Methylation of Histone H3 at Lysine 9 Targets Programmed DNA Elimination in Tetrahymena

Cell, Vol. 110, 701–711, September 20, 2002, Copyright 2002 by Cell Press

Methylation of Histone H3 at Lysine 9 Targets Programmed DNA Elimination in Tetrahymena Sean D. Taverna,2 Robert S. Coyne,2 and C. David Allis1 Department of Biochemistry and Molecular Genetics University of Virginia Health System Charlottesville, Virginia 22908

Summary Histone H3 lysine 9 methylation [Me(Lys9)H3] is an epigenetic mark for heterochromatin-dependent gene silencing, mediated by direct binding to chromodomain-containing proteins such as Heterochromatin Protein 1. In the ciliate Tetrahymena, two chromodomain proteins, Pdd1p and Pdd3p, are involved in the massive programmed DNA elimination that accompanies macronuclear development. We report that both proteins bind H3(Lys9)Me in vitro. In vivo, H3(Lys9)Me is confined to the time period and location where DNA elimination occurs, and associates with eliminated sequences. Loss of parental Pdd1p expression drastically reduces H3(Lys9)Me. Finally, tethering Pdd1p is sufficient to promote DNA excision. These results extend the range of H3(Lys9)Me involvement in chromatin activities outside transcriptional regulation and also strengthen the link between heterochromatin formation and programmed DNA elimination. Introduction Eukaryotic DNA is packaged with histones into nucleosomes, the fundamental subunit of chromatin. The structure and composition of chromatin regulates accessibility of DNA to proteins involved in DNA-templated processes such as transcription, recombination, replication, and repair (Wolffe, 1998; Kornberg and Lorch, 1999). Chromatin structure is modulated in part by enzymes that covalently modify histones at defined positions, primarily in the N- or C-terminal tails; such modifications include acetylation, methylation, phosphorylation, and ubiquitination (van Holde, 1988; Wolffe and Hayes, 1999). Several lines of evidence suggest certain histone modifications may present platforms for the binding of particular regulatory proteins and complexes (Strahl and Allis, 2000; Jenuwein and Allis, 2001). The existence of such a “histone code” would allow chromosomal regions to be epigenetically “marked”, for example, as transcriptionally active or repressed. The discovery of enzymes that specifically methylate histone lysine or arginine residues and proteins that bind these modified sites has revealed histone methylation to be of major importance in long-range chromatin regulatory processes (Lachner and Jenuwein, 2002). For example, a 20 kb region constituting the silent mating type locus 1 2

Correspondence: [email protected] These authors contributed equally to this work.

of Schizosacchromyces pombe is packaged into nucleosomes preferentially methylated at lysine 9 of histone H3 [Me(Lys9)H3] and unmethylated at lysine 4 of H3, whereas in adjacent chromosomal regions, the opposite pattern predominates (Noma et al., 2001). Likewise, a 16 kb region of condensed chromatin at transcriptionally silent regions of the chicken ␤-globin locus contains high levels of Me(Lys9)H3, but adjacent active regions are enriched in Me(Lys4)H3 (Litt et al., 2001). The chromodomain of HP1 specifically binds Me (Lys9)H3, but not Me(Lys4)H3 or unmodified H3 (Jacobs et al., 2001; Lachner et al., 2001; Bannister et al., 2001). Mutations within the chromodomain that disrupt this binding abolish HP1’s gene silencing activity (Platero et al., 1995). Drosophila HP1 and its S. pombe homolog Swi6 colocalize in vivo with Me(Lys9)H3 to heterochromatic regions (James et al., 1989; Nakayama et al., 2000). Recently, the structure of the HP1 chromodomain complexed to Me(Lys9)H3 peptides was solved (Jacobs and Khorasanizadeh, 2002; Nielsen et al., 2002) providing insight into how both the methyl modification and the H3 peptide contribute to overall binding interaction. These studies make clear predictions as to what other chromodomains may bind to Me(Lys9)H3, including two proteins from the ciliated protozoan Tetrahymena thermophila. Although unicellular, Tetrahymena contains two functionally distinct nuclei. The macronucleus (MAC) is responsible for transcriptional activity and is therefore analogous in function to a somatic nucleus of a metazoan. The role of the transcriptionally silent germline micronucleus (MIC) becomes apparent during the sexual process of conjugation. The newly formed zygotic MIC divides twice mitotically to produce four genetically identical nuclei, two of which develop into new MACs (the parental MAC degenerates). During differentiation of these nuclei, several thousand internal eliminated sequences (IESs) are precisely excised, comprising 10%– 15% of the germline genome (reviewed in Yao et al., 2002). Previously, we identified two chromodomain-containing proteins from Tetrahymena, Pdd1p and Pdd3p (for Programmed DNA Degradation). Both are abundant polypeptides that are expressed only during conjugation and colocalize with eliminated DNA (Madireddi et al., 1996; Nikiforov et al., 2000). Disruption of the parental PDD1 gene results in aberrant DNA elimination and lethality (Coyne et al., 1999). Because some chromodomain proteins bind Me(Lys9)H3, we have explored the correlation of this modification with programmed DNA elimination. Here we report that Pdd1p and Pdd3p bind Me(Lys9)H3 in vitro, and that Me(Lys9)H3 is associated with eliminated DNA in vivo. Furthermore, we show that tethering Pdd1p to an otherwise inactive IES is sufficient to promote its excision. We propose that Me(Lys9)H3 acts as a developmental mark for large-scale DNA elimination that is “read” by Pdd1p and/or Pdd3p-containing complexes.

Cell 702

Figure 1. Chromodomains of Pdd1p and Pdd3p Bind Me(Lys9)H3 Peptides In Vitro (A) Alignment of the first chromodomain of Pdd1p and only chromodomain of Pdd3p with the chromodomain of Drosophila HP1. The three aromatic residues (shown in red with asterisks) are predicted to form a “cage” enclosing the methylammonium group of Me(Lys9)H3 (Jacobs and Khorasanizadeh, 2002). Other residues identical or similar in all three sequences are shaded in black or gray, respectively. (B) The Tetrahymena chromodomains in (A) were expressed in E. coli (tagged with 6⫻ His), incubated with fluorescently labeled peptides (see Experimental Procedures), and analyzed by fluorescence anisotropy. (C) shows binding constants calculated from the data in (B).

Results The First Chromodomain of Pdd1p and the Only Chromodomain of Pdd3p Bind Me(Lys9)H3 Methylation at lysine 9 of histone H3 produces a binding site for the chromodomain of HP1 (Jacobs et al., 2001; Bannister et al., 2001; Nakayama et al., 2001). Recently, the structure of the chromodomain interacting with Me(Lys9)H3 was solved and suggests two aspects of this interaction to be critical for stability. The first is an “induced fit” ␤ sandwich composed of residues 5–10 of the H3 tail and ␤ strands 1 and 5 of the chromodomain. The second is a “cage” of three aromatic side chains into which the methylammonium group of di- or trimethylated lysine is inserted (Jacobs and Khorasanizadeh, 2002; Nielsen et al., 2002). Previously, we identified three Pdd proteins in Tetrahymena that play a central role in the heterochromatin assembly and DNA elimination events that occur during macronuclear development. Notably, Pdd1p contains two obvious chromodomains and Pdd3p one. Figure 1A shows alignment of the first chromodomain of Pdd1p (Pdd1p-CD1) and that of Pdd3p (Pdd3p-CD1), with that of Drosophila HP1, highlighting the conserved aromatic residues of the proposed methylammonium cage.

Based on this alignment, we and others predicted these Pdd chromodomains also have the potential to bind Me(Lys9)H3 (Jacobs and Khorasanizadeh, 2002). To evaluate this potential, we expressed Pdd1p-CD1 (residues 46–103) and Pdd3p-CD1 (residues 14–70) in E. coli downstream of a 6⫻-His tag. Chromodomains were purified by nickel-affinity chromatography and incubated with fluorescently labeled unmodified or methylated peptides representing the H3 tail. Binding was measured by fluorescence anisotropy (Figure 1B). Pdd1p-CD1 and Pdd3p-CD1 recognized H3 peptide methylated at lysine 9 with dissociation constants of 7 ␮M and 4 ␮M, respectively. Remarkably, these values are approximately the same as the affinity determined for the chromodomain of Drosophila HP1 (5 ␮M) under similar assay conditions (Jacobs and Khorasanizadeh, 2002). Unmodified H3 peptides, as well as H3 peptides methylated on lysines 4 or 27, yielded only weak or no binding (Figure 1C). Methylation of Histone H3 on Lysine 9 Occurs only during Conjugation In previous work (Strahl et al., 1999; Briggs et al., 2001), we showed that Me(Lys9)H3 is undetectable in vegetatively growing Tetrahymena. However, the data de-

Histone H3 Methylation Marks DNA for Elimination 703

Figure 2. Developmental Regulation of H3 Lys9 Methylation (A) Nuclear events of Tetrahymena conjugation. Zygotic MICs (small open circles) divide twice mitotically, and two products develop into MAC anlagen (large black circles), as parental MACs (large open circles) degenerate. Formation of Pddp-containing heterochromatin-like structures (small black circles) in the anlagen and DNA elimination occur between 9 and 14 hr. (B and C) Whole-cell proteins were separated by SDS-PAGE (12% for histones, 8% for Pdd1p), transferred to membranes, and probed with either ␣-Me(Lys9)H3, ␣-Pdd1p, or ␣-H4. Asterisk in (C) denotes proteolytic fragment of H4 derived from “old” MAC (Lin et al., 1991). (D) Antibodies were preincubated with the indicated peptide before probing acid-soluble extracts from unit gravity-purified 9 hr anlagen.

scribed above suggests that direct engagement of H3(Lys9)Me by Pddps might occur in vivo in a pathway independent of transcriptional regulation. Since detectable expression of Pddps is confined to Tetrahymena sexual conjugation, we investigated the methylation status of H3 during this developmental process. Synchronous conjugation can be induced by mixing prestarved cell populations of different mating types (Figure 2A). Western analysis of whole-cell extracts (Figure 2B) shows that the peak abundance of Pdd1p occurs at 9 hr after mixing and that Me(Lys9)H3 peaks at the same time. An expanded time course shows the highly regulated appearance and disappearance of Me (Lys9)H3 (Figure 2C), with the modification first appearing at 7.5 hr, peaking at 9 hr and gradually disappearing by 14 hr. To confirm the specificity of ␣-Me (Lys9)H3 antibody, nuclear extracts from 9 hr pairs were subjected to Western analyses using antibody preincubated with modified or unmodified H3 peptides (Figure 2D). Only the H3 peptide methylated at lysine 9 successfully competed away this signal. Methylation of Histone H3 on Lysine 9 Colocalizes with Specialized Pdd1p-Containing DNA Elimination Structures The peak abundance of Me(Lys9)H3 occurs in a stage wherein conjugating pairs contain three forms of nuclei: old parental macronuclei (OM), micronuclei (MIC), and macronuclear anlagen (AN) (Figure 3A). Since Pdd1p is confined to anlagen at this stage, we sought to determine if Me(Lys9)H3 is as well. Nuclei of mating Tetrahymena were collected at 9 hr, purified by sedimentation, and histone-enriched extracts were subjected to Western analysis. Immunoblotting with ␣-Me(Lys9)H3 revealed that, similar to Pdd1p, Me(Lys9)H3 is highly enriched in anlagen, but not detectable in MICs or old MACs (Figure 3A). Within anlagen, Pdd1p displays a dynamic distribution. Initially uniform, the protein coalesces into distinct, spherical structures that also contain Pdd2p and Pdd3p,

as well as MIC-limited DNA sequences (Madireddi et al., 1994, 1996; Smothers et al., 1997a; Nikiforov et al., 2000). These structures eventually concentrate at the periphery of the nucleus and disappear. The timing of DNA excision (10–14 hr) corresponds with the development of the Pddp structures (Yokoyama and Yao, 1982; Austerberry et al., 1984; Saveliev and Cox, 1995). Accordingly, if Me(Lys9)H3 “marks” DNA destined for excision, we would predict it to localize to these structures as well. We performed indirect immunofluorescence analysis of Tetrahymena during conjugation. As predicted, during early stages of MAC development (7 hr), a strong analgen-specific signal was detected using ␣-Me(Lys9)H3 (Figure 3B). Furthermore, the localization of Me(Lys9)H3 during the course of MAC development, including the observation of spherical structures at the nuclear periphery, was indistinguishable from that observed for Pdd1p. Peptide competition experiments demonstrate that staining is specific for Me(Lys9)H3 (data not shown). To determine if the peripherally located structures were actually the same as Pddp structures, coimmunostaining was performed using both ␣-Me(Lys9)H3 and ␣-Pdd1p antibodies. Figure 3C shows complete correspondence between the localization of Me(Lys9)H3 and Pdd1p, providing strong evidence for their in vivo functional association. Interestingly, Me(Lys4)H3, previously described in Tetrahymena (Strahl et al., 1999) and commonly associated with transcriptional activation, was seemingly excluded from Pddp structures (data not shown). Pdd1p and H3 Lys9 Methylation Physically Associate with Unique MIC-Limited Sequences Previously, we used chromatin immunoprecpitation (ChIP) coupled with DNA hybridization to demonstrate that Pddp-associated chromatin is enriched with a repetitive micronuclear-limited sequence (Tt2512), but not the MAC-destined sequences rDNA or ␤-tubulin 2 (Smothers et al., 1997b; Nikiforov et al., 2000). Here, we

Cell 704

Figure 3. Me(Lys9)H3 Colocalizes with Pdd1p-Containing DNA Elimination Structures (A) The three types of nuclei present in 9 hr conjugants—micronuclei (Mi), macronuclear anlagen (An), and old parental macronuclei (OM)—were purified by unit gravity sedimentation. Nuclear proteins were separated by SDS-PAGE, immunoblotted, and probed as in Figure 2B. (B) Mating Tetrahymena cells from the indicated time points were processed for immunofluorescence using ␣-Me(Lys9)H3 antibodies and staining with DAPI. (C) 9 hr mating cells were processed as in (B), but also double stained with mouse ␣-Pdd1p. Triangles indicate MAC anlagen. Arrows indicate Pddp-containing intranuclear structures. In (B) and (C), the bars represent 5 micrometers.

extend these studies to address the levels of Pdd1p and two histone modifications at several unique IESs and mac-retained sequences, including the extensively characterized M/R chromosomal locus (Figure 4A). Chromatin from 9 hr mating cells was crosslinked and sheared to average lengths of 500 to 1000 bp. Aliquots were then immunoprecipitated with antibodies against Pdd1p, Me(Lys9)H3, Acetyl(Lys9)H3, or general H4. After reversal of crosslinking, the DNA was amplified with each of the primer pairs, as well as primers for the unlinked MAC-destined ␤-tubulin 1 (BTU1) gene as an internal control for reaction conditions and loading. In parallel, input DNA was amplified with the same primers, as a control for the amplification efficiency of each primer pair. All reactions were performed in triplicate; typical results are shown in Figure 4B. Fluorescence intensities were normalized as described in Noma et al. (2001) and the results are graphically displayed in Figures 4C and 4D. Interestingly, both Me(Lys9)H3 and Pdd1p were enriched at the MIC-limited M and R elements, but not at the MAC-destined interregion between them (Figure 4C). Conversely, both M and R have a negative association with a histone modification associated with transcriptionally active chromatin [Acetyl(Lys 9)H3], relative to the BTU1 locus (Figure 4C). Antibodies to general H4 were detected at similar levels across the entire M/R locus. Similar results were obtained with chromatin from

13 hr mating cells (data not shown). Thus, Me(Lys9)H3 and Pdd1p are physically associated with unique MIClimited sequences before the bulk of DNA elimination is known to occur and remain at those regions throughout the elimination process. To confirm that these patterns are general, the 9 hr ChIP experiments were repeated with four other unlinked IESs and three MAC-retained sequences adjacent to IESs (Chau and Orias, 1996; Katoh et al., 1993; Heinonen and Pearlman, 1994). Figure 4D shows that, in each case, Pdd1p and Me(Lys9)H3 were enriched at the IESs relative to the MAC-retained sequences. H3 Lys9 Methylation Is Greatly Reduced in Parental Pdd1p Knockout Strains Early in conjugation, PDD1 transcripts are provided by the parental MAC, but following anlagen formation, transcriptional activation of the zygotic gene occurs. We previously created parental gene “knockout” strains that lack the early expression (Coyne et al., 1999). Although Pdd1p is present at near wild-type levels by the normal time of DNA rearrangement, such strains have severe defects in the establishment of Pddp structures and undergo greatly reduced DNA elimination. These observations led us to hypothesize that a developmental delay of Pdd1p accumulation results in the failure to completely establish a specialized chromatin structure necessary for elimination.

Histone H3 Methylation Marks DNA for Elimination 705

Figure 4. Adjacent Eliminated and Retained Chromatin Regions Show Distinct Histone Modification Patterns (A) Schematic of the M/R chromosomal locus. The M element undergoes one of two alternative excision events (0.6 or 0.9 kb; represented by open boxes) with a common right junction. M is separated from the 1.1 kb R element by 2.7 kb of mac-retained sequence. Arrows indicate the approximate locations of the three primer pairs used for PCR. (B) Representative results of PCR before (Total) or after chromatin immunoprecipitation (ChIP) from 9 hr conjugating cells with the indicated antibodies. PCR products were separated by 2% agarose gel electrophoresis and stained with ethidium bromide. The upper band in each panel results from amplification of the control BTU1 locus, the lower band (ⵑ200 bp) from the region shown in (A). (C and D) Normalized quantitation of three independent ChIP PCR trials. Bars indicate standard errors. (C) shows results for the M/ R locus, (D) for four unlinked IESs and three mac-retained regions near IESs. The ARP1 amplified sequence is 550 bp from the mse2.9 IES, within an ORF. CAM1 and CAM2 are about 350 and 2300 bp, respectively, from a 1.4 kb IES upstream of the calmodulin gene. Neither CAM1 nor CAM2 is known to be transcribed.

Since the results described above strongly suggest that Me(Lys9)H3 is an important component of this specialized chromatin structure, we examined whether levels or localization of Me(Lys9)H3 are affected by disruption of the parental PDD1 gene. Figure 5A shows Western analysis of 9 hr mating cells lacking parental PDD1 expression. Levels of Me(Lys9)H3 are drastically reduced, although longer exposures suggest minimal levels of this antigen are still present (data not shown). In contrast, Me(Lys4)H3 does not appear to be affected. We also failed to observe Me(Lys9)H3 by immunofluorescence analysis of knockout mating cells (Figure 5B). Tethering of Pdd1p Is Sufficient to Promote DNA Excision If the role of Me(Lys9)H3 is to provide an epigenetic mark for binding of Pdd1p to MIC-limited DNA sequences, then it may be possible to bypass this requirement by artificially tethering Pdd1p to a particular DNA sequence, in much the same way other chromodomain proteins have been tethered to sites that nucleate transcriptionally repressive chromatin structure (Muller, 1995; Poux et al., 2001). We designed such an experi-

ment, making use of a transformation assay that has been extensively employed in the study of IES excision. The transformation vector pD5H8 (Godiska et al., 1993) contains a cloned copy of the MIC rDNA locus. When this vector is introduced into 10 hr conjugating Tetrahymena, cis-acting signals promote the formation of a mature palindromic rDNA minichromosome, which is then differentially amplified to about 10,000 copies per cell. The rDNA allele in pD5H8 contains a mutation conferring resistance to paromomycin and other sequence differences conferring a replication advantage, allowing selection of transformants and complete replacement of endogenous rDNA. If a functional IES is cloned downstream of the rDNA transcription unit, it will also be processed with fairly high efficiency (Godiska and Yao, 1990; Chalker et al., 1999). Using this system, it has been shown that the cisacting DNA sequences necessary for deletion of the M IES include at least two sequence elements located within the 0.6 kb region common to both alternative rearrangements (Yao et al., 2002). Artificial deletion of these elements or replacement with non-IES DNA results in a loss of excision capability. We replaced the

Cell 706

Figure 5. Loss of Parental Pdd1p Expression Drastically Reduces Me(Lys9)H3, but Not Me(Lys4)H3 (A) Whole-cell proteins from wild-type mating cells or mating cells carrying a parental PDD1 gene disruption were separated by SDSPAGE (12%), immunoblotted, and probed with the indicated antibodies. (B) 9 hr conjugants were processed for immunofluorescence with modification-specific antibodies. Triangles indicate MAC anlagen.

entire 0.6 kb region (within pD5H8) with 4 copies of the E. coli LexA operator to produce pMLexAop (Figure 6A). When this plasmid is transformed into wild-type conjugants, no M IES excision is detectable, as expected (Figure 6B). The endogenous M element is not detected under these conditions because of its far lower copy number.

To tether Pdd1p to the modified M element, we fused the coding region of PDD1 to the E. coli LexA gene with flanking upstream and downstream PDD1 genomic sequences (Figure 6A). Using biolistic transformation, the endogenous MAC PDD1 gene was replaced by the fusion allele in vegetatively growing Tetrahymena strain CU428. Because the MAC divides amitotically, unequal

Figure 6. Tethering of PDD1p Is Sufficient to Promote DNA Elimination (A) Schematic of Pdd1p tethering experiment. Left third: 592 bp of M IES were replaced by 228 bp containing 4⫻-repeated LexA operator (hatched box), and inserted into transformation vector pD5H8 to create pMLexAop. Middle third: MAC PDD1 gene was replaced by homologous recombination with LexAPDD1 fusion allele in strain CU428. After making fusion allele homozygous (represented by shaded MAC), strain was mated to wild-type strain CU427 and transformed with pMLexAop. Right third: Transformants were analyzed by PCR with primers flanking M element. (B) PCR products were separated by 2% agarose gel electrophoresis and stained with ethidium bromide. Fourteen independent pMLexAop transformants resulting from mating of CU428: LexA-PDD1 and two controls from mating of CU428 are shown.

Histone H3 Methylation Marks DNA for Elimination 707

Figure 7. Heterogeneity of Deletion Junctions Independent PCR products representing four each of the approximately 228 and 545 bp excision events shown in Figure 6 were cloned and sequenced to determine the precise junctions. (A) and (B) show the sequences bordering the smaller and larger excisions, respectively. Direct repeats constituting the predominant junctions in wildtype M element excision are underlined. The detected junctions are indicated by the lines above or below the sequence. Where the precise junction is ambiguous due to a direct repeat, brackets delineate the extent of the repeat.

allelic segregation gives rise to homozygous sublines by the process of phenotypic assortment (Bruns and Cassidy-Hanley, 2000). Nearly complete gene replacement was confirmed in three independent sublines by PCR analysis (data not shown). Each subline was mated to a wild-type strain (we previously demonstrated exchange of Pdd1p between partners) and the conjugants transformed with pMLexAop. Analysis of the transformants shows that, when the LexA-Pdd1p fusion protein is provided by the parental MAC of one mating partner, excision of the modified M element occurs with fairly high efficiency (Figure 6B). Sequencing of several of these excision PCR products confirms that they are produced from the MLexAop element rather than the endogenous M element. Also, the junction sequences show that the excision events are considerably less precise than with wild-type M element DNA (Figure 7). Discussion Extending the Heterochromatin-DNA Elimination Link Previously, we showed that the chromodomain proteins Pdd1p and Pdd3p associate with eliminated DNA in electron-dense heterochromatin-like structures (Madireddi et al., 1996; Nikiforov et al., 2000). Furthermore, eliminating parental expression of Pdd1p or the colocalized Pdd2p disrupts DNA elimination (Nikiforov et al., 1999; Coyne et al., 1999). These results led us to suggest a parallel between programmed DNA elimination in Tetrahymena and the formation of transcriptionally repressed heterochromatin in other organisms (reviewed in Coyne et al., 1996; Yao et al., 2002). In this report, we show that the same epigenetic mark that characterizes transcriptionally repressed heterochromatin, Me(Lys9)H3, also plays a central role in programmed DNA elimination. The first chromodomain of Pdd1p and only chromodomain of Pdd3p bind specifically to Me(Lys9)H3 peptides with affinities equivalent to the chromodomain of HP1 (Jacobs and Khorasanizadeh, 2002). Temporally, Me(Lys9)H3 appears only after the formation of anlagen, and its peak abundance corresponds

to when Pddps and DNA destined for elimination begin to form heterochromatin-like structures. Both Pdd1p and Me(Lys9)H3 peak about 3 hr before most programmed DNA elimination events are observed (Yokoyama and Yao, 1982; Austerberry et al., 1984; Saveliev and Cox, 1995), and then decline. No histone demethylase activity has yet been observed, but the decrease in Me(Lys9)H3 could possibly result from such an activity or the turnover of methylated H3 and/or replacement with the variant histone hv2, which is very actively synthesized and targeted to anlagen during this period (Allis and Wiggins, 1984). At a finer resolution, ChIP analyses reveal that both Pdd1p and Me(Lys9)H3 associate with IESs but not adjacent chromatin that is retained in the MAC. Disruption of the parental PDD1 gene causes drastic reduction in H3 lysine 9 methylation. Finally, tethering of Pdd1p is sufficient to induce elimination of an otherwise defective IES. The precise mechanism by which the tethering promotes excision is uncertain. Because the event occurs only in the rare transformed cells, we were unable to determine if LexA-Pdd1p recruits an H3-methylase to the modified IES or directly recruits the excision machinery to unmethylated chromatin, substituting for a Pdd1p-Me(Lys9)H3 interaction. In either case, our results strongly support a central role for Me(Lys9)H3 in targeting Pddp-containing DNA excision machinery to sites of programmed elimination in Tetrahymena. The Methyl-Acetyl Balance Such a model is consistent with the recent observation that treatment of conjugating Tetrahymena with the histone deacetylase inhibitor trichostatin A interferes with Pdd1p localization and DNA elimination (Duharcourt and Yao, 2002). Acetylation at lysine 9 of H3 inhibits methylation of this residue (Rea et al., 2000; Nakayama et al., 2001), and thus deacetylation is a likely prerequisite to DNA elimination. Indeed, our ChIP analyes generally show a negative association between H3 lysine 9 acetylation and eliminated chromatin (Figure 4). A functional interaction between deacetylation and methylation has been observed in transcriptional repression as well (Czermin et al., 2001; Hwang et al., 2001). Since Me(Lys9)H3 might be assumed to precede

Cell 708

Pdd1p binding to chromatin, our finding that the absence of parentally expressed Pdd1p drastically reduces the level of Me(Lys9)H3 was unexpected. One possible explanation would be that the reduction or delay in Pdd1p accumulation prevents histone deacetylation, thus blocking methylation. Another possibility is that stabilization and/or propagation of the methyl mark requires binding by the chromodomain of Pdd1p. Either of these mechanisms would imply an exquisitely dosage-sensitive response to Pdd1p levels and/or developmental timing. Other chromatin-associated proteins are known to exhibit highly dosage-sensitive phenotypes, as for example, in the case of position effect variegation (Weiler and Wakimoto, 1995).

tion, and that accumulation of these RNAs is Pdd1pdependent. It is interesting to note that some chromodomains have RNA binding activity (Akhtar et al., 2000) and that Pdd1p’s third globular domain, characterized in one study as a divergent chromodomain (Callebaut et al., 1997), shares sequence similarity with an RNA binding region of the protein SUI1 (Aravind and Koonin, 1999, and our unpublished observations). If an RNA/ Pdd1p association (direct or indirect) between three and seven hours of conjugation is required for targeting of methyltransferase activity to IESs, this would explain the drastic loss of Me(Lys9)H3 in mating cells carrying a parental PDD1 knockout, which completely lack Pdd1p during this period.

How Is Methylation Targeted? Our proposal that localization of Pddps is targeted by Me(Lys9)H3 simply begs the question of how methylation is targeted specifically to IESs. It has been estimated that about 6000 DNA elimination events may occur during Tetrahymena MAC development (reviewed in Coyne et al., 1996; Yao et al., 2002). Coordinated excision of these elements, but not adjacent MACretained sequences, must require a marking mechanism of considerable specificity. Again, we can return to the parallel phenomenon of transcriptional repression for potential explanations. Repression of euchromatic genes by HMTases and HP1 can be mediated by association with gene-specific transcriptional corepressors, such as Rb (Nielsen et al., 2001) and members of the KRAB family (Schultz et al., 2002). Potentially, factors that bind IESs could target the modification of their associated chromatin. In fact, cis-acting sequences essential for elimination have been localized within the M and R elements and are potential binding sites for such putative factors (described in Yao et al., 2002). However, the boundaries of such sequences have been difficult to define, and to date, no recognizable shared sequences have been identified between multiple IESs. Another possibility is that non-coding RNA molecules play a role in IES recognition. Increasing evidence points to the involvement of non-coding RNAs in the regulation of chromatin structure (reviewed in Zamore, 2002; Matzke et al., 2001). Several lines of evidence suggest a possible involvement of RNA in ciliate programmed DNA elimination as well. Chalker and Yao (2001) have shown that bidirectional transcription of IESs occurs between approximately three and nine hours of conjugation. Inhibiting this transcription perturbs IES excision. Although the transcripts apparently vanish prior to DNA elimination, it is possible they have become digested into small fragments undetectable by the blotting methods used, as routinely found with small interfering RNAs. In addition, numerous studies have documented the ability of the parental MAC to influence processing of DNAs in the developing anlagen. Because this influence is sequence-specific and transmissible through the cytoplasm, an RNA component has been hypothesized to be involved (Meyer and Garnier, 2002). Indeed, the accompanying report by Mochizuki et al. (2002 [this issue of Cell]) provides strong evidence that small, MICspecific RNAs play a role in programmed DNA elimina-

Experimental Procedures Cell Culture T. thermophila strains CU 427 and CU 428 were used as wild-type. These were provided by Dr. Peter Bruns (Cornell University, Ithaca, NY). Conjugation was performed as previously described (Allis and Dennison, 1982). Preparation of Nuclei Nuclei were isolated as described previously (Gorovsky et al., 1975), except that the nucleus isolation buffer contained 1 mM phenylmethylsulfonyl fluoride (PMSF) and 10 mM sodium butyrate, but not spermidine. Purification of nuclei by sedimentation was performed as previously described (Allis and Dennison, 1982). Electrophoresis and Immunoblotting SDS-PAGE was performed as previously described (Laemmli, 1970; Guttman et al., 1980). Approximately 1 ⫻ 105 cells equivalent of protein was electrophoresed per lane. Purified nuclei were acidextracted with 0.4 N H2SO4 and precipitated with 20% TCA. Immunoblot analyses were done as previously described (Madireddi et al., 1994). Balanced protein loads were ensured by staining immunoblots with Ponceau S. Immune sera were diluted in TBST as follows: ␣-Pdd1p 1:2000, ␣-Me(Lys9)H3 1:2000, ␣-general H4 1:3000, and ␣-Me(Lys4)H3 1:3000. Immunoreactivity was detected using an ECL Plus kit (Amersham). For competition experiments, 1 ␮g/ml of peptide was added to the primary antibody prior to incubation. Immunofluorescence Mating cells were fixed in Schaudin’s fixative and processed as described previously (Madireddi et al., 1994). Digital images were collected using either SPOT or Openlab software on a Zeiss Axioskop 2 Plus microscope. Codon Alteration and Expression of Recombinant T. thermophila Chromodomains To convert all TAG and TAA glutamine codons in the PDD1 gene to the conventional CAG and CAA codons, we performed DNA mutagenesis as previously described (Nikiforov et al., 2000). The Pdd3p chromodomain was cloned from the full-length clone (Nikiforov et al., 2000) into the pET-11a vector (Novagen). Both Pdd1p and Pdd3p chromodomains were cloned with an N-terminal 6⫻ His tag using primer pairs: for PDD1 5⬘-GGAATTCCATATGAAAAAACACCACCAC CACCACCACGAGGATCAATATGAA -3⬘ and 5⬘-CCGGGATCCTCAT TTTTGCTTCTTTTC-3⬘; for PDD3 5⬘-GGAATTCCATATGAAAAAAC ACCACCACCACCACCACCAACAAGAATATGAA -3⬘ and 5⬘-CCGGG ATCCTCAGTTTTTCTTCTTAAA-3⬘. Induction and purification of recombinant proteins was performed as previously described (Jacobs et al., 2001). Fluorescence Anisotropy Protein concentrations were determined by absorbance spectroscopy using predicted extinction coefficients (for Pdd1p CD1, ⑀280 ⫽ 20910 M⫺1 cm⫺1; for Pdd3p CD1, ⑀280 ⫽ 22190 M⫺1 cm⫺1). Peptide concentrations were estimated from the mass, except in tyrosine-

Histone H3 Methylation Marks DNA for Elimination 709

containing peptides, where absorbance spectroscopy using the extinction coefficient, ⑀280 ⫽ 1280 M⫺1 cm⫺1, was used. Fluorescence anisotropy analysis was performed as described earlier (Jacobs et al., 2001). Peptides were synthesized at the Baylor College of Medicine Protein Core Facility (Houston, TX). The peptide nomenclature was as follows: Unmodified, NH2-ARTKQTARKSTGG KAY-COOH; Methyl-K9 H3, NH2-ARTKQTARK(Me3)STGGKAY-COOH; Methyl-K4 H3, NH2-ARTK(Me2)QTARKSTGGKAPRKQLC-COOH; Methyl-K27 H3, NH2-APRKQLATKAAR K(Me2)SAPSTY-COOH. Fluorescence anisotropy binding assays (Klotz, 1997) were performed on a Beacon 2000 fluorescence polarization system (PanVera). Peptides were N-terminally labeled with fluorescein succinimidyl ester. Binding assays were performed in 20 mM sodium phosphate buffer containing 25 mM NaCl and 2 mM DTT in the presence of 120 nM fluorescein-labeled peptide. Chromatin Immunoprecipitation Analyses Cells were crosslinked, harvested, and washed as described previously (Dedon et al., 1991), except that cells were resuspended at 5 ⫻ 107 cells/mL in SDS resuspension buffer and disrupted by sonication as previously described (Chadee et al., 1999). Cell lysates were immunoprecipitated using ␣-Pdd1p, ␣-Me (Lys9) H3, ␣-general H4, and ␣-Ac H3 (recognizing H3 acetylated at either Lys-9 or Lys14). Immunoprecipitation and washing of the protein A bound-chromatin were also done as described previously (Boggs et al., 2002). PCR was performed using Hot Star Taq (Qiagen, Inc.). Sequences of the primers are available from the authors upon request. PCR products were resolved on 2% agarose gels containing 0.5 ␮g/ml ethidium bromide. Fluorescence intensities were quantified using Alpha Imager 2000 software (Alpha Innotech). Tetrahymena Genetic Manipulations The LexA-PDD1 fusion was constructed from three PCR-amplified fragments. The upstream region of PDD1 was amplified using the primers 5⬘-GATACCGCGGTTAAATAGCAAAACAAGCTACTC-3⬘ and 5⬘-TGGCTCTAGACAAAGCATTAATCAATAGTT-3⬘ and inserted into the SacI and XbaI sites of pBluescript-SK (Stratagene, Inc.). The coding plus downstream regions of PDD1 were amplified with 5⬘CGCGGGATCCTATGTTTACTGTGAAATCACT-3⬘ and 5⬘-GCATGG TACCCATGTGCATGAAGAATAAGAATACTTTG-3⬘ and inserted into the KpnI and BamHI sites of the above plasmid. The complete E. coli LexA coding region was amplified with 5⬘-TGGCTCTAGAAT GAAAGCGTTAACGGCCAG-3⬘ and 5⬘-CGGGATCCAGCCAGTCGC CGTTGCGAA-3⬘ from the plasmid pLex2A (a gift of S. Hollenberg) and inserted into the BamHI and XbaI sites of the second plasmid to generate pBSLexA-PDD1. This plasmid was digested with KpnI and SacI to release the transforming fragment, which was cotransformed into CU428 cells with HindIII-digested pRpL29A (Yao and Yao, 1991) using standard biolistic procedures (Bruns and CassidyHanley, 2000). Transformants were selected by resistance to cycloheximide and successful cotransformation was confirmed by PCR with the primers 5⬘-GAGATTTCATCCGAATTATTTGAGTC-3⬘ and 5⬘CTCTTCAGTATCAGAGCTATAGTC-3⬘, which flank the LexA insertion site. To select for phenotypic assortants with increased copy number of the LexA-PDD1 allele, several single cells from three independent transformant lines were cloned into media droplets, allowed to proliferate and the PCR repeated. Single cells from the most promising subclones were then recloned and the process repeated until essentially complete replacement of the endogenous allele was achieved. Plasmid p⌬6K is a derivative of pH8-M2A (Godiska and Yao, 1990; R. Godiska, C.-H. Yao and M.-C. Yao, personal communication) in which 592 bp internal to the M element are deleted and replaced with a unique KpnI site. Into this site was ligated a 210 bp PCR product containing four copies of the 36 bp LexA operator, derived from p5H18-34⌬Spe (a gift of S. Hollenberg) to generate the plasmid pMLexAop. CU428 cells carrying the LexA-PDD1 fusion (or not, as a control) were mated to CU427 and transformed with pMLexAop by standard electroporation (Gaertig and Kapler, 2000). Transformants were selected by resistance to paromomycin. Aliquots of transformed cells were transferred to a 10⫻ volume of 50 mM KCl, 10 mM Tris, [pH 9.0], 0.1% Triton X-100, 2 mM MgCl2, 0.1 ␮g/ml Proteinase K, and

incubated for 30 min at 55⬚C and 10 min at 95⬚C. The lysate was PCR amplified using primers 5⬘-TACGATAGATCGACTGACGG and 5⬘-GTGGGGAGGGAGAAGGATTCAAC-3⬘, which flank the M element. Rearranged products were gel-purified, cloned into pCR4TOPO (Invitrogen Corp.), and sequenced using an Applied Biosystems DNA sequencer. Acknowledgments This work was supported by grants from the National Institutes of Health to C.D.A. (GM-53512) and from the National Science Foundation to R.S.C. (MCB-9808573 and MCB-0096270). We are grateful to Dr. Mikhail Nikiforov for his many helpful suggestions; Drs. Sepideh Khorasanizadeh, Donna Cassidy-Hanley, and Heather McDonald for helpful advice; Joseph Fontana for assistance with the LexA-PDD1 construction; Dr. Wolfgang Fischle for preparation of fluoresceinlabeled peptides; and Drs. Meng-Chao Yao and Larry Klobutcher for providing constructs and unpublished sequence information. R.S.C. wishes to gratefully acknowledge Dr. Peter Bruns, in whose laboratory at Cornell University part of this work was conducted. Received: June 24, 2002 Revised: August 13, 2002 Published online: September 6, 2002 References Akhtar, A., Zink, D., and Becker, P.B. (2000). Chromodomains are protein-RNA interaction modules. Nature 407, 405–409. Allis, C.D., and Dennison, D.K. (1982). Identification and purification of young macronuclear anlagen from conjugating cells of Tetrahymena thermophila. Dev. Biol. 93, 519–533. Allis, C.D., and Wiggins, J.C. (1984). Histone rearrangements accompany nuclear differentiation and dedifferentiation in Tetrahymena. Dev. Biol. 101, 282–294. Aravind, L., and Koonin, E.V. (1999). Novel predicted RNA-binding domains associated with the translation machinery. J. Mol. Evol. 48, 291–302. Austerberry, C.F., Allis, C.D., and Yao, M.C. (1984). Specific DNA rearrangements in synchronously developing nuclei of Tetrahymena. Proc. Natl. Acad. Sci. USA 81, 7383–7387. Bannister, A.J., Zegerman, P., Partridge, J.F., Miska, E.A., Thomas, J.O., Allshire, R.C., and Kouzarides, T. (2001). Selective recognition of methylated lysine 9 on histone H3 by the HP1 chromo domain. Nature 410, 120–124. Boggs, B.A., Cheung, P., Heard, E., Spector, D.L., Chinault, A.C., and Allis, C.D. (2002). Differentially methylated forms of histone H3 show unique association patterns with inactive human X chromosomes. Nat. Genet. 30, 73–76. Briggs, S.D., Bryk, M., Strahl, B.D., Cheung, W.L., Davie, J.K., Dent, S.Y., Winston, F., and Allis, C.D. (2001). Histone H3 lysine 4 methylation is mediated by Set1 and required for cell growth and rDNA silencing in Saccharomyces cerevisiae. Genes Dev. 15, 3286–3295. Bruns, P.J., and Cassidy-Hanley, D. (2000). Biolistic transformation of macro- and micronuclei. Methods Cell Biol. 62, 501–512. Callebaut, I., Courvalin, J.C., Worman, H.J., and Mornon, J.P. (1997). Hydrophobic cluster analysis reveals a third chromodomain in the Tetrahymena Pdd1p protein of the chromo superfamily. Biochem. Biophys. Res. Commun. 235, 103–107. Chadee, D.N., Hendzel, M.J., Tylipski, C.P., Allis, C.D., Bazett-Jones, D.P., Wright, J.A., and Davie, J.R. (1999). Increased Ser-10 phosphorylation of histone H3 in mitogen-stimulated and oncogenetransformed mouse fibroblasts. J. Biol. Chem. 274, 24914–24920. Chalker, D.L., and Yao, M.C. (2001). Nongenic, bidirectional transcription precedes and may promote developmental DNA deletion in Tetrahymena thermophila. Genes Dev. 15, 1287–1298. Chalker, D.L., La Terza, A., Wilson, A., Kroenke, C.D., and Yao, M.C. (1999). Flanking regulatory sequences of the Tetrahymena R deletion element determine the boundaries of DNA rearrangement. Mol. Cell. Biