Two means of transcriptional reactivation within ... - Wiley Online Library

The Plant Journal (2003) 33, 743–749

Two means of transcriptional reactivation within heterochromatin Aline V. Probst1,, Paul F. Fransz2, Jerzy Paszkowski1 and Ortrun Mittelsten Scheid1 1 Friedrich Miescher Institute for Biomedical Research, PO Box 2543, CH-4002 Basel, Switzerland, and 2 Swammerdam Institute for Life Sciences, University of Amsterdam, Kruislaan 318, 1098 SM Amsterdam, the Netherlands Received 1 October 2002; revised 19 November 2002; accepted 27 November 2002. For correspondence (fax þ41 61 697 39 76; e-mail [email protected]).

Summary DNA methylation levels and specific histone modifications of chromatin in interphase nuclei are taken as an indicator of transcriptional activity or silencing. Arabidopsis mutants impaired in maintenance of transcriptional gene silencing (TGS) alleviate TGS with or without affecting DNA methylation. Mutant ddm1, representing the first type, lacks a chromatin remodeling factor that regulates histone and DNA methylation. Mutant mom1, representing the second type, is affected in a different but still unknown silencing mechanism. Both classes of mutation have been studied mainly for their effects on specific loci. Here, we describe the cytological analysis of chromatin in ddm1 and mom1 mutants. The ddm1 mutation causes a striking decondensation of centromeric heterochromatin, a re-distribution of the remaining methylation of DNA, and a drastic change in the pattern of histone modification. A complex transgenic locus, which underwent stable inactivation and became heterochromatin-like, follows similar structural alterations. In contrast, nuclear organization in mom1 appears unaltered, demonstrating an involvement of MOM1 in transcriptional regulation within a heterochromatic environment. Keywords: gene silencing, heterochromatin, mom1, ddm1, histone modifications, nuclear organization.

Introduction Chromatin of interphase nuclei can be distinguished in decondensed, potentially active euchromatin, and condensed late-replicating heterochromatin (Goldman et al., 1984) with repressed transcription. In plants and mammals, heterochromatin contains hypermethylated DNA, packaged into tightly arranged nucleosomes with characteristically modified histones. Within heterochromatin, histone H3 is often methylated at Lysine 9 (Peters et al., 2001), whilst histone H4 is hypoacetylated (Belyaev et al., 1996; Jasencakova et al., 2000). Heterochromatin predominantly consists of centromeric and pericentromeric DNA and follows a certain spatial arrangement characteristic for the organism. In mammalian cells, centromeres are usually found at the nuclear periphery and in the vicinity of the nucleolus, but their positions can change during the cell cycle and are proposed to reflect the functional or developmental capacity of a specific cell (Haaf and Schmid, 1991). For example, in some species interphase chromosomes are polarized, with centromeres and telomeres at ß 2003 Blackwell Publishing Ltd

opposite poles of the nucleus (Rabl configuration; Rabl, 1885). This is characteristic for salivary gland cells of Drosophila (Hochstrasser et al., 1986) and some monocotyledons plants like wheat (Abranches et al., 1998). In contrast, in Arabidopsis thaliana, centromeric and pericentromeric DNA as well as rDNA repeats are organized in DAPI-bright chromocenters (Maluszynska and HeslopHarrison, 1991). Furthermore, localization studies of the silenced brown gene (bwD) in Drosophila (Csink and Henikoff, 1998) and of lymphocyte-specific genes inactivated during B-cell development (Brown et al., 1997), suggest that genes entering the proximity of centromeric heterochromatin can loose their transcriptional activity. Thus, centromeric heterochromatin may provide a foundation for functional compartments required for the repression of genes, and therefore spatial nuclear organization must be considered as an important factor influencing epigenetic states of gene expression. 743

744 Aline V. Probst et al. Characterization of Arabidopsis mutations interfering with the maintenance of transcriptional gene silencing (TGS) and causing a reduction in DNA methylation level has revealed a number of genes involved, such as DNA methyltransferases (Finnegan and Kovac, 2000; Lindroth et al., 2001), a histone methyltransferase (Jackson et al., 2002), and the DDM1 gene encoding a SWI2/SNF2-like chromatin remodeling factor (Jeddeloh et al., 1999; Vongs et al., 1993). In addition to reduced DNA methylation levels, the ddm1 mutant also has altered patterns of H3K9 methylation (Gendrel et al., 2002; Johnson et al., 2002). Because the status of H3 methylation influences general levels of DNA methylation in Neurospora (Tamaru and Selker, 2001) and methylation specifically at CpNpG sites in Arabidopsis (Jackson et al., 2002), DDM1 may control methylation directly and indirectly through changes in histone methylation (Gendrel et al., 2002; Johnson et al., 2002). In contrast to mutations involving DNA/histone modifications, a mutation in the MOM1 gene (Amedeo et al., 2000), encoding a large nuclear protein with limited homology to SWI2/ SNF2, does not affect DNA methylation. The mechanism of epigenetic regulation mediated by MOM1 still remains an enigma. Effects of mutations releasing TGS have been assessed mainly at selected target loci, and changes in DNA methylation and histone methylation have been revealed by Southern blotting and/or chromatin immunoprecipitation (Gendrel et al., 2002; Johnson et al., 2002; Vongs et al., 1993). However, the relation of these local effects to more general structural changes in chromatin and/or nuclear architecture is not well documented. Using cytological methods, we have visualized complex structural alterations of the entire nucleus and determined the chromosomal regions that are most strongly affected. We document chromatin changes associated with TGS release by the ddm1 mutation for a previously silent transgenic locus and heterochromatic chromosomal regions. Furthermore, we examined cytological consequences of the mom1 mutation, which releases silencing without any detectable influence on DNA methylation at the target genes.

Results and discussion Organization of endogenous and acquired heterochromatin in wild-type nuclei To assess possible global changes in chromatin organization associated with TGS release by the ddm1 and mom1 mutations, we first examined nuclei of transgenic Arabidopsis line A, the genetic background of the mom1 and ddm1-5 (former som8) mutant alleles studied here (Amedeo et al., 2000; Mittelsten Scheid et al., 1998). Line A has a transgenic insert composed of approximately 15

re-arranged plasmid copies containing the hygromycin phosphotransferase (HPT) gene under the control of the CaMV 35S promoter. The transcriptional activity of the HPT locus was lost in the progeny of the primary transgenic plant. The locus became hypermethylated and progressively silenced upon further inbreeding (Mittelsten Scheid et al., 1991, 1998). We combined DAPI staining of DNA and fluorescent in situ hybridization (FISH) with the following probes: pAL1 (180-bp repeat), which labels the core centromeric regions of all chromosomes (Haupt et al., 2001; Murata et al., 1994), transcriptionally silent information (TSI) specific to pericentromeric repeats (Steimer et al., 2000), ribosomal DNA, and a telomeric probe (Weiss and Scherthan, 2002). The bright DAPI-stained heterochromatic chromocenters containing pAL1 sequences (Figure 1a–c) predominantly aligned at the nuclear periphery. Those chromocenters that contain the nucleolus organizer regions (NORs), were found close to the nucleolus (Figure 2c) and telomeres were clustered around this subnuclear compartment (Armstrong et al., 2001; Figures 1b and 2c). Such an arrangement is characteristic for Arabidopsis, but differs from monocots like wheat (Abranches et al., 1998) or from Saccharomyces cerevisiae, where the telomeres are clustered at the nuclear periphery (Hediger and Gasser, 2002). Although the images of line A and the corresponding wild type (Figure 1) were very similar, we detected one or two additional DAPI-bright spots in nuclei of line A plants, hemizygous or homozygous for the transgenic locus respectively (Figure 1c). To investigate whether they reflected a chromatin condensation acquired de novo by the silent transgenic locus, we combined DAPI staining with twocolor FISH using HPT and pAL1 as probes. The HPT signals indeed co-localized with the two neo-chromocenters (Figure 1c). This implies that the insertion of a complex transgene and its consecutive silencing triggered the formation of an additional, neo-heterochromatic knob. It has been observed that formation of repressive chromatin at silenced transgenes is enhanced by neighboring heterochromatic regions (Dorer and Henikoff, 1994). However, the HPT locus seems to be located distant to the pre-existing heterochromatin in interphase nuclei (Figures 1c and 2a) and on chromosome spreads (Figure 1d). Therefore, the HPT silencing and the heterochromatinization were acquired and maintained autonomously. In addition to reactivation of transgenic loci, TGS mutations cause transcriptional activation of a specific class of silent endogenous repeats, TSI. These repeats are located in pericentromeric regions of all chromosomes, as visualized on pachytene chromosomes and confirmed by localization on the Arabidopsis physical map (Steimer et al., 2000). FISH of wild-type and line A nuclei with the TSIspecific probe showed these sequences to be associated with the chromocenters also in interphase nuclei, in close vicinity to the centromeric repeats (Figure 2b). ß Blackwell Publishing Ltd, The Plant Journal, (2003), 33, 743–749

Transcriptional reactivation within heterochromatin Organization of endogenous and acquired heterochromatin in mutant nuclei As the overall structure of the nucleus of the transgenic line A was identical to the wild type except for the additional heterochromatic knob, we examined the consequences of ddm1 and mom1 mutations on the nuclear organization and structure of the reactivated transgenic locus using mutant alleles in a background isogenic to line A. We chose a previously described mom1 T-DNA insertion allele (Amedeo et al., 2000) and the likely null allele ddm1-5 (former som8) with an insertion of 82 bp in the second exon, which should result in premature termination after 30 codons (Jeddeloh et al., 1999). DAPI staining of the ddm1-5 nuclei revealed a striking decondensation of chromocenters. The number of clearly stained chromocenters was reduced and those visible were considerably smaller than in line A (Figure 2d–f). FISH analysis of centromeric repeats revealed their expansion in the form of loops into territory usually occupied by euchromatin (Figure 2d). The neo-heterochromatin associated with the transgenic HPT locus was likewise decondensed (Figure 2d). Moreover, the TSI templates lost the tight association with the centromeric DNA. This was clearly visible also in nuclei that still had remnants of chromocenters (Figure 2e), suggesting that decondensation and looping out of the pericentromeric repeats precedes a similar process occurring at centromeres, and results in disappearance of chromocenters. Interestingly, telomeres were largely unaffected and remained clustered around the nucleolus, and the rDNA was associated with remnants of chromocenters where these were still present (Figure 2f). It has been suggested previously that the ddm1 mutation predominantly influences DNA and H3K9 methylation of centromeric regions (Johnson et al., 2002). This is in accordance with our observation of prevailing structural changes of these regions. Importantly, both processes are not directly linked to transcriptional activation, since transcripts of 180-bp repeats were absent even in ddm1 mutants (Gendrel et al., 2002; Johnson et al., 2002). In contrast to ddm1, the nuclear structure of the mom1 mutant in general, and of the pericentromeric heterochromatin in particular, was not altered (Figure 2g–i). In addition, and despite reactivation of the HPT gene, the heterochromatic nature of the transgenic locus was also retained (Figure 2g). The localization of the transcriptionally activated TSI DNA in relation to the core centromeric 180-bp repeats and the chromocenters also remained unchanged: TSI repeats were still closely associated with centromeric sequences and were part of the DAPI-bright chromocenters (Figure 2h). Telomeres were also located around the nucleolus, and the structure of the rDNA repeats was not altered (Figure 2i). Thus, in contrast to ddm1-5, release of transcriptional silencing of TSI and HPT in the mom1 mutant is not connected with ß Blackwell Publishing Ltd, The Plant Journal, (2003), 33, 743–749

745

detectable structural changes of associated heterochromatin. The inactivation of B-cell-specific genes during development is linked to a change in nuclear organization involving an enhanced centromeric association (Brown et al., 1997). We hypothesized that a similar mechanism, requiring MOM1 protein, might be necessary to maintain the silenced state in line A. For this reason, we compared the frequency of centromeric association of the HPT locus in approximately 300 nuclei of line A and mom1. However, no significant differences could be found (data not shown). Nevertheless, it is still possible that MOM1 contributes to the relocation of the silent transgenic loci to pre-existing heterochromatin or has a part in positioning silent loci into the vicinity of a distinct nuclear domain, e.g. the nuclear periphery or the nucleolus. This would be detectable only in nuclei with a conserved three-dimensional structure, but not revealed in nuclear spreads examined here. Distribution of DNA methylation In plants and mammals, heterochromatic DNA contains a significant amount of 5-methylcytosine (Johnson et al., 2002). This modification is believed to regulate transcriptional repression via the interaction with 5mC-binding proteins; that, in turn, recruit factors responsible for modification of chromatin structures (Bird, 2002). ddm1 mutants are characterized by a substantial hypomethylation of the repetitive pericentromeric sequences and of the reactivated transgenic loci (Vongs et al., 1993). Surprisingly, individual genes sometimes acquire more methylation in the ddm1 mutant background. This suggests that, in addition to the global reduction of methylation levels, DDM1 also influences the chromosomal distribution of the remaining methylation. By immunolocalization of modified cytosines, the DNA in chromocenters in line A was found to be hypermethylated, while the remaining DNA showed low methylation (Figure 3a). The additional small heterochromatic knobs associated with the transgenic loci were also hypermethylated(Figure 3a,arrowheads).Inddm1-5nuclei,methylation signals were reduced and rather uniformly distributed (Figure 3b). The methylation changes visualized by immunostainingwereclearlycorrelatedwiththedecondensationof the centromeric heterochromatin and its relocation away from the chromocenter remnants (Figure 3b). Although the mom1 mutation does not alter the methylation of reactivated target sequences like HPT (Amedeo et al., 2000) or TSI (Steimer et al., 2000) or modify methylation of centromeric repeats, it is still possible that other chromosomal regions suffer methylation changes. In clear contrast to the distorted methylation patterns in the ddm1-5 mutant, immunostaining revealed no changes in methylation level or its distribution in mom1 (Figure 3c), corroborating previous observations (Amedeo et al., 2000).

746 Aline V. Probst et al. Distribution of histone modifications As the loss of the MOM1 protein affected neither global organization of heterochromatin nor DNA methylation, we examined particular modifications of histones associated with transcriptional activity or silencing. Transcriptionally active euchromatin is marked by acetylated histones and silent heterochromatin by de-acetylated histones, respectively (Struhl, 1998). We analyzed H4 acetylation in line A and the two TGS mutants by immunostaining with an antibody directed against a tetra-acetylated isoform of H4. Euchromatin of line A was intensely stained (Figure 3d), while the chromocenters lacked any signal. In contrast, nuclei of ddm1 plants were almost evenly stained (Figure 3e). Since ddm1 nuclei have dispersed heterochromatic regions, these could appear intermingled with euchromatin that gives immunostaining signals. Alternatively, the loss of DDM1 might have an indirect influence on the distribution ofacetylated histonesand letthemassociate also with centromeric and pericentromeric DNA. In contrast to ddm1 and despite the release of silencing, the loss of MOM1 had no effect on the distribution of histone acetylation (Figure 3f). Whereas histone acetylation marks transcriptionally active euchromatin, methylation of histone H3 at position K9 denotes silent heterochromatin. We labeled nuclei of line A and the two TGS mutants with an antibody directed against di-methylated H3K9. Methylated histones in nuclei of line A were clustered at chromocenters while the remaining nucleus showed only a weak signal (Figure 3g). Histone methylation therefore appears to be a specific mark for constitutive heterochromatin in Arabidopsis thaliana. In contrast to line A, H3K9 methylation signals at the chromocenter remnants were dispersed and weak in ddm1 (Figure 3h). In the mom1 mutant, although TSI sequences as part of the chromocenters were transcribed, H3K9 methylation remained specifically localized to the chromocenters, as in line A (Figure 3i).

In contrast to H3K9met, the methylation of histone H3 at lysine 4 (H3K4 met) marks euchromatin (Jenuwein and Allis, 2001). Immunostaining of line A for H3K4met revealed a pattern similar to that of acetylated histones H4, namely uniform euchromatic staining and exclusion of the chromocenters and the nucleolus (Figure 3j). In ddm1, an increase in H3K4met, detected by chromatin immunoprecipitation, has been observed at selected transposons and single copy genes (Gendrel et al., 2002). A global enrichment in H3K4met could not be clearly visualized by immunostaining of the ddm1 nuclei (Figure 3k). Interestingly, the chromocenter remnants remained unlabeled. Therefore, H3K4met does not automatically replace H3K9met in these chromosomal regions. In addition to patterns of H3K9met, also H3K4 methylation remained unaltered in the mom1 mutant (Figure 3l). Heterochromatic transcription in mom1 Our studies revealed drastic changes in nuclear organization, DNA methylation, and histone modifications resulting from the ddm1 mutation. The observed changes in chromatin structure may be indirectly caused by the loss and redistribution of DNA methylation. Alternatively, loss of DDM1 protein and its chromatin remodeling role (Brzeski and Jerzmanowski, 2002) could prevent other chromatinmodifying enzymes from accessing their targets. Importantly, parallel observations of mom1 nuclei suggest that release of TGS can also be achieved without gross changes in chromatin structure, DNA methylation, or histone modifications. The results of recent epistatic analysis of the ddm1/mom1 relationship suggest that each of the mutations interferes with a distinct level of epigenetic regulation (Mittelsten Scheid et al., 2002). Therefore, mechanisms of epigenetic control, mediated by MOM1, may not involve the popular epigenetic tools examined here. However, there are examples of active genes residing within

Figure 1. Organization of heterochromatin and formation of new transgenic heterochromatin. Nuclear spreads of wild-type (ecotype Zu¨ rich) and transgenic line A, stained with DAPI (black and white in left panel, blue in merged image on the right panel), and hybridized with fluorescent probes for centromeric repeats (pAL1: (a–c), red), telomeric repeats (b, green), or HPT vector DNA ((c–d), green). The transgenic HPT locus forms an independent heterochromatic knob ((c–d), white arrowheads) that localizes apart from centromeric heterochromatin, as revealed in interphase nuclei (c) and on chromosome spreads of an early prophase (d). Figure 2. Organization of heterochromatin in line A and two mutants impaired in the maintenance of transcriptional gene silencing (TGS). Interphase nuclear spreads of transgenic line A, ddm1-5, and mom1-1, stained with DAPI (black and white in left panel, blue in merged image on the right panel) and hybridized with fluorescent probes for centromeric repeats (pAL1: (a–h), red); HPT vector DNA (a,d,g, green); pericentromeric TSI repeats (b,e,h, green); telomeric repeats (c,f,i, green); or rDNA repeats (c,f,i, red). Note that part of the rDNA in wild type as well as in the two mutant loops out of the chromocenter into the nucleolus. The transgenic locus is marked with white arrowheads in the left panel of (a) and (g). The ddm1-5 mutant had been backcrossed to wild type to promote recovery from the progressive developmental effects, and the illustrated donor plant for the nuclei, though homozygous for the mutation, is hemizygous for the transgenic locus. Figure 3. DNA and chromatin modifications in transgenic line A and mutants ddm1-5 and mom1-1. (a,b,c) Distribution of DNA methylation revealed by DAPI stain (blue) and FISH with probes for centromeric repeats (pAL1, red, left panel) and immunodetection with an antibody specific for 5-methylcytosine (green, right panel). The newly formed heterochromatic knob is marked by white arrowheads in (a) and (c). (d,e,f) Distribution of histone acetylation revealed by DAPI stain (blue, left panel) and immunodetection with an antibody specific for tetra-acetylated histones (green, middle panel). Right panels show merged images. (g–l) Distribution of histone methylation revealed by DAPI stain (blue, left panel) and immunodetection (green, middle panel) with an antibody specific for histone H3 di-methylated at lysine 9 (g,h,i) or histone H3 di-methylated at lysine 4 (j,k,l). Right panels show merged images. Images (d–l) show single layers selected from de-convoluted image stacks.

ß Blackwell Publishing Ltd, The Plant Journal, (2003), 33, 743–749

Transcriptional reactivation within heterochromatin

Figure 1.

Figure 2.

Figure 3.


747

748 Aline V. Probst et al. heterochromatin that resist inactivation (Carrel et al., 1999; Tanaka et al., 1983). Regulatory mechanisms for such genes are not well understood. Transcriptional activation in the mom1 mutant, occurring without obvious signs of conversion into euchromatin, resembles such ‘heterochromatic transcription’. Thus, MOM1 may contribute to the regulation of transcription in usually silent, heterochromatic regions of chromosomes.

ated H3 (1 : 500), or lysine9-di-methylated H3 (1 : 100) in blocking solution or 1% BSA in PBS (1 h, 378C or overnight at 48C). Detection was carried out with an anti-rabbit FITC-coupled antibody (Molecular Probes, 1 : 100, 378C, 40 min) in 0.5% BSA in PBS. DNA was counterstained with DAPI in Vectashield. Images were analyzed with a Deltavision Deconvolution Microscope. The WoRx software was applied for de-convolution of the image stacks, and single layers were chosen for illustration.

Acknowledgements Experimental procedures Plant material Plants were grown in soil in a growth chamber under short-day conditions (12 h light, 218C; 12 h dark, 168C).

Fluorescence in situ hybridization Young rosette leaves (1–1.5 cm) were fixed in 3 : 1 ethanol acetic acid and stored at 208C. Interphase nuclear spreads, as well as FISH and immunocytochemical detection were carried out as described (Fransz et al., 1998) with the following modifications: the pAL1 repeat was cloned into pBluescript vector and labeled probes were generated by PCR with 0.1 mM dATP, dCTP, dGTP; 0.065 mM dTTP; and 0.035 mM biotin-dUTP (Roche). TSI repeats were labeled with the help of the Dig-DNA labeling mix (Roche) by PCR using primers 5-GTTAATCCAAGTAGCTGACTCTCC-3 and 5TTTAACAACTAAGGTTCCTG-3. The amplified sequence corresponds to region 68442–68869 of BAC F7N22 (GenBank AF058825). Probes for the HPT locus were obtained by labeling the pGL2 plasmid (used to generate the transgenic strain) with the digoxygenin-dUTP nick translation kit (Roche), rDNA probes were obtained with the biotin nick translation kit (Roche) using 18S- and 25S-rDNA-containing plasmids. Telomere-specific labeling with digoxygenin was achieved in a primer extension PCR with 5TTTAGGG-3 and 5-CCCTAAA-3 oligonucleotides. Before hybridization, the tissue on the slides was treated with pepsin (10 mg ml1, 378C, 20 min). One microliter of the PCR reaction and/or 3 ml of the nick translation mix were added to 20 ml hybridization mix. After hybridization for about 15 h in a wet chamber, slides were washed for 5 min in 2 SSC, 5 min in 0.1 SSC, 3 min in 2 SSC at 428C, and 5 min in 2 SSC/0.1% Tween 20 at RT and for telomere detection, 2 5 min in 0.3 SSC, 3 min in 2 SSC at 428C, and 5 min in 2 SSC/0.1% Tween 20 at RT. For the simultaneous detection of cytosine methylation, the digoxygenin-detecting antibody was replaced by 5-methyl C-specific antibody (Eurogentec) in a 1 : 50 dilution in TNB. DNA was counterstained with DAPI in Vectashield (Vector Laboratories). Images were analyzed with a Leitz DMR Fluorescence Microscope and documented with a SPOT RT camera (Diagnostic Instruments). Images were merged and processed using Adobe Photoshop 6.0.

Immunostaining Protoplasts were isolated from young leaves of line A, mom1, and ddm1-5 plants and fixed according to http://stein.cshl.org/atir/biology/protocols/cshl-course/7-gene_expression.html. Following rehydration in PBS, slides were blocked in 2% BSA in PBS (30 min, 378C) and incubated with antibodies (all from Upstate Biotechnology) against tetra-acetylated H4 (dilution 1 : 100), lysine4-di-methyl-

We thank Dr Hanna Weiss for providing rDNA vectors and telomere primers, as well as Drs. Ingo Schubert, Patrick King, Fang-Lin Sun, and Lukas Landmann for critical reading of the manuscript.

References Abranches, R., Beven, A.F., Aragon-Alcaide, L. and Shaw, P.J. (1998) Transcription sites are not correlated with chromosome territories in wheat nuclei. J. Cell Biol. 143, 5–12. Amedeo, P., Habu, Y., Afsar, K., Mittelsten Scheid, O. and Paszkowski, J. (2000) Disruption of the plant gene MOM releases transcriptional silencing of methylated genes. Nature, 405, 203– 206. Armstrong, S.J., Franklin, F.C. and Jones, G.H. (2001) Nucleolusassociated telomere clustering and pairing precede meiotic chromosome synapsis in Arabidopsis thaliana. J. Cell Sci. 114, 4207–4217. Belyaev, N.D., Keohane, A.M. and Turner, B.M. (1996) Histone H4 acetylation and replication timing in Chinese hamster chromosomes. Exp. Cell Res. 225, 277–285. Bird, A. (2002) DNA methylation patterns and epigenetic memory. Genes Dev. 16, 6–21. Brown, K.E., Guest, S.S., Smale, S.T., Hahm, K., Merkenschlager, M. and Fisher, A.G. (1997) Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin. Cell, 91, 845–854. Brzeski, J. and Jerzmanowski, A. (2002) Deficient in DNA methylation 1 (DDM1) defines a novel family of chromatin remodeling factors. J. Biol. Chem. 278, 823–828. Carrel, L., Cottle, A.A., Goglin, K.C. and Willard, H.F. (1999) A firstgeneration X-inactivation profile of the human X chromosome. Proc. Natl. Acad. Sci. USA, 96, 14440–14444. Csink, A.K. and Henikoff, S. (1998) Large-scale chromosomal movements during interphase progression in Drosophila. J. Cell. Biol. 143, 13–22. Dorer, D.R. and Henikoff, S. (1994) Expansions of transgene repeats cause heterochromatin formation and gene silencing in Drosophila. Cell, 77, 993–1002. Finnegan, E.J. and Kovac, K.A. (2000) Plant DNA methyltransferases. Plant Mol. Biol. 43, 189–201. Fransz, P., Armstrong, S., Alonso-Blanco, C., Fischer, T.C., TorresRuiz, R.A. and Jones, G. (1998) Cytogenetics for the model system Arabidopsis thaliana. Plant J. 13, 867–876. Gendrel, A.V., Lippman, Z., Yordan, C., Colot, V. and Martienssen, R.A. (2002) Dependence of heterochromatic histone H3 methylation patterns on the Arabidopsis gene DDM1. Science, 297, 1871–1873. Goldman, M.A., Holmquist, G.P., Gray, M.C., Caston, L.A. and Nag, A. (1984) Replication timing of genes and middle repetitive sequences. Science, 224, 686–692. Haaf, T. and Schmid, M. (1991) Chromosome topology in mammalian interphase nuclei. Exp. Cell Res. 192, 325–332. ß Blackwell Publishing Ltd, The Plant Journal, (2003), 33, 743–749

Transcriptional reactivation within heterochromatin Haupt, W., Fischer, T.C., Winderl, S., Fransz, P. and Torres-Ruiz, R.A. (2001) The centromere1 (CEN1) region of Arabidopsis thaliana: architecture and functional impact of chromatin. Plant J. 27, 285–296. Hediger, F. and Gasser, S.M. (2002) Nuclear organization and silencing: putting things in their place. Nat. Cell Biol. 4, E53–E55. Hochstrasser, M., Mathog, D., Gruenbaum, Y., Saumweber, H. and Sedat, J.W. (1986) Spatial organization of chromosomes in the salivary gland nuclei of Drosophila melanogaster. J. Cell Biol. 102, 112–123. Jackson, J.P., Lindroth, A.M., Cao, X. and Jacobsen, S.E. (2002) Control of CpNpG DNA methylation by the KRYPTONITE histone H3 methyltransferase. Nature, 416, 556–560. Jasencakova, Z., Meister, A., Walter, J., Turner, B.M. and Schubert, I. (2000) Histone H4 acetylation of euchromatin and heterochromatin is cell cycle dependent and correlated with replication rather than with transcription. Plant Cell, 12, 2087–2100. Jeddeloh, J.A., Stokes, T.L. and Richards, E.J. (1999) Maintenance of genomic methylation requires a SWI2/SNF2-like protein. Nat. Genet. 22, 94–97. Jenuwein, T. and Allis, C.D. (2001) Translating the histone code. Science, 293, 1074–1080. Johnson, L.M., Cao, X. and Jacobsen, S.E. (2002) Interplay between two epigenetic marks: DNA methylation and histone H3 lysine 9 methylation. Curr. Biol. 12, 1360–1367. Lindroth, A.M., Cao, X., Jackson, J.P., Zilberman, D., McCallum, C.M., Henikoff, S. and Jacobsen, S.E. (2001) Requirement of CHROMOMETHYLASE3 for maintenance of CpXpG methylation. Science, 292, 2077–2080. Maluszynska, J. and Heslop-Harrison, J.S. (1991) Localization of tandemly repeated DNA sequences in Arabidopsis thaliana. Plant J. 1, 159–166.


749

Mittelsten Scheid, O., Paszkowski, J. and Potrykus, I. (1991) Reversible inactivation of a transgene in Arabidopsis thaliana. Mol. Gen. Genet. 228, 104–112. Mittelsten Scheid, O., Afsar, K. and Paszkowski, J. (1998) Release of epigenetic gene silencing by trans-acting mutations in Arabidopsis. Proc. Natl. Acad. Sci. USA, 95, 632–637. Mittelsten Scheid, O., Probst, A.V., Afsar, K. and Paszkowski, J. (2002) Two regulatory levels of transcriptional gene silencing in Arabidopsis. Proc. Natl. Acad. Sci. USA, 99, 13659–13662. Murata, M., Ogura, Y. and Motoyoshi, F. (1994) Centromeric repetitive sequences in Arabidopsis thaliana. Jpn. J. Genet. 69, 361– 370. Peters, A.H., O’Carroll, D., Scherthan, H. et al. (2001) Loss of the Suv39h histone methyltransferases impairs mammalian heterochromatin and genome stability. Cell, 107, 323–337. Rabl, C. (1885) Ueber Zellteilung. Morphol. Jahrbuch. 10, 214–330. Steimer, A., Amedeo, P., Afsar, K., Fransz, P., Mittelsten Scheid, O. and Paszkowski, J. (2000) Endogenous targets of transcriptional gene silencing in Arabidopsis. Plant Cell, 12, 1165–1178. Struhl, K. (1998) Histone acetylation and transcriptional regulatory mechanisms. Genes Dev. 12, 599–606. Tamaru, H. and Selker, E.U. (2001) A histone H3 methyltransferase controls DNA methylation in Neurospora crassa. Nature, 414, 277–283. Tanaka, K., Appella, E. and Jay, G. (1983) Developmental activation of the H-2K gene is correlated with an increase in DNA methylation. Cell, 35, 457–465. Vongs, A., Kakutani, T., Martienssen, R.A. and Richards, E.J. (1993) Arabidopsis thaliana DNA methylation mutants. Science, 260, 1926–1928. Weiss, H. and Scherthan, H. (2002) Aloe spp. – plants with vertebrate-like telomeric sequences. Chromosome Res. 10, 155–164.