Epigenetic regulation of transcription in intermediate
heterochromatin
Yoshiki Habu
1,2+, Olivier Mathieu
3w, Muhammad Tariq
2z, Aline V. Probst
2,3y, Chotika Smathajitt
2,3, Tong Zhu
4J& Jerzy Paszkowski
2,31National Institute of Agrobiological Sciences, Tsukuba, Japan,2Friedrich Miescher Institute for Biomedical Research, Basel,
Switzerland,3Laboratory of Plant Genetics, University of Geneva, Geneva, Switzerland, and4Torrey Mesa Research Institute,
San Diego, California, USA
Constitutive heterochromatin is a compact, transcriptionally inert structure formed in gene-poor and repeat- and transposon-rich regions. In Arabidopsis, constitutive heterochromatin is charac-terized by hypermethylated DNA and histone H3 dimethylated at lysine (K) 9 (H3K9me2) together with depletion of histone H3 dimethylated at lysine 4 (H3K4me2). Here, we describe loci with intermediate properties of heterochromatin in which transcrip-tion downregulatranscrip-tion is inherited in a manner similar to constitutive heterochromatin, although the loci are associated with opposing histone marks—H3K4me2 and H3K9me2. In the ddm1 (decrease in DNA methylation 1) mutants, their transcrip-tional activation is accompanied by the expected shift in the H3 modifications—depletion of H3K9me2 and enrichment in H3K4me2. In mom1 (Morpheus’ molecule 1) mutants, however, a marked increase in transcription is not accompanied by detectable changes in the levels of H3K4me2 and H3K9me2. Therefore, transcriptional regulation in the intermediate hetero-chromatin involves two distinct epigenetic mechanisms. Interest-ingly, silent transgenic inserts seem to acquire properties characteristic of the intermediate heterochromatin.
Keywords: Arabidopsis; epigenetics; intermediate heterochromatin; MOM1; DDM1
EMBO reports (2006) 7, 1279–1284. doi:10.1038/sj.embor.7400835
INTRODUCTION
In Arabidopsis, mutations in MOM1 (Morpheus’ molecule 1) activate transcription of selected silent chromosomal and trans-genic loci without apparent changes in DNA methylation (Amedeo et al, 2000). Cytological analysis has shown that mom1 nuclei show normal compaction of pericentromeric constitutive heterochromatin containing hypermethylated DNA associated with histones modified by repressive marks such as H3 dimethylated at lysine (K) 9 (H3K9me2) and depletion of histone H3 dimethylated at lysine 4 (H3K4me2; Probst et al, 2003). By contrast, in ddm1 (decrease in DNA methylation 1) mutants, much broader transcriptional activation is associated with intensive DNA demethylation (Vongs et al, 1993; Mittelsten Scheid et al, 1998), replacement of H3K9me2 by H3K4me2 and markedly distorted nuclear morphology resulting from relaxation of centromeric and pericentromeric heterochromatin (Gendrel et al, 2002; Mittelsten Scheid et al, 2002; Probst et al, 2003). Therefore, although both mutations impair gene silencing, the mechanisms are different. Epistatic analyses have suggested the involvement of MOM1 and DDM1 in distinct silencing pathways (Mittelsten Scheid et al, 2002), but detailed studies of changes in DNA and histone modifications at reactivated genes were restricted by the complexity of the silencing targets of MOM1. Most are present as multiple copies that are simultaneously but possibly not identically subjected to the release of silencing and to the associated modifications in DNA and chromatin.
To avoid this problem, we searched for single-copy loci reactivated in both mom1 and ddm1 mutants.
