Kcnq1ot1/Lit1 noncoding RNA mediates transcriptional silencing by ...

Kcnq1ot1/Lit1 noncoding RNA mediates transcriptional silencing by targeting to the perinucleolar region

Faizaan Mohammad1# Radha Raman Pandey

1# Takshi Nagano

3,

Lyubomira Chakalova3, Tanmoy Mondal

1, Peter Fraser

3 and

Chandrasekhar Kanduri1, 2,

*

1Department of Genetics and Pathology, Dag Hammarskölds Väg 20, 75185

Rudbeck Laboratory, Uppsala University, Uppsala, Sweden 2

Department of Genetics and Development, Norbyvägen18A, S-75236, Uppsala

University, Uppsala, Sweden 3 Laboratory of Chromatin and Gene Expression, Babraham Institute, Babraham

Research Campus, Cambridge, CB2243AT , United Kingdom

* = Address for correspondence

Email: [email protected]

Telephone: 0046739600450

Fax: 004618558931

# = These authors have contributed equally to this manuscript

Running title: An antisense RNA mediates Perinucleolar localization

ACCEPTED

Copyright © 2008, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.Mol. Cell. Biol. doi:10.1128/MCB.02263-07 MCB Accepts, published online ahead of print on 25 February 2008

on February 13, 2018 by guest

http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Abstract 1

2

The Kcnq1ot1 antisense non-coding RNA has been implicated in long-range bidirectional 3

silencing, but the underlying mechanisms remain enigmatic. Here we characterize a 4

domain at the 5’ end of the Kcnq1ot1 RNA that carries out transcriptional silencing of 5

linked genes using an episomal vector system. The bidirectional silencing property of 6

Kcnq1ot1 maps to a highly conserved repeat motif within the silencing domain, which 7

directs transcriptional silencing by interaction with chromatin resulting in histone H3 8

lysine9-trimethylation. Intriguingly, the silencing domain is also required to target the 9

episomal vector to the perinucleolar compartment, during mid S phase. Collectively, our 10

data unfold a novel mechanism by which an antisense RNA mediates transcriptional gene 11

silencing of chromosomal domains by targeting them to distinct nuclear compartments, 12

known to be rich in heterochromatic machinery. 13

14

15

16

17

18

19

20

21

22

Keywords: Antisense RNA/Chromatin/Epigenetics/Gene silencing/Kcnq1ot1 23

24

25

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Introduction 26

Recent in silico-based analyses of the mammalian genome have predicted that 27

approximately 60- 70% of the genome is transcribed. However, only ∼1.5% of the 28

genome encodes protein, but the rest of the mammalian genome appears to be 29

transcribing non-protein-coding RNAs (19). Interestingly, a detailed analysis of the 30

mouse transcriptome by the FANTOM3 consortium indicated that more than 72% of all 31

43,553 genome-mapped transcription units overlap transcripts encoded from the opposite 32

strand, with the majority of transcripts being noncoding. This suggests that antisense 33

transcription is extremely widespread in mammals (14). 34

An accumulating weight of new evidence from a variety of model systems indicates that 35

noncoding RNAs play an important role in epigenetically controlled gene expression (2, 36

25, 28, 36). Noncoding RNAs can be classified into two groups: housekeeping and 37

regulatory noncoding RNAs. Housekeeping noncoding RNAs, which include tRNAs, 38

rRNAs and snRNAs, have a range of functions that are necessary for cell viability. 39

Regulatory noncoding RNAs have been implicated in phenomena that have an impact on 40

cellular differentiation and development in both plants and animals. 41

Based on size, regulatory noncoding RNAs can be further grouped into two classes: short 42

and long regulatory noncoding RNAs. Short regulatory noncoding RNAs are specifically 43

21 to 31 nucleotides and include miRNAs siRNAs and piRNAs, which regulate gene 44

silencing at the transcriptional and/or post-transcriptional level (2). Long regulatory 45

noncoding RNAs, whose lengths range in size from 100 bp to several hundred kilobases, 46

have been shown to participate in several important biological functions. Among the long 47

noncoding RNAs, Xist, an X-inactivation-specific transcript, and Tsix, its antisense 48

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

counterpart, have been well studied for their functional role in mammalian dosage 49

compensation, an epigenetic process by which levels of X-linked gene products are 50

maintained at an equal ratio between the sexes (11, 15). During early embryonic 51

development, the X-inactivation process is initiated with the onset of Xist gene 52

transcription, followed by coating of this RNA all along the future inactive X-53

chromosome. This triggers step-wise recruitment of the heterochromatin machinery, 54

which leads to silencing of the chromosome in cis. However, on the future active X-55

chromosome, Tsix antagonize Xist-mediated X-chromosome inactivation by 56

epigenetically regulating the Xist locus (21, 27, 30). 57

One of the salient features of the gene clusters exhibiting parent of origin-specific 58

expression patterns is the prevalence of noncoding RNAs at these clusters (22). Of 59

particular interest is that the noncoding antisense RNAs encoded from the differentially 60

methylated imprinting control regions (ICR) are generally long, ranging in size from 50 61

kb to several hundred kilo bases, and are reciprocally imprinted to their sense 62

counterparts. These long antisense transcripts have been functionally implicated in long-63

range bidirectional control of gene expression at imprinted clusters (18, 26, 29, 33). This 64

unique bidirectional control of the expression of flanking genes by the long antisense 65

RNAs raises the important question of how these RNAs differ in their execution of 66

silencing mechanisms, as compared to the antisense RNAs, such as Tsix, whose effects 67

are primarily restricted to overlapping genes. 68

In this investigation to address the functional role of long antisense noncoding RNAs, we 69

have exploited an imprinted cluster located at the distal end of mouse chromosome 7. 70

This cluster is divided into two sub-expression domains: the H19/Igf2 domain and the 71

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Kcnq1 domain. Imprinting in the H19/Igf2 domain is primarily regulated by a chromatin 72

insulator located at the 5´end of the H19 gene (1, 9, 12, 31). Imprinted gene regulation in 73

the Kcnq1 imprinted domain seems to require the presence of a long noncoding antisense 74

transcript, Kcnq1ot1, whose expression on the paternal chromosome has been linked to 75

the bidirectional repression in cis of eight maternally expressed genes spread over a 76

mega-base region (8, 18, 23, 33). The promoter of the Kcnq1ot1 transcript maps to the 77

Kcnq1 imprinting control region (ICR) in intron 10 of the Kcnq1 gene. The Kcnq1 ICR is 78

methylated on the maternal chromosome, but is unmethylated on the paternal 79

chromosome, which encodes the Kcnq1ot1 antisense transcript. It has been recently 80

documented that the imprinted silencing of the Kcnq1 domain involves histone 81

modifications (13, 16, 34). However, the specific mechanisms underlying the Kcnq1ot1-82

mediated transcriptional silencing are yet unclear. 83

Previously, we have shown that a 3.6 kb Kcnq1 ICR fragment, containing the Kcnq1ot1 84

antisense promoter and 1.7 kb of downstream sequence, is sufficient to silence flanking 85

genes, and more importantly, that the transcriptional elongation of the 1.7 kb sequence 86

results in the efficient silencing of flanking genes in an episome-based system (13, 32). 87

Based on these observations, we posit that the 1.7 kb Kcnq1ot1 sequence may harbor the 88

crucial information required for bidirectional silencing, and that this region silences 89

flanking genes more efficiently when it is part of long transcripts than it does as part of 90

short antisense transcripts. In this investigation, we sought to address the mechanisms by 91

which Kcnq1ot1 executes long-range bidirectional silencing. The data presented here 92

document that Kcnq1ot1 harbors a silencing domain that carries out transcriptional 93

silencing in an orientation-dependent and position independent manner downstream of 94

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

the antisense promoter, through promoting its interaction with the chromatin. 95

Interestingly, this domain also targets the flanking sequences to the perinucleolar 96

compartment, a distinct nuclear space enriched with factors involved in replication and 97

heterochromatin formation. These results unfold a mechanism by which the silencing 98

domain of Kcnq1ot1 initiates transcriptional silencing by recruiting repressive chromatin 99

machinery and spreading it bidrectionally over flanking chromosomal regions. This is 100

clonally maintained through subsequent cell divisions by targeting the flanking sequences 101

to the perinucleolar compartment. 102

Results 103

Functional characterization of a silencing domain at the 5´end of the Kcnq1ot1 104

antisense RNA 105

Previous investigations from our lab have implicated a 1.7 kb sequence (1897-3625, 106

Kcnq1ot1 transcription start site at 1897), downstream of the Kcnq1ot1 promoter in the 107

