes

have analyzed coding genes. To cover the expression of the

whole genome upon stress, we monitored transcription using

strand-specific tiling arrays (David et al., 2006). The number of

coding genes induced upon stress was 343 at 0.4 M NaCl

(15 min) and 294 at 1.2 M NaCl (100 min) using a stringent

threshold (see Experimental Procedures). Expression of 56%

and 84% of the stress-induced genes depended on Hog1 at

0.4 M and 1.2 M NaCl, respectively. Overall, the number of

stress-responsive genes was similar to that in previous reports

(Capaldi et al., 2008; Gasch et al., 2000; Nadal-Ribelles et al.,

2012; Posas et al., 2000).

Remarkably, in addition to coding genes, up to 173 lncRNAs

were strongly induced upon treatment with 0.4 M NaCl and up

to 216 with 1.2 M NaCl (Figure 1). Almost a hundred of them

were shared between the two stress conditions (Figure S1

available online). The average length of these stress-induced

lncRNAs is about 843 nt (Figure S1). Expression of 50%

and 91% of the lncRNAs induced by treatment with 0.4 M

and 1.2 M NaCl, respectively, depended on the presence of

Hog1 (http://steinmetzlab.embl.de/francescData/arrayProfile/

index.html) (Figure 1B). Some overlapped with previously anno-

tated cryptic unstable transcripts (CUTs) or stable unannotated

transcripts (SUTs) (Wu et al., 2012; Xu et al., 2009). However,

most of them were not expressed in the absence of RRP6,

TRF4, or XRN1 and were present only upon stress (Figures 1C

and S1). Thus, Hog1 mediates the expression of a set of

stress-inducible lncRNAs.

Hog1 Associates with the Promoters of Stress-InducedlncRNAs and Stimulates RNA Pol II Recruitment andGene ExpressionHog1 associates with chromatin of stress-responsive genes

upon stress (Alepuz et al., 2001, 2003; Cook and O’Shea,

2012; Pokholok et al., 2006; Proft et al., 2006). Actually, Hog1

is present in at least 80% of the Hog1-induced genes upon

stress (Nadal-Ribelles et al., 2012). We also found that Hog1

is present in �63% of Hog1-dependent lncRNA promoters,

whereas it is recruited to <30% of Hog1-independent lncRNAs

(Figure 2A). Genome-wide association of RNA Pol II showed

that it strongly associates with stress-responsive loci upon

stress (Cook and O’Shea, 2012; Nadal-Ribelles et al., 2012).

RNA Pol II was also recruited at the stress-induced lncRNAs

(2.3-fold increase) upon stress. In contrast, RNA Pol II was not

recruited at the Hog1-dependent lncRNA promoters in a hog1

strain (Figures 2B and S2A). Therefore, Hog1 associates to and

stimulates the recruitment of RNA Pol II at the promoters of

stress-induced lncRNAs.

Once recruited to stress-responsive genes, Hog1 mediates

chromatin remodeling (Mas et al., 2009; Pelet et al., 2011).

Genome-wide micrococcal nuclease (MNase) digestion of chro-

2 Molecular Cell 53, 1–13, February 20, 2014 ª2014 Elsevier Inc.

matin and deep sequencing (MNase-seq) showed that upon

osmostress, Hog1 mediates dramatic change of nucleosome

occupancy (Nadal-Ribelles et al., 2012). We analyzed the chro-

matin organization at genes that expressed Hog1-induced

lncRNAs in antisense orientation and found that regions beyond

the transcription termination site (TTS) also suffered strong

Hog1-dependent chromatin remodeling (Figure 2C). Although

levels of nucleosome occupancy decreased slightly upon stress

in hog1D cells, this decrease was clearly more prominent in the

wild-type strain. In contrast, no changes in chromatin structure

were observed in promoters of lncRNAs that do not respond to

stress (Figure S2B).

Expression from stress-responsive promoters can be quanti-

tatively measured by the fusion to quadruple Venus (qV)

fluorescent protein (Pelet et al., 2011; Regot et al., 2013). To

characterize one of these lncRNA promoters further, we fused

the 30 untranslated region (30 UTR) of CDC28 in both the sense

and the antisense orientation to qV-yellow fluorescent protein

(YFP) and assessed gene expression by flow cytometry in

wild-type and hog1 strains. Expression of qV-YFP was induced

upon stress, depending on the presence of Hog1 and only

when placed in the antisense orientation (Figure 2D). Therefore,

the 50 regions of stress-induced lncRNAs behave as bona fide

stress-responsive promoters.

Induction of CDC28 lncRNA Expression Promotes theInduction of CDC28 Gene Expression upon StressTo functionally characterize the role of stress-induced lncRNAs,

we asked whether there is a correlation between expression of

the sense and antisense-induced transcription. We found a

correlation for only a few relevant cases (8 out of 91; Hog1-

dependent lncRNAs at 0.4 M NaCl) in which an increase

of the antisense was associated with an increase in the

sense transcript (Figure 3A). Tiling arrays are not very sensi-

tive for slight increases on transcription; thus, we analyzed

the expression of the 91 genes with stress-induced lncRNAs

from previous run-on assays and found that 41 out of 91 genes

displayed a positive correlation (Romero-Santacreu et al.,

2009).

One of the genes with a clear correlation of the sense and

lncRNA expression was CDC28 (Figure 3B). CDC28 encodes

the main CDK that drives progression of the cell cycle in

yeast. We found that upon osmostress, there is an increase

in CDC28 expression that was not observed in a hog1 strain

(Figure 3B). Systematic insertion analysis at the 30 UTR of

CDC28 showed that insertion of a KanR marker at 180 nt

downstream of the TTS (lncRNAD) abolished expression of

the lncRNA. In this strain, the induction of CDC28 upon osmo-

stress was impaired (Figures 3B and S3A). Notably, the rrp6

mutation did not alter the induction of the CDC28 sense tran-

scription (Figure S3B). Thus, the presence of the stress-induc-

ible CDC28 lncRNA correlates with induction of the CDC28

gene expression.

We then assessed whether it was the expression of the

lncRNA from the 30 UTR or solely the presence of the

lncRNA that induced CDC28 expression. We created a strain

containing a CDC28::GFP that recapitulated CDC28 gene ex-

pression (CDC28::GFP) and then abolished lncRNA expression

http://steinmetzlab.embl.de/francescData/arrayProfile/index.html


A

B

C

Figure 1. Hog1 Controls Transcription of a Set of lncRNAs upon Stress(A) Osmostress induces expression of a class of lncRNAs. Expression data from tiling arrays of several stress-induced lncRNAs for theWatson (W, top) and Crick

(C, bottom) strands. Normalized signal intensities from duplicate hybridizations of wild-type (WT) and hog1D strains under basal conditions (YPD), treated with

0.4 M NaCl for 15 min or treated with 1.2 M NaCl for 100 min (y axis) are shown. Genomic coordinates and gene annotation, transcript boundaries (red lines) and

transcription start site (arrows) are depicted.

(B) Hog1 regulates lncRNA transcription upon stress. lncRNAs induced at least 2-fold upon osmostress (0.4 M NaCl, left; 1.2 M NaCl, right) grouped into three

categories according to the dependence of their expression in hog1D: strong (<50% of wild-type), moderate (50%–75%), none (>75%).

(C) Stress-induced lncRNAs are a family of lncRNAs. WT, hog1D, trf4D, rrp6D, and xrn1D mutant strains were subject to osmostress (0.4 M NaCl), and the

indicated lncRNAs were assessed. See also Figure S1.