RESULTS AND DISCUSSION
We carried out a microarray analysis with the Affymetrix Arabidopsis Genome (8K) chip (Zhu et al, 2001), using RNA samples prepared from wild-type plants, mom1, ddm1 and their parental transgenic line (Mittelsten Scheid et al, 1998; Amedeo et al, 2000), and sought single-copy genes that were co-regulated in mom1 and ddm1. Among 21 genes co-regulated in mom1 and ddm1 mutants, 19 were upregulated and two were downregulated more than threefold (supplementary Table I and supplementary Results and Discussion online). The gene encoding Cyclophilin 40 (CyP40, At2g15790) was one of the strongest upregulated genes in
Received 14 July 2006; revised 7 September 2006; accepted 15 September 2006; published online 3 November 2006
+Corresponding author. Tel: þ 81 29 838 8401; Fax: þ 81 29 838 7073;
E-mail: habu@affrc.go.jp
1National Institute of Agrobiological Sciences, Kannondai 2-1-2, 305-8602 Tsukuba,
Japan
2Friedrich Miescher Institute for Biomedical Research, CH-4058 Basel, Switzerland 3Laboratory of Plant Genetics, University of Geneva, CH-1211 Geneva 4, Switzerland 4Torrey Mesa Research Institute, 3115 Merryfield Row, San Diego, California 92121,
USA
w
Present address: UMR 6547—CNRS, Universite´ Blaise Pascal, 24 Avenue des Landais, 63177 Aubie´re Cedex, France
zPresent address: Center for Molecular Biology Heidelberg, University of Heidelberg,
Im Neuenheimer Feld 282, D-69120, Heidelberg, Germany
y
Present address: Institute Curie/Research section, UMR 218 du CNRS, 26 rue d’Ulm, 75248 Paris Cedex 05, France
JPresent address: Syngenta Biotechnology Inc., 3054 Cornwallis Road, Research
both mutants (23-fold on average; Fig 1A; supplementary Table I online). CyP40 encodes a peptidylprolyl cis–trans isomerase—an evolutionarily conserved immunophilin associated with the heat-shock protein 90 chaperone complex (Pratt et al, 2001). CyP40 seems to be required for the vegetative phase change in Arabidopsis, and mutants in this gene are defective in vegetative, but not reproductive, maturation (Berardini et al, 2001). As a single-copy gene activated in both mutants, we chose CyP40 as a model target to compare the molecular mechanisms and possible consequences of the two pathways of silencing release.
The predicted 1.5-kb messenger RNA of CyP40 (Berardini et al, 2001) is in low abundance in wild-type leaves, but a prominent CyP40-specific signal of approximately 7 kb was clearly visible on RNA blots from ddm1 and mom1 (Fig 1A). As the 7-kb RNA must cover a region extending beyond CyP40 (Fig 1B), we investigated transcriptional activation in regions flanking this gene. Probes for a gene (At2g15780) downstream of CyP40 did not detect RNA accumulation in the mutants (data not shown). By contrast, a probe specific for a Mutator-like element (MULE-F19G14; Singer et al, 2001) upstream of CyP40 showed a pattern of RNA accumulation that was similar to that visualized using the CyP40 probe (Fig 1A). The results indicated that MULE-F19G14 and CyP40 are simultaneously activated in mom1 and ddm1, and that the 7-kb transcript includes the sequences of both MULE-F19G14 and CyP40 (supplementary Fig 1 online). A probe for a CyP40 intron gave no signals on RNA blots (data not shown), indicating that transcripts accumulated in mom1 and ddm1 are spliced into the CyP40 exonic region.
Analysis of the products of 50 rapid amplification of cloned
ends (RACE), primed with an oligo-dT or gene-specific primers, indicate that the main transcription start site of the 7-kb transcript is located in a putative intron of a hypothetical gene in MULE-F19G14, with the direction of transcription towards CyP40, transcribing the non-coding strand of MULE-F19G14 (Fig 1B; supplementary Fig 1 online). An identical transcription start site was mapped in both ddm1 and mom1. As the 7-kb transcript was also detected by a probe corresponding to a region between the main transcription start site and CyP40 (data not shown), the transcript apparently covers the entire region downstream of the transcription start site and extends into CyP40 (Fig 1B).
We did not detect transcripts carrying the coding strand of MULE-F19G14 in the mutants. It has been reported that a MULE accumulated transcripts in both directions on general activation of transposons in ddm1 (Lippman et al, 2004). This indicates that remnants of transposons could produce aberrant transcripts using their cryptic transcription start sites on activation of other related transposons. Therefore, we examined whether other transposons known to be under epigenetic control are activated in mom1. We assessed transcription of At Mu1 and Tar17—a type II and a type I transposon, respectively—both of which have been reported to be activated in ddm1 (Hirochika et al, 2000; Singer et al, 2001). Reverse transcription–PCR (RT–PCR; for AtMu1) and RNA blot analysis (for Tar17) showed that apparently neither transposon was activated in mom1, whereas both were activated in ddm1, as reported (Fig 1C,D). Therefore, activation of MULE-F19G14 in mom1 is not accompanied by a general reactivation of trans-posons residing in constitutive, pericentromeric heterochromatin. The expression profile of the microarray supports this idea (supplementary Tables I,II online), suggesting that release of
silencing of the transposon-related MULE-F19G14 by the mom1 mutation is a rather exceptional and specific event. A similar observation is described in the accompanying paper of Vaillant et al (2006).
A homology search and Southern blots using MULE-F19G14-specific probes showed that an approximately 2-kb region around
CyP40 MULE-F19G14 9.0 5.0 3.0 2.0 1.0
CyP40 exon RNA
Col-0 ddm1-2 mom1-2 Col-0 ddm1-2 mom1-2 Col-0 ddm1-2 mom1-2
Deduced 7-kb transcript
Probes for RNA blots Regions analysed in ChIP Probes for small RNA blots
AtMu1 wt ddm1 mom1-2 wt ddm1-2 mom1-2 RT 0.4 kb 1 kb 0.2 kb Tar17 + − + − + − MULE-F19G14 D C A B
Fig 1|Accumulation of transcripts from MULE-F19G14–CyP40 and transposons in mom1 and ddm1. (A) RNA blot analysis of MULE-F19G14/CyP40 transcripts. Total RNAs extracted from Col-0 (wild type), ddm1-2 and mom1-2 were probed with CyP40 and MULE-F19G14 probes. Positions of the size markers are shown on the right in kilobases (kb). (B) A chromosomal region containing CyP40 and MULE-F19G14. Filled and open boxes represent exons and introns, respectively. Grey boxes represent terminal inverted repeats of MULE-F19G14. The direction and position of annotated genes are indicated by arrows. A deduced structure of the 7-kb transcript is shown by bars connected with dashed lines corresponding to introns. The positions of probes for RNA blots and ChIP are indicated at the bottom. The main transcription start site observed in mom1 and ddm1 is indicated by a hooked arrow. (C) Accumulation of the transcripts of AtMu1. Reverse transcription– PCR (RT–PCR) was carried out using RNA samples for Col-0 (wild type, wt), ddm1-2, and mom1-2 with ( þ ) or without () RT. (D) Accumulation of transcripts of Tar17. RNA blot analysis was carried out using total RNA of Col-0 (wild type (wt)), ddm1-2 and mom1-2. ChIP, chromatin immunoprecipitation; CyP40, Cyclophilin 40; ddm1, decrease in DNA methylation 1; mom1, Morpheus’ molecule 1; MULE, Mutator-like element.
the transcription start site of the 7-kb transcript is unique in the genome of Arabidopsis (Fig 2A). This allowed us to examine the details of the DNA and histone modifications in this region in the wild type and in both mutants. Conventional analysis, using methylation-sensitive restriction enzymes, indicated that the region surrounding the transcription start site of the 7-kb transcript is hypermethylated at CpG sequences in the wild-type plants and becomes hypomethylated in ddm1 (Fig 2B). Compared with the wild type, mom1 also showed a partial reduction in DNA methylation in several of the examined sites. Consistent with this observation, further analysis using bisulphite genomic sequencing showed that there is a partial reduction in DNA methylation in mom1 compared with ddm1 (Fig 2C). Overall, the density of CpG and CpNpG sequences is rather low in the region surrounding the transcription start site of the 7-kb transcript, and the frequency of methylated cytosine is low even in the wild type (Fig 2C).
The methylation status of H3K4 and H3K9 was analysed using chromatin immunoprecipitation (ChIP) with primers designed to amplify a region surrounding the transcription start site of the 7-kb transcript and the coding region of CyP40 (Figs 1B,2C). Other chromosomal and transgenic targets, such as the TSI (transcrip-tionally silent information; Steimer et al, 2000) and HPT loci (hygromycin phosphotransferase; Mittelsten Scheid et al, 1998), which are silent in the wild type but activated in mom1, were also examined. As a control, we assayed the status of H3 methylation in active genes and in loci encoding pericentromeric repeats related to transposons that remain silent in mom1 but are transcribed in ddm1. Notably, in the wild-type plants, we detected the presence of both H3K9me2 and H3K4me2 at all mom1-regulated loci (MULE-F19G14, CyP40, TSI and HPT; Fig 2D,E). As CyP40 is expressed in many plant tissues (Berardini et al, 2001), the presence of H3K9me2 at high levels was unexpected, and is in contrast with other expressed genes (actin2/7 and tubulin8) that are clearly depleted of H3K9me2 and enriched with H3K4me2 (Fig 2D). Moreover, endogenous loci that remain silent in mom1 but are reactivated in ddm1 are markedly enriched with H3K9me2 and depleted of H3K4me2 (Fig 2D; Gendrel et al, 2002; Lippman et al, 2003; data not shown).
In Arabidopsis, most H3K9me2 is confined to constitutive heterochromatin associated with pericentromeric transposons and centromeric repeats (Lippman & Martienssen, 2004); however, a small proportion of transposons in heterochromatin regions seem to associate with both H3K4me2 and H3K9me2 (Gendrel et al, 2002; Lippman et al, 2003). It is unclear why CyP40, as an active gene, retains a partially heterochromatic character similar to these transposons, pericentromeric TSI sequences or complex transgenic loci that underwent recent, de novo transcrip-tional silencing. Similar multivalent states of histone modi-fication have also been reported for developmentally regulated genes in other higher eukaryotes (Sims et al, 2002; Bernstein et al, 2006). Although retention of opposing H3 methylation marks might be required for MOM1-silencing targets, this characteristic seems to be insufficient as loci such as AtSN1 and AtMu1 meet these criteria (Lippman et al, 2003), but are not reactivated in mom1 mutants (Figs 1C,2F). Therefore, the intermediate hetero-chromatin state of MOM1 targets might be marked with further unknown characters that make a clear distinction between MOM1 targets and a subset of transposons possessing
a similar multivalent state of histone modification, but remain silent in mom1.
Although we do not exclude a possibility in which DNA methylation at specific cytosine residues is important for the maintenance of silencing, it is apparent that despite the drastic activation of transcription in mom1, the intermediate H3 methylation status remains unchanged (Fig 2D,E), suggesting that the transcription activation (Fig 1A) and the partial DNA demethylation (Fig 2B,C) do not depend on, or lead to, changes in histone H3 methylation at lysine 4 or 9. Moreover, the methylation status of H3K27 silent and active chromatin marks also remained unchanged in mom1 (Fig 2D). Data in an accompanying paper (Vaillant et al, 2006) also show that the silent 5S ribosomal RNA cluster was activated in mom1 and that properties related to the DNA and histone modifications are similar to those of the MULE-F19G14/CyP40.
In mom1 mutants, transcriptional reactivation occurs uniformly in all plant tissues (Amedeo et al, 2000; Tariq et al, 2002), arguing against a chimeric situation owing to the presence of a mixture of alleles containing either H3K9me2 or H3K4me2. It should be noted that mom1 mutants have been propagated through several generations, indicating that the active state of transcription persisting for several generations does not alter the fundamental properties of the intermediate heterochromatin.
As small RNAs have been shown to be involved in the regulation of heterochromatin features (Grewal & Rice, 2004; Lippman & Martienssen, 2004) by providing a feedback loop for controlling its silencing properties, we examined their possible regulatory role for targets in the intermediate heterochromatin. However, we were not able to detect small RNAs for MULE-F19G14/CyP40 under the conditions in which changes in the accumulation of small RNAs derived from the 5S RNA genes in ddm1 could be clearly visualized (Elmayan et al, 2005; Fig 1B; data not shown). A similar, small-RNA-independent mechanism of silencing has also been reported for the transgenic locus L5, which is one of the targets of MOM1 (Elmayan et al, 2005).
Thus, MOM1 seems to operate largely independently of histone–DNA methylation and small RNA signals, but its activity seems to be restricted to loci with intermediate heterochromatin properties. However, histone and/or DNA methylation signals can be involved in the control of transcription in the intermediate and constitutive heterochromatin, as illustrated by experiments with ddm1. Therefore, it can be suggested that maintenance of marks at levels characteristic of the intermediate heterochromatin is required for the regulatory activity of MOM1. As the MOM1 protein has no motifs that are known to be responsible for binding to modified histones (Amedeo et al, 2000), it remains unclear how MOM1 recognizes these properties and operates in the inter-mediate heterochromatin.
The results presented here provide a first glimpse of the characteristics and regulatory aspects of intermediate hetero-chromatin in plants, which seems to be heritably maintained in a state between euchromatin and heterochromatin. It is attractive to speculate that such intermediate heterochromatin could provide particular epigenetic flexibility. Moreover, it is conceivable that formation of intermediate heterochromatin contributes to the early steps in defence against foreign DNA insertions, as illustrated by transgenic and endogenous loci that require MOM1 for their silencing.
m d
w w d m w d m w d m w d m w d m w d m
SspI
HincII HpaII NheI SacII
Aci I Ava I − Sp Ac HcScAcAc NhAv Hp Sp Sc Hp Hp Ac Ac 100 bp F6D8 F1N21 T5E15 T2L5 MULE-F19G14 1 kb TCGACAATTGATGAGTTGACGGTACAGGAACGTACCGCGGGTGAGAAACCGAGCGGTCATGGTTTGGCCATGAGCTGCTTGGGCGGTCATCAGAACGTGT GGATGAGAGTGAGCCGACCTTACTTGTGGCCAAGAGTTGCCCACGTCCATGCTTATCCTTATCCCATGACTCATGTGTTTATCCATATCCTTATCCTATC CAAAACTCATCATTATCCTTATCCGTTAGTAACTTGGTGACCAAGTTTCGTAGCTTAGCCCCTAAGTTTCGCCAACTATATAAACCCCTTCTCATTTCAT TCCAAACCATCACTCACCACTTCACAAACCCACCTAAAGTCACTTAGGAAAAACGCTAGCATTTTAGGAAGCTCTCGGGAAGTTCCAAGTCCGTTCTGCC AACGAACTGTGAGTTTCCGGTCGTGTGAAGTAGAACACGATTAGGAATTTTAAATATTTATTTATTTATTTAATTTATGCATGAGGTTGATATTTTATAC TTGGTATAAATTATGCATAGATTAGGATTAGCTTATTATACGTTTTATCGTATAATCGGCTTAAGTATTTTTAGGATAGAT Col-0 mom1-2 ddm1-2 mom1-2 Transcription Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes No Yes No ddm1-2 mom1-2 ddm1-2 ddm1-2 mom1-2 IP C m C m C m MULE-F19G14 CyP40 TSI Ta2 At4g03760 tubulin8 actin2/7
Mock H3K4me2 H3K9me2 Mock H3K4me2 H3K9me2 IP C d C d C d
Mock H3K27me3 H3K27me2 IP C m C m C m
IP A m A m A m
Mock H3K4me2 H3K9me2 Transcription Mock H3K4me2 H3K9me2 Transcription
HPT actin2/7 IP A m A m A m AtSN1 E F D C A B
METHODS
Plant materials. mom1-1 and ddm1-5 (som8) have been described previously (Mittelsten Scheid et al, 1998; Amedeo et al, 2000). mom1-2 was obtained from the SALK/SAIL collection (SAIL_610_G01) and ddm1-2 was provided by E. Richards (Vongs et al, 1993).
DNA and RNA isolation and gel blot analyses. Genomic DNA was isolated from soil-grown seedlings using ISOPLSNT II (Nippongene, Tokyo, Japan). DNA blot analysis was carried out as described previously (Amedeo et al, 2000). Total RNA was isolated from soil-grown seedlings using Trizol (Invitrogen, Carlsbad, CA, USA) or ISOGEN (Nippongene). RNA blot analysis was carried out as described previously (Amedeo et al, 2000). Microarray. Information on the microarray, probe preparation, data processing and summaries of the expression profiles are shown in the supplementary Materials and methods and supplementary Table I online, and discussed in the supplementary Results and Dis-cussion online. The original data are available in the Gene Expres-sion Omnibus in NCBI under the accesExpres-sion number GSE5771. Cloning and sequencing of transcripts accumulated in mom1 and ddm1. Details of RT–PCR and 50 RACE that were used to
characterize the 7-kb transcripts that accumulated in mom1-2 and ddm1-2 are shown in supplementary Fig 1 online. The com-binations and sequences of the primers are shown in supplementary Fig 1 and supplementary Table II online, respectively.
Bisulphite genomic sequencing. Bisulphite genomic sequencing was carried out as described previously (Amedeo et al, 2000), except that only the sense strand for the transcription was analysed. Primers used are listed in supplementary Table II online. Chromatin immunoprecipitation. ChIP was carried out as descri-bed previously (Mathieu et al, 2005). Primers for CyP40 are shown in supplementary Table III online. Primers to amplify other genes were same as those described previously (Mathieu et al, 2005). Supplementary information is available at EMBO reports online (http://www.emboreports.org).
ACKNOWLEDGEMENTS
We thank S. Lienhard, K. Hioki, R. Nigorikawa, Y.M. Li and S. Ando for technical assistance, T. Kakutani for the Tar17 probe, and D. Schu¨beler, M. Goldschmidt-Clermont and A. Blonstein for comments on the
manuscript. This work was supported by the Swiss National Foundation (grant 3100A0-102107), The Epigenome—Network of Excellence (LSHG-CT-2004-503433) and an internal grant by the National Institute of Agrobiological Sciences, Japan.
REFERENCES
Amedeo P, Habu Y, Afsar K, Mittelsten Scheid O, Paszkowski J (2000) Disruption of the plant gene MOM releases transcriptional silencing of methylated genes. Nature 405: 203–206
Berardini TZ, Bollman K, Sun H, Poething RS (2001) Regulation of vegetative phase change in Arabidopsis thaliana by cyclophilin 40. Science 291: 2405–2407
Bernstein BE et al (2006) A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125: 315–326 Elmayan T, Proux F, Vaucheret H (2005) Arabidopsis RPA2: a genetic link
among transcriptional genesilencing, DNA repair, and DNA replication. Curr Biol 15: 1919–1925
Gendrel A-V, Lippman Z, Yordan C, Colot V, Martienssen RA (2002) Dependence of hetrochromatic histone H3 methylation patterns on the Arabidopsis gene DDM1. Science 297: 1871–1873
Grewal SIS, Rice JC (2004) Regulation of heterochromatin by histone methylation and small RNAs. Curr Opin Cell Biol 16: 230–238 Hirochika H, Okamoto H, Kakutani T (2000) Silencing of retrotransposons
in Arabidopsis and reactivation by the ddm1 mutation. Plant Cell 12: 357–368
Lippman Z, Martienssen RA (2004) The role of RNA interference in heterochromatic silencing. Nature 431: 364–370
Lippman Z, May B, Yordan C, Singer T, Martienssen RA (2003) Distinct mechanisms determine transposon inheritance and methylation via small interfering RNA and histone modification. PLoS Biol 1: 420–428 Lippman Z et al (2004) Poles of transposable elements in hetrochromatin and
epigenetic control. Nature 430: 471–476
Mathieu O, Probst AV, Paszkowski J (2005) Distinct regulation of histone H3 methyaltion at lysines 27 and 9 by CpG methylation in Arabidopsis. EMBO J 24: 2783–2791
Mittelsten Scheid O, Afsar K, 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 AV, Afsar K, Paszkowski J (2002) Two regulatory levels of transcriptional gene silencing in Arabidopsis. Proc Natl Acad Sci USA 99: 13659–13662
Pratt WB, Krishna P, Olsen LJ (2001) Hsp90-binding immunophilins in plant: the protein movers. Trends Plant Sci 6: 54–58
Probst AV, Fransz PF, Paszkowski J, Mittelsten Scheid O (2003) Two means of transcriptional reactivation within heterochromatin. Plant J 33: 743–749
Sims III RJ, Nishioka K, Reinberg D (2002) Histone lysine methylation: a signature for chromatin function. Trends Genet 19: 629–639 Fig 2|DNA and histone methylation in MULE-F19G14/CyP40 and other representative genes. (A) Similarity in the DNA sequence of MULE-F19G14
with other MULEs. The structure and positions of annotated genes are indicated as in Fig 1B. Regions in MULE-F19G14 showing sequence similarities to other MULEs (expected value cutoff ¼ 3 in BLASTN) are shown as bars at the bottom, with the corresponding names of BACs carrying the MULEs. (B) DNA blot analysis of a region surrounding the transcription start site of the 7-kb transcript with methylation-sensitive restriction enzymes. Genomic DNAs were digested with SspI (methylation insensitive) followed by digestion with methylation-sensitive restriction enzymes. Recognition sites for the restriction enzymes around the main transcription start site (hooked arrow) of the 7-kb transcript are shown at the bottom. Av, AvaI; d, ddm1-2; Hc, HincII; Hp, HpaII; m, mom1-2; Nh, NheI; Sc, SacII; Sp, SspI; w, Col-0. (C) Bisulphite genomic sequencing of a region surrounding the main transcription start site of the 7-kb transcript. The percentage of methylated cytosine at a particular site is indicated in grey. Positions of cytosines in CpG and CpNpG are indicated by filled and open squares, respectively. Positions of primers, used for ChIP (D), are indicated by arrows. The transcription start site of the 7-kb transcript is indicated by the hooked arrow. (D) ChIP analysis of the methylation status of histone H3. ChIP analysis using antibodies recognizing H3K4me2, H3K9me2, H3K27me2 and H3K27me3 was carried out using chromatin extracts from wild-type plants (C, Col-0), mom1-2 (m) and ddm1-2 (d). Target loci that were analysed are listed on the left (supplementary Table III online); IP, input. (E) ChIP analysis of the methylation status of histones H3 associated with the transgenic HPT locus. A, line A (the wild-type parent of mom1-1; Mittelsten Scheid et al, 1998); IP, input; m, mom1-1. (F) ChIP analysis of the H3 methylation status of histones H3 associated with the AtSN1 transposon. ChIP, chromatin immunoprecipitation; ddm1, decrease in DNA methylation 1; H3K4me2, histone H3 dimethylated at lysine 4; H3K27me2, histone H3 dimethylated at lysine 27; H3K27me3, histone H3 trimethylated at lysine 27; HPT, hygromycin phosphotransferase; mom1, Morpheus’ molecule 1; MULE, Mutator-like element. BACs, bacterial artificial chromosomes; CyP40, Cyclophilin 40.
Singer T, Yordan C, Martienssen RA (2001) Robertson’s Mutator transposons in A. thaliana are regulated by the chromatin-remodeling gene Decrease in DNA methylation (DDM1). Genes Dev 15: 591–602
Steimer A, Amedeo P, Afsar K, Fransz P, Mittelsten Scheid O, Paszkowski J (2000) Endogenous targets of transcriptional gene silencing in Arabidopsis. Plant Cell 12: 1165–1178
Tariq M, Habu Y, Paszkowski J (2002) Depletion of MOM1 in non-dividing cells of Arabidopsis plants releases transcriptional gene silencing. EMBO Rep 3: 951–955
Vaillant I, Schubert I, Tourmente S, Mathieu O (2006) MOM1 mediates DNA-methylation-independent silencing of repetitive sequences in Arabidopsis. EMBO Rep [doi:10.1038/sj.embor.7400791]
Vongs A, Kakutani T, Martienssen RA, Richards EJ (1993) Arabidopsis DNA methylation mutants. Science 260: 1926–1928
Zhu T, Budworth P, Han B, Brown D, Chang HS, Zou G, Wang X (2001) Towards elucidating the global gene expression patterns of developing Arabidopsis: parallel analysis of 8300 genes by high-density oligonucleotide probe array. Plant Physiol Biochem 39: 221–242