3.6 kb Kcnq1 ICR, in bidirectional silencing of flanking reporter genes. To define the 108

sequences that are critical for bidirectional silencing, we initially generated serial 109

deletions encompassing the region that encodes the 1.7 kb antisense transcription unit in 110

the 3.6 Kb Kcnq1 ICR using a PCR-based strategy (Figure 1A). We inserted all these 111

modified ICRs between the H19 and Hygromycin reporter genes in the PH19 episome. 112

In all of these constructs, the ICR is oriented in such a way that the Kcnq1ot1 antisense 113

promoter faces the H19 promoter, thus the H19 and Hygromycin gene promoters become 114

overlapping and nonoverlapping promoters, respectively, to the Kcnq1ot1 promoter 115

(Figure 1A). We transfected all of the episomal constructs into the human placenta-116

derived JEG-3 cell line and analyzed the activity of Kcnq1ot1, H19 and Hygromycin 117

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

genes using an RNase protection assay (RPA). The Hygromycin gene activity was also 118

measured by counting the Hygromycin resistant colonies obtained after selection with 119

hygromycin. 120

Analysis of the serial deletions indicated that a fragment encompassing the antisense 121

promoter and downstream 600 bp region (see PS1050 in Figure 1A) could not silence the 122

Hygromycin reporter gene. However, Hygromycin gene silencing was detected with a 123

fragment encompassing both the antisense promoter and downstream 1.7 kb region (see 124

PS1947 in Figure 1A), suggesting that a 1.1 kb region (2479-3625; Kcnq1ot1 125

transcription start site at 1897) may play an important role in silencing (Figure 1B & 126

Figure S1A). 127

When we analyzed the fragments that harbored serial deletions for their ability to support 128

promoter activity in a promoter-less Luciferase construct, we detected significant 129

Luciferase activity from the fragments that did not encompass the 1.1 kb region, but not 130

from the fragment that included it (compare PGL_Basic742 and PGL_Basic1050 with 131

PGL_basic1947 in Figure 1C). Interestingly, however, all of these serial deletions could 132

produce RNA in the episomal context (Figure S1B) as well as in the Luciferase 133

constructs (Figure S1C). Lack of Luciferase activity in PGL_Basic1947 could be due to 134

retention of the fused transcript (contains the 1.1 kb portion of Kcnq1ot1+Luciferase 135

gene) in the nuclear compartment and/or interference in the translation of the fused 136

transcript due to occupancy of bulky heterochromatin complex in the 1.1 kb portion of 137

the fused transcript. We presume that these results collectively indicate that the 1.1 kb 138

region in the 1.7 kb Kcnq1ot1 sequence holds crucial information required for 139

bidirectional silencing. 140

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

To further define the minimal region required for bidirectional silencing, we created 141

selective internal deletions in the 1.1 kb Kcnq1ot1 sequence using an approach as 142

described in the experimental procedures. To incorporate internal deletions into the 1.1 143

kb sequence, we used a larger version of the Kcnq1 ICR that contained a 3.9 kb Kcnq1ot1 144

sequence (PS6 in Figure 2A). In the PS6 episome, the Kcnq1ot1 transcript encoded from 145

the ICR runs on through the H19 coding gene and is truncated at the SV40 polyA 146

sequence inserted 11.4 kb downstream of the Kcnq1ot1 start site (Figure 2A, data not 147

shown). We inserted each of the modified ICRs, carrying various internal deletions, 148

between the H19 and Hygromycin reporter genes in the PH19 episome (Figure 1A & 2A) 149

and performed silencing assay as described above. 150

Selective deletion of an 890 bp sequence (2514-3402, PS6A1 in Figure 2A) from the 151

fine-mapped 1.1 kb region, 617 bp downstream of the antisense transcription start site, 152

resulted in incomplete silencing of the H19 gene, whereas Hygromycin gene was 153

significantly increased. We observed a slight increase in the activity of the Kcnq1ot1 154

promoter in PS6A1 (Figure S2A). We have further narrowed down the 890 bp silencing 155

domain to 450 bp by creating several deletions of 150 bp starting from 5’ end of the 890 156

bp fragment (PS6A2-PS6A4 in Figure 2A & S2A) and found no discernible differences 157

between PS6A1 and PS6A2-PS6A4, as the Hygromycin, H19 and Kcnq1ot1 gene 158

activities more or less remained the same between these constructs (Figure 2A & S2A). 159

If the 890 bp fragment was a silencing domain, one would expect a complete activation 160

of the H19 and Hygromycin genes on deletion of this fragment rather than the complete 161

activation only of the Hygromycin gene. Previously, by analyzing the kinetics of 162

silencing and heterochromatin formation on the H19 and Hygromycin genes in relation to 163

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Kcnq1ot1 transcription, we have shown that the silencing of the nonoverlapping 164

Hygromycin gene occurs primarily due to Kcnq1ot1-mediated heterochromatin 165

formation, whereas the overlapping H19 gene silencing occurs by both the Kcnq1ot1 166

transcription-mediated occlusion of the basal transcription machinery and by 167

heterochromatin formation (13). So we presume that specific activation of the 168

Hygromycin gene in the 890 bp deletion could be due to the loss of the Kcnq1ot1-169

mediated heterochromatin formation, while the incomplete silencing of the H19 gene 170

could be as a result of partial occlusion of the basal transcription machinery due to 171

Kcnq1ot1 transcription. 172

To further confirm these observations, we have deleted the 890 bp Kcnq1ot1 silencing 173

domain (henceforth the 890 bp Kcnq1ot1 silencing domain is known as Kcnqot1 SD) in 174

the PS4polyA4.9 construct (contains 4.9 kb Kcnq1ot1 sequence followed by SV40 PolyA 175

sequence), wherein we have previously shown that the 4.9 kb long episome encoded 176

Kcnq1ot1 transcript is capable of inducing moderate silencing of H19 and Hygromycin 177

genes. In this construct the Kcnq1ot1 transcript does not overlap with the H19 Gene (13). 178

Interestingly, we did not see silencing of the H19 as well as Hygromycin genes in 179

PS4polyA4.9A1 when compared to PS4polyA4.9, suggesting that the incomplete 180

silencing of the H19 gene in PSA1-PSA4 is mainly due to occlusion of basal transcription 181

machinery (see PS4polyA4.9A1 and PS4polyA4.9 in Fig.2A & S2A). 182

To further reinforce this statement, we used chromatin immunoprecipitation (ChIP) 183

assays to determine the levels of the major constituents of the pre-initiation complex, 184

RNA PolII and TFIIB, on both the H19 and the Hygromycin promoters. As shown in 185

Figure S3A, we found significant reduction of both RNA PolII and TFIIB on the H19 186

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

promoter, but not on the Hygromycin promoter, in PS6A1. This suggests that moderate 187

silencing of the H19 gene in PS6A1 is primarily due to occlusion of the pre-initiation 188

complex from its promoter by the antisense transcription machinery. In addition, we have 189

analyzed two positions in the H19 coding regions (H19c1 and H19c2) to check for the 190

loading efficiency of basal transcription machinery over the non-promoter regions, which 191

revealed that the basal transcription machinery is specifically loaded over the promoters 192

(Fig. S3B). 193

The Kcnqot1 SD induces gene silencing in an orientation and transcription-dependent, 194

but position-independent manner downstream of the Kcnq1ot1 promoter 195

We next addressed whether change in position and/or orientation of the Kcnqot1 SD 196

relative to the Kcnq1ot1 promoter affects its bidirectional silencing property. First, we 197

wanted to investigate whether transcription through the Kcnqot1 SD is a prerequisite for 198

the silencing process. To address this issue, we moved the Kcnqot1 SD from its natural 199

position to 1.9 kb upstream of the antisense promoter. In this configuration, the Kcnqot1 200

SD is not transcribed by the Kcnq1ot1 promoter (PS6A11 & PS6A12 in Figure 2B). As 201

can be seen from Figure 2B & S2B, the Kcnq1 ICR in this configuration could not silence 202

the Hygromycin gene, suggesting that transcription through the Kcnqot1 SD is crucial for 203

bidirectional silencing. However, when we moved the Kcnqot1 SD from its natural 204

position to 3.3 kb downstream of the antisense promoter in correct orientation, we 205

observed significant silencing of both the H19 and Hygromycin genes, indicating that the 206

Kcnqot1 SD has all the information required for bidirectional silencing and that it carries 207

out silencing in a distance-independent manner, when present downstream of the 208

antisense promoter (PS6A7 in Figure 2B & S2B). 209

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

We next addressed the effect of the orientation of the Kcnqot1 SD on bidirectional 210

silencing, by inverting the Kcnqot1 SD at its native position, as well as at other positions. 211

The Kcnq1ICR with the inverted Kcnqot1 SD no longer brought about silencing of the 212

Hygromycin gene as compared to the wild type Kcnq1 ICR, suggesting that the sequence 213

of the transcribed Kcnqot1 SD is crucial for bidirectional silencing (see PS6A5 & PS6A6 214

in Figure 2A-B & S2A-B). Moreover, when we placed the Kcnqot1 SD in both 215

orientations 3.3 kb downstream of the Kcnq1ot1 transcription start site in PS6A5, we 216

observed bidirectional silencing only with its native orientation but not with the opposite 217

orientation (see PS6A5, PS6A8 & PS6A9 in Figure 2A & 2B). In addition, when we 218

replaced the Kcnqot1 SD with a neutral fragment in the Kcnq1 ICR, we could not detect 219

any silencing by the Kcnq1 ICR (PS6A10 in Figure 2B & S2B). Taken together, these 220

experiments provide a strong support for the functional role of the Kcnq1ot1 antisense 221

RNA in bidirectional silencing. 222

We next wanted to investigate whether the loss of bidirectional silencing upon Kcnqot1 223

SD deletion could be due to disturbance in its localization or stability of the transcript. To 224

this end, we first characterized the localization of episome-encoded Kcnq1ot1 transcript 225

with respect to nuclear and cytoplasmic compartments. We found that the Kcnq1ot1 226

transcript is exclusively localized in the nuclear compartment. We, however, noted that 227

the deletion of the Kcnqot1 SD had no effect on its nuclear localization as well as 228

stability of the transcript (Figure 2D-F), suggesting that the lack of bidirectional silencing 229

in the PS6A1 construct is not due to change in the half-life or nuclear specific 230

localization of the PS6A1 episome-encoded Kcnq1ot1 transcript. 231

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Functional role of conserved sequence motifs in the Kcnq1ot1 antisense RNA 232

We next sought to address the functional role of specific sequences within the Kcnqot1 233

SD that are critical for bidirectional silencing. A previous bioinformatic study identified 234

five evolutionarily conserved 30 bp repeats sequences (MD1 repeats) at the 5’ end of the 235

Kcnq1ot1 transcript (17). Interestingly, these repeat sequences map between the 236

Kcnq1ot1 promoter and the Kcnqot1 SD identified in this study. Although PS6A1 has all 237

the five 30 bp repeats intact, there was no detectable silencing of the Hygromycin gene, 238

suggesting that these repeats do not play any functional role in the silencing process. This 239

observation is consistent with a recently published report that targeted deletion of these 240

conserved repeats from the Kcnq1 ICR in the mouse had no effect on the imprinting of 241

flanking genes (18). 242

In another recent bioinformatic study, several evolutionarily conserved motifs were 243

identified in the Kcnq1 ICR (24). Interestingly, the A1 (TCCGAGTY) & A2 244

(YGYGGTTCYGAG) conserved motifs identified in this study map to the 3´end of the 245

Kcnqot1 SD. We thought to explore the functional role of A1 and A2 motifs in the 246

Kcnq1ot1-mediated bidirectional silencing. When we mutated 5 conserved residues in the 247

A2 motif, we observed moderate loss of silencing of the Hygromycin gene (PS6 & 248

PS6A15 in Figure 2C & S2C). However, when we introduced mutations into the 249

conserved residues of the A1 motif (PS6A16 in Figure 2C), we could not detect any loss 250

of silencing. To gain further insights into the functional role of the A2 motif in the 251

bidirectional silencing, we generated episome constructs by inserting Kcnq1ot1 252

sequences with (PS6A13) and without (PS6A14) A2 motif into PS6A1. As can be seen 253

from figure 2C & S2C, we noticed significant activation of Hygromycin gene in PS6A14 254

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

but not in PS6A13, implicating a critical role for the A2 motif in long-range 255

transcriptional silencing. 256

Preliminary results from our lab indicated that Kcnq1ot1 antisense RNA spans more than 257

90 kb in length in vivo. A search for the sequence motifs that show homology to the A2 258

motif in the 90 kb long Kcnq1ot1 transcription unit revealed none, indicating that A2 259

could represent the most important functional sequence in the antisense RNA. To 260

address this issue further, we have PCR amplified overlapping Kcnq1ot1 sequences 261

spanning about 52 kb region of Kcnq1ot1 (each fragment ranging in size from 3 to 5 kb 262

in length) and cloned into PS6A1 (Figure 3A). We have analyzed the effect of these 263

fragments on the silencing by measuring the Hygromycin gene activity. Interestingly, 264

none of the fragments encompassing the 52 kb Kcnq1ot1 transcription unit showed 265

silencing activity that is comparable to the Kcnqot1 SD, indicating that the 890 bp 266

sequence could be one of the main functional sequences mediating the long-range 267

transcriptional gene silencing (Figure 3B). We did not see any significant effect on the 268

Kcnq1ot1 promoter activity due to insertion of these fragments in PS6A1 (Figure S4) 269

Flanking the Kcnqot1 SD with the Xist chromatin attachment region increases the 270

efficiency of bidirectional silencing 271

Previously, by selectively incorporating deletions into Xist in an ES cell model system, it 272

has been shown that the chromatin localizing property of the Xist maps to several regions 273

in a functionally redundant fashion ((35); Figure 4A). Here, we test whether the 274

efficiency of silencing by the Kcnqot1 SD increases if it is flanked with a region of Xist 275

that shows high chromatin localizing activity. We constructed an episomal plasmid, 276

wherein a chromatin localizing region of Xist was inserted flanking the Kcnqot1 SD, and 277

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

the SV40 PolyA sequence (PS6xc4.4) was inserted at the end of Xist chromatin localizing 278

region such that the length of the encoded transcript was around 6.25 kb. Similarly, we 279

also constructed an episomal plasmid containing only Xist chromatin localizing region 280

and the SV40 polyA at the end, but lacking the Kcnqot1 SD (PS6A1xc4.4). In both 281

PS6xc4.4 and PS6A1xc4.4 episomes, Kcnq1ot1 does not overlap the overlapping H19 282

reporter gene. We compared the reporter gene activity in the PS6xc4.4 with a control 283

episomal plasmid, PS4∆H19polyA9.2 that encodes a similar length of Kcnq1ot1 as in 284

PS6xc4.4 (see (13) for details). As can be seen from Figure 4A-B & S5A, the efficiency 285

of bidirectional silencing by the Kcnq1ot1 silencing domain was significantly enhanced 286

when it is flanked with chromatin localizing region (compare PS6xc4.4 with 287

PS4∆H19polyA9.2 in Fig. 4B). However, we could not detect any silencing if the 288

chromatin localizing region was transcribed in the absence of the Kcnqot1 SD (compare 289

PS6A1xc4.4 with PS6xc4.4 in Figure 4B) or it was transcribed along with the Kcnqot1 290

SD positioned in an inverted orientation (data not shown), suggesting that the Kcnqot1 291

SD is a critical regulator of bidirectional silencing and its efficiency of silencing 292

significantly increases with the inclusion of the chromatin localizing regions of Xist. This 293

can be compared to the previously demonstrated effect of a 950 bp Xist silencing domain, 294

an A-rich repeat sequence located at the 5´end of Xist transcript, which displayed similar 295

functional features in our episome-based system i.e., it induces bidirectional silencing in 296

an orientation-dependent manner (Figure 4C & S5B). 297

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

The Kcnqot1 SD silences the flanking reporter genes by spreading repressive 298

epigenetic modifications 299

Previously we have documented that Kcnq1ot1 silences the flanking reporter genes 300

through spreading epigenetic modifications in cis (13). In this investigation, we addressed 301

whether the acquisition of epigenetic modifications by the flanking chromatin due to the 302

antisense RNA involves the functional role of the Kcnqot1 SD. To this end, we 303

investigated the chromatin structure on both the H19 and Hygromycin promoters by 304

analyzing the levels of histone H3 with trimethylated (H3K9me3) and acetylated 305

(H3K9ac) lysine 9, using ChIP assay on JEG-3 cells transfected with PS6 and PS6A1. 306

Strikingly, both the H19 and Hygromycin promoters were enriched in H3K9me3 when 307

fully silenced (PS6; Figure 5), but not when silencing is lost due to deletion of the 308

Kcnqot1 SD (PS6A1; Figure 5). Interestingly, we observed marked increase in the levels 309

of H3K9ac on both the H19 and Hygromycin promoters in PS6A1 (PS6A1; Figure 5), 310

suggesting that the Kcnqot1 SD silences the flanking genes by spreading epigenetic 311

modifications specific to inactive chromatin. 312

The Kcnq1ot1 mediates transcriptional silencing by targeting the episomes to 313

perinucleolar region 314

The above observations clearly indicate that the Kcnqot1 SD mediates transcriptional 315

silencing of flanking genes by regulating chromatin structure, probably through recruiting 316

heterochromatic machinery. Several studies have found a link between gene silencing and 317

recruitment to nuclear heterochromatin compartments (3, 4, 6, 7, 10). We were therefore 318

keen to examine whether Kcnqot1 SD-mediated transcriptional silencing correlates with 319

positioning of the episomal sequences in distinct nuclear compartments. We first wanted 320

