XIST-induced silencing of flanking genes is achieved by additive action of repeat a monomers in human somatic cells
© Minks et al.; licensee BioMed Central Ltd. 2013
Received: 6 March 2013
Accepted: 18 June 2013
Published: 1 August 2013
The establishment of facultative heterochromatin by X-chromosome inactivation requires the long non-coding RNA XIST/Xist. However, the molecular mechanism by which the RNA achieves chromosome-wide gene silencing remains unknown. Mouse Xist has been shown to have redundant domains for cis-localization, and requires a series of well-conserved tandem ‘A’ repeats for silencing. We previously described a human inducible XIST transgene that is capable of cis-localization and suppressing a downstream reporter gene in somatic cells, and have now leveraged these cells to dissect the sequences critical for XIST-dependent gene silencing in humans.
We demonstrated that expression of the inducible full-length XIST cDNA was able to suppress expression of two nearby reporter genes as well as endogenous genes up to 3 MB from the integration site. An inducible construct containing the repeat A region of XIST alone could silence the flanking reporter genes but not the more distal endogenous genes. Reporter gene silencing could also be accomplished by a synthetic construct consisting of nine copies of a consensus repeat A sequence, consistent with previous studies in mice. Progressively shorter constructs showed a linear relationship between the repeat number and the silencing capacity of the RNA. Constructs containing only two repeat A units were still able to partially silence the reporter genes and could thus be used for site-directed mutagenesis to demonstrate that sequences within the two palindromic cores of the repeat are essential for silencing, and that it is likely the first palindrome sequence folds to form a hairpin, consistent with compensatory mutations observed in eutherian sequences.
Silencing of adjacent reporter genes can be effected by as little as 94 bp of XIST, including two ‘monomers’ of the A repeat. This region includes a pair of essential palindromic sequences that are evolutionarily well-conserved and the first of these is likely to form an intra-repeat hairpin structure. Additional sequences are required for the spread of silencing to endogenous genes on the chromosome.
KeywordsX-chromosome inactivation XIST Long non-coding RNA Eutherian dosage compensation Gene silencing
To ensure dosage compensation of X-linked genes between males and females, eutherian females silence one X chromosome . The minimal region required for X-chromosome inactivation contains the non-coding (nc) RNA gene XIST, which is expressed solely from the inactive X chromosome . Experiments in mice have shown that Xist is both required and sufficient for inactivation; however, the mechanism by which the XIST/Xist RNA causes chromosome-wide gene silencing remains elusive (reviewed in ). XIST localizes in cis to the chromatin of the inactive X chromosome , suggesting a potential role in targeting silencing complexes to the chromosome. The alternatively spliced and polyadenylated RNA is over 15 kb long in all eutheria examined. Overall the gene is only weakly conserved among mammals, but its regions of repetitive sequences called repeat A to F show better conservation [4, 5]. Additionally, exon 4 of XIST/Xist is well-conserved, and shows homology with the protein-coding Lnx3 gene, from which the Xist gene may have evolved by the addition of sequences from transposable elements [6, 7]. Intriguingly, in marsupials, Lnx3 remains protein-coding and Rsx3 encodes an RNA that is similar to XIST in that the long non-coding, repeat-rich RNA is transcribed from and associates with the inactive X chromosome . While there is no sequence conservation between Rsx3 and XIST, both are able to silence in cis, and show regions of putative stem-loop structure, supporting the idea that these long ncRNAs may be serving as adapter molecules containing different protein-recognition motifs to recruit components of the gene-silencing machinery to the inactive X chromosome.
As X-chromosome inactivation is a developmental process, most studies of Xist function have been undertaken in mice, where embryonic stem (ES) cells or embryos can be analyzed during the inactivation process. Human ES cells have demonstrated considerable epigenetic instability (for example, ) and studies of human embryos are necessarily restricted ([10, 11]). However, the potential differences in the inactivation process between mice and humans, suggested by both differences in the regulation of the XIST gene and the number of genes escaping inactivation (reviewed in ), led us to develop an inducible model to study human XIST action . Induced expression of XIST in the immortal HT1080 fibrosarcoma cell line is able to induce some features of an inactive X, including XIST localization, silencing of a co-integrated reporter gene, depletion of repetitive (CoT1) RNA, and the acquisition of some heterochromatic histone modifications associated with the inactive X.
