regulation of imprinted x-chromosome inactivation in mice by tsix

INTRODUCTION

During early embryogenesis of female mammals, one of thetwo X chromosomes becomes transcriptionally silenced (Lyon,1961). While the choice of which X chromosome to beinactivated in the embryonic (epiblast) lineage is random,the paternally derived X chromosome (Xp) is preferentiallyinactivated in the extra-embryonic lineages in mouse (Takagiand Sasaki, 1975). The mechanism that controls this complexchromosome-wide epigenetic process is believed to consist ofmultiple steps, including (1) counting of the number of Xchromosomes, (2) choosing which X chromosome to beinactivated, (3) initiation of inactivation, (4) spreading of theinactive state along the X chromosome and (5) its maintenance.

It is known that a cytogenetically identified region, X-chromosome inactivation center (Xic), is essential forX-inactivation, from which X-inactivation initiates andpropagates in both directions along the chromosome (Russelland Montgomery, 1965). The Xist gene, mapped in the Xicregion, is exclusively expressed from the inactive Xchromosome in female somatic cells (Brown et al., 1991;Borsani et al., 1991; Brockdorff et al., 1991), and is thought toplay a role as a functional RNA. Xist is paternally expressedfrom four-cell stage onwards, and is thought to be responsible

for imprinted X-inactivation in the extra-embryonic lineagesthat takes place afterwards, while it is randomly expressedfrom either allele in the embryonic lineage, reflecting randomX-inactivation (Kay et al., 1993; Kay et al., 1994). Targetedmutagenesis of the Xistgene demonstrated that Xistis essentialfor X-inactivation to occur in cis (Penny et al., 1996;Marahrens et al., 1997). Fluorescent in situ hybridization(FISH) analyses detecting XistRNA revealed that there are twodistinct patterns for Xistexpression (Lee et al., 1996; Panningand Jaenisch, 1996; Panning et al., 1997; Sheardown et al.,1997). One form is a pinpoint signal that represents a nascentunstable RNA characteristic of the transcriptionally active Xchromosome in undifferentiated male and female embryonicstem (ES) cells and epiblast cells. The other form is anaccumulated signal occupying a relatively large domain in anucleus, which is a processed stable RNA that coats theinactivated X chromosome in female somatic cells anddifferentiated female ES cells.

It has been shown that YAC and cosmid-derived transgenescontaining the Xistgene appear to have the properties of Xic(Lee and Jaenisch, 1997; Herzing et al., 1997; Heard et al.,1999). However, Simmler et al. have suggested that anothergenetically defined locus lying in Xic, X chromosomecontrolling element (Xce), which is known to influence the

1275Development 128, 1275-1286 (2001)Printed in Great Britain © The Company of Biologists Limited 2001DEV4536

In mammals, X-chromosome inactivation is imprinted inthe extra-embryonic lineages with paternal X chromosomebeing preferentially inactivated. In this study, weinvestigate the role of Tsix, the antisense transcript fromthe Xist locus, in regulation of Xist expression and X-inactivation. We show that Tsix is transcribed from twoputative promoters and its transcripts are processed.Expression of Tsix is first detected in blastocysts and isimprinted with only the maternal allele transcribed. Theimprinted expression of Tsix persists in the extra-embryonic tissues after implantation, but is erased inembryonic tissues. To investigate the function of Tsixin X-inactivation, we disrupted Tsixby insertion of an IRESβgeocassette in the second exon, which blocked transcripts fromboth promoters. While disruption of the paternal Tsixallele

has no adverse effects on embryonic development,inheritance of a disrupted maternal allele results in ectopicXist expression and early embryonic lethality, owing toinactivation of both X chromosomes in females and singleX chromosome in males. Further, early developmentaldefects of female embryos with maternal transmission ofTsix mutation can be rescued by paternal inheritance of theXist deletion. These results provide genetic evidence thatTsix plays a crucial role in maintaining Xistsilencing in cisand in regulation of imprinted X-inactivation in the extra-embryonic tissues.

Key words: X-inactivation, Dosage compensation,Xist, Tsix,Genomic imprinting, Mouse

SUMMARY

Regulation of imprinted X-chromosome inactivation in mice by Tsix

Takashi Sado 1,2,*, Zhenjuan Wang 3, Hiroyuki Sasaki 1,4 and En Li 3,5,*1Division of Human Genetics, National Institute of Genetics, 1111, Yata, Mishima, 411-8540, Japan2Department of Biosystems Science, The Graduate University for Advanced Studies, 1111, Yata, Mishima, 411-8540, Japan3Cardiovascular Research Center, Massachusetts General Hospital-East, 149, 13th Street, Charlestown, MA 02129, USA4Department of Genetics, The Graduate University for Advanced Studies, 1111, Yata, Mishima, 411-8540, Japan5Department of Medicine, Harvard Medical School, Charlestown, MA 02129, USA*Authors for correspondence (e-mail: [email protected] or [email protected])

Accepted 23 January; published on WWW 22 March 2001

1276

probability of which X chromosome in female cells is to beinactivated (Cattanach et al., 1972), is segregated from the Xistgene and located 3′ to Xist (Simmler et al., 1993). It is,therefore, less likely that the Xistgene is solely responsible forall the function of Xic. Moreover, Clerc and Avner haverecently shown that deletion of a 65 kb genomic region 3′ toXist resulted in aberrant expression of Xistand subsequent X-inactivation in differentiating ES cells (Clerc and Avner, 1998),suggesting the presence of an element in the 3′ region, whichis involved in regulation of Xist expression.

Several lines of evidence suggest that in the extra-embryonictissues, the maternal X chromosome (Xm) is imprinted to resistthe inactivation cue rather than Xp being imprinted to becomeinactivated (Lyon and Rastan, 1984; Shao and Takagi, 1990; Gotoand Takagi; Okamoto et al., 2000). In this study, we attempted toidentify an element in Xic region responsible for the imprintinglaid on Xm. We report the isolation of a processed antisense RNAat the Xistlocus and effects of its targeted mutagenesis on X-inactivation and embryonic development in mice.

MATERIALS AND METHODS

RACE, RT-PCR and sequencing5′ and 3′RACE systems (Life Technologies) were used. Total RNAwas prepared as described (Chomczyinski and Sacchi, 1987). 5′ and3′ RACE were performed on RNA prepared from the placenta ofE12.5 male embryos and female ES cells, respectively, according tomanufacturer’s instruction with following gene-specific primers(GSPs). For the isolation of clone24: GSP1, cgg ggt ttt gga tac tta cct;GSP2, cgg ggc ttg gtg gat gga aat; GSP3, gag tta ttg cac tac ctg gaa.For the isolation of clone21b and 1b: GSP1, tgc gct gcg cta gct aaa;GSP2, tgc ggg att cgc ctt gat tt; GSP3, gcc aag gag tgc tcg tat ta. Forthe isolation of clone10: GSP1: cag gta gtg caa taa ctc aca a; GSP2,tga aaa ggc agg taa gta tcc a. For the isolation of clone4/4: GSP1: caggta gtg caa taa ctc aca a; GSP2, aca tga cgc gtg aat ttc at. For theisolation of clone1/23: GSP1: cag gta gtg caa taa ctc aca a; GSP2, ccacat gaa aga gat cag ac. The RACE products were subsequently clonedinto pCRII vector (Invitrogen) for sequencing.

For the isolation of rt-clones, RT-PCR was carried out on cDNAconverted from RNA prepared from female ES cells with randomhexamer (Sado et al., 1996), and the amplified products were clonedinto pCRII vector (Invitrogen) for sequencing. Primers used were asfollows: 21b80F, cct gca agc gct aca cac tt; 21b352R, gga gag cgc atgctt gca at; 1b36F, gcc agc acg tgt gct agc tt. Thermal condition forPCR consisted of a 5 minutes denature at 94°C, followed by 35 cyclesof 94°C for 30 seconds, 55°C for 30 seconds, and 1 minute at 72°C,followed by a 5 minutes extension at 72°C.