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

to confirm that episome-encoded Kcnq1ot1 RNA can be detected in the vicinity of 321

episomal DNA sequences. To this end, we performed combined RNA/DNA fluorescent 322

in situ hybridization (FISH) with a Kcnq1ot1 RNA probe and an episomal DNA probe on 323

JEG-3 cells, transfected transiently with PS6A13 (with A2 motif) for four days. We 324

found that RNA signals colocalize with DNA signals indicating that episomal sequences 325

can be visualized by either RNA or DNA FISH (Figure S6A-B). We then evaluated the 326

nuclear localization of episomal sequences in JEG-3 cells, propagated with the PS6, 327

PS6A1, PS6A5, PS6A13 and PS6A14 (without A2 motif) episomes by RNA and DNA 328

FISH (Figure 6). We noted that four days after transfection with PS6A13, many 329

transfected cells displayed clusters of episomal signals either surrounding nucleoli or 330

enriched near the nuclear periphery (Figure 6A). Similarly, after 30 days under 331

hygromycin selection, in many cells carrying the PS6 construct, episomal signals were 332

located in close proximity to nucleoli or the nuclear periphery (Figure 6B). We scored the 333

number of cells (n=104) displaying different types of distribution of RNA FISH signals 334

in JEG-3 cells, transiently transfected with PS6A13. We found that in 34% of nuclei, the 335

Kcnq1ot1 signals were surrounding nucleoli (Figure 6A). Further 17% of the nuclei 336

showed RNA signals dispersed at the nuclear periphery. We also assessed the nuclear 337

positioning of the DNA FISH signals in JEG-3cells, stably propagated with PS6 and 338

PS6A1. PS6 episomes were located in the vicinity of nucleoli in 43% of the nuclei in the 339

cellular population (n= 101, Figure 6B), or localized at the nuclear periphery in other 340

29% of nuclei (Figure 6B). By contrast, in the vast majority of JEG-3 cells, stably 341

propagated with PS6A1, the episomal FISH signals did not occupy perinucleolar or 342

perinuclear positions (Figure 6C). In these cells, association with nucleoli and perinuclear 343

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

distribution was found only in 11% and 8% of all nuclei, respectively. These results show 344

that silencing of the episomal genes is indeed associated with positioning of the episomes 345

in close proximity to heterochromatin. Importantly, we did not detect specific enrichment 346

of PS6A5 (having the silencing domain in reverse orientation) episomes at the nucleolar 347

periphery (data not shown). This observation indicates that the silencing domain in the 348

Kcnq1ot1 transcript, rather than the DNA sequence, is crucial for nucleolar targeting. 349

We wanted to examine the association of Kcnq1ot1 RNA transcripts with the nucleolus in 350

more detail. We analyzed JEG-3 cells stably propagated with the PS6A13 and PS6A14 351

episomes. We performed RNA FISH, in combination with immunostaining for a 352

nucleolar marker, nucleophosmin. Consistent with our earlier findings, we demonstrated 353

that PS6A13-encoded Kcnq1ot1 RNA is in direct contact with nucleophosmin-stained 354

nucleoli (Figure 6D-E) and the percentage of cells that show nucleolar contact in 355

nucleophosmin stained cells is more than 40% and is similar to what has been obtained 356

with RNA-FISH. We next performed combined RNA/DNA immuno FISH on PS6A13 357

stably propagated JEG-3 cells, to check whether RNA and DNA signals that are in 358

contact with nucleoli colocalize with each other. In combined RNA/DNA immuno FISH, 359

following hybridization with a digoxigenin-labeled Kcnq1ot1 probe overnight, the RNA 360

FISH signals were visualized by immunodetection of digoxigenin with fluorescent 361

antibodies. After that, the cells are re-fixed and treated with RNase. The RNA FISH 362

signals survive the RNase treatment because the Kcnq1ot1 RNA is protected by the probe 363

and additional layers of antibodies. Chromatin is then denatured followed by a second 364

round of hybridization with a biotin-labeled episome probe. This probe hybridizes to 365

episomal DNA sequences and is subsequently visualized by fluorescent detection of 366

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

biotin. These experiments confirmed that nucleoli-associated DNA and RNA FISH 367

signals colocalize with each other (Figure 6F). In combination, our results show that the 368

Kcnqot1 SD plays a crucial role in targeting the episomes to the vicinity of nucleolus. 369

In a recent investigation, using BrdU labeling and PCNA immunostaining, it has been 370

documented that Xist–mediated inactive X-chromosome contact with the nucleolus 371

occurs predominantly during mid-to-late S phase (37). Using the same methodology, we 372

wanted to investigate whether localization of Kcnq1ot1 DNA FISH signals in the 373

nucleolar periphery correspond to any particular phase of the cell cycle. Interestingly, we 374

found that 41% of Kcnq1ot1 RNA-FISH signals were localized in the vicinity of 375

nucleolus around mid S phase of the cell cycle. Although we do see nucleolar localization 376

during early (8%) and late (14%) S phase but lower than that of mid S phase (41%), 377

implying that the silencing domain mediated nucleolar contact is dependent on the phase 378

of the cell cycle (Figure 7A-B). 379

The Kcnq1ot1 SD mediates the chromatin interaction of the Kcnq1ot1 antisense 380

transcript 381

Our combined RNA/DNA FISH indicated that the RNA and DNA signals were co-382

localized with each other, giving an indication that the transcribed Kcnq1ot1 RNA might 383

be in direct contact with chromatin. To address whether Kcnq1ot1 RNA is in direct 384

contact with chromatin, we have used chromatin-associated RNA immunoprecipitation 385

(ChRIP). In this technique, we immunopurified the chromatin using a histone H3 386

antibody and extracted RNA from the immunopurified chromatin, followed by DNase I 387

treatment and reverse transcription of the purified RNA. We found that chromatin 388

purified from cells transfected with the PS6xc4.4, showed 5-6 fold enrichment of 389

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Kcnq1ot1 in comparison to the chromatin purified from the cells transfected with the 390

PS6. This suggests that the Kcnq1ot1 encoded from the PS6 associates with chromatin, 391

albeit with lower affinity than the Kcnq1ot1 containing chromatin-localizing signals from 392

Xist (compare PS6 with PS6xc4.4 in Figure 7C). Surprisingly, we found that the 393

chromatin association of the Kcnq1ot1 encoded from PS6 was dependent on the Kcnqot1 394

SD, as the loss of this region from the transcript resulted in a significant decrease in the 395

level of its association with chromatin (compare PS6A1 with PS6 in Figure 7C). 396

Moreover, when we truncated Kcnq1ot1 to 1.7 kb length, we also found a significant 397

decrease of the Kcnq1ot1 RNA levels in the purified chromatin, suggesting that the 398

length of the Kcnq1ot1 determines its association with the chromatin (compare 399

PS4polyA1.7 with PS6 in Figure 7C). 400

Discussion 401

Several models have been proposed to explain the mode of action of long noncoding 402

RNAs. For example, in Drosophila it has been documented that expression of noncoding 403

transcripts from TRE elements has been linked to the recruitment of Trithorax protein 404

Ash1 to the downstream cis regulatory elements (28). Similarly, onset of Xist expression, 405

followed by its coating on one of the two X-chromosomes, during early female 406

embryonic development, coincides with the recruitment of transcriptional repressors and 407

inactivation of the X-chromosome. It has been shown that A repeat-rich silencing domain 408

at the 5’ end of Xist, and chromatin localizing domains, spread in a redundant fashion all 409

over the gene body of Xist, play a crucial role in the Xist-mediated X-chromosome 410

inactivation (35). On the other hand, mechanisms underlying the antisense noncoding 411

RNAs functions have not been investigated in greater detail. It is yet unclear as to 412

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

whether antisense RNA itself or its transcriptional process as such plays an important role 413

in the transcriptional silencing. However, recent studies on Tsix have provided some 414

important insights into antisense RNA-mediated transcriptional silencing. It has been 415

shown that the production of Tsix antisense RNA has been linked to Xist promoter 416

regulation on the future active X-chromosome. Using RNA immunoprecipitation and 417

chromatin immunoprecipitation experiments, it has been demonstrated that Tsix 418

associates with DNMT-3B and that the promoter region of Xist is also enriched with 419

DNMT3B, indicating that Tsix may be involved in Xist promoter regulation through 420

DNMT3B deposition (30). Interestingly, Tsix antisense RNA effects are restricted to its 421

overlapping sense counter part, Xist. Unlike Tsix, the other long functional antisense 422

RNAs, such as Kcnq1ot1 and Air, are implicated in the transcriptional silencing of not 423

only overlapping genes but also nonoverlapping genes. However, the mechanisms 424

underlying the mode of action of these RNAs have so far not been investigated in greater 425

detail. We discuss the possible implications of our findings in understanding the 426

mechanisms underlying the Kcnq1ot1-mediated long-range transcriptional silencing. 427

One of the most important observations of the present investigation is the identification 428

of a Kcnq1ot1 silencing domain (Kcnq1ot1 SD) at the 5´end of the Kcnq1ot1 antisense 429

transcript. This domain silences genes in an orientation-dependent and position-430

independent manner only when it is located downstream of the antisense promoter but 431

not when it is placed upstream of the antisense promoter. Lack of silencing by the 432

Kcnq1ot1SD upstream of the antisense promoter rules out the possibility that the cis 433

acting DNA sequences of the Kcnq1ot1SD, independently and/or in coordination with the 434

antisense promoter, may take part in silencing, and confirm that antisense transcription 435

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

through the Kcnq1ot1 SD is prerequisite to inducing long-range bidirectional silencing. 436

The orientation-dependent and position-independent silencing by the Kcnq1ot1 SD 437

downstream of the antisense promoter suggests that the primary sequence of the 438

transcribed RNA is crucial for the silencing process and that it has all the information 439

required for inducing the silencing process. Interestingly, a scan for sequences over 52 kb 440

of the more than 90 kb long Kcnq1ot1, revealed no sequences that show silencing activity 441

in a manner similar to the Kcnq1ot1 SD. Based on our observations, we rule out the 442

possible role of transcriptional interference in the Kcnq1ot1 mediated silencing pathway 443

in the episomal context, as we found significant activation of the Hygromycin gene 444

despite an increased expression of mutant Kcnq1ot1 in the PS6A1 when compared to the 445

PS6. In addition, the latter data also rules out the functional role of dsRNA-mediated 446

RNA interference in the bidirectional silencing pathway. To our knowledge, uncovering 447

of functional sequences in Kcnq1ot1 represents first characterization of functional 448

sequences in an antisense RNA. 449

Another important observation that we provide in these investigations is the Kcnq1ot1 450

SD-mediated localization of the episomal DNA sequences in the vicinity of the nucleolar 451

compartment. Perinuclear and perinucleolar compartments have long been suggested to 452

harbor protein factors involved in repressive chromatin formation. Interestingly, the 453

perinucleolar localization of the episomes by the silencing domain occurs as early as 4-5 454

days after transient transfection and their co-localization with the nucleolus increased in 455

the cell lines containing stably propagated episomes, suggesting that the perinucleolar 456

localization of the episomes may be crucial for both establishment and maintenance of 457

Kcnq1ot1 silencing domain-mediated transcriptional silencing. The observations that I) 458

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

RNA and DNA signals colocalize in the perinucleolar region, II) the silencing domain 459

mediates transcriptional silencing through spreading repressive chromatin structures by 460

promoting chromatin interaction, and III) lack of perinucleolar contacts of the episomes 461

as well as loss of repressive chromatin structures over neighboring chromatin regions of 462

the episomes, when Kcnq1ot1 was transcribed in the absence of the Kcnq1ot1 SD, 463

suggest that the Kcnq1ot1 SD establishes the transcriptional silencing through recruiting 464

ribonucleoprotein (RNP) complexes and promoting chromatin interaction, and it 465

maintains the repressive chromatin further through successive cell divisions by 466

translocating the linked episomal sequences to the nucleolar periphery. This specific 467

localization of the episomes by the Kcnq1ot1 SD occurs during mid S phase, indicating 468

that the Kcnq1ot1 SD contact with the nucleolus is regulated in a cell cycle-dependent 469

manner. 470

Interestingly, however, the functional features of the Kcnq1ot1 SD resembles closely that 471

of Xist silencing domain (Xist SD), which was shown to mediate transcriptional silencing 472

on the inactive X-chromosome in an orientation and position-independent manner in an 473

ex vivo ES cell model system (35). Moreover, the Xist SD also displays similar silencing 474

features in the episome-based system, documenting the specificity and reliability of using 475

an episomal-based system in addressing the intricate molecular pathways underlying the 476

function of long antisense noncoding RNAs. Most importantly, perinucleolar targeting of 477

episomal sequences by the Kcnq1ot1 SD is reminiscent of Xist-mediated perinucleolar 478

localization of the inactive X-chromosome. It has been shown that the Xic transgenes 479

maintains the transcriptional silencing of flanking autosomal genes by targeting them to 480

the nucleolar periphery (37). It is yet not clear whether the Xist SD has any functional 481

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

role in perinucleolar targeting of the inactive X-chromosome. However, it has been 482

shown that the Xist SD is required for relocating the linked genes into Xist RNA silent 483

domain (5), indicating that the Xist SD may also have a functional role in perinucleolar 484

targeting of X-linked genes. Both Xist and Kcnq1ot1 silencing domains hardly show any 485

homology at the primary sequence level, indicating similarity at the functional level 486

and/or target-specific interactions. 487

Yet another interesting observation of the current investigation is the characterization of 488

the functional role of conserved motifs in the Kcnq1ot1-mediated silencing process. 489

Mutation or selective exclusion of the previously characterized highly conserved motif 490

A2 results in a relaxation of the silencing of flanking genes, indicating that the A2 motif 491

plays a critical role in the Kcnq1ot1-mediated silencing. More importantly, sequences 492

encompassing the 12 bp A2 motif also play a crucial role in targeting the linked episomes 493

to the nucleolar periphery. M-fold predictions on a sequence containing the wild type A2 494

motif revealed a putative stem loop structure, with the stem mainly formed by the A2 495

motif. However, we did not observe proper stem loop structure with the sequence 496

containing the mutant A2 motif, indicating that secondary structure attained by the A2 497

motif may play a crucial role in the silencing process (data not shown). Although in 498

silico-based m-fold predictions emphasize the functional role of RNA secondary 499

structures in the silencing process, we cannot exclude the possibility that the crucial cis 500

acting elements within the A2 motif may be involved in the formation of repressive 501

chromatin through recruiting transcription factors. We could not, however, detect any 502

sequence motif that shows homology with the A2 motif in the 90 kb long Kcnq1ot1 503

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

transcript, suggesting that the A2 motif itself may be playing a crucial role in the 504

Kcnq1ot1-mediated transcriptional silencing. 505

It is interesting to note that when we flanked the Kcnq1ot1 SD with a chromatin-506

localizing region of Xist, we not only observed a significant increase in the silencing 507

activity, but also increased chromatin interaction of Kcnq1ot1. These observations will 508

have implications in understanding why Xist transcript mediated transcriptional silencing 509

on the inactive X-chromosome is spread over most of the chromosome, in comparison to 510

Kcnq1ot1-mediated silencing, whose effects are restricted only to a 900 kb region. We 511

presume that the limited silencing by Kcnq1ot1 could be due to the lack of high affinity 512

chromatin localizing regions, such as those seen in Xist. Although the Kcnq1ot1 SD 513

mediates chromatin localization, it seems that the affinity is insufficient for chromosome-514

wide silencing. The Xist SD, on the other hand, mediates chromosome-wide 515

transcriptional silencing with the help of chromatin localizing regions which are 516

distributed throughout the transcript in a functionally redundant fashion (35). 517

From the above observations, it appears that the transcriptional silencing that we see in 518

the PS6 episome is mediated by the antisense RNA per se and that in part it mimics the 519

mode of action of Xist. In this model, the Kcnq1ot1 antisense RNA is expressed and coats 520

the chromatin of flanking sequences in cis, followed by the recruitment of 521

heterochromatinization machinery, which in turn targets the linked chromosomal regions 522

to perinucleolar vicinity in order to maintain the repressive chromatin through successive 523

cell divisions (Figure 8). Although, we have shown that repressive chromatin marks over 524

the flanking sequences correlate with its chromatin localizing activity, the ChRIP 525

methodology can not resolve whether the RNA localizes the entire episome in cis or if it 526

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

remain attached at the site of antisense transcription through the Kcnq1ot1 SD. 527

Interestingly, a recent report using high-resolution fiber RNA-FISH and RNA-TRAP 528

demonstrated that the human KCNQOT1/LIT1 transcript coats the neighboring regions of 529

chromatin containing the SLC22A18/IMPT1 and CDKN1C/p57KIP2 genes (20). Taken 530

together, these results show that Kcnq1ot1 mediated transcriptional silencing mimics in 531

part the Xist-mediated X-chromosome inactivation. 532

Taken together, here we provide evidence that demonstrate a critical role for the antisense 533