Molecular Cell



(CDC28::GFP lncRNAD) (Figure S3C). We then expressed

CDC28 and its terminator region from a plasmid. This permitted

us to distinguish transcription from the plasmid or the endoge-

nous locus within the same cell. Remarkably, in response to

osmostress, we could detect an increase in expression of

CDC28 sense from the plasmid, but we did not observe an


A B

C

D

Figure 2. Hog1 Binds and Recruits RNA Pol II at

Genes with lncRNA

(A) Hog1 associates with lncRNA promoters. The percent-

age of Hog1 binding at the Hog1-dependent or -indepen-

dent lncRNAs determined by ChIP-seq (0.4 M NaCl,

p < 0.05).

(B) Hog1 stimulates Pol II recruitment to lncRNA promoters.

Distribution of RNA Pol II binding (normalized reads, TRPKs)

was determined by ChIP-seq (0.4 M NaCl for 10 min) for

Hog1-dependent lncRNAs in wild-type (WT) and hog1D.

***p < 0.001 (t test).

(C) Hog1 mediates changes in chromatin architecture at

lncRNA promoters. Distribution of nucleosome hits (RPKMs;

reads per kilobase per million) expanding 1 kb up and

downstream from the TTS of wild-type and hog1D mutant

strains under basal conditions (dark blue and red) and 0.4 M

NaCl (light blue and yellow). Plot represents coverage of

reads of approximately 90 genes: Hog1-dependent (upper

graph), Hog1-independent (lower graph), and non-stress

responsive (see Figure S2). Dotted black line marks TTS.

(D) Antisense-oriented 30 UTR of CDC28 is an osmores-

ponsive promoter. The 30 UTR of CDC28 fused to the qV-

YFP in the antisense orientation (promoter lncRNA) or in

its natural orientation (terminator sense). Fluorescence in-

tensity was measured by flow cytometry.

Molecular Cell




B

0

1

2

3

4

5

6

7

-2 2 4 6log2 sense (WTNaCl-WTcontrol)

log2

lncR

NA

(WT N

aCl-W

T con

trol

)

0

CDC28

lncRNA

ENO1

wt0 5 10 15 30NaCl (min)

CDC28 sense

hog1Δ0 5 10 15 30

lncRNAΔ0 5 10 15 30

1 1.4 1.9 2.9 1.6 1 1.1 1 1.2 1.2 1 1 0.8 0.8 0.8

A

1 15 27 35 9 1 1 1 1 1 1 1 1 1 1

Figure 3. CDC28 Induction Correlates with lncRNA Expression

(A) Expression of sense (x axis) versus lncRNA (y axis), in a log2 scale, in

response to stress in Hog1-dependent lncRNAs. Highlighted dots represent

genes with positive (red), negative (orange), or no (black) correlation.

(B) Hog1 and the CDC28 lncRNA are required for CDC28 expression. CDC28

sense and CDC28 lncRNA transcripts were assessed in WT, hog1D, and

lncRNAD strains upon osmostress. See Figure S3 for systematic insertion of a

KanR marker at the 30 UTR region of CDC28. Normalized quantification of

CDC28 sense is shown. Quantifications were done by normalizing by the

loading control gene and the expression in the absence of stress.

Molecular Cell



increase of the CDC28::GFP, thus indicating a cis effect of the

lncRNA (Figure S3C).

Hog1 Associates with the 30 UTR and the +1 NucleosomeRegions in CDC28 to Promote Chromatin RemodelingTo characterize the mechanism by which Hog1 induces sense

and antisense transcription, we monitored Hog1 association in

CDC28. Hog1 associated with the 30 UTR of CDC28 (the pro-

moter region of the CDC28 lncRNA) upon stress. Strikingly,

Hog1 also associated with a region close to the TSS (region +41

to +125) corresponding to the CDC28 +1 nucleosome. By

contrast, in cells unable to induce CDC28 lncRNA, Hog1 associ-

ation with the TSS was abolished completely and only slightly

reduced at the 30 UTR (Figure 4A).

The presence of Hog1 at the region corresponding to the +1

nucleosome around the TSS led us to analyze the chromatin

architecture by MNase digestion. We found that the chromatin

at the 30 UTR of CDC28 changed upon osmostress, as expected

for a stress-responsive promoter. Strikingly, the region corre-

sponding to the +1 nucleosome at the 50 region of CDC28 was

also strongly remodeled upon stress (Figure 4B). The eviction of

the +1 nucleosome was observed in neither hog1- nor lncRNA-

deficient cells (Figure 4B). Therefore, Hog1 association and re-

modeling at the +1 nucleosome region of CDC28 occur in

response to osmostress and in the presence of CDC28 lncRNA.

The RSC Chromatin Remodeling Complex MediatesChromatin Remodeling at the CDC28 +1 NucleosomeRegion upon StressHog1 stimulates chromatin remodeling at specific stress-

responsive loci by recruiting the RSC complex (Mas et al.,

2009). Induction of the CDC28 lncRNA in cells deficient in the

RSC complex (rsc9ts) under nonpermissive temperature was

similar to that of wild-type but reduced in a SAGA mutant (Fig-

ures 4C and S4A). Thus, RSC is not necessary for lncRNA

expression. In clear contrast, rsc9ts mutant cells did not induce

CDC28 expression upon stress. This suggests a key role of

RSC for the increase of CDC28 sense upon stress.

Then, we assessed the recruitment of RSC and Hog1 to

various regions of CDC28 before and after the addition of

NaCl. We found that RSC associates with 30 UTR and +1 nucle-

osome regions of CDC28 in response to stress only in the pres-

ence of Hog1 (Figure 4D). Induction of lncRNA and CDC28

required active Hog1, since a catalytically inactive Hog1 was

unable to associate to chromatin and promote RSC association

(Figures S4B–S4D). By contrast, association of Hog1 was not

altered in the rsc9ts mutant under nonpermissive temperature

(Figure S4E), suggesting that Hog1 mediates the recruitment of

RSC at CDC28 to remodel chromatin upon stress. Correspond-

ingly, chromatin remodeling at the +1 nucleosome, assessed by

MNase digestion, was impaired in the rsc9ts mutant under

nonpermissive temperature (Figure 4E). Thus, recruitment of

RSC by Hog1 is essential to mediate chromatin reorganization

at the +1 nucleosome region and CDC28 gene induction.

The Expression of Both the CDC28 lncRNA and Hog1 IsRequired for CDC28 Induction upon StressIn the absence of CDC28 lncRNA expression, the presence of

Hog1 at the 30 UTR of CDC28 is not sufficient to increase

CDC28 expression. Then, we asked whether the expression of

the CDC28 lncRNA alone was sufficient for CDC28 induction.

Thus, we inserted an inducible GAL1 promoter in the anti-

sense orientation at the endogenous 30 UTR of CDC28

(CDC28::pGAL1). Expression from the GAL1 promoter is driven

by the Gal4-ER-VP16 activator in the presence of b-estradiol

(Louvion et al., 1993). Although the presence of estradiol strongly

induced expression of theCDC28 lncRNA, this was not sufficient

to stimulate sense transcription (Figure 5A). We then monitored

Hog1 recruitment and found that the presence of estradiol did

not mediate Hog1 recruitment in CDC28 (Figure 5B). Corre-

spondingly, chromatin remodeling at the +1 nucleosome did

not occur by the sole induction of the CDC28 lncRNA from the

GAL1 promoter in the presence of the Gal4-ER-VP16 activator

(Figure 5C). Thus, the induction of the CDC28 lncRNA alone is

not sufficient to mediate chromatin remodeling and CDC28

gene induction.

Then, we assessed whether the recruitment of Hog1 together

with the CDC28 lncRNA expression from the GAL1 promoter

could induce CDC28 gene expression. Expression from

CDC28::pGAL1 was then driven by Gal4DBD-Msn2 activator.

Msn2 mediates the recruitment of Hog1 to Msn2-dependent

genes (Alepuz et al., 2001). Tethering Msn2 to the Gal4-binding

domain stimulated stress-inducible transcription of the lncRNA

in the CDC28::pGAL1 strain upon stress and restored CDC28


MN

ase

prot

ectio

n (A

.U.)

wt

+1

controlNaCl

lncRNAΔ

ATG TAA

0.5

1.0

1.5

0.5

1.0

1.5

0.5

1.0

1.5 hog1Δ

ACDC28

JHGFA C EB D I K

KANR

elat

ive

Hog

1-H

A bi

ndin

g

wtlncRNAΔ

HGFA C E JB D I K0

1

2

3

4

5

B

wt rsc9ts

lncRNA

ENO1

NaCl (min)

CDC28 sense

0 5 10 15 300 5 10 15 30

1 1.1 1.6 2.8 0.7 1 1.1 1.1 0.8 1.2

controlNaCl

rsc9ts25ºC 37ºC

0

0.5

1.0

1.5

+1 n

ucle

osom

eev

ictio

n (A

.U.)

C

D E

012345

0

1

2

3

Rsc

1 -H

A bi

ndin

g

wt hog1Δ wt hog1Δ

controlNaCl

+1 nucleosome 3’UTR

1 18 26 33 15 1 29 39 15 16

KANKAN Figure 4. Binding of Hog1 Induces lncRNA and Promotes Chromatin

Remodeling by the RSC Chromatin Remodeler

(A) Hog1 binds at the 30 UTR and at the +1 nucleosome region. Graphical

representation of the CDC28 locus with the insertion of a KanR marker at the

30 UTR region to disrupt expression of lncRNA (lncRNAD). Hog1 association at

the CDC28 locus was assessed by ChIP at the indicated regions and strains.

(B) Induction of CDC28 lncRNA by Hog1 promotes chromatin remodeling

at the +1 nucleosome. Nucleosome positioning was assessed by MNase

digestion in wild-type (WT), hog1D, and lncRNAD strains under control (black)

or osmostress (0.4 M NaCl for 10 min, blue). The normalized nucleosome

occupancy of one representative experiment is shown (x axis).

(C) CDC28 induction depends on RSC activity. Wild-type (WT) and rsc9ts

strains were grown at restrictive temperature (37�C) and subjected to osmo-

stress. CDC28 sense and lncRNA transcripts were detected by northern blot

for the indicated times.

(D) Hog1 mediates the recruitment of RSC to CDC28. Rsc1 binding was

assessed by ChIP at the +1 nucleosome region (amplicon D) and 30 UTR(amplicon J) of CDC28 in wild-type and hog1D strains that were (or were not)

subjected to osmostress (0.4 M NaCl, 5 min).

(E) RSC is essential tomediate chromatin reorganization. Remodeling at the +1

nucleosome in a Rsc9ts strain was assayed byMNase digestion after 10 min of

NaCl. Bars represent the average of the stressed (black bars) compared to the

unstressed (white bars) cells ± SD. See also Figure S4.

Molecular Cell




induction upon stress (Figure 5D). We then inserted a terminator

downstream of the GAL1 promoter (IT) (Kopcewicz et al., 2007;

Loya et al., 2012) that permitted transcription initiation but pre-

vented the generation of the lncRNA. Here, the Gal4DBD-Msn2

didnotpromoteCDC28 sense induction (Figure5E).Correspond-

ingly, Hog1 was recruited at the GAL1 promoter in the 30 UTRregion of CDC28 as well as at the +1 region of CDC28 (Fig-

ure 5F), whereas it was not recruited at the +1 region in the IT

construct (Figure 5G). We then monitored chromatin remodeling

at the +1 region. In contrast to Gal4-ER-VP16 activator, expres-

sion of the CDC28 lncRNA from the GAL1 promoter by the

Gal4DBD-Msn2 activator caused remodeling of the +1 nucleo-

some upon stress (Figure 5H). Thus, induction of the CDC28

lncRNA and the recruitment of Hog1 at the +1 region are required

for chromatin remodeling at the 50 region of CDC28 and CDC28

gene expression.

The Establishment of Gene Looping Permits theRecruitment of Hog1 at the +1 Nucleosome Region andInduction of CDC28

The absence of Hog1 recruitment and remodeling at the 50 regionof CDC28 in cells deficient in lncRNA induction prompted us to

assess whether the presence of Hog1 at this region was

mediated by gene looping formation (O’Sullivan et al., 2004;

Tan-Wong et al., 2012). Gene loop formation depends on the

essential protein Ssu72 (Ansari and Hampsey, 2005). Expres-

sion of SSU72 under GAL1 is repressed in the presence of

glucose (YPD). Cells were grown in galactose, shifted to glucose,

and subjected to osmostress. Depletion of Ssu72 did not alter

induction of the CDC28 lncRNA but prevented induction of

the CDC28 (Figure 6A). Similar results were obtained in a

sua7-1 mutant (Singh and Hampsey, 2007) with impaired gene

looping (Figure S5A). Then, we assessed the recruitment of

Ssu72 and found that there was a clear Hog1-dependent in-

crease in Ssu72 binding upon stress at both the 30 UTR and +1

nucleosome regions (Figure S5B). Thus, the enhanced recruit-

ment of Ssu72 in response to stress does not alter CDC28

pGAL1CDC28

Gal4DBD

Msn2 ZFΔ

Hog1

D X

CDC28 sense

lncRNA

ENO1

NaCl (min)

1 0.9 2.3 1.8 1

0 5 15 3010

CDC28::pGAL1

0 5 15 3010Ø Gal4DBD-Msn2

1 0.9 1.1 1.1 1.1

Prot

ein

bind

ing

Ø Ø

D X

0

1

Gal4DBD-Msn2

Gal4DBD-Msn2

2

3 controlNaCl

0

0.5

1.0

1.5

Ø Gal4DBD-Msn2

CDC28::pGAL1

+1 n

ucle

osom

eev

ictio

n (A

.U.)

A

D

GF

CB

controlβ-E

Ø0

0.5

1.0

1.5

CDC28::pGAL1

Gal4DBD-ER-VP16

+1 n

ucle

osom

eev

ictio

n (A

.U.)

Hog

1-H

A bi

ndin

g

D XCDC28::pGAL1 +

Gal4DBD-ER-VP16

0

0.5

1.0

1.5

controlβ-Eβ-E + NaCl

controlNaCl Hog1NaCl Rsc1

H

0

1

2

3

Ø Ø

D X

Gal4DBD-Msn2

Gal4DBD-Msn2

controlNaCl

Hog

1 -H

A bi

ndin

g

DCDC28

Gal4DBD

Msn2 ZFΔ

Hog1

pGAL1*Y

STOP

CDC28 sense

lncRNA

ENO1

NaCl (min) 0 5 15 3010

CDC28::IT::pGAL1*

0 5 15 3010Ø Gal4DBD-Msn2

E

1 24 19 13 71 1 1 1 1

1 0.7 1 1.1 0.61 0.8 0.6 0.7 1.1

1 1 1 1 11 1 1 1 1

Gal4DBD

β-estradiolER

VP16AD

DpGAL1CDC28

X

CDC28 sense

wt +Gal4DBD-ER-

VP16CDC28::pGAL1 +

Gal4DBD-ER-VP16

lncRNA

ENO1

β-E0 5 15 30Time (min)

β-E0 5 15 30

β-E+NaCl0 5 15 30

1 1 0.8 1 1 1.1 1 1.1

1 1 1 1 1 4 12 45 1 1 25 40

1 1.1 1 1.2

Figure 5. Antisense Transcription and Hog1 Recruitment Are Required to Induce CDC28 Expression

(A) lncRNA alone is not sufficient for CDC28 expression. Graphical representation of the CDC28 locus with inducible lncRNA expression achieved by insertion of

the pGAL1 promoter (CDC28::pGAL1). Gal4-ER-VP16 activator induces lncRNA in the presence of b-estradiol (black bars) or b-estradiol and NaCl (gray bars).

CDC28 sense and lncRNA transcripts were assessed by northern blot.

(B and C) We measured Hog1 association (5 min) by ChIP (B) and +1 nucleosome eviction (10 min) by MNase (C) in the indicated strains upon induction with

b-estradiol.

(D) Hog1 and lncRNA induction are necessary for CDC28 induction. Graphical representation of the CDC28::pGAL1 locus induced by the Gal4-Msn2DBD.

Transcript levels were followed as in (A).

(E) Graphical representation of the CDC28::pGAL1 strain containing an internal terminator (IT).

(F) Hog1 and Rsc1 recruitment (5 min) in cells expressing and empty or Gal4-Msn2DBD.

(G) Presence of the lncRNA is required for Hog1 recruitment at the +1 nucleosome. Hog1 recruitment (5 min) of cells in the IT strain.

(H) +1 nucleosome eviction (10 min) was assessed in CDC28::pGAL1 cells harboring empty vector (Ø) or Gal4-Msn2DBD upon stress. Normalized quantification

of CDC28 is shown. Error bars represent SD.

Molecular Cell




A

+1 nucleosome 3’UTRB

controlNaCl

+1 n

ucle

osom

eev

ictio

n (A

.U.)

GAL1::SSU72YPGal YPD

0

0.5

1.0

1.5D

O1

CDC28

AciI

TO2

F

A B C D E F G H I J0

1

2

3

4

0

1

2

3

4

YPGal YPD

GAL1::SSU72

YPGal YPD

C

Hog

1-H

A bi

ndin

g

Rsc

1-H

A bi

ndin

g

E

GAL1::SSU72 (YPD)

wthog1ΔlncRNAΔ

RN

A Po

l II b

indi

ng

0

0.5

1.0

NaCl

YPGalYPD

- + - + - + - + ND Dwt hog1Δ YPGal YPD

GAL1::SSU72

NaCl

T+O1

T+O2

TEL

sua7-1wt- + - + ND D

lncRNA

ENO1

NaCl (min)

CDC28 sense

YPGal YPDGAL1::SSU72

0 5 10 15 300 5 10 15 30

1 1.7 1.8 3.2 2 1 1.2 1.5 1.4 1

1 22 28 32 15 1 19 21 30 28

YPD0 5 10 15 30

wt

1 1.7 2 2.6 1.4

1 33 39 41 20

+1 nucleosome

GAL1::SSU72

Figure 6. Gene Looping Allows Recruitment

of Hog1 at the +1 Nucleosome Region as

well as Induction of CDC28

(A) CDC28 induction depends on Ssu72. Wild-type

and GAL1::SSU72 strains grown as stated were

subjected to osmostress. CDC28 sense and

lncRNA transcripts were detected.

(B) Hog1 binding at the +1 nucleosome of CDC28

depends on Ssu72. The GAL1::SSU72 strain was

grown as stated, and Hog1 recruitment was

analyzed by ChIP after 5 min of NaCl at the indi-

cated regions (as in Figure 4A).

(C) Binding of RSC at the +1 nucleosome requires

gene looping. Recruitment of Rsc1 was assessed

by ChIP at the indicated regions.

(D) Chromatin remodeling at the +1 region depends

on Ssu72. The GAL1::SSU72 strain was grown as

stated, and remodeling at the +1 nucleosome was

assayed by MNase digestion after 10 min of NaCl.

Bars represent the average of the stressed (black

bars) compared to the unstressed (white bars)

cells ± SD.

(E) Recruitment of RNA Pol II requires gene looping.

Binding of RNA Pol II at the +1 nucleosome was

assessed by ChIP. Levels in the wild-type upon

stress were used as a reference for the other indi-

cated strains.

(F) Hog1 induces physical interaction between

the +1 nucleosome region and 30 UTR of CDC28

in response to stress. CDC28 locus is depicted

along with the positions of AciI cleavage sites

(red lines) and primer position in 3C analysis

(arrows). 3C analysis at the indicated strains and

conditions in the presence (+) or absence (�) of

osmostress (0.4 M NaCl, 10 min) is shown. Ampli-

fication of tandem primer pairs is shown, and the

TEL region was used as loading control. Non-

digested (ND) and digested (D) chromatin samples

from the treated WT strain were used as an internal

control of 3C specificity. See also Figures S5

and S6.

Molecular Cell



lncRNA expression, but it is essential for the increase of CDC28

expression.

The lack of CDC28 induction in the absence of Ssu72 and in

the sua7-1 mutant suggested that gene looping might mediate

the transfer of activities from the 30 UTR to +1 nucleosome


regions of CDC28. Association of Hog1

at the 30 UTR region of CDC28 upon

stress was not altered by the absence of

Ssu72. In clear contrast, the absence of

Ssu72 (YPD) completely abrogated the

association of Hog1 at the +1 nucleosome

region of CDC28 (Figure 6B). Correspond-

ingly, recruitment of Rsc1 was also abol-

ished at the +1 region in the absence of

Ssu72 (Figure 6C), which prevented chro-

matin remodeling at the +1 region upon

stress (Figure 6D). Moreover, the increase

in RNA Pol II association upon stress at

the CDC28 50 region observed in the

wild-type strain was abolished in cells deficient in hog1,

CDC28 lncRNA, and ssu72 (Figure 6E). Thus, gene looping

mediates the recruitment of Hog1 at the 50 region of CDC28 to

induce chromatin remodeling and RNA Pol II recruitment upon

stress.

Molecular Cell



To further confirm the establishment of gene looping between

the 30 UTR and promoter regions, we applied the 3C assay (see

Experimental Procedures). We found that there was a clear

increase in gene looping formation upon stress between the

30 UTR and the +1 nucleosome regions, as detected by the pres-

ence of O1-T PCR products. The O1-T PCR product was ligation

dependent (D), and it was not detected when an alternative

region (O2) was assessed (Figures 6F and S5C). Notably, the

increase in gene looping formation upon stress was dependent

on Hog1 and abolished in the absence of SSU72 (YPD) and

sua7-1 mutant cells (Figures 6F and S5C), but not in a rsc9ts

mutant (Figure S5D). Therefore, gene looping is critical for the

recruitment of Hog1 from 30 UTR to the +1 nucleosome region

of CDC28 to promote chromatin remodeling and induce

CDC28 gene expression.

An lncRNA in MMF1 Induces Hog1 Recruitment andChromatin Remodeling at the +1 Nucleosome RegionTo assess whether genes other than CDC28 displayed a similar

regulatory mechanism, we choseMMF1 because it expresses a

strong, stress-induced lncRNA (Figure S6A) and, albeit not seen

in the tiling arrays due to insufficient sensitivity, it was reported

to be induced upon stress by Hog1 in run-on and dynamic

transcriptome analysis (DTA) experiments (Miller et al., 2011;

Romero-Santacreu et al., 2009). We created a mutant in anti-

sense MMF1 lncRNA expression (Figure S6A). Hog1 was re-

cruited at the 30 UTR and the 50 region, but not in the body of

theMMF1 gene, depending on the presence of the lncRNA (Fig-

ure S6B). Nucleosome eviction occurred in response to stress in

a Hog1- and lncRNA-dependent manner (Figure S6C). Then, we

performed 3C experiments in wild-type, hog1, and SSU72 shut-

off system (pGAL1::SSU72). Notably, in the absence of stress,

we could already detect gene looping between the P and T

regions. But most remarkably, the wild-type strain showed an

increase of P-T association in response to stress, which was

fully dependent on the presence of Hog1 and gene looping

(Figure S6D). Thus, albeit some particularities, MMF1 seems

to stimulate chromatin remodeling via Hog1 and lncRNA ex-

pression as in CDC28.

We then asked whether the 30 UTR region of CDC28 could

confer osmoinduction in a nonosmoresponsive gene. We re-

placed the 30 UTR region of a non-stress responsive gene,

MBA1, with the 30 UTR region of CDC28 (300 bp downstream

of STOP codon). We choseMBA1 because its expression under

normal conditions and the length of the gene are similar to those

of CDC28. Replacement of the MBA1 terminator by CDC28 led

to the transcription of an lncRNA from the 30 UTR in MBA1,

and most important, it conferred Hog1-dependent osmoinduc-

tion (Figure S6E), suggesting that the role of CDC28 30 UTR is

to confer stress-inducible regulation of gene expression from

the terminator region of the gene.

Stress-Induced CDC28 lncRNA Results in an Increase ofCdc28 that Permits Cells to Reenter the Cell Cycle MoreEfficiently in Response to StressWe then asked whether an increase of CDC28 mRNA results

in an increase of Cdc28 protein production upon stress.

Endogenous [35S]methionine Cdc28 protein increased �2-fold

in response to stress in wild-type cells, whereas no increase

was observed in a CDC28 lncRNA-deficient strain (Figure 7A).

Thus, stress-induced Cdc28 lncRNA expression leads to in-

creased levels of Cdc28 kinase.

Hog1 promotes an immediate, but transient, cell cycle delay

that permits stress adaptation (Clotet and Posas, 2007; Duch

et al., 2012). The increase in Cdc28 levels occurred when cells

were already recovering from the initial arrest caused by stress.

Thus, we hypothesized that this increase of Cdc28 can serve to

accelerate cell cycle reentry after stress. To test this hypothesis,

we assessed the exit from the arrest caused by osmostress in a

phase of cell cycle with high Cdc28 by synchronizing cells using

a temperature-sensitive allele of cdc15 (cdc15ts) (see Experi-

mental Procedures). Cells deficient in CDC28 lncRNA were

able to arrest and exit the cell cycle from cdc15ts synchronization

as efficiently as wild-type in the absence of stress. In contrast,

compared to the wild-type, cells deficient inCDC28 lncRNA pro-

duction delayed cell cycle reentry by approximately 20 min upon

stress (Figure 7B). Notably, overexpression of CDC28 from a

plasmid can suppress the delay on cell cycle progression

observed upon osmostress inCDC28 lncRNAD cells (Figure S7).

Thus, stress-induced CDC28 lncRNA results in an increase of

Cdc28 that permits cells to reenter the cell cycle more efficiently

in response to stress.

DISCUSSION

ASpecific Set of Hog1-Dependent lncRNAs Is Induced inResponse to OsmostressStress-activated protein kinases regulate gene expression to

maximize cellular adaptation to environmental stresses (de

Nadal et al., 2011; Weake and Workman, 2010). In yeast, activa-

tion of Hog1 leads to major changes in gene expression. Here,

we provide evidence that in addition to controlling expression

of coding genes, Hog1 also induces a dedicated set of stress-

responsive lncRNAs. Upon osmostress, about 200 lncRNAs

are rapidly induced. The induction of these stress-induced

lncRNAs dependsmostly on the presence of Hog1. Correspond-

ingly, similar to osmoresponsive genes, Hog1 associates to the

promoters of lncRNAs upon stress and stimulates RNA Pol II

recruitment and chromatin remodeling. In fact, fusing the pro-

moter of one of these lncRNAs (CDC28) to a GFP reporter

showed that expression occurred only upon stress, in antisense

orientation, and depending on Hog1. This observation is remark-

able, since most of the described antisense transcripts have

been proposed to arise from bidirectional promoters (Tan-

Wong et al., 2012; Xu et al., 2009). The fact that this terminator

can function as a heterologous promoter suggests that a

different transcriptional unit recruited to this region that is inde-

pendent of the neighboring genes must exist. Accordingly, a

recent study by chromatin immunoprecipitation-exonuclease

(ChIP-exo) precisely positioned distinct transcriptional machin-

eries at bidirectional promoters, supporting the idea of unique

transcription units (Rhee and Pugh, 2012).

Most of the stress-induced lncRNAs are transcribed in

response to osmostress. Except for SUTs, which are stable

transcripts, the rest of lncRNAs are only detectable in strains

with deleted components of the nuclear or cytosolic exosome


A

B

Cdc

28p

leve

lsFo

ldin

duct

ion

Time (min)

wtlncRNAΔ

0

1

2

0 20 40 60 80 100

2C

wt lncRNAΔ

control NaCl

cdc15ts

control NaCl1C

Tim

e (m

in)

95

85

75

65

55

45

35

0 88

74

52

40

59

41

65

83

90

88

88

82

70

48

40

86

76

57

44

46

54

60

80

89

89

89

88

88

78

71

C

Figure 7. Changes of Cdc28 Levels Promote Cell Cycle Reentry

upon Stress

(A) Cdc28 protein levels increase upon stress. 35S-labeled Cdc28 was

immunoprecipitated for the indicated lengths of time and strains. Cdc28 levels

were normalized against total protein.

(B) Increased Cdc28 restores cell cycle progression. The indicated cdc15ts

strains were synchronized at anaphase and released under control or stress

Molecular Cell




(CUTs and XUTs) (Xu et al., 2009), when gene looping is impaired

(Ssu72-restricted transcripts), or by deletion of SET3 (Kim et al.,

2012; Tan-Wong et al., 2012). Expression analysis of some

representative Hog1-dependent lncRNAs showed that they

were not expressed under basal conditions in the absence of

RRP6, XRN1, or TRF4. Transcription was only induced upon

stress, but stability was altered in these mutants. Thus, Hog1

regulates the transcription of a distinctive class of stress-

induced lncRNAs whose induction might have relevant implica-

tions for proper cellular adaptation.

The CDC28 lncRNA and Hog1 Induce ChromatinRemodeling and CDC28 Expression via Gene LoopingTo unravel the biological function of the stress-induced lncRNAs,

we investigated whether there was correlation between the

expression of sense and lncRNAs. Overall, this was not evident

except for in the cases of some genes. Remarkably, there was

a positive correlation between the induction of CDC28 and an

lncRNA in CDC28 expressed in antisense orientation. CDC28

expression was dependent on Hog1, since there was induction

of neither the lncRNA nor the CDC28 sense in hog1 cells. This

posed the question of how the SAPK and the induction of an

lncRNA lead to increased gene expression. In clear contrast to

typical osmoresponsive genes in which the SAPK associates

all along the gene (Proft et al., 2006), Hog1 associated at the 30

region of CDC28, which corresponds to the promoter region of

the CDC28 lncRNA, and at a region surrounding the +1 nucleo-

some ofCDC28. Transcription ofCDC28 is not controlled by any

of the transcription factors targeted by Hog1, thus opening the

possibility that Hog1 uses the 30 UTR region to mediate its asso-

ciation to the +1 nucleosome region to promote gene expres-

sion. This interesting Hog1-binding pattern resembles some of

the features of osmoresponsive genes in which Hog1 recruit-

ment at the ORFs depends on the 30 UTR region (Proft et al.,

2006).

Cells deficient in lncRNA still recruit Hog1 at the 30 UTR, butnot at the +1 nucleosome region in CDC28. Remarkably, these

cells induce neither chromatin remodeling nor CDC28 gene in-

duction. On the other hand, when CDC28 lncRNA was induced

by a heterologous activator that does not promote recruitment

of Hog1 at the 30 UTR, the SAPK did not bind at the +1 nucleo-

some region and cells could induce neither chromatin remodel-

ing nor gene expression. Therefore, the combination of the

induction of CDC28 lncRNA transcription and the recruitment

of Hog1 is necessary for gene induction. Correspondingly, the

combination of artificial tethering of Hog1 to a strain containing

the GAL1 at the CDC28 30 UTR, together with the expression

of the lncRNA, allows chromatin remodeling and induction

of CDC28. RSC mediates chromatin remodeling at specific

stress-responsive genes (Mas et al., 2009). Here, we found

that expression of the CDC28 lncRNA was affected not by

conditions. Cell cycle progression was analyzed by fluorescence-activated

cell sorting (FACS), and percentage of cells in G2/M is shown. See also

Figure S7.

(C) Schematic representation of Hog1-mediated gene looping between the

30 UTR and the +1 nucleosome region at the CDC28 locus in response to

osmostress.

Molecular Cell



depletion of RSC, but by depletion of SAGA. In contrast, RSC

was absolutely necessary for chromatin remodeling at the +1

nucleosome region andCDC28 gene induction. Thus, the target-

ing of RSC by Hog1 at the +1 nucleosome region is required for

gene induction.

Unlike osmoresponsive genes in which Hog1 travels with elon-

gating polymerase (Proft et al., 2006), the Hog1-binding pattern

at the CDC28 locus suggested that Hog1 could reach the 50 endof the gene without traveling through the coding region. In yeast,

gene looping has been shown to juxtapose promoter-terminator

regions during active transcription (O’Sullivan et al., 2004).

Indeed, osmostress stimulates Hog1-mediated gene looping in

CDC28. Looping can be prevented by impairing expression of

SSU72 or in cells containing the sua7-1 mutation (Ansari and

Hampsey, 2005; Singh and Hampsey, 2007). Depletion of

Ssu72 or sua7-1 mutation did not alter induction of CDC28

lncRNA but completely abolished CDC28 gene induction, most

likely because Hog1 and RSC cannot be transferred from the

30 UTR to the +1 nucleosome position.

Altogether, these data suggest the following tentative model

for the induction of CDC28 by Hog1 (Figure 7C). In response

to osmostress, Hog1 associates at the 30 UTR region of

CDC28 and induces lncRNA transcription. Once antisense tran-

scription is induced, gene looping is established and Hog1 is

transferred to the +1 nucleosome region in CDC28. The recruit-

ment of Hog1 serves to target the RSC chromatin remodeler,

which remodels the +1 region, thus permitting an increase of

the transcription of the CDC28 gene. It is worth noting that

another example has recently demonstrated that DNA looping

facilitates targeting of chromatin remodeling complexes (Yadon

et al., 2013). Taken together, the regulation of CDC28 transcrip-

tion by the induction of a stress-responsive lncRNA provides a

paradigm by which an lncRNA mediates gene induction

through changes in chromatin architecture.

Induction of the CDC28 lncRNA Controls Cell CycleReentry upon StressThe CDC28 gene encodes the main CDK kinase (CDK1) that

drives progression of the cell cycle in yeast. Cdc28 is regulated

by several mechanisms, including cyclin association and CDK

inhibitors (Bloom and Cross, 2007). However, the increase in

transcription of CDC28 observed upon stress was unexpected

since transcription of CDC28 was assumed to be constant

(Spellman et al., 1998). The increase in CDC28 transcription re-

sulted in an increase of de novo synthesis of Cdc28. In response

to osmostress, Hog1 mediates a rapid, but transient, arrest of

cell cycle progression to allow adaptation (Clotet and Posas,

2007; Duch et al., 2012; Saito and Posas, 2012). One of the

mechanisms under the control of Hog1 consists in the downre-

gulation of Cdc28 activity, which seems to be contradictory

with an increase of Cdc28 protein. Nevertheless, the increase

in Cdc28 protein levels occurred when cells started to recover

from stress; thus, we postulated that this increase in Cdc28 pro-

tein levels should have an effect during the recovery phase.

Indeed, cells deficient in CDC28 lncRNA arrested similar to

wild-type upon stress but reentered cell cycle less efficiently,

suggesting that the increase in Cdc28 permits a faster recovery

of the cell cycle delay caused by stress. Although cell cycle

reentry delay in CDC28 lncRNAD cells might seem modest

(20 min), this lapse of time has proven to be important to maxi-

mize cell survival upon stress (Escote et al., 2004; Duch et al.,

2013). Therefore, Hog1 is able to induce a cell cycle delay and

promote the recovery by controlling transcriptionalCDC28mod-

ulation, thus achieving a different temporal outcome.

In summary, we present here a mechanism of Cdc28 regula-

tion through a stress-inducible lncRNA production that is able

to alter cell cycle progression in response to environmental

challenges. Cdc28 regulation provides a mechanism by which

an lncRNA together with a SAPK can mediate gene induction

through changes of chromatin architecture. Moreover, this

study provides insights into how lncRNAs might affect the

regulation of gene expression through chromatin changes in

eukaryotic cells.

EXPERIMENTAL PROCEDURES

Yeast Strains and Plasmids

Full list and description of strains and plasmids used in this study is included in

the Supplemental Experimental Procedures.

Tiling Array

Wild-type (BY4741) and hogD cells were grown to mid-log phase and sub-

jected (or not) to mild osmostress (0.4 M NaCl for 15 min) or hyper osmostress

(1.2 M NaCl for 100 min). Hybridization of tiling array was performed as

described (David et al., 2006).

Definition of Stress-Induced lncRNAs

Stress-induced lncRNAs were defined as upregulated, with a minimum

change of 2-fold in response to stress (at the indicated osmolarity) in the

wild-type (BY4741) strain. Hog1 dependence was determined by the percent-

age of expression in a hog1Dmutant with respect to the wild-type strain. Table

S1 provides the entire list of the osmoresponsive lncRNA, their relationship

with the sense transcript, and HOG1 dependence.

ChIP-Seq and MNase-Seq

Wild-type and hog1D mutant S. cerevisiae strains were grown to mid-log

phase and exposed to 5 min of osmostress (0.4 M NaCl) for Hog1 immunopre-

cipitation and 10 min for RNA Pol II immunoprecipitation and nucleosome

positioning. ChIP and MNase protocols were performed, and purified DNA

was sequenced as described. Enrichment of Hog1 and RNA Pol II was done

by running the Pyicos enrichment protocol comparing untreated to treated

samples (Nadal-Ribelles et al., 2012).

MNase Nucleosome Mapping

Spheroplasts and digestion with MNase were done with the indicated strains

subjected (or not) to osmostress (0.4 M NaCl for 10 min) or treated with

b-estradiol (100 nM, 10 min). For the analysis of CDC28, DNA was used in a

real-time PCR.

ChIP Assays

Chromatin immunoprecipitation was done as described previously (Zapater

et al., 2007). Briefly, indicated yeast cultures were grown to mid-log phase

and exposed (or not) to osmotic stress (0.4 M NaCl, 5 min) or treated with

b-estradiol (100 nM, 5 min). Real-time PCR of the indicated regions was

performed.

3C Analysis

Cells were grown to mid-log phase before being subjected (or not) to osmo-

stress (0.4 M NaCl, 10 min). 3C analysis was performed as described previ-

ously, with minor modifications indicated in the Supplemental Experimental

Procedures.


Molecular Cell



Metabolic Labeling

Briefly, indicated PHO85-TAP strains were grown in YPD and shifted to MET

media for 2 hr before being stressed. A mixture of 35S[methionine] and 0.4 M

NaCl was added simultaneously.

Flow Cytometry

Flow cytometry experiments were performed as described (Pelet et al., 2011).

Cells were stressed for 45 min with 0.4 M NaCl. To study cell cycle progres-

sion, cdc15ts (cdc15-2) cells were synchronized at 37�C (incubated for 2 hr)

and released at 25�C to allow cell cycle progression.

ACCESSION NUMBERS

The link http://steinmetzlab.embl.de/francescData/arrayProfile/index.html

directs to an interface to visualize array expression data. Raw array data are

available from ArrayExpress under accession number E-MTAB-1686.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures,

seven figures, and one table and can be found with this article online at

http://dx.doi.org/10.1016/j.molcel.2014.01.006.

ACKNOWLEDGMENTS

We thank L. Subirana, S. Obejas, and A. Fernandez for technical support; Nuria

Conde for ChIP-seq analysis; and Javier Jimenez and Alba Duch for cell cycle

analysis. We thank Christoph Schuller for helpful discussions throughout the

study and the Gal4DBD-Msn2 plasmid. We thank Dr. S. Buratowski and

Dr. N. Proudfoot for strains. M.N.-R. is a recipient of a FIS fellowship. This

work was supported by grants from the Spanish Ministry of Economy and

Competitiveness (BFU2012-33503 and FEDER to F.P., BFU2011-26722 to

E.d.N.), the Fundacion Marcelino Botın (FMB), and the Consolider Ingenio

2010 programme CSD2007-0015 (to F.P.). This work was supported by the

National Institutes of Health and Deutsche Forschungsgemeinschaft

(L.M.S.). F.P. and E.d.N. are recipients of an ICREA Academia award (Genera-

litat de Catalunya). The authors declare no competing financial interest.

Received: July 26, 2013

Revised: October 31, 2013

Accepted: December 31, 2013

Published: February 6, 2014

REFERENCES

Adrover, M.A., Zi, Z., Duch, A., Schaber, J., Gonzalez-Novo, A., Jimenez, J.,

Nadal-Ribelles, M., Clotet, J., Klipp, E., and Posas, F. (2011). Time-dependent

quantitative multicomponent control of the G1-S network by the stress-acti-

vated protein kinase Hog1 upon osmostress. Sci. Signal. 4, ra63.

Alepuz, P.M., Jovanovic, A., Reiser, V., and Ammerer, G. (2001). Stress-

induced map kinase Hog1 is part of transcription activation complexes. Mol.

Cell 7, 767–777.

Alepuz, P.M., de Nadal, E., Zapater, M., Ammerer, G., and Posas, F. (2003).

Osmostress-induced transcription by Hot1 depends on a Hog1-mediated

recruitment of the RNA Pol II. EMBO J. 22, 2433–2442.

Ansari, A., and Hampsey, M. (2005). A role for the CPF 30-end processing

machinery in RNAP II-dependent gene looping. Genes Dev. 19, 2969–2978.

Bloom, J., and Cross, F.R. (2007). Multiple levels of cyclin specificity in cell-

cycle control. Nat. Rev. Mol. Cell Biol. 8, 149–160.

Capaldi, A.P., Kaplan, T., Liu, Y., Habib, N., Regev, A., Friedman, N., and

O’Shea, E.K. (2008). Structure and function of a transcriptional network acti-

vated by the MAPK Hog1. Nat. Genet. 40, 1300–1306.

Clotet, J., and Posas, F. (2007). Control of cell cycle in response to

osmostress: lessons from yeast. Methods Enzymol. 428, 63–76.


Cook, K.E., and O’Shea, E.K. (2012). Hog1 controls global reallocation of RNA

Pol II upon osmotic shock in Saccharomyces cerevisiae. G3 (Bethesda) 2,

1129–1136.

David, L., Huber, W., Granovskaia, M., Toedling, J., Palm, C.J., Bofkin, L.,

Jones, T., Davis, R.W., and Steinmetz, L.M. (2006). A high-resolution map of

transcription in the yeast genome. Proc. Natl. Acad. Sci. USA 103, 5320–5325.

de Nadal, E., and Posas, F. (2010). Multilayered control of gene expression by

stress-activated protein kinases. EMBO J. 29, 4–13.

De Nadal, E., Zapater, M., Alepuz, P.M., Sumoy, L., Mas, G., and Posas, F.

(2004). The MAPK Hog1 recruits Rpd3 histone deacetylase to activate osmor-

esponsive genes. Nature 427, 370–374.

de Nadal, E., Ammerer, G., and Posas, F. (2011). Controlling gene expression

in response to stress. Nat. Rev. Genet. 12, 833–845.

Duch, A., de Nadal, E., and Posas, F. (2012). The p38 and Hog1 SAPKs control

cell cycle progression in response to environmental stresses. FEBS Lett. 586,

2925–2931.

Duch, A., Felipe-Abrio, I., Barroso, S., Yaakov, G., Garcıa-Rubio, M., Aguilera,

A., de Nadal, E., and Posas, F. (2013). Coordinated control of replication and

transcription by a SAPK protects genomic integrity. Nature 493, 116–119.

Escote, X., Zapater, M., Clotet, J., and Posas, F. (2004). Hog1 mediates cell-

cycle arrest in G1 phase by the dual targeting of Sic1. Nat. Cell Biol. 6, 997–

1002.

Gasch,A.P.,Spellman,P.T.,Kao,C.M.,Carmel-Harel,O.,Eisen,M.B., Storz,G.,

Botstein, D., and Brown, P.O. (2000). Genomic expression programs in the

responseof yeast cells toenvironmental changes.Mol. Biol.Cell11, 4241–4257.

Guttman, M., and Rinn, J.L. (2012). Modular regulatory principles of large non-

coding RNAs. Nature 482, 339–346.

Hainer, S.J., Pruneski, J.A., Mitchell, R.D., Monteverde, R.M., and Martens,

J.A. (2011). Intergenic transcription causes repression by directing nucleo-

some assembly. Genes Dev. 25, 29–40.

Hohmann, S., Krantz, M., and Nordlander, B. (2007). Yeast osmoregulation.

Methods Enzymol. 428, 29–45.

Houseley, J., Rubbi, L., Grunstein, M., Tollervey, D., and Vogelauer, M. (2008).

A ncRNA modulates histone modification and mRNA induction in the yeast

GAL gene cluster. Mol. Cell 32, 685–695.

Jacquier, A. (2009). The complex eukaryotic transcriptome: unexpected

pervasive transcription and novel small RNAs. Nat. Rev. Genet. 10, 833–844.

Kim, T., Xu, Z., Clauder-Munster, S., Steinmetz, L.M., and Buratowski, S.

(2012). Set3 HDAC mediates effects of overlapping noncoding transcription

on gene induction kinetics. Cell 150, 1158–1169.

Klopf, E., Paskova, L., Sole, C., Mas, G., Petryshyn, A., Posas, F.,

Wintersberger, U., Ammerer, G., and Schuller, C. (2009). Cooperation between

the INO80 complex and histone chaperones determines adaptation of stress

gene transcription in the yeast Saccharomyces cerevisiae. Mol. Cell. Biol.

29, 4994–5007.

Kopcewicz, K.A., O’Rourke, T.W., and Reines, D. (2007). Metabolic regulation

of IMD2 transcription and an unusual DNA element that generates short tran-

scripts. Mol. Cell. Biol. 27, 2821–2829.

Kuehner, J.N., and Brow, D.A. (2008). Regulation of a eukaryotic gene by GTP-

dependent start site selection and transcription attenuation. Mol. Cell 31,

201–211.

Louvion, J.F., Havaux-Copf, B., and Picard, D. (1993). Fusion of GAL4-VP16 to

a steroid-binding domain provides a tool for gratuitous induction of galactose-

responsive genes in yeast. Gene 131, 129–134.

Loya, T.J., O’Rourke, T.W., and Reines, D. (2012). A genetic screen for termi-

nator function in yeast identifies a role for a new functional domain in termina-

tion factor Nab3. Nucleic Acids Res. 40, 7476–7491.

Margaritis, T., Oreal, V., Brabers, N., Maestroni, L., Vitaliano-Prunier, A.,

Benschop, J.J., van Hooff, S., van Leenen, D., Dargemont, C., Geli, V., and

Holstege, F.C. (2012). Two distinct repressive mechanisms for histone 3 lysine

4 methylation through promoting 30-end antisense transcription. PLoS Genet.

8, e1002952.



Molecular Cell



Martınez-Montanes, F., Pascual-Ahuir, A., and Proft, M. (2010). Toward a

genomic view of the gene expression program regulated by osmostress in

yeast. OMICS 14, 619–627.

Mas, G., de Nadal, E., Dechant, R., Rodrıguez de la Concepcion, M.L., Logie,

C., Jimeno-Gonzalez, S., Chavez, S., Ammerer, G., and Posas, F. (2009).

Recruitment of a chromatin remodelling complex by the Hog1 MAP kinase

to stress genes. EMBO J. 28, 326–336.

Miller, C., Schwalb, B., Maier, K., Schulz, D., Dumcke, S., Zacher, B., Mayer,

A., Sydow, J., Marcinowski, L., Dolken, L., et al. (2011). Dynamic transcriptome

analysis measures rates of mRNA synthesis and decay in yeast. Mol. Syst.

Biol. 7, 458.

Molin, C., Jauhiainen, A., Warringer, J., Nerman, O., and Sunnerhagen, P.

(2009). mRNA stability changes precede changes in steady-state mRNA

amounts during hyperosmotic stress. RNA 15, 600–614.

Nadal-Ribelles, M., Conde, N., Flores, O., Gonzalez-Vallinas, J., Eyras, E.,

Orozco, M., de Nadal, E., and Posas, F. (2012). Hog1 bypasses stress-medi-

ated down-regulation of transcription by RNA polymerase II redistribution

and chromatin remodeling. Genome Biol. 13, R106.

O’Sullivan, J.M., Tan-Wong, S.M., Morillon, A., Lee, B., Coles, J., Mellor, J.,

and Proudfoot, N.J. (2004). Gene loops juxtapose promoters and terminators

in yeast. Nat. Genet. 36, 1014–1018.

Pelet, S., Rudolf, F., Nadal-Ribelles, M., de Nadal, E., Posas, F., and Peter, M.

(2011). Transient activation of the HOG MAPK pathway regulates bimodal

gene expression. Science 332, 732–735.

Pinskaya, M., Gourvennec, S., and Morillon, A. (2009). H3 lysine 4 di- and tri-

methylation deposited by cryptic transcription attenuates promoter activation.

EMBO J. 28, 1697–1707.

Pokholok, D.K., Zeitlinger, J., Hannett, N.M., Reynolds, D.B., and Young, R.A.

(2006). Activated signal transduction kinases frequently occupy target genes.

Science 313, 533–536.

Posas, F., Chambers, J.R., Heyman, J.A., Hoeffler, J.P., de Nadal, E., and

Arino, J. (2000). The transcriptional response of yeast to saline stress.

J. Biol. Chem. 275, 17249–17255.

Proft, M., Mas, G., deNadal, E., Vendrell, A., Noriega, N., Struhl, K., and Posas,

F. (2006). The stress-activated Hog1 kinase is a selective transcriptional elon-

gation factor for genes responding to osmotic stress. Mol. Cell 23, 241–250.

Regot, S., de Nadal, E., Rodrıguez-Navarro, S., Gonzalez-Novo, A., Perez-

Fernandez, J., Gadal, O., Seisenbacher, G., Ammerer, G., and Posas, F.

(2013). The Hog1 stress-activated protein kinase targets nucleoporins to con-

trol mRNA export upon stress. J. Biol. Chem. 288, 17384–17398.

Rhee, H.S., and Pugh, B.F. (2012). Genome-wide structure and organization of

eukaryotic pre-initiation complexes. Nature 483, 295–301.

Romero-Santacreu, L., Moreno, J., Perez-Ortın, J.E., and Alepuz, P. (2009).

Specific and global regulation of mRNA stability during osmotic stress in

Saccharomyces cerevisiae. RNA 15, 1110–1120.

Saito, H., and Posas, F. (2012). Response to hyperosmotic stress. Genetics

192, 289–318.

Singh, B.N., and Hampsey, M. (2007). A transcription-independent role for

TFIIB in gene looping. Mol. Cell 27, 806–816.

Sole, C., Nadal-Ribelles, M., Kraft, C., Peter, M., Posas, F., and de Nadal, E.

(2011). Control of Ubp3 ubiquitin protease activity by the Hog1 SAPK modu-

lates transcription upon osmostress. EMBO J. 30, 3274–3284.

Spellman, P.T., Sherlock, G., Zhang, M.Q., Iyer, V.R., Anders, K., Eisen, M.B.,

Brown, P.O., Botstein, D., and Futcher, B. (1998). Comprehensive identifica-

tion of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by

microarray hybridization. Mol. Biol. Cell 9, 3273–3297.

Tan-Wong, S.M., Zaugg, J.B., Camblong, J., Xu, Z., Zhang, D.W., Mischo,

H.E., Ansari, A.Z., Luscombe, N.M., Steinmetz, L.M., and Proudfoot, N.J.

(2012). Gene loops enhance transcriptional directionality. Science 338,

671–675.

Thiebaut, M., Colin, J., Neil, H., Jacquier, A., Seraphin, B., Lacroute, F., and

Libri, D. (2008). Futile cycle of transcription initiation and termination modu-

lates the response to nucleotide shortage in S. cerevisiae. Mol. Cell 31,

671–682.

Uhler, J.P., Hertel, C., and Svejstrup, J.Q. (2007). A role for noncoding tran-

scription in activation of the yeast PHO5 gene. Proc. Natl. Acad. Sci. USA

104, 8011–8016.

van Werven, F.J., Neuert, G., Hendrick, N., Lardenois, A., Buratowski, S., van

Oudenaarden, A., Primig, M., and Amon, A. (2012). Transcription of two long

noncoding RNAs mediates mating-type control of gametogenesis in budding

yeast. Cell 150, 1170–1181.

Warringer, J., Hult, M., Regot, S., Posas, F., and Sunnerhagen, P. (2010). The

HOG pathway dictates the short-term translational response after hyperos-

motic shock. Mol. Biol. Cell 21, 3080–3092.

Weake, V.M., and Workman, J.L. (2010). Inducible gene expression: diverse

regulatory mechanisms. Nat. Rev. Genet. 11, 426–437.

Wei, W., Pelechano, V., Jarvelin, A.I., and Steinmetz, L.M. (2011). Functional

consequences of bidirectional promoters. Trends Genet. 27, 267–276.

Wu, J., Delneri, D., and O’Keefe, R.T. (2012). Non-coding RNAs in

Saccharomyces cerevisiae: what is the function? Biochem. Soc. Trans. 40,

907–911.

Xu, Z., Wei, W., Gagneur, J., Perocchi, F., Clauder-Munster, S., Camblong, J.,

Guffanti, E., Stutz, F., Huber, W., and Steinmetz, L.M. (2009). Bidirectional pro-

moters generate pervasive transcription in yeast. Nature 457, 1033–1037.

Xu, Z., Wei, W., Gagneur, J., Clauder-Munster, S., Smolik, M., Huber, W., and

Steinmetz, L.M. (2011). Antisense expression increases gene expression vari-

ability and locus interdependency. Mol. Syst. Biol. 7, 468.

Yadon, A.N., Singh, B.N., Hampsey, M., and Tsukiyama, T. (2013). DNA loop-

ing facilitates targeting of a chromatin remodeling enzyme. Mol. Cell 50,

93–103.

Zapater, M., Sohrmann, M., Peter, M., Posas, F., and de Nadal, E. (2007).

Selective requirement for SAGA in Hog1-mediated gene expression depend-

ing on the severity of the external osmostress conditions. Mol. Cell. Biol. 27,

3900–3910.


Documents

Control of Cdc28 CDK1 by a Stress-Induced lncRNAet al., 2010). Thus, Hog1 plays a key role in the regulation of mRNA biogenesis (de Nadal et al., 2011; de Nadal and Posas, 2010; Martı´nez-Montan˜e´s