Previous studies in mice targeted a panel of truncated inducible Xist transgenes to the single X chromosome in a male ES cell and demonstrated that redundant sequences were involved in localization of the mouse Xist RNA to the chromosome , with a construct containing only approximately 3 kb of Xist cDNA, including the well-conserved A repeat region, able to localize to and repress the single X chromosome. Furthermore, chromosomal silencing was fully compromised when the 5’ region encompassing repeat A was deleted , but concatamers of a synthetic version of these repeats were able to replace the A repeat region. A near normal complement of 7.5 repeats or an increase to 12 repeats fully recapitulated silencing, while 5.5 repeats showed less silencing and 4 repeats were only minimally active . Therefore, in mice the A repeats are necessary for silencing, but additional redundant domains of Xist are involved in localization to the chromosome and the presence of different domains supports models that the RNA is serving as an adaptor to bring different epigenetic silencing proteins to the inactive X.
A number of chromatin remodeling proteins associate with the inactive X chromosome resulting in the acquisition of many histone modifications characteristic of heterochromatin (reviewed in ). The binding of many of these proteins is Xist-dependent; and it has been shown that the A repeat region interacts in vitro and in vivo with components of PRC2 [15–17]. Surprisingly, however, a silencing defective Xist RNA that lacks the repeat A region is still able to recruit PRC2, PRC1, SAF-A, ASH2L and macroH2A1 to the inactive X chromosome in ES cells (reviewed in ). In contrast, a similar deletion in transgenic mice failed to produce Xist RNA, suggesting an important regulatory role for the repeat A region . Furthermore, interaction with the transcriptional repressor YY1  has been shown to occur at the mouse C repeat region and while a direct interaction with the A repeat region has been reported for the splicing factor ASF/SF2, this has been proposed to have a role in enabling proper processing of the Xist RNA facilitating choice of the future inactive X chromosome . Therefore, despite the growing body of literature on XIST/Xist-interacting partners and identification of a critical role for the A repeat region, understanding how XIST/Xist expression leads to gene silencing remains elusive. Contributing to the challenge is the large size of the XIST RNA, and that the monitoring of silencing at distal sites requires both silencing and spread of the RNA along the chromosome.
The palindromic nature of the repeat A core sequences suggests their involvement in forming a distinct secondary RNA structure, and several alternative but mutually exclusive structures have been suggested. The first model proposed that each of the two palindromes forms a hairpin and thus the repeat A region of XIST RNA folds into a two-hairpin 8- or 9-mer . This structure was supported by the abrogation of silencing activity in a construct with two base alterations that would disrupt the putative first hairpin. However, an in vitro analysis of repeat A structure by fluorescence resonance energy transfer, as well as sensitivity to RNases that specifically digest single- or double-stranded RNA regions, proposed an alternative structure. The first palindrome was suggested to engage in pairing between two separate monomers, rather than within repeat A monomers, and the model proposed that the second palindrome did not form a defined structure . Recently, a third option, supported by nuclear magnetic resonance analyses of repeat A monomer and dimer structures, suggested that under in vitro conditions, the first palindrome forms a hairpin, while the second palindrome engages in pairing between repeat A units [21, 22].
Our previously reported inducible transgenic system in the immortal fibrosarcoma line HT1080 provides a tractable system to study the RNA sequences involved in gene repression by XIST . Here, we focus on refining the minimal XIST sequence necessary for cis-regulated silencing, independent of the developmental signals that establish mono-allelic XIST expression in females. We demonstrate silencing of reporter genes by expression of less than 100 bp of XIST containing two consensus repeat A monomers.