RNA itself. In sum, mechanistic insights provided in this study represent a first step in 534

understanding the multilayered silencing pathway mediated by the long noncoding 535

antisense RNAs. 536

Experimental Procedures 537

Cell culture, transfection and transcriptional inhibitors: The human placenta derived 538

JEG-3 cells and the human liver-derived Hep-3B cells were maintained in MEM 539

(Invitrogen) as previously described (13). The transfection of plasmid DNA into the JEG-540

3 and Hep-3B cells was performed using a standard CaCl2 method. 541

Please see supplementary information for other experimental procedures. 542

Acknowledgements 543

544

We gratefully acknowledge Prof Wolf Reik for providing the Kcnq1ot1 RNA probe for 545

our studies. We greatly appreciate Dr Joanne Whitehead for critical review of the 546

manuscript. We thank Prof Rolf Ohlsson, Dept of Development and Genetics, Uppsala 547

University, Sweden; Johan Ericsson, Ludwig Institute for Cancer Research, Uppsala, 548

Sweden, and Thierry Grange, Institut Jacques Monod, Paris for help and suggestions. 549

This work was supported by the grants from Swedish Research Council (VR-NT) and the 550

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Swedish Cancer Research foundation (Cancerfonden) to CK. CK is a Senior Research 551

Fellow supported by Swedish Medical Research Council (VR-M). 552

References 553

1. Bell, A. C., and G. Felsenfeld. 2000. Methylation of a CTCF-dependent 554

boundary controls imprinted expression of the Igf2 gene. Nature 405: 482-5. 555

2. Bernstein, E., and C. D. Allis. 2005. RNA meets chromatin. Genes Dev 556

19:1635-55. 557

3. Brown, K. E., S. Amoils, J. M. Horn, V. J. Buckle, D. R. Higgs, M. 558

Merkenschlager, and A. G. Fisher. 2001. Expression of alpha- and beta-globin 559

genes occurs within different nuclear domains in haemopoietic cells. Nat Cell 560

Biol 3:602-6. 561

4. Brown, K. E., J. Baxter, D. Graf, M. Merkenschlager, and A. G. Fisher. 1999. 562

Dynamic repositioning of genes in the nucleus of lymphocytes preparing for cell 563

division. Mol Cell 3:207-17. 564

5. Chaumeil, J., P. Le Baccon, A. Wutz, and E. Heard. 2006. A novel role for 565

Xist RNA in the formation of a repressive nuclear compartment into which genes 566

are recruited when silenced. Genes Dev 20:2223-37. 567

6. Csink, A. K., and S. Henikoff. 1996. Genetic modification of heterochromatic 568

association and nuclear organization in Drosophila. Nature 381:529-31. 569

7. Dernburg, A. F., J. W. Sedat, and R. S. Hawley. 1996. Direct evidence of a role 570

for heterochromatin in meiotic chromosome segregation. Cell 86:135-46. 571

8. Fitzpatrick GV, S. P., Higgins MJ. 2002. Reginal loss of imprinting and growth 572

deficiency in mice with targeted deletion of KvDMR1. Nat Genet 32:426-431. 573

9. Hark, A. T., C. J. Schoenherr, D. J. Katz, R. S. Ingram, J. M. Levorse, and S. 574

M. Tilghman. 2000. CTCF mediates methylation-sensitive enhancer-blocking 575

activity at the H19/Igf2 locus. Nature 405:486-9. 576

10. Harmon, B., and J. Sedat. 2005. Cell-by-cell dissection of gene expression and 577

chromosomal interactions reveals consequences of nuclear reorganization. PLoS 578

Biol 3:e67. 579

11. Heard, E., P. Clerc, and P. Avner. 1997. X-chromosome inactivation in 580

mammals. Annu Rev Genet 31:571-610. 581

12. Kanduri, C., V. Pant, D. Loukinov, E. Pugacheva, C.-F. Qi, A. Wolffe, R. 582

Ohlsson, and A. Lobanenkov. 2000. Functional interaction of CTCF with the 583

insulator upstream of the H19 gene is parent of origin-specific and methylation-584

sensitive. Curr Biol 10:853-856. 585

13. Kanduri, C., N. Thakur, and R. R. Pandey. 2006. The length of the transcript 586

encoded from the Kcnq1ot1 antisense promoter determines the degree of 587

silencing. Embo J 25:2096-106. 588

14. Katayama, S., Y. Tomaru, T. Kasukawa, K. Waki, M. Nakanishi, M. 589

Nakamura, H. Nishida, C. C. Yap, M. Suzuki, J. Kawai, H. Suzuki, P. 590

Carninci, Y. Hayashizaki, C. Wells, M. Frith, T. Ravasi, K. C. Pang, J. 591

Hallinan, J. Mattick, D. A. Hume, L. Lipovich, S. Batalov, P. G. Engstrom, 592

Y. Mizuno, M. A. Faghihi, A. Sandelin, A. M. Chalk, S. Mottagui-Tabar, Z. 593

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Liang, B. Lenhard, and C. Wahlestedt. 2005. Antisense transcription in the 594

mammalian transcriptome. Science 309:1564-6. 595

15. Lee, J. 2003. Molecular Links between X-Inactivation and Autosomal Imprinting: 596

X-Inactivation as a Driving Force for the Evolution of Imprinting? Curr. Biol. 597

13:R242-54. 598

16. Lewis, A., K. Mitsuya, D. Umlauf, P. Smith, W. Dean, J. Walter, M. Higgins, 599

R. Feil, and W. Reik. 2004. Imprinting on distal chromosome 7 in the placenta 600

involves repressive histone methylation independent of DNA methylation. Nat 601

Genet 36:1291-5. 602

17. Mancini-DiNardo, D., S. J. Steele, R. S. Ingram, and S. M. Tilghman. 2003. A 603

differentially methylated region within the gene Kcnq1 functions as an imprinted 604

promoter and silencer. Hum Mol Genet 12:283-94. 605

18. Mancini-Dinardo, D., S. J. Steele, J. M. Levorse, R. S. Ingram, and S. M. 606

Tilghman. 2006. Elongation of the Kcnq1ot1 transcript is required for genomic 607

imprinting of neighboring genes. Genes Dev 20:1268-82. 608

19. Mattick, J. S., and I. V. Makunin. 2006. Non-coding RNA. Hum Mol Genet 15 609

Spec No 1:R17-29. 610

20. Murakami, K., M. Oshimura, and H. Kugoh. 2007. Suggestive evidence for 611

chromosomal localization of non-coding RNA from imprinted LIT1. J Hum 612

Genet. 613

21. Navarro, P., S. Pichard, C. Ciaudo, P. Avner, and C. Rougeulle. 2005. Tsix 614

transcription across the Xist gene alters chromatin conformation without affecting 615

Xist transcription: implications for X-chromosome inactivation. Genes Dev 616

19:1474-84. 617

22. O'Neill, M. J. 2005. The influence of non-coding RNAs on allele-specific gene 618

expression in mammals. Hum Mol Genet 14 Spec No 1:R113-20. 619

23. Pandey, R. R., M. Ceribelli, P. B. Singh, J. Ericsson, R. Mantovani, and C. 620

Kanduri. 2004. NF-Y regulates the antisense promoter, bidirectional silencing, 621

and differential epigenetic marks of the Kcnq1 imprinting control region. J Biol 622

Chem 279:52685-93. . 623

24. Paulsen, M., T. Khare, C. Burgard, S. Tierling, and J. Walter. 2005. 624

Evolution of the Beckwith-Wiedemann syndrome region in vertebrates. Genome 625

Res 15:146-53. 626

25. Prasanth, K. V., and D. L. Spector. 2007. Eukaryotic regulatory RNAs: an 627

answer to the 'genome complexity' conundrum. Genes Dev 21:11-42. 628

26. Rougeulle, C., and E. Heard. 2002. Antisense RNA in imprinting: spreading 629

silence through Air. Trends Genet 18:434-7. 630

27. Sado, T., Y. Hoki, and H. Sasaki. 2005. Tsix silences Xist through modification 631

of chromatin structure. Dev Cell 9:159-65. 632

28. Sanchez-Elsner, T., D. Gou, E. Kremmer, and F. Sauer. 2006. Noncoding 633

RNAs of trithorax response elements recruit Drosophila Ash1 to Ultrabithorax. 634

Science 311:1118-23. 635

29. Sleutels, F., R. Zwart, and D. P. Barlow. 2002. The non-coding Air RNA is 636

required for silencing autosomal imprinted genes. Nature 415:810-3. 637

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

30. Sun, B. K., A. M. Deaton, and J. T. Lee. 2006. A transient heterochromatic state 638

in Xist preempts X inactivation choice without RNA stabilization. Mol Cell 639

21:617-28. 640

31. Szabó, P., S.-H. Tang, A. Rentsendorj, G. Pfeifer, and J. Mann. 2000. 641

Maternal-specific footprints at putative CTCF sites in the H19 imprinting control 642

region give evidence for insulator function. Curr Biol 10:607-610. 643

32. Thakur, N., M. Kanduri, C. Holmgren, R. Mukhopadhyay, and C. Kanduri. 644

2003. Bidirectional silencing and DNA methylation sensitive methylation 645

spreading properties of the KcnqI ICR map to the same regions within the KcnqI 646

imprinting control region. J. Biol. Chem. 273:9514-9. 647

33. Thakur, N., V. K. Tiwari, H. Thomassin, R. R. Pandey, M. Kanduri, A. 648

Gondor, T. Grange, R. Ohlsson, and C. Kanduri. 2004. An antisense RNA 649

regulates the bidirectional silencing property of the kcnq1 imprinting control 650

region. Mol Cell Biol 24:7855-62. 651

34. Umlauf, D., Y. Goto, R. Cao, F. Cerqueira, A. Wagschal, Y. Zhang, and R. 652

Feil. 2004. Imprinting along the Kcnq1 domain on mouse chromosome 7 involves 653

repressive histone methylation and recruitment of Polycomb group complexes. 654

Nat Genet 36:1296-300. 655

35. Wutz, A., T. P. Rasmussen, and R. Jaenisch. 2002. Chromosomal silencing and 656

localization are mediated by different domains of Xist RNA. Nat Genet 30:167-657

74. 658

36. Yang, P. K., and M. I. Kuroda. 2007. Noncoding RNAs and intranuclear 659

positioning in monoallelic gene expression. Cell 128:777-86. 660

37. Zhang, L. F., K. D. Huynh, and J. T. Lee. 2007. Perinucleolar targeting of the 661

inactive X during S phase: evidence for a role in the maintenance of silencing. 662

Cell 129:693-706. 663

664

Figure Legends 665

Figure 1. Identification of a silencing domain at the 5´end of the Kcnq1ot1 antisense 666

RNA. A) shows schematic diagrams of the PH19 and the 3.6 kb Kcnq1 ICR deletion 667

fragments. The PH19 is the parent episome construct, which contains H19 (box with long 668

horizontal stripes) and Hygromycin (box with diagonal stripes) as reporter genes under 669

the regulation of SV40 enhancer (filled circle). PS742/1050/1947 are derivatives of the 670

PH19 episome and contain various portions of 3.6 kb Kcnq1 ICR covering the 1.7 kb 671

Kcnq1ot1 sequence (box with vertical stripes). Arrow represent transcription start site. 672

The SV40 PolyA sequence, depicted as round-headed arrow, inserted 9.2 kb downstream 673

of the transcription start site. (B) The Hygromycin gene activity in PS742, PS1050 and 674

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

PS1947 episome constructs was analyzed by counting the hygromycin resistant cell 675

colonies and is represented as percent expression in relation to the parent construct PH19, 676

The data represents means + standard deviation (SD) of a minimum of six independent 677

experiments. C) Shows the schematic drawings of PGL-Basic and its derivatives, PGL-678

Basic742/1050 /1947. A bar graph showing the relative Luciferase activity of PGL-679

Basic742/1050 /1947 plasmids in JEG-3 cells is shown below. The strength of the 680

promoter activity of the constructs is presented relative to the PGL-BasicCycB2. The data 681

represent means + SD of three independent experiments each performed in duplicate. 682

Figure 2. The Kcnq1ot1 SD at the 5´end of the Kcnq1ot1 antisense RNA silences 683

flanking genes in an orientation and transcription-dependent manner. (A-C) Bar graphs 684

show the activity of Hygromycin and H19 genes, analyzed for each construct. The maps 685

alongside the bar graphs depict the wild-type and mutant forms of the 5.8 kb Kcnq1 ICR 686

which were inserted into the PH19 episome between the H19 and Hygromycin genes. 687

The activity of reporter genes in each construct is shown in percent expression levels and 688

their activities are presented relative to control vectors (PH19 for H19 and Hygromycin 689

genes). The expression levels of H19 were quantified by RPA eight days after transient 690

transfection of wild type and mutant constructs into the JEG-3 cell line. The levels were 691

normalized against total input RNA (using GAPDH control) and episome copy numbers 692

(by Southern hybridization of genomic DNA from the transfected cells with an episome-693

specific probe and an internal β-actin control probe). Hygromycin gene activity was 694

analyzed by counting the hygromycin resistant cell colonies after selection with 695

hygromycin. RPA for H19: mean + SD of three independent experiments; hygromycin 696

resistant colony counts: mean + SD of a minimum of six independent experiments. (D) 697

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

shows that episome encoded Kcnq1ot1 is specifically localized in the nuclear 698

compartment and that selective deletion of the silencing domain from Kcnq1ot1 or 699

truncation of Kcnq1ot1 has no effect on its nuclear localization. E & F) show that the 700

half-life of the episome coded Kcnq1ot1 transcript is 3.2 hours (E) and that selective 701

deletion of silencing domain has no significant effect on the stability of Kcnq1ot1 (F). 702

Figure 3. The search for a Kcnq1ot1 fragment with bidirectional silencing features over 703

52 kb portion of the Kcnq1ot1 transcription unit revealed none of the fragments that have 704

similarities with the Kcnq1ot1 SD. 3A) physical maps depicting the Kcnq1/Kcnq1ot1 705

locus, overlapping fragments of 3-5 kb of Kcnq1ot1 transcription unit used in episome-706

silencing assays, and a diagrammatic presentation of the cloning of the Kcnq1ot1 gene 707

fragments into the PS6A1 episome. 3B) bar graph shows the percentage activity of the 708

Hygromycin gene in PS6A1Kcnq1ot1Frg1-15, analyzed by counting the hygromycin 709

resistant cell colonies. The data represents means + SD of a minimum of three 710

independent experiments. 711

Figure 4. High affinity chromatin attachment regions of the Xist transcript increase the 712

efficiency of bidirectional silencing. (A) A schematic representation of the Xist RNA. 713

The silencing domain containing the A-rich repeats is located at the 5´end of the 714

transcript. The boxes below the map indicate the chromatin localizing regions that have 715

previously been mapped in (35). Figure B-C) schematic maps depicted alongside the bar 716

graphs are the wild type and mutant Kcnq1 ICR fragments (the Kcnq1ot1 SD is either 717

flanked with a high affinity Xist chromatin localizing region or replaced with the Xist SD) 718

were inserted into the PS6 or PS6A1 episomes. The schematic map of the PS4∆H19 719

polyA9.2 construct alongside the bar graph is shown along with the reporter genes. The 720

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

PS6Xist SD+/- represent that the Xist SD was inserted in both positive (+) and negative 721

orientations (-) in PS6A1. Bar graphs show the percentage activity of Hygromycin and 722

H19 genes in the episomes containing wild type and mutant Kcnq1 ICR fragments, 723

calculated as described in Figure 2. The data represents means + SD of three independent 724

experiments 725

Figure 5. Kcnq1ot1 SD mediates transcriptional silencing through regulating the 726

chromatin structure. Quantitative analysis of H3K9Me3 and H3K9Ac profiles using ChIP 727

over the H19 and Hygromycin promoters on crosslinked chromatin obtained from JEG-3 728

cells transfected with PS6, PH19 and PS6A1 episomes. Bar graphs shows percentage 729

enrichment and the data were quantified as described in Experimental Procedures. The 730

data represents mean + SD of 3 independent ChIP experiments. 731

Figure 6. The Kcnq1ot1 SD localizes the linked episomal sequences to the perinucleolar 732

and perinuclear regions. (A) RNA FISH on JEG-3 cells transiently transfected with 733

PS6A13 showing perinucleolar localization of episomes. Kcnq1ot1 probe is detected in 734

green, DAPI staining is in blue. Nucleoli are visible as darker areas in the DAPI staining 735

pattern. Scalebar, 5 µm. (B) and (C) DNA FISH on JEG-3 cells propagated with PS6 (B) 736

or PS6A1 (C) for 30 days under hygromycin selection. Episomal DNA FISH signals are 737

detected in green, DAPI staining is in blue. Scale is as in (A). Representative examples of 738

cells with PS6 signals in close proximity to a nucleolus (B, cell on the left), the nuclear 739

periphery (B, cell on the right), and PS6A1 signals positioned away from these nuclear 740

compartments (C). (D) and (E) RNA immuno-FISH on JEG-3 cells stably propagated 741

with PS6A13 (D) or PS6A14 (E). Green represents episomal-encoded RNA, red 742

corresponds to nucleolar marker nucleophosmin. The outlines of the cells are shown as 743

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

dotted lines. Scale is as in (A). (F) RNA/DNA immuno-FISH on JEG-3 cells stably 744

propagated with PS6A13. Episomal-encoded RNA is in green, DNA is in red, 745

nucleophosmin staining is in blue. The outlines of the cells are shown as dotted lines. 746

Scalebar, 5 µm. 747

Figure 7. Perinucleolar localization of Kcnq1ot1 occurs during mid-S phase. A) 748

Simultaneous detection of episomal DNA by DNA FISH (green) and replication foci by 749

immunostaining for PCNA (red). B) Bar graph showing the percentage nucleolar 750

localization of PS6A13 episomes in early, mid and late S-phase of the cell cycle. The 751

data represent means + SD of three independent experiments, (n=>100). C) Kcnq1ot1 752

interacts with chromatin. Real time quantification of reverse transcribed immuno-purified 753

chromatin-associated RNA suggests that Kcnq1ot1 interacts with chromatin and this 754

interaction is mediated by the Kcnq1ot1 SD. 755

Figure 8. The transcriptional silencing by Kcnq1ot1 requires the linked sequences at the 756

perinucleolar region. Production of Kcnq1ot1 on the paternal chromosome results in the 757

recruitment of heterochromatic machinery by the Kcnq1ot1SD. Subsequently, the 758

heterochromatic machinery associated with the antisense RNA (shown in green 759

squiggles) mediates the chromatin interaction in a manner analogous to the Xist 760

transcript, resulting in the spreading of heterochromatic histone modifications (red blobs) 761

in a Kcnq1ot1-dependent manner in cis. Then the Kcnq1ot1 transcript and associated 762

heterochromatic machinery guides the linked chromosomal domain to the perinucleolar 763

region during mid S phase in order to maintain bidirectional transcriptional silencing of 764

flanking genes (red and green arrow heads indicate silenced and active status of a gene, 765

respectively) through successive cell divisions. 766

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

PH19

Orip

(origin of replication)H19 Hygrom SV40 enhancer

20 40 60 80 100 120 140 160 180 200

PGL-Basic

PGL-Basic742

PGL-Basic1050

PGL-Basic1947

Relative Luciferase activity (fold activation)

20 40 60 80 100

Relative no. of Hygromycin-resistant

colonies (%)

PH19

PS742

PS1050

PS1947

3.6 kb Kcnq1 ICR

1947

1050

742

Figure 1

A

B

PGL-BasicCycB2

PS4polyA9.2

PolyA

(1.7 kb Kcnq1ot1 sequence; 1897-3625)

PGL-Basic

Kpn1 BglII

PGL-Basic1947/

PGL-Basic1050/

PGL-Basic742

Luciferase

1947

1050

742

++

PS1947/

PS1050/PS742

Luciferase Kcnq1 ICR

deletion Fragments

H19 Hygrom

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

20 40 60 80 100

20 40 60 80 100

20 40 60 80 100

Relative no. Hygromycin-resistant

colonies (%)

20 40 60 80 100

PS6A1

PS6A2

PS6A3

PS6A4

PH19

PS6A0

PS6

Percentage activity of H19

PS6

PH19


PS6A6

PS6A7

PS6A8

PS6A9 PS6A10

PS6A11

PS6A12

Orip H19 Hygrom

Kcnq1 ICR

SV40 enhancer

PS6A5

PolyA

PS6

PS6A15

PS6A16

CTTGGATCCGAGTTGGGTGTGTGTATGTGTGGGTCTACAGTGGTTCCGAGTCTAGAGT T TT TTTT CT (A1 motif )

A1mt

(A2 motif )

A2mt

20 40 60 80 100


PS6

PH19


colonies (%)


colonies (%)

Figure 2

A)

C)

B)

Kcnq1ot1 (11.4 kb)

PS6A13

PS6A14

20 40 60 80100120

PS4polyA4.9A1

Nuclear-

specific

human

rRNA

GAPDH

Kcnq1ot1

PS

4poly

A1.7

PS

6

-RN

A

PS

4poly

A1

.7

PS

6

Nuclear Fraction Cytoplasmic Fraction

PS

6A

1

PS

6A

1

PS

4poly

A1

.7

PS

6

PS

4poly

A1.7

PS

6

Nuclear Fraction Cytoplasmic Fraction

PS

6A

1

PS

6A

1

D)

-20

0

20

40

60

80

100

120

0 2 4 6 8 10

ActD treatments (in hours)

Fra

ctio

n of

RN

A r

emai

ned

Kcnq1ot1

0

20

40

60

80

100

120

0 2 4 6 8 10

ActD treatment (in hours)

Fra

ctio

n R

NA

rem

aine

d

Kcnq1ot11

Kcnq1ot1 stability in PS6A1

Kcnq1ot1 stability in PS6E)

F)

PS4polyA4.9

3402 2514

1897

3402

3402

3402 2943

2806

2650

1897 5.8 kb Kcnq1 ICR (PS6)(1.7 kb Kcnq1ot1 sequence; 1897-3625)

(1.1 kb Kcnq1ot1 sequence; 2479-3625)

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

20 40 60 80 100 120 140

Percentage activity of Hygromycin

PS

6A

1

Kcnq1ot1

frgs (1

-15)

PS

6A

1K

cnq1ot1

Frg

s (1

-15)

Kcnq1ot1

Kcnq1ot1

frgs (1

-15)

PS6A1Kcnq1ot1Frg2

PS6A1Kcnq1ot1Frg1

PS6A1Kcnq1ot1Frg3

PS6A1Kcnq1ot1Frg4

PS6A1Kcnq1ot1Frg5

PS6A1Kcnq1ot1Frg6

PS6A1Kcnq1ot1Frg7

PS6A1Kcnq1ot1Frg8

PS6A1Kcnq1ot1Frg9

PS6A1Kcnq1ot1Frg10

PS6A1Kcnq1ot1Frg11

PS6A1Kcnq1ot1Frg12

PS6A1Kcnq1ot1Frg13

PH19

PS6

PS6A1

52 k

bK

cnq1

AB

PS6A1Kcnq1ot1Frg14

PS6A1Kcnq1ot1Frg15

90

kb

Fig

ure

3

ACCEPTED on February 13, 2018 by guest

http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

20 40 60 80 100 20 40 60 80 100


colonies (%)

PS6A1

PH19

PS6

Percentage activity of H19Kcnq1 ICR

PolyA

PS6

20 40 60 80 100 20 40 60 80 100


colonies (%)

PH19

PS6

Percentage activity of H19Kcnq1 ICR

PolyA

PS6

PS6A5

PS6A1xc4.4

PS6xc4.4

PS6A1Xist SD+

PS6A1Xist SD-

A -repeat rich

silencing domainChromatin localizing region

Xist

xc4.4

Figure 4

A)

B)

C)

H19 HygromOrip

Kcnq1ot1(11.4)

Kcnq1ot1(11.4)

= Chromatin localizing region (Xist)

= Xist silencing domain(Xist SD)

= SV40 PolyA

= Kcnq1ot1 silencing domain

PS4 H19polyA9.2

Kcnq1ot1 (6.2 kb)

Kcnq1ot1 (6.25 kb)

5’ 3’

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Figure 5

0

5

10

15

20

25

30

35

PS

6P

H1

9

PS

6P

H1

9

PS

6P

H1

9

PS

6A

1

PS

6P

H19

PS

6P

H1

9

PS

6P

H1

9

H3K9Me3 H3K9Ac Serum H3K9Me3 H3K9Ac Serum

H19 promoter Hyg Promoter

PS

6A

1

PS

6A

1

PS

6A

1

PS

6A

1

PS

6A

1

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Figure 6

B C

D E

F

A

ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Figure 7 20 40 60 80 1

00

Early S

ph

ase

Mid

S p

ha

se

Late

S p

ha

se

A)

B)

20 40 60 80 100

Percentage enrichment

RN

aseA

PH19

PS4PA1.7

PS6

PS6A1

PS6xc4.4

(Kcnq1ot1)C)ACCEPTED


http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Paternal

locus

Maternal locus

Kcnq1ot1 gene

Kcnq1ot1 gene

Nucleolus

Kcnq1ot1 RNA

Heterochromatin complex

Repressive histone

modifications

Active gene

Repressed gene

Figure 8

CpG methylation

ACCEPTED on February 13, 2018 by guest

http://mcb.asm

.org/D

ownloaded from

http://mcb.asm.org/

Kcnq1ot1/Lit1 noncoding RNA mediates transcriptional silencing by ...

Documents

Transcript of Kcnq1ot1/Lit1 noncoding RNA mediates transcriptional silencing by ...