Sequencing was performed with an ABI PRISM BigDye TerminatorCycle Sequencing Ready Reaction Kit (Applied Biosystems)

Allelic expression assay of TsixBy comparing genomic sequences of each exon between C57Bl/6 andJF1, a single nucleotide polymorphism was identified in exon 4. TotalRNA were prepared from eight-cell embryos, blastocysts, the embryoproper and placenta at E11.5 derived from crosses between C57Bl/6and JF1, which were converted into cDNA at 50°C using SuperscriptII (Life Technologies) and a strand-specific primer for either Tsix orXist, as described by the manufacturer. Total RNA (2.5 µg) of theembryo proper and placenta was used for each cDNA synthesis. TotalRNA isolated from 30 eight-cell embryos or 20 blastocysts weredivided into three aliquots for Tsix-specific, Xist-specific cDNAsynthesis, with an RT-minus control in each attempt. Following PCRon thus generated cDNA, the amplified products were subjected to

direct sequencing. Thermal conditions were as already describedabove with 40 cycles instead of 35. Primer sequences used are asfollows: a primer for Tsix-specific cDNA synthesis (R1910J), cat cggggc tgt gga tac ct; a primer for Xist-specific cDNA synthesis(F1063AS), gca caa ccc cgcaaa tgc ta; for PCR (700P2), cgg ggc ttggtg gat gga aat; AS1634F, gcg taa ctg gct cga gaa ta.

Targeted disruptionOn the basis of a 94 kb sequence (GenBank Accession NumberX99946) reported by Simmler et al. (Simmler et al., 1996), restrictionmaps of the genomic regions containing exons 1, 2, and 3 wereproduced. For the isolation of relevant genomic clones, a 129/Sv BAClibrary (Genome Systems) was screened using either exon 1 sequenceor exon 2 sequence as a probe. For the construction of pSS1∆2.7, aSacI-SacI fragment was cloned into pBluescriptII SK (Stratagene), fromwhich all of the internal four NheI-NheI fragments across exon 1 wereremoved and replaced with a Pgk-neocassette, so as to be transcribedin an opposite direction to Tsix. For the construction of pAA2∆1.7, aSpeI-ApaI fragment was cloned into pBluescriptII SK, from which a1.7 kb SmaI-SalI fragment that contained exons 2 and 3 was removedand an IRESβgeocassette (Mountford et al., 1994) was placed in thesame direction as the transcription of the antisense RNA. The linearizedtargeting constructs thus produced were electroporated into J1 male EScells, which were subsequently selected in the presence of G418 asdescribed (Li et al., 1992). Appropriate targeting events were verifiedby Southern hybridization. Chimeric males were generated followinginjection of each type of targeted ES cells into blastocysts.

Northern hybridizationPoly(A)+RNA was isolated from male and female ES cells usingOligotex-dT30 (Qiagen). About 5 µg of highly enrichedpoly(A)+RNA were run on a 1% agarose gel containing formaldehydeand blotted onto HybondN+ (Amersham Pharmacia). Hybridizationwas performed in standard hybridization buffer at 65°C in thepresence of a radiolabeled RNA probe prepared by in vitrotranscription using T7 RNA polymerase (Promega). Washing wascarried out according to manufacturer’s instruction.

Histology and X-gal stainingEmbryos were fixed in Bouin’s fixative without taking out from thedecidium at E6.5, E7.5 and E8.5, followed by dehydration andembedding in paraffin. Each specimen was serially sectioned at 6 µm,dewaxed and stained with Hematoxylin and Eosin. Fibroblasts grownon culture dishes or embryos were fixed and stained for β-galactosidase activity as described (Sado et al., 2000).

RNA FISHPreparations for FISH on E7.5 embryos were made according to themethods previously described (Takagi et al., 1982). A probe wasprepared by nick translation with Cy3-dUTP (Amersham Pharmacia)from an equimolar mixture of a series of XistcDNA clonesencompassing exons 1-7. Hybridization and following washes werecarried out as described (Lawrence et al., 1989) with minormodifications. Following examination of XistRNA signal, preparationsused for RNA FISH were re-probed with either X- or Y-specific paintingprobes (Cambio), as described in the manufacturer’s instructions fordetermining the sex of embryos. Slides were examined with a ZeissAxioplan microscope, and images were captured with a PhotometicsCCD camera coupled to IPLab software (Signal Analytics).

RESULTS

Identification of a processed antisense RNA fromthe Xist locusAs antisense RNA is often associated with imprinted genes

T. Sado and others

1277Role for Tsix in X-inactivation

(Moore et al., 1997; Wutz et al., 1997; Rougeulle et al., 1998),we surmised that there might be transcription occurring inan opposite direction to Xist in the extra-embryonic tissues,where the paternally derived X chromosome is preferentiallyinactivated. Accordingly, we examined this possibility byperforming 5′RACE on RNA prepared from the placenta atembryonic day 12.5 (E12.5) with primers annealing tosequences complementary to the 5′ region of Xist. This attemptallowed us to isolate a fragment extended along Xist(clone24),whose 5′end was at 1109 of the Xist cDNA sequence reportedby Brockdorff et al. (Brockdorff et al., 1992; GenBankAccession Number L04961; Fig. 1A). Further 5′RACE wascarried out using another set of GSPs designed according tothe sequence of clone24 and two additional clones weresubsequently obtained (clone21b and 1b). Sequencing ofclone21b and 1b revealed that they contained not only a 919bp sequence common to Xist, which corresponds to a region1020-1938 in Xistrelative to the transcription start site, butdifferent extra sequences of 220 bp and 177 bp, respectively,both of which were not present in Xist. Homology search witha 220 bp sequence in clone21b and a 177 bp sequence in 1brevealed the presence of identical sequences in a 94 kb regionlocated 3′to Xist (Fig. 1A), which had been fully sequencedby Simmler et al. (Simmler et al., 1996; GenBank AccessionNumber X99946). The position of 220 bp and 177 bpsequences were 77705-77924 and 61493-61669, respectively,according to X99946, strongly suggesting that they coded foralternative exons of the antisense transcript of Xist. To confirmthis, reverse transcription polymerase chain reaction (RT-PCR)was conducted with a primer specific to either the 220 bpsequence in 21b (21b80F) or the 177 bp sequence in 1b

(1b36F) in combination with one designed in a sequenceshared with Xist(21b352R; Fig. 1). Both RT-PCR producedmultiple bands including one with an expected size on RNAprepared from male and female ES cells as well as the placenta(data not shown). Sequencing of the amplified products (clonert34, rt35, rt37, rt49, and rt50) verified that they were indeedsplicing variants transcribed from the Xistlocus, which alsoallowed us to identify various splicing events and an additionalexon (Fig. 1A). Exon-intron boundaries in these 5′RACE andRT-PCR products precisely followed the GT-AG rule in allinstances. None of these splicing variants contained a longopen reading frame, suggesting that they did not encode aprotein. Additional 5′RACE from 177 bp segment in 1b did notextend the exon by more than 8 bp.

A northern blot containing poly(A)+RNA isolated from maleand female ES cells was hybridized with a clone rt50-derivedRNA probe specific to the antisense transcript but not Xist,which displayed that the major transcripts were about 4.3 and2.7 kb in length (Fig. 1B). Probing with the opposite strand didnot yield a specific hybridization signal (data not shown;also see Fig. 3B). To determine the transcription unit of theantisense RNA, we employed 3′RACE on RNA prepared fromfemale ES cells and identified three different 3′ends, all ofwhich were followed by a poly(A) tail (clone10, 4/4, and 1/23in Fig. 1A). The position of each 3′ end relative to thetranscription start site of Xistwere –171, −1375 and –1738,respectively. Therefore, the transcription unit of the antisenseRNA spanned about 50 kb covering the entire Xist gene. AnRNA probe generated from clone4/4 failed to detect the lower2.7 kb band on the northern blot (Fig. 1B), indicating that itrepresented RNA terminated at the 3′end identified in clone10.

Fig. 1. (A) Genomic structure of Tsix. The location of each exon relative to Xist(Brockdorff et al., 1992; Hong et al., 1999) and Tsx(Simmler etal., 1996) are shown under the map. A 65 kb region deleted by Clerc and Avner (Clerc and Avner, 1998) is also shown on the top of the map.Clones obtained by RACE and RT-PCR are aligned under the map. Clones, 10, 4/4, and 1/23 were isolated by 3′RACE (open boxes) and 24,21b and 1b were by 5′RACE (filled boxes). (A)n represents a poly(A) tail found in 3′RACE clones. The exon 2′is an isoform of exon 2, whosesplicing donor site appears 17 bp upstream from the end of exon2, which was found in rt34 and 49. Primers used for RT-PCR were shown onthe right of each clone and their positions are indicated below the Tsixexons. (B) Northern analysis of Tsixexpression in male and female EScells. Northern blot containing 5 µg of poly(A)+RNA was serially hybridized with a RNA probe prepared from either clone rt50 or 4/4. β-actinwas a loading control.

1278

The presence of antisense RNA at the Xistlocus has beenshown previously (Lee et al., 1999), which is termed Tsix.Although Lee et al. claimed that Tsix is a 40 kb RNA withoutintrons, our results unequivocally demonstrate that Tsixwas, atleast in part, subject to processing. The functional significanceof the processing of both Xist and Tsixtranscripts remains tobe determined.

Imprinted expression of Tsix in blastocysts andextra-embryonic tissuesTo elucidate whether Tsixis involved in imprinted X-inactivationin the extra-embryonic lineages, allelic expression profile wasstudied in the embryo proper and placenta. We took advantageof a single nucleotide polymorphism found in exon 4 betweenC57Bl/6 (M. m. domesticus) and JF1 (M. m. molossinuss) (Fig.2A). Following the reciprocal crosses between C57Bl/6 and JF1,total RNA was prepared from the embryo proper and placentaat E11.5. Although expression levels of Tsix appeared muchlower in the embryo proper than in the placenta (data not shown),an expected fragment was successfully amplified from bothRNA by RT-PCR with extended cycles. Direct sequencingdemonstrated that in male, Tsixexpressed in the embryo properand placenta was certainly derived from the single maternal Xchromosome (Fig. 2A). In female, however, although bothparental alleles were expressed in the embryo proper, only the

maternal transcript was detected in the placenta in either cross(Fig. 2A). By E15.5, Tsixexpression in the embryo proper andplacenta became downregulated and barely detectable (data notshown). The allelic expression pattern was further analyzed ineight-cell embryos and blastocysts. With a common primer setacross the polymorphism, RT-PCR was performed to amplifyeach of Tsixand Xiston the strand-specifically primed cDNAprepared from pooled eight-cell embryos, which, although Xistexpression was detected as previously shown (Kay et al., 1993),failed to amplify the Tsixsequence (Fig. 2B). No amplificationoccurred even after the second round of PCR with another 40cycles (data not shown), indicating Tsixexpression had not yetbeen started at the eight-cell stage. On the other hand, theexpected fragments were successfully amplified from pooledblastocysts from both templates. Direct sequencing revealed thatXist was derived from the paternal allele, while Tsix was fromthe maternal allele (Fig. 2C). It seemed that Tsixexpressionbegan around the time of blastocyst formation, when preferentialpaternal X-inactivation first occurs in the trophectoderm that hasjust differentiated. This suggests that Tsixplay a role introphectoderm formation and/or in imprinted X-inactivationoccurring in this tissue.

Targeted disruption of the Tsix gene Based on the positions of each exon, we created two different

T. Sado and others

Fig. 2. Expression of Tsixin the embryonic and extra-embryonic tissues, and preimplantation embryos. (A) Allelic expression profile of Tsix inthe embryo proper and placenta at E11.5. In female, both parental transcripts of Tsixwere detected in the embryo proper, while only thematernal allele was expressed in the placenta. B6JF, C57Bl/6×JF1 (female is written first and male second); JFB6, JF1×C57Bl/6; Emb, embryoproper; Pl, placenta; M, male; F, female. (B) Expression of Tsixand Xistin pooled eight-cell embryos. RT-PCR was performed with the sameprimer set to amplify each of Tsixand Xistsequences on strand-specifically primed cDNA. Tsixwas not expressed in eight-cell embryos.Control was a reverse transcriptase minus reaction with a Tsix-specific primer. (C) Allelic expression profile of Tsixand Xistin pooledblastocysts. RT-PCR products amplified from the strand-specifically primed cDNA were sequenced. While Xistwas paternally expressed (Kayet al., 1993), Tsixwas transcribed from only the maternal allele in blastocysts.


targeting constructs to disrupt the Tsix gene in two differentways. In one construct (pSS1∆2.7), putative exon 1 and its 5′region, which is located about 28 kb away from the recentlyrevised 3′end of Xist(Hong et al., 1999), was deleted andreplaced with a Pgk-neocassette (Fig. 3A). We expected thatremoval of these sequences would abolish transcription itselfat the Tsix locus. In the other construct (pAA2∆1.7), anIRESβgeocassette (Mountford et al., 1994) was placed in exon2 so as not to disturb the transcription from the promoters buteliminate Tsix RNA (Fig. 3A). Making use of these twoconstructs, targeted disruption was carried out in J1 male EScells (Li et al., 1992) and multiple targeted clones were isolatedfor each targeting constructs.

Following confirmation of the accurate targeting events (Fig.3A), we examined expression of Tsixin the knockout cell linesby northern blotting with an RNA probe prepared from clone10 (see Fig. 1A), which specifically detects Tsix (Fig. 3B).Unexpectedly, #218 cell line involving pSS1∆2.7 (TsixSS1∆2.7)still retained Tsixexpression at a level comparable to theparental J1 cells, while expression of Tsixwas not detected incell lines #6, involving pAA2∆1.7 (TsixAA2∆1.7). These wereconfirmed in multiple cell lines in both cases by RT-PCR aswell as northern analysis (data not shown). It seemed,therefore, likely that there were, at least, two distinct promotersfor Tsix transcription and a major one would be located in thevicinity of exon 2. Deletion of exon 1 and the putative firstpromoter probably did not affect this second promoter,

allowing Tsix expression from theTsixSS1∆2.7 allele. A strongsplicing acceptor site present in pAA2∆1.7, however, wouldtrap transcripts originated from both promoters, resulting inessentially no mature TsixRNA produced from the TsixAA2∆1.7

allele. Interestingly, when the northern blot was reprobed withan RNA probe specific to Xist, significant expression of Xistwas detected in #6 cell lines but not in #218 and the parentalJ1 cells (Fig. 3B). FISH on multiple ES cell lines carrying theTsixAA2∆1.7 allele demonstrated that although the majority ofnuclei were negative for Xist expression, a small proportion ofcells showed accumulation of XistRNA as seen in femalesomatic cells (detailed analysis will appear elsewhere). Thisimplied that elimination of TsixRNA somehow impaired theregulatory mechanism of Xistexpression.

Maternal transmission of the Tsix -disrupted X AA2D1.7

is embryonic lethalThe knockout ES cell lines, #218 and #6, were injected intoblastocysts to create germline chimeras, and the mutatedX-chromosome derived from either #218 (XSS1∆2.7) or #6(XAA2∆1.7) was transmitted to female offspring from thechimeric males. Females heterozygous for XSS1∆2.7 transmittedit to both male and female offspring, from which femaleshomozygous for XSS1∆2.7 were produced. Both hemizygousmale (XSS1∆2.7Y) and homozygous female (XSS1∆2.7XSS1∆2.7)were apparently healthy and fertile, and no discerniblephenotypic abnormality was found. In contrast, females

Fig. 3.Targeted disruption of the Tsixgene in two different ways. (A) pSS1∆2.7 wasused to replace putative exon 1 and the neighboring sequences with a Pgk-neocassette,whereas pAA2∆1.7 was designed to interfere the production of TsixRNA by placing apromoterless IRESβgeocassette in exon 2. The precise targeting events were confirmedby Southern blotting (PstI digestion for #218 and KpnI for #6) with external probes Aand D. Targeting schemes (Lee and Lu, 1999; Debrand et al., 1999) are also shown forcomparison. (B) Expression of Tsixand Xistin the disrupted ES cells. Cell line #6harbored TsixAA2∆1.7 and #218 harbored Tsix SS1∆2.7. J1 was a parental cell line. Anorthern blot containing 5 µg of poly(A)+RNA was hybridized with a RNA probespecific to either Tsixor Xist. Gapdwas a loading control.

1280

heterozygous for XAA2∆1.7 (XXAA2∆1.7) (the maternallyderived X chromosome is written first and the paternallyderived one the second) mated with wild-type males deliveredtotally 59 males and 45 females, out of which only one maleand one female had XAA2∆1.7 transmitted (Table 1). This raisedthe possibility that maternal transmission of XAA2∆1.7 causedpredominant lethality during embryogenesis.

PCR genotyping of 42 embryos recovered from five littersat E6.5 and E7.5 revealed that XAA2∆1.7 had been reasonablytransmitted from XXAA2∆1.7 mothers to both male (14 out of20) and female (8 out of 22) embryos. Gross morphology ofXAA2∆1.7X/XAA2∆1.7Y embryos at E6.5 showedsome variations, that is, some appeareddevelopmentally retarded and others wereindistinguishable from wild-type littermates(Fig. 4A). By E7.5, the differences in grossmorphology became quite obvious betweenwild-type embryos and XAA2∆1.7X/XAA2∆1.7Ylittermates. Although variation amongXAA2∆1.7X/XAA2∆1.7Y embryos was stillobserved, all of them were invariablysmaller than wild-type egg-cylinder stageembryos. The variations found amongXAA2∆1.7X/XAA2∆1.7Y embryos were notrelated to the sex of embryos. Some ofXAA2∆1.7X/XAA2∆1.7Y embryos at E8.5appeared developmentally arrested at or priorto E7.5, while others developed an empty yolksac with little or no embryonic tissues (Fig.4C). In contrast, female embryos that inheritedXAA2∆1.7 from the chimeric father wereindistinguishable from wild-type malelittermates at E8.5 (data not shown), suggestingthat the paternally derived TsixAA2∆1.7 had nodeleterious effect on embryonic development.

To further characterize the embryonicdefects, we performed histology on embryosrecovered from XAA2∆1.7X females at E6.5,E7.5 and E8.5. At E6.5, the presumptiveXAA2∆1.7X/XAA2∆1.7Y embryos weredevelopmentally retarded and failed to form awell organized egg-cylinder though all theprimitive tissues, including the ectoplacental

T. Sado and others

Fig. 4.Phenotypic comparison betweenXAA2∆1.7X/X AA2∆1.7Y and wild-type embryos.Gross morphology of (A) E6.5, (B) E7.5 and (C)E8.5 embryos upon maternal transmission ofXAA2∆1.7. Embryos marked with an asterisk werewild type and remaining all carried maternallyderived Tsix AA2∆1.7. Histology of (D) E6.5 wild-type embryo, (E,F) E6.5 presumptiveXAA2∆1.7X/XAA2∆1.7Y embryos, (G) E7.5 wild-typeembryo, (H,I) E7.5 presumptiveXAA2∆1.7X/XAA2∆1.7Y embryos, (J,K) E8.5presumptive XAA2∆1.7X/XAA2∆1.7Y embryos. Allwere photographed at the same magnification. a,aminotic cavity; al, allantois; ch, chorion; ec,ectoplacental cavity; ee, embryonic ectoderm; epc,ectoplacental cone; exe, extra-embryonic ectoderm;m, mesoderm; ve, visceral endoderm. Scale bars:1 mm in A-C; 0.1 mm in D-K.

Table 1. No. of pups born to the chimeric male or femalesheterozygous for TsixAA2∆1.7

B6 ×chimera XXAA2∆1.7 × B6

Female Male Total Female Male Total

Wild type 1* 86 87 44 58 102w/XAA2∆1.7 87 1‡ 88 1 1 2Total 88 87 175§ 45 59 104

(19 litters) (24 litters)

*XO female.‡XXAA2∆1.7Y male.§All pups but one were embryonic stem cell-derived.


cone, extra-embryonic ectoderm, visceral endoderm andembryonic ectoderm were formed (Fig. 4E,F). At E7.5,XAA2∆1.7X/XAA2∆1.7Y embryos were shown to containmesodermal tissues localized predominantly at the embryonicand extra-embryonic junction (Fig. 4H,I). It appeared that cellsof the ectoplacental cone failed to expand into the decidualtissue, but instead were accumulated in the proximal region ofthe egg cylinder. The ectoplacental cavity, exocoelom andamniotic cavity were all small, which was probably due tofailure of the egg-cylinder to elongate distally. Morphology ofXAA2∆1.7X/XAA2∆1.7Y embryos was variable at E8.5. Some ofthem retained the same morphology as the ones at E7.5 (Fig.4J). Others grew larger and formed the yolk sac withdisorganized embryonic tissues in it (Fig. 4K). The embryoslacked discernible patterning and were characterized by thepresence of disorganized allantois and chorion and dead cells

in the exocoelom (Fig. 4K). These results indicate thatmaternal transmission of XAA2∆1.7 leads to abnormal formationof the extra-embryonic tissues and growth arrest of theembryos prior to or during gastrulation.

Non-random X-inactivation in somatic cells deficientfor TsixTo examine whether each of these mutations affected randomX-inactivation occurring in female somatic cells, we producedfemales carrying the X-linked lacZtransgene (XH253) incombination with either of XSS1∆2.7 or XAA2∆1.7. XH253 hadbeen extensively studied and shown to be a useful marker tomonitor the activity of the X chromosome carrying thetransgene (Tan et al., 1993; Tam et al., 1994; Sado et al., 2000).Fibroblasts derived from a tail tip of each female were stainedfor β-galactosidase activity. As shown in Fig. 5A,B, about 50%

Fig. 5. Impaired random X-inactivation in female fibroblastsand embryos heterozygous for XAA2∆1.7. (A) An Xchromosome carrying the lacZtransgene (XH253) had beenintroduced in female mice in combination with one of eitherTsixmutation. Tail fibroblasts from these females doublyheterozygous for one of Tsixmutation and XH253 were stainedfor β-galactosidase activity. (B) Proportion of fibroblastspositive and negative for lacZexpression. (C) Upper panel,gross morphology of male (left; XAA2∆1.7Y) and female (twomiddle: XAA2∆1.7XH253) mutant embryos and a wild-typelittermate (right); lower panel, histology of three mutant femaleembryos stained for β-galactosidase activity. The embryonictissues were uniformly positive for β-galactosidase. Note that a

subset of cells in the extra-embryonic tissues were also β-galactosidase positive. ch, chorion; ee, embryonic ectoderm; epc, ectoplacental cone;m, mesoderm. Scale bars in C: 1 mm in the upper panel and 0.1 mm in the lower panels.

1282

of fibroblasts derived from an XH253X female were positive forβ-galactosidase, manifesting random X-inactivation. Theproportion of β-galactosidase-positive cells in XH253XSS1∆2.7

fibroblasts were almost the same as XH253X, indicating thatrandom X-inactivation was not affected by this mutation. Incontrast, almost every XH253XAA2∆1.7 fibroblast was β-galactosidase-positive as seen in XH253Y, which demonstratedthat XAA2∆1.7 had been inactivated in every cell. No β-galactosidase-staining was detected in XXAA2∆1.7 cells,although XAA2∆1.7 carried the IRESβgeocassette, presumablyreflecting almost no expression of Tsixin adult somatic cells(data not shown).

Embryos recovered at E7.5 from XXAA2∆1.7 females crossedwith XH253Y males were stained for β-galactosidase activity.None of XAA2∆1.7Y embryos were positive for β-galactosidase,indicating that β-galactosidase expression in females came fromXH253 rather than from IRESβgeoon XAA2∆1.7. Fig. 5C showsgross morphology of male and female mutant embryos carryingXAA2∆1.7 and histology of three female mutant embryos, inwhich most of cells consisting of the embryonic ectoderm andmesoderm were positive for β-galactosidase (>95%, more than150 cells were counted for each embryo). No abnormal celldeath was observed in the embryonic tissues, arguing againstthe possibility of cell selection. It is likely that elimination ofTsix RNA in the epiblast lineage disrupted the mechanism forrandom X-inactivation in female embryos with the mutated Xchromosome being exclusively inactivated. These findingswere consistent with skewed X-inactivation observed indifferentiated female ES cells heterozygous for Tsix∆CpG (Leeand Lu, 1999) and the transgenic male ES cell lines carryingXic with the DXPas34-deletion at the ectopic site (Debrand etal., 1999). It is, therefore, most probable that Tsix is involvedin the choice mechanism in the epiblast lineage. It is worthmentioning that a small fraction of extra-embryonic cells inXAA2∆1.7XH253 were positive for β-galactosidase (Fig. 5C),indicative of an active paternal XH253 and an inactive maternalXAA2∆1.7, because two active X chromosomes are incompatiblewith development of the extra-embryonic lineages (Shao andTakagi, 1990; Goto and Takagi, 1998) (see Discussion).

Ectopic accumulation of Xist RNA inXAA2∆1.7X/XAA2∆1.7Y embryosTo elucidate the possible role of Tsix in X-inactivation, RNAFISH was carried out to analyze the expression of Xist inXAA2∆1.7X/XAA2∆1.7Y embryos. Fig. 6 shows FISH detectingXist RNA on the cells of whole egg-cylinder at E7.5. It wasevident that accumulation of XistRNA occurred in a subset ofinterphase nuclei and the metaphase chromosome spread inXAA2∆1.7Y embryos (Fig. 6A,E). The coating of the Xchromosome by XistRNA was verified by X chromosomepainting (Fig. 6B,F). Similar ectopic accumulations wereobserved in XAA2∆1.7X female embryos, in which Xist RNAspecifically co-localized with both X chromosomes in afraction of cells (Fig. 6C,D,G,H). The frequency of cellsshowing the aberrant coating of Xist RNA with the Xchromosomes was 8-21% (the number of embryos examinedwere six and five for male and female, respectively). Theectopic expression of Xistwas never observed in wild-typelittermates in both sexes. Our results, therefore, demonstratedthat the maternal inheritance of XAA2∆1.7 induced the ectopicaccumulation of XistRNA irrespective of the number of Xchromosomes in a subset of cells in mutant embryos. Althoughit was important to determine what type of cells showed suchectopic Xist accumulation, it was practically difficult toseparate embryonic and extra-embryonic tissues fromdevelopmentally retarded and disorganized XAA2∆1.7X/XAA2∆1.7Y embryos. However, given the fact that the paternaltransmission did not compromise embryonic development andTsix is maternally expressed in the placenta, it seemedreasonable to assume that those cells showing aberrantaccumulation of XistRNA were derivatives of the extra-embryonic tissues. These results, therefore, suggest thatmaternal inheritance of XAA2∆1.7 may lead to aberrant X-inactivation in both male and female embryos.

Genetic complementation between the maternalXAA2∆1.7 and the paternal X ∆Xist

A possible role of Tsixtranscript is to maintain stable silencingof the maternal Xistto ensure inactivation of only the paternal

T. Sado and others

Fig. 6.Accumulation of ectopicXist RNA in E7.5 embryos uponmaternal transmission ofTsixAA2∆1.7. (A,C,E,G) RNAFISH with a Cy3-labeled probeprepared from XistcDNAclones. (A,C) Accumulation ofectopic XistRNA in a nucleus inmale (A) and female (C)embryos. (E,G) Coating of Xchromosome with XistRNA onmetaphase spreads in male (E)and female (G) embryos.(B,D,F,H) Either X or Ychromosome painting or both.(B) Y chromosome painting onthe nucleus in (A). (D) X chromosome painting on the nucleus in (C). (F) X and Y chromosome painting of the metaphase spread in (E). (H) Xchromosome painting on the mataphase spread in (G).


X chromosome in the extra-embryonic tissues. In femaleembryos with maternal inheritance of XAA2∆1.7, the normallysilent maternal Xistbecame transcriptionally active, leading toinactivation of both X chromosomes in the extra-embryoniccells and embryonic lethality. Conversely, it has been shownthat paternal inheritance of a disrupted Xist gene (X∆Xist)precludes inactivation of the paternal X, leading to two activeX chromosomes in the extra-embryonic tissues and earlyembryonic lethality (Marahrens et al., 1997). Based on theseresults, we predicted that female embryos with a maternalXAA2∆1.7 and a paternal X∆Xist might survive as they wouldhave an active paternal X chromosome and an inactivematernal X chromosome in the extra-embryonic tissues,opposite to the imprinted X inactivation in wild-type embryos.

To test this hypothesis, we crossed XAA2∆1.7X mice withX∆XistY mice. From a total of nine litters of mice, we recoveredfour female mice and 22 male mice at birth (Table 2). Aspredicted, all male mice were found to be wild type and thefour female mice were XAA2∆1.7X∆Xist by Southern blotanalysis (data not shown). No viable XAA2∆1.7X and XX∆Xist

mice were found. This result suggested that not allXAA2∆1.7X∆Xist mice were rescued and survived to birth, someprobably died in utero. We then dissected four litters ofembryos at E8.5 and E9.5 and found that of 13 XAA2∆1.7X∆Xist

female embryos, seven were significantly smaller than theirwild-type male littermates while the remaining six embryoswere grossly normal in size and morphology(Fig. 7A,B; Table 2). However, the smallerembryos apparently underwent gastrulationnormally, as they showed morphologycomparable with normal E8.5 embryos (Fig.7B). From the same litters, nine XAA2∆1.7Xembryos recovered were severely impaired,lacking discernible embryonic tissues (Fig.7D) (except one growth retarded embryoshown in Fig. 7C). The trophoblast tissueswere present, but much smaller than thoseseen in normal embryos (Fig. 7D).Interestingly, no XX∆Xist embryos wererecovered from these litters, indicating thatXX∆Xist embryos died at earlier stages thanXAA2∆1.7X embryos. These results indicatethat most XAA2∆1.7X∆Xist mice are partiallyrescued when compared with their littermatesXAA2∆1.7X and XX∆Xist, and some arecompletely viable and normal.

DISCUSSION

In this study, we identified processedantisense transcripts from the Xistlocus inplacenta and ES cells, which are subjectto complex alternative splicing. Thesetranscripts coincide with the Tsixtranscriptidentified by Lee et al. (Lee et al., 1999) bymeans of RNA FISH with strand-specificprobes and strand-specific RT-PCR on EScells. Tsix was described as a continuous 40kb RNA without introns, whose transcriptionstarted at 15 kb downstream of the Xist gene

(Lee et al., 1999). Our northern analysis, however, indicatesthat Tsix is, at least in part, subject to processing that gives riseto the major transcripts of 4.3 kb and 2.7 kb in length. Weshowed that Tsixexpression is first detected in the blastocystand is imprinted with the maternal allele expressed and thepaternal allele silent, opposite to the imprinted expression ofXist. Targeted disruption of Tsixin two different regions in thisstudy revealed that Tsixhad two putative promoters, locatedupstream of exons 1 and 2, and the majority of the transcriptswere derived from the latter. Although deletion of exon 1(TsixSS1∆2.7) did not appear to have discernible effect on X-

Table 2. Offspring from intercrosses between XAA2∆1.7Xand XXist∆Y mice

Genotype

Stage XY XAA2∆1.7XXist∆ XAA2∆1.7Y XX Xist∆ Early aborted

P1-20 22 4(1) 0 0 -E9.5 7 9(5) 8(8) 0 6E8.5 2 4(2) 1(1) 1 3

The numbers in the parentheses indicate the number of abnormal embryosor pups. Of four XAA2∆1.7XXist∆ pups, one was growth retarded and died at 3weeks of age. One of the XAA2∆1.7Y embryos developed to the somite stage,while the others died during gastrulation. One XXXist∆ embryo died beforegastrulation. The early aborted decidua contained no embryonic tissue or yolksac membrane, presumably accounting for the missing XXXist∆ embryos.

Fig. 7. Gross morphology of E9.5 embryos from XAA2∆1.7X mice mated with X∆XistYmice. (A) A normal wild-type mouse. (B) A partially rescued XAA2∆1.7X∆Xist embryo thatunderwent relatively normal gastrulation though growth retarded. (C) In rare events, anXAA2∆1.7X embryo developed to the early somite stage. (D) In most cases, the XAA2∆1.7Xembryos were growth arrested prior to gastrula. Note no XX∆Xist embryos were recoveredfrom these crosses (Table 2), presumably owing to poor development of their trophoblastand early lethality. Al, allantois; em, embryo; h, heart; tb, trophoblast; ys, yolk sac.

1284

inactivation and embryonic development, disruption oftranscripts from both promoters by insertion of a promoterlessIRESβgeo cassette in exon 2 (TsixAA2∆1.7) perturbed bothrandom X-inactivation in the embryonic tissues and theimprinted X-inactivation in the extra-embryonic tissues.

The most striking finding in our study is that maternaltransmission of XAA2∆1.7 resulted in embryonic lethality in bothsexes, while paternal transmission of XAA2∆1.7had no deleteriouseffect on embryonic development. AsTsix is exclusivelyexpressed from the maternal allele in blastocysts and the placenta,our result provides direct evidence that the expression of thematernal Tsixallele is essential for embryonic development. Theloss of Tsixexpression from Xm in the extra-embryonic lineagesappeared to have caused ectopic accumulation of Xist RNA in asubset of cells in XAA2∆1.7X/XAA2∆1.7Y embryos (Fig. 6), whichprobably led to inactivation of both X chromosomes in femalesand a single X chromosome in males after all, resulting in celldeath due to functional nullisomy for the X chromosome in theextra-embryonic tissues. Since Tsixand Xistare reciprocallyimprinted in the extra-embryonic tissues with respect to theparental origin of the X chromosome, our results suggest thatimprinted expression of Tsix in the extra-embryonic lineagesrenders Xm extremely resistant to X-inactivation in theseparticular lineages. Although the affected embryos were severelyimpaired in postimplantation growth, the primitive germ layertissues such as trophoblast and primitive endoderm, andembryonic ectoderm and mesoderm were formed. It is, therefore,unlikely that differentiation into the trophectoderm in a blastocystimmediately induces the ectopic inactivation of Xm carryingTsixAA2∆1.7 in XAA2∆1.7X/XAA2∆1.7Y embryos. As embryos thatfail to inactivate all but one X chromosome barely form the extra-embryonic tissues and mesoderm (Shao and Takagi, 1990; Gotoand Takagi, 1998), the presence of these tissues in XAA2∆1.7Xembryos suggests that imprinted paternal X-inactivation shouldhave occurred at the appropriate time in XAA2∆1.7X embryos.

In contrast to the imprinted X-inactivation in the extra-embryonic tissues, X-inactivation is random in the embryonictissues, probably owing to erasure of imprinting of both Xistand Tsix in the epiblast. Both alleles of Xist and Tsixmayexhibit basal levels of expression in the epiblast as seen inundifferentiated ES cells (Lee et al., 1999; Debrand et al.,1999). It is possible that once multipotential epiblast cellsbegin to differentiate, one of the two Tsixalleles is repressedin a random fashion, perhaps, by de novo methylation, whichwould then allow accumulation of stable Xist RNA in cis,eventually leading to X-inactivation. On the other Xchromosome, as in differentiating ES cells (Lee et al., 1999;Debrand et al., 1999), expression of Tsix is sustained for awhile after differentiation to suppress the production of stableXist RNA. We have shown that elimination of Tsix RNAwithout perturbation of the counting mechanism causesnonrandom X-inactivation, rendering the XAA2∆1.7

chromosome uniformly inactive in female embryos (Fig. 5).Once X-inactivation is established by the epigeneticmodification such as DNA methylation and histonedeacetylation, Tsixmay not be required anymore for repressionof Xist in cis and eventually disappears. In other words, Tsixmay be important only for the period in which differentialexpression of Xistis established during the reversible phase ofX-inactivation (Wutz and Jaenisch, 2000).

Genetic complementation between mutations of the two

reciprocally imprinted genes, Xistand Tsix, supports themodel that coordinated expression of Xist and Tsixregulatesimprinted X-inactivation in the extra-embryonic tissues.Disruption of imprinted X-inactivation is detrimental to thedevelopment of the extra-embryonic tissues. XXXist∆ embryoswith two active X chromosomes are characterized by extremelypoor formation of the extra-embryonic tissues (Marahrens etal., 1997). Interestingly, the extra-embryonic tissues andmesoderm are formed in the XAA2∆1.7X embryos, allowingthem to survive longer than XX∆Xist embryos. It is probablethat Xp is inactivated normally in XAA2∆1.7X embryos, and thatsubsequent activation of the maternal Xist gene leads todelayed inactivation of Xm. In rare cases, the maternal Xistmaybe activated in blastocysts before Xp undergoing inactivation,allowing Xp to respond to counting and to reverse imprintedX-inactivation (see below). Importantly, most XAA2∆1.7X∆Xist

embryos were partially rescued to the early somite stage (somedeveloped to adult). As ectopic inactivation of Xm takes placelater than imprinted inactivation of Xp, the successful rescueof females carrying the maternal XAA2∆1.7 and paternal X∆Xist

would depend on the proportion of extra-embryonic cellscarrying the inactivated Xm before the detrimental effect of twoactive X chromosome becomes severe.

Lee has recently described the effect of Tsixdeletion(Tsix∆CpG) on X-inactivation and embryonic development(Lee, 2000). She showed that maternal transmission ofTsix∆CpG resulted in ectopic X-inactivation in male embryosand inactivation of both X chromosomes in female embryos,largely in agreement with our findings described in this study.The major differences between the two studies lie in theconstruction of the mutant Tsixallele and the severity of itseffect on embryonic development. Lee found that about 18%of mice with maternal transmission of Tsix∆CpG survived toterm. She proposed that imprinting of X-inactivation was notabsolute, which allowed (though less frequent), inactivation ofXm, leaving Xp active in the extra-embryonic lineages (whichwould explain the infrequent survivors carrying the maternallyinherited Tsix∆CpG). However, this assumption does notaccommodate with complete lethality of female embryos withpaternally derived X∆Xist (Marahrens et al., 1997). In contrast,we recovered only two pups who carried maternallytransmitted XAA2∆1.7 among 102 wild-type littermates (2%).We speculate that the minor transcript derived from theupstream promoter, which is produced by alternative splicingconnecting exon 1 directly to exon4 (like 5′ RACE clone1b) ora similar event, might be responsible for maternal transmissionof Tsix∆CpG and TsixAA2D1.7 to offspring. Frequency ofsurvivors with each mutation may represent the efficiency toperturb the production of such minor transcripts.

Unlike XAA2∆1.7Y embryos, which die during earlydevelopment, XpO embryos are not lost during gestation in anuterine environment of XX mothers, although developmentalretardation is a common feature among them (Tada et al., 1993;Jamieson et al., 1998). Based on the different behavior of thesingle X chromosome in XAA2∆1.7Y and XpO embryos, despitethe common absence of Tsixexpression, we propose thatXp, but not Xm, is capable of responding to the countingmechanism in the extra-embryonic tissues. In normal femaleembryos, the counting mechanism senses the presence of twoX chromosomes in a cell and allows Xp to keep expressing Xistto initiate X-inactivation (Fig. 8). In XpO embryos, it senses

T. Sado and others


the presence of only one X chromosome and extinguishes Xistexpression to prevent inactivation of a single X chromosome(Fig. 8). However, when the dormant maternal Xist allelebecomes transcritionally activated in XAA2∆1.7Y, because Xm isinsensitive to the counting, it cannot shut off ectopic Xistexpression, leading to inappropriate X-inactivation (Fig. 8).Although similar event also occurs predominantly inXAA2∆1.7X (causing inactivation of both X chromosomes; Fig.8), if activation of the maternal Xisttakes place early enoughto allow the counting mechanism to switch off Xist expressionfrom Xp, such cells would manage to achieve dosagecompensation and contribute to the extra-embryonic tissues asseen in Fig. 5. Okamoto et al. have recently described properdosage compensation in the extra-embryonic tissues ofXpXp/XpY androgenones (Okamoto et al., 2000), supportingthe notion that Xp is responsive to the counting mechanism thatis operative in the extra-embryonic lineages. Conversely, thefailure of Xm to be inactivated in the extra-embryonic tissuesof the embryos with supernumerary Xm (Shao and Takagi,1990; Goto and Takagi, 1998) or with Xist-deficient Xp(Marahrens et al., 1997) is not due to the lack of counting butto the differential sensitivity between Xm and Xp to it.

We have demonstrated that although Tsix has not yetbeen transcribed at eight-cell stage embryos, they maintainimprinted Xist expression, keeping the maternal allele silent,which suggests that Xist expression is regulated by somemechanism other than Tsixbefore blastocyst formation. Mostprobably, however, Tsixexpression stabilizes imprintedexpression of Xistafter initiation of X-inactivation in thetrophectoderm. The epigenetic mechanism underlying the

differences between Xm and Xp is not fully understood.Differential methylation of Xist 5′ region may be responsiblein part for the imprinted expression of Xist. However, Whatcontrols imprinted expression of Tsix remains to bedetermined. We showed that exon 2 is associated with a regiondifferentially methylated between the active and inactive Xchromosomes (Courtier et al., 1995), which is partly includedin the deleted region in the TsixAA2∆1.7 allele. The deletion inTsixAA2∆1.7 did not appear to block Tsix transcription per se asthe IRESβgeo cassette was expressed in male ES cells –detected by northern analysis (data not shown) – but it didblock Tsix transcripts running across the Xist gene, which isprobably responsible for the loss of Tsixfunction. Furthermore,the role of exon 1 in regulation of Tsixexpression and functionis unclear. The detailed analysis of the genomic regionspanning exons 1 and 2 will shed light on our understandingof still complicated mechanism of X-inactivation.

We thank York Marahrens and Rudolf Jaenisch for kindly providingus with the Xistknockout mice, and Seong-Seng Tan for the XH253

transgenic mice. We also thank Hong Lei for excellent technicalassistance, Hideyuki Beppu for helping with dissection of mouseembryos and Nobuo Takagi for valuable discussions. This work wassupported by grants from NIH (GM52106 and CA82389) to E. L. andfrom the Ministry of Education, Science, Sports and Culture of Japanto H. S. and T. S.

REFERENCES

Borsani, G., Tonolorenzi, R., Simmler, M. C., Dandolo, L., Arnaud, D.,Capra, V., Grompe, M., Pizzuti, A., Muzny, D., Lawrence, C. et al.(1991). Characterization of a mouse gene expressed from the inactive Xchromosome. Nature351, 329-331.

Brockdorff, N., Ashworth, A., Kay, G. F., Cooper, P., Smith, S., McCabe,V. M., Norris, D. P., Penny, G. D., Patel, D. and Rastan, S. (1991).Conservation of position and exclusive expression of mouse Xist from theinactive X chromosome. Nature351, 329-331.

Brown, C. J., Ballabio, A., Rupert J. L., Lafreniere, R. G., Grompe, M.,Tonlorenzi, R. and Willard, H. F. (1991). A gene from the region of thehuman X inactivation centre is expressed exclusively from the inactive Xchromosome. Nature349, 38-44.

Cattanach, B. M. and Williams, C. E. (1972). Evidence of non-random X-chromosome activity in the mouse. Genet. Res.15, 189-195.

Chomczynski, P. and Sacchi, N. (1987). Single-step method of RNA isolatedby acid guanidium thiocyanate-phenol-chloroform extraction. Anal.Biochem.162, 156-159.

Clerc, P. and Avner, P. (1998). Role of the region 3′ to Xist exon6 in thecounting process of X-chromosome inactivation. Nat. Genet.19, 249-253.

Courtier, B., Heard, E. and Avner, P. (1995). Xcehaplotypes show modifiedmethylation in a region of the active X chromosome lying 3′to Xist. Proc.Natl. Acad. Sci. USA92, 3531-3535.

Debrand, E., Chureau, C., Arnaud, D., Avner, P. and Heard, E. (1999).Functional analysis of the DXPas34locus, a 3′ regulator of Xistexpression.Mol. Cell. Biol.19, 8513-8525.

Goto, Y. and Takagi, N. (1998). Tetraploid embryos rescue embryoniclethality caused by an additional maternally inherited X chromosome in themouse. Development125, 3353-3363.

Heard, E., Mongelard, F., Arnaud, D. and Avner, P. (1999). Xistyeastartificial chromosome transgenes function as X-inactivation centers only inmulticopy arrays and not as single copies. Mol. Cell. Biol.19, 3156-3166.

Herzing, L. B. K., Romer, J. T., Horn, J. M. and Ashworth, A. (1997). Xisthas properties of the X-chromosome inactivation centre. Nature 386, 271-275.

Hong, Y.-K., Ontiveros, S. D., Chen, C. and Strauss, W. M. (1999). A newstructure for the murine Xistgene and its relationship to chromosomechoice/counting during X-chromosome inactivation. Proc. Natl. Acad. Sci.USA96, 6829-6834.

Kay, G. F., Penny, G. D., Patel, D., Ashworth, A., Brockdorff, N. and

Fig. 8.A model of Tsix-mediated regulation of Xistsilencing in cisand imprinted X-inactivation in the extra-embryonic lineages. Xistisexpressed from the Xp by default in preimplantation embryos.Paternal expression of Xistfrom counting-sensitive Xp can be turnedoff according to the number of X chromosome which an individualcell senses. Elimination of TsixRNA on Xm allows ectopicexpression of Xistin cisafter blastocyst formation. As Xm isinsensitive to the counting, even though Xp has undergone X-inactivation, Xm is unable to shut off ectopic expression of maternalXist, leading to delayed X-inactivation in the extra-embryoniclineages.

Xist (on)

Tsix (on)

Xist (on) Xist (off)

Xist (on)

Tsix (off)

no Xist

Tsix (off)

no Xist

Tsix (on)

no Xist

no Tsix

no Xist

no TsixXist (on)

XmXp

XpO

XmXp∆Xist

Xm XpAA2∆1.7 ∆Xist

Inactivated

Active

Active

No Inactivation

Active

No Inactivation

Delayed Ectopic Inactivatioin

Xist (on)

Xist (on)

Xist (on)

Xist (off)

XpXpInactivated

Active

no Tsix

Xist (on) Xist (on)

no TsixXist (on)

Xm XpAA2∆1.7 Inactivated

Delayed Ectopic Inactivatioin

no Tsix

Xm Y AA2∆1.7

no TsixXist (on)

Delayed Ectopic InactivatioinXist (off)

Xist (off)

Xist (off)

Xist (off) Xist (off)

Xist (off) Xist (off)

Preimplantation embryos(before Blastocyst Formation) Extraembryonic Lineages

1286

Rastan, S. (1993). Expression of Xist during mouse development suggestsa role in the initiation of X chromosome inactivation. Cell 72, 171-182.

Kay, G. F., Barton, S., Surani, M. A. and Rastan, S. (1994). Imprinting andX chromosome counting mechanisms determine Xist expression in earlymouse development. Cell77, 639-650.

Jamieson, R. V., Tan, S.-S. and Tam, P. P. L. (1998). Retardedpostimplantation development of XO mouse embryos: impact of the parentalorigin of the monosomic X chromosome. Dev. Biol.201, 13-25.

Lawrence, J. B., Singer, R. H. and Marselle, L. M. (1989). Highly localizedtracks of specific transcripts within interphase nuclei visualized by in situhybridization. Cell 57, 493-502.

Lee, J. T., Strauss, W. M., Dausman, J. A. and Jaenisch, R. (1996). A 450kb transgene displays properties of the mammalian X-inactivation center.Cell 86, 83-94.

Lee, J. T. and Jaenisch, R. (1997). Long-range cis effects of ectopic X-inactivation centres on a mouse autosome. Nature386, 275-279

Lee, J. T. and Lu, N. (1999). Targeted mutagenesis of Tsixleads to nonrandomX inactivation. Cell99, 47-57.

Lee, J. T., Davidow, L. S. and Warshawsky, D. (1999). Tsix, a gene antisenseto Xist at the X-inactivation centre. Nat. Genet.21, 400-404.

Lee, J. T. (2000). Disruption of imprinted X inactivation by parent-of-origineffects at Tsix. Cell103, 17-27.

Li, E., Bestor, T. H. and Jaenisch, R. (1992). Targeted mutation of the DNAmethyltransferase gene results in embryonic lethality. Cell69, 915-926.

Lyon, M. F. (1961). Gene action in the X-chromosome of the mouse (Musmusculus L). Nature190, 372-373.

Lyon, M. F. and Rastan, S. (1984). Paternal source of chromosome imprintingand its relevance for X chromosome inactivation. Differentiation26, 63-67.

Marahrens, Y., Panning, b., Duasman, J., Strauss, W. and Jaenisch, R.(1997). Xist-deficient mice are defective in dosage compensation but notspermatogenesis. Genes Dev.11, 156-166.

Mountford, P., Zenvnik, B., Düwel, A., Nichols, J., Li, M., Dani, C.,Robertson, M., Chambers, I. and Smith, A. (1994). Dicistronic targetingconstructs: reporters and modifiers of mammalian gene expression. Proc.Natl. Acad. Sci. USA91, 4303-4307.

Moore, T., Costancia, M., Zubair, M., Bailleul, B., Feil, R., Sasaki, H. andReik, W. (1997). Multiple imprinted sense and antisense transcripts,differential methylation and tandem repeats in a putative imprinting controlregion upstream of mouse Igf2. Proc. Natl. Acad. Sci. USA94, 12509-12514.

Okamoto, I. Tan, S.-S. and Takagi, N. (2000). X-chromosome inactivationin XX androgenetic mouse embryos surviving implantation. Development127, 4137-4145.

Panning, B. and Jaenisch, R. (1996). DNA hypomethylation can activate Xistexpression and silence X-linked genes. Genes Dev.10, 1991-2002.

Panning, B., Dausman, J. and Jaenisch, R. (1997). X chromosomeinactivation is mediated by XistRNA stabilization. Cell90, 907-916.

Penny, G. D., Kay, G. F., Sheardown, S. A., Rastan, S. and Brockdorff, N.(1996). Requirement for Xist in X chromosome inactivation. Nature 379,131-137.

Russell, L. B. and Montgomery, C. S. (1965). The use of X-autosometranslocations in locating the X-chromosome inactivation center. Genetics52, 470-471.

Rougeulle, C., Cardoso, C., Fontes, M., Colleaux, L. and Lalande, M.(1998). An imprinted antisense RNA overlapsUBE3A and a secondmaternally expressed transcript. Nat. Genet.19, 15-16.

Sado, T., Tada, T. and Takagi, N. (1996). Mosaic methylation of Xist genebefore chromosome inactivation in undifferentiated female mouseembryonic stem and embryonic germ cells. Dev. Dyn.205, 421-434.

Sado, T., Fenner, M. H., Tan, S.-S., Tam, P., Shioda, T. and Li, E. (2000).X-inactivation in the mouse embryo deficient for Dnmt1: Distinct effect ofhypomethylation on imprinted and random X-inactivation. Dev. Biol.255,294-303.

Shao, C. and Takagi, N. (1990). An extra maternally derived X chromosomeis deleterious to early mouse development. Development110, 969-975.

Sheardown, S. A., Duthie, S. M., Johnston, C. M., Newall, A. E. T.,Formstone, E. J., Arkell, R. M., Nesterova, T. B., Alghisi, G.-C., Rastan,S. and Brockdorff, N. (1997). Stabilization of XistRNA mediates initiationof X chromosome inactivation. Cell91, 99-107.

Simmler, M.-C., Cattanach, B. M., Rasberry, C., Rougeulle, C. and Avner,P. (1993). Mapping the murine Xce locus with (CA)n repeats. Mamm.Genome4, 523-530

Simmer, M.-C., Cunningham, D. B., Clerc, P., Vermat, T., Caudron, B.,Cruaud, C., Pawlac, A., Szpirer, C., Weissenbach, J., Claverie, J.-M. andAvner, P. (1996). A 94 kb genomic sequence 3′ to the murine Xistgenereveals an AT rich region containing a new tistis specific gene Tsx. Hum.Mol. Genet.5, 1713-1726.

Tada, T., Takagi, N. and Adler, I.-D. (1993). Parental imprinting on themouse X chromosome: effects on the early development of XO, XXY andXXX embryos.Genet. Res.62, 139-148.

Takagi, N. and Sasaki, M. (1975). Preferential inactivation of the paternallyderived X chromosome in the extraembryonic membranes of the mouse.Nature256, 640-642.

Takagi, N., Sugawara, O. and Sasaki, M. (1982). Regional and temporalchanges in the pattern of X-chromosome replication during the early post-implantation development of the female mouse. Chromosoma85, 275-286.

Tam, P. P. L., Williams, E. A. and Tan, S.-S. (1994). Expression of an X-linked HMG-lacZtransgene in mouse embryos: implication of chromosomeimprinting and lineage-specific X-chromosome activity. Dev. Genet.15,491-503.

Tan, S.-S. Williams, E. A. and Tam, P. P. L. (1993). X-chromosomeinactivation occurs at different times in different tissues of the post-implantation mouse embryo. Nat. Genet.3, 170-174.

Wutz., A., Smrzka, O. W., Schweifer, N., Schellander, K., Wagner, E. F.and Barlow, D. P. (1997). Imprinted expression of the Igf2r gene dependson an intronic CpG island. Nature389, 745-749.

Wutz, A. and Jaenisch, R. (2000). A shift from reversible to irreversible Xinactivation is triggered during ES cell differentiation. Mol. Cell5, 695-705.

T. Sado and others

regulation of imprinted x-chromosome inactivation in mice by tsix

Documents