Results and discussion
Repeat A is sufficient for XIST-dependent reporter gene silencing
To confirm that silencing results from an XIST RNA-related, sequence-specific effect rather than transcriptional interference, we demonstrated that transcription of the DOX-induced XIST transgenes ceased before the reporter construct. While some transcripts were present downstream of the polyadenylation site, transcription was completely absent at a site approximately 2 kb 5’ of the EGFP promoter (Figure 1C). Our conclusion that silencing is not due to transcriptional interference is further supported by XIST-dependent attenuation of the expression of the hygromycin resistance gene (Hyg) located upstream of XIST and absence of gene silencing with vector lacking XIST sequences (Figure 1D).
Endogenous gene silencing induced by full-length XIST
In order to explore whether XIST RNA is able to induce silencing of the endogenous genes flanking the integration site, we identified the FRT integration site into which subsequent single-copy integrations were directed. DNA-FISH using the XIST cDNA identified the full-length transgene as integrated onto the der(11)t(3;11) of 46,XY,del(1)(p21),i(3)(p10),i(3)(q10),der(4)t(1;4)(p21;p16),der(5)t(5;5)(p15;?),der(11)t(3;11)(q11;q25) cells. We used inverse PCR from primers in the pFRT/lacZeo plasmid to identify the 3q FRT integration site as just downstream of the CLDN1 gene (Figure 1A). Low expression levels of the CLDN1, TMEM207 and LEPREL1 genes prevented a reliable analysis of these adjacent genes by qRT-PCR. Using qRT-PCR following induction of full-length XIST, we observed significant silencing of CLDN16, a gene located approximately 100 kb downstream of XIST (Figure 1D). Neither the construct consisting only of repeat A, nor the construct containing a deletion of repeat A showed significant silencing of CLDN16 upon induction, although there was a non-significant reduction with the repeat A-containing construct. IL1RAP, which is located a further 120 kb downstream (that is, 220 kb from XIST) did not show significant XIST-induced silencing, although there was a non-significant drop in expression. The decrease in CLDN16 transcription is consistent with the almost complete silencing of the cis-located allele; however, attempts to confirm silencing of the XIST-associated allele by FISH failed, presumably due to the relatively low expression levels of CLDN16. In order to examine whether one allele of endogenous genes was being silenced, we identified more distal genes that contained an expressed polymorphism, and thus provided an opportunity to probe allelic silencing. At the DNA level these genes show an allelic ratio of approximately 66%, consistent with the presence of a single allele on the der(11)t(3;11) and the alternate allele in two copies on the isochromosome 3q. Upon DOX treatment there was a significant decrease in relative expression of the single allele for BCL6, LPP and OPA1 (Figure 1D), shown as a change upon DOX induction, relative to the expression in cells without DOX treatment, as there can be variations in allelic expression levels. Similar to the q-PCR results with CLDN16, constructs containing XIST lacking repeat A, or no XIST (vector sequences only) showed no change in allelic ratio upon DOX induction; however, in these cell lines the DNA ratio showed an equivalent allelic DNA ratio, reflecting karyotypic instability of the HT1080 line. There was a significant allelic silencing of BCL6 with the construct containing only repeat A; however, the reduced silencing seen with this construct suggests that additional sequences are required for the spread of the XIST-induced silencing effect beyond the immediate XIST domain.
As repeat A binds the polycomb group 2 proteins that are responsible for trimethylation of H3K27, we asked whether there would be a differential ability of full-length versus repeat A alone to recruit H3K27me3. However, we did not observe any H3K27me3 enrichment by ChIP at the EGFP, Hyg or the CLDN16 promoters (Additional file 1: Figure S1). H3K27me3 is a mark of the inactive X, and has been shown to be enriched at the promoters of inactivated genes ; however, given that the silencing we have observed in this system is reversible ( and data not shown), it is perhaps not surprising that this heritable mark of silent chromatin is not recruited. Similarly, we had previously shown that there was no recruitment of DNA methylation in this reversible system . A similar inducible transgene in mice had identified a developmental window during which inactivation could occur , yet we observe induction of silencing in our somatic cell model; possibly reflecting a more epigenetically dynamic state to these cancer-derived cells, or differences in the genes being examined, as we observed variability between genes in their ability to be silenced. By recapitulating XIST-induced gene silencing, but not requiring sequences involved in the spread of XIST, the A repeat construct exposes the most basal aspects of XIST silencing function. To identify the minimal functional unit for silencing, we further dissected the repeat A sequences.
Repeat A monomers contribute additively to silencing
These observations provide strong evidence that the silencing of an adjacent EGFP reporter is achieved through an additive effect of repeat A monomers, with even a 2-mer repeat A inducing partial EGFP silencing. The number of repeat A units was previously reported to correlate with the ability of Xist to induce silencing in differentiating mouse ES cells . Also, in agreement with a previous report on mouse Xist , artificial repeat A retains full silencing potential when compared to human repeat A, suggesting that neither sequence variation within the CG-rich core nor the varying length of the T-rich spacers in individual repeat A monomers is essential for XIST function. The remarkable ability of a construct with only two repeats to silence EGFP in a reproducible and statistically significant fashion provided us with a well-defined template for further dissection of the relationship between repeat A sequence and its silencing ability.
Measured by flow cytometry, the mutants predicted to enforce pairing within each monomer functioned better than those enforcing the interaction between the monomers; although none of the four mutants silenced EGFP as efficiently as the canonical repeat A 2-mer (Figure 4B), suggesting that there may be more complex structures involved. While the differences in EGFP expression were relatively subtle due to a limited silencing effect of the repeat A 2-mer, they were highly statistically significant and equivalent results were obtained for two single-cell clones of independent integrations and a total of seven biological replicates. More representative structures containing greater than two repeat units were not tested as they could not be predicted to reliably form only a single thermodynamically favored structure. However, given the number of eutherian genome sequences that have now been assembled, we turned instead to a characterization of the full repeat A sequences that are available in genome databases.
Survey of repeat A mutations shows strong preference for stem 1 and mild preference for stem 2 formation
Taking advantage of the increasing number of sequenced mammalian genomes, we generated an alignment of repeat A sequences from 27 mammalian species (Additional file 2: Figure S2A). Repeat A consists of 24 bp-long CG-rich core sequences separated by approximately 20 to 50 bp-long T-rich spacers. The CG-rich core is formed by two palindromes, each of which is broken by four bp-long sequences. As expected, repeat A was well conserved, in particular within the CG-rich core sequences (Additional file 2: Figure S2B). Interestingly, 22/27 mammalian XIST sequences contained either eight or nine monomers of repeat A, and at least one of the remaining five was incomplete across the region, supporting the need for eight monomers to achieve full XIST functionality.
Several secondary structures of repeat A have been proposed based on analysis of repeat A mutants , NMR data, [21, 22] and RNase footprinting and fluorescence resonance energy transfer data . The first palindrome was suggested to form either a hairpin by pairing within each monomer [14, 21, 22] or, alternatively, to pair between monomers . Both our targeted mutations in an artificial 2-mer repeat A construct, as well as our assessment of evolutionary sequence conservation, support the intra-repeat pairing model, consistent with outcomes observed in mice , that the first palindrome indeed forms a stem to expose the ‘ATCG’ tetraloop. The mutations we introduced to the second palindrome also resulted in a complete loss of silencing by XIST (Figure 3), supporting the importance of these sequences; however, these mutations did not directly address secondary structure formation. While the second palindrome was proposed to pair within each monomer to form the second stem-loop , recent studies suggest that the secondary structure may rather involve pairing between individual repeat A monomers [21, 22] or with the T-rich spacers . Our assessment of evolutionary sequence conservation provides evidence in favor of second stem-loop formation, though the frequency of compensatory mutations is less striking than observed for stem-loop 1.
We utilized a single-copy FRT integration site to generate DOX-inducible XIST cDNA integrations allowing the delineation of the repeat A monomers as the minimal functional unit that additively contributes to gene silencing. The ability of only two copies of repeat A to reproducibly silence a flanking EGFP reporter gene allowed for further dissection of repeat A sequence to elucidate the relationship between repeat A structure and function. Disruption of either the putative stems or loops of the repeat A abrogated silencing, and mutations of the first palindrome to enforce pairing within a repeat, or between the first and second repeat, supported models that the first palindrome forms a hairpin. An evolutionary analysis of sequence changes within the palindromes allowed assessment of intra- versus inter-repeat pairing in full-length XIST sequences. Again, the model of intra-repeat pairing was favored. The intricate set of events that ultimately lead to X-chromosome inactivation in female mammals remains a vanguard to mammalian epigenetic research. By focusing only on the ability to silence a proximal reporter we have reduced the complexity of deciphering the critical roles of XIST. We demonstrate that a mere 94 bp-long sequence of repeat A can silence flanking reporter genes, but not more distal endogenous genes that are silenced by induction of the full-length XIST RNA. Further data on the relationship of repeat A sequence and function will provide a foundation for elucidating the yet unclear connection between the sequence of long non-coding RNAs, like XIST/Xist, and their ability to silence chromatin.
The artificial repeat A construct and its shorter derivatives and mutants were synthesized by GeneArt (now Life Technologies Inc, Burlington, ON, Canada) and cloned into the pcDNA5/FRT/TO plasmid (Life Technologies Inc.) using standard techniques. HT1080 cells were transfected as described previously . Mfold server version 2.3 was used to predict secondary RNA structures (http://mfold.rna.albany.edu) of new constructs.
Clones harboring single-copy FLP-mediated integrations of XIST constructs into HT1080 fibrosarcoma cell lines were generated and cultured as described previously . The XIST transgenes were induced by doxycycline (1 μg/mL) and cell culture medium was changed every 24 hours.
Identification of the transgene integration site
Inverse PCR utilizing primers complementary to a sequence within the integrated pEGFP-N1 plasmid (Life Technologies) was used to identify the precise integration site of the XIST - EGFP transgene in the HT1080 2-3-0.5 + 3#4 cell line. PCR primer sequences are listed in Additional file 5: Supplementary methods.
RNA was isolated from frozen cell pellets by TRIZOL (Life Technologies Inc.) and treated with DNase I (Roche Diagnostics, Laval, QC Canada) according to the manufacturers’ recommendations. Following phenol-chloroform extraction, RNA concentration was assessed by spectrophotometry and 0.5 to 2.5 μg of RNA was reverse-transcribed by M-MLV reverse transcriptase (Invitrogen). Fermentas HS Taq (Thermo Scientific, Waltham, MA USA) and EvaGreen (Biotium Inc., Hayward CA USA) were used in quantitative PCR under the following conditions: 5 minutes 95°C, 40x (15 sec. 95°C, 30 sec. 60°C, 60 sec. 72°C). PCR primer sequences are listed in Additional file 5: Supplementary methods.
HT1080 cell pellets were washed with PBS and resuspended in 0.5 mL of PBS supplemented with 10% FCS. A total of 30,000 events were recorded using an LSRII flow cytometer (BD Biosciences, Mississauga, ON, Canada). Mean fluorescence intensity of EGFP was assessed by using a combination of 488 nm laser excitation and a 530/30 nm bandpass filter.
Allelic discrimination by pyrosequencing
A total of 2 μL of cDNA was added to a standard 25-μL pyrosequencing reaction containing 1 × PCR buffer (QIAGEN, Valencia, CA, USA), 0.2 mM dNTPs, 0.625 unit Hot Start Taq DNA polymerase (QIAGEN), 0.25 μM forward primer and 0.25 μM reverse primer. PCR conditions were: 95o for 15 minutes, 35 cycles of 94° for 30 sec, 56.3° or 58.3° for 30 sec (see Supplementary table), 72° for 30 sec, and finally 72° for 10 minutes. Template preparation for pyrosequencing was done according to the manufacturer’s protocol, using 10 to 15 μl of PCR products.
Analysis of repeat A core in mammals
Repeat A sequences in a panel of mammalian species were identified using a combination of BLAST, BLAT and in silico PCR searches of mammalian genomes available through NCBI (http://blast.ncbi.nlm.nih.gov) and ENSEMBL (http://www.ensembl.org/Multi/blastview) databases and the UCSC genome browser (http://genome.ucsc.edu). A table listing the accession numbers or genomic locations of repeat A sequences is provided in Additional file 5: Supplementary methods. Sequences were aligned using clustalw2 (http://www.ebi.ac.uk/Tools/msa/clustalw2) and screened to exclude all non-bona fide repeat A CG-rich core sequences from further analyses. CG-rich core sequences that contained bases deviating from the canonical sequence of either stem 1 or stem 2 were identified. Finally, we tested whether such a mutation was reciprocated by a mutation within the same repeat A unit, or in all other repeats of that species.
Enhanced Green Fluorescent Protein gene
Fetal calf serum
This work was funded by a CIHR operating grant (MOP-13690) to CJB.
- Lyon MF: Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature. 1961, 190: 372-373.View ArticlePubMedGoogle Scholar
- Brown CJ, Ballabio A, Rupert JL, Lafreniere RG, Grompe M, Tonlorenzi R, Willard HF: A gene from the region of the human X inactivation centre is expressed exclusively from the inactive X chromosome. Nature. 1991, 349: 38-44.View ArticlePubMedGoogle Scholar
- Wutz A: Gene silencing in X-chromosome inactivation: advances in understanding facultative heterochromatin formation. Nat Rev Genet. 2011, 12: 542-553.View ArticlePubMedGoogle Scholar
- Brown CJ, Hendrich BD, Rupert JL, Lafreniere RG, Xing Y, Lawrence J, Willard HF: The human XIST gene: analysis of a 17 kb inactive X-specific RNA that contains conserved repeats and is highly localized within the nucleus. Cell. 1992, 71: 527-542.View ArticlePubMedGoogle Scholar
- Brockdorff N, Ashworth A, Kay GF, McCabe VM, Norris DP, Cooper PJ, Swift S, Rastan S: The product of the mouse Xist gene is a 15 kb inactive X-specific transcript containing no conserved ORF and located in the nucleus. Cell. 1992, 71: 515-526.View ArticlePubMedGoogle Scholar
- Duret L, Chureau C, Samain S, Weissenbach J, Avner P: The Xist RNA gene evolved in eutherians by pseudogenization of a protein-coding gene. Science. 2006, 312: 1653-1655.View ArticlePubMedGoogle Scholar
- Elisaphenko EA, Kolesnikov NN, Shevchenko AI, Rogozin IB, Nesterova TB, Brockdorff N, Zakian SM: A dual origin of the Xist gene from a protein-coding gene and a set of transposable elements. PLoS One. 2008, 3: e2521.PubMed CentralView ArticlePubMedGoogle Scholar
- Grant J, Mahadevaiah SK, Khil P, Sangrithi MN, Royo H, Duckworth J, McCarrey JR, VandeBerg JL, Renfree MB, Taylor W, Elgar G, Camerini-Otero RD, Gilchrist MJ, Turner JM: Rsx is a metatherian RNA with Xist-like properties in X-chromosome inactivation. Nature. 2012, 487: 254-258.PubMed CentralView ArticlePubMedGoogle Scholar
- Hoffman LM, Hall L, Batten JL, Young H, Pardasani D, Baetge EE, Lawrence J, Carpenter MK: X-inactivation status varies in human embryonic stem cell lines. Stem Cells. 2005, 23: 1468-1478.View ArticlePubMedGoogle Scholar
- van den Berg IM, Laven JS, Stevens M, Jonkers I, Galjaard RJ, Gribnau J, van Doorninck JH: X chromosome inactivation is initiated in human preimplantation embryos. Am J Hum Genet. 2009, 84: 771-779.PubMed CentralView ArticlePubMedGoogle Scholar
- Okamoto I, Patrat C, Thepot D, Peynot N, Fauque P, Daniel N, Diabangouaya P, Wolf JP, Renard JP, Duranthon V, Heard E: Eutherian mammals use diverse strategies to initiate X-chromosome inactivation during development. Nature. 2011, 472: 370-374.View ArticlePubMedGoogle Scholar
- Yang C, Chapman AG, Kelsey AD, Minks J, Cotton AM, Brown CJ: X-chromosome inactivation: molecular mechanisms from the human perspective. Hum Genet. 2011, 130: 175-185.View ArticlePubMedGoogle Scholar
- Chow JC, Hall LL, Baldry SE, Thorogood NP, Lawrence JB, Brown CJ: Inducible XIST-dependent X-chromosome inactivation in human somatic cells is reversible. Proc Natl Acad Sci U S A. 2007, 104: 10104-10109.PubMed CentralView ArticlePubMedGoogle Scholar
- Wutz A, Rasmussen TP, Jaenisch R: Chromosomal silencing and localization are mediated by different domains of Xist RNA. Nat Genet. 2002, 30: 167-174.View ArticlePubMedGoogle Scholar
- Kanhere A, Viiri K, Araújo CC, Rasaiyaah J, Bouwman RD, Whyte WA, Pereira CF, Brookes E, Walker K, Bell GW, Pombo A, Fisher AG, Young RA, Jenner RG: Short RNAs are transcribed from repressed polycomb target genes and interact with polycomb repressive complex-2. Mol Cell. 2010, 38: 675-688.PubMed CentralView ArticlePubMedGoogle Scholar
- Maenner S, Blaud M, Fouillen L, Savoye A, Marchand V, Dubois A, Sanglier-Cianférani S, Van Dorsselaer A, Clerc P, Avner P, Visvikis A, Branlant C: 2-D structure of the A region of Xist RNA and its implication for PRC2 association. PLoS Biol. 2010, 8: e1000276.PubMed CentralView ArticlePubMedGoogle Scholar
- Zhao J, Sun BK, Erwin JA, Song JJ, Lee JT: Polycomb proteins targeted by a short repeat RNA to the mouse X chromosome. Science. 2008, 322: 750-756.PubMed CentralView ArticlePubMedGoogle Scholar
- Hoki Y, Kimura N, Kanbayashi M, Amakawa Y, Ohhata T, Sasaki H, Sado T: A proximal conserved repeat in the Xist gene is essential as a genomic element for X-inactivation in mouse. Development. 2009, 136: 139-146.View ArticlePubMedGoogle Scholar
- Jeon Y, Lee JT: YY1 tethers Xist RNA to the inactive X nucleation center. Cell. 2011, 146: 119-133.PubMed CentralView ArticlePubMedGoogle Scholar
- Royce-Tolland ME, Andersen AA, Koyfman HR, Talbot DJ, Wutz A, Tonks ID, Kay GF, Panning B: The A-repeat links ASF/SF2-dependent Xist RNA processing with random choice during X inactivation. Nat Struct Mol Biol. 2010, 17: 948-954.PubMed CentralView ArticlePubMedGoogle Scholar
- Duszczyk MM, Zanier K, Sattler M: A NMR strategy to unambiguously distinguish nucleic acid hairpin and duplex conformations applied to a Xist RNA A-repeat. Nucleic Acids Res. 2008, 36: 7068-7077.PubMed CentralView ArticlePubMedGoogle Scholar
- Duszczyk MM, Wutz A, Rybin V, Sattler M: The Xist RNA A-repeat comprises a novel AUCG tetraloop fold and a platform for multimerization. RNA. 2011, 17: 1973-1982.PubMed CentralView ArticlePubMedGoogle Scholar
- Calabrese JM, Sun W, Song L, Mugford JW, Williams L, Yee D, Starmer J, Mieczkowski P, Crawford GE, Magnuson T: Site-specific silencing of regulatory elements as a mechanism of X inactivation. Cell. 2012, 151: 951-963.PubMed CentralView ArticlePubMedGoogle Scholar
- Wutz A, Jaenisch R: A shift from reversible to irreversible X inactivation is triggered during ES cell differentiation. Mol Cell. 2000, 5: 695-705.View ArticlePubMedGoogle Scholar
- Zuker M: Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 2003, 31: 3406-3415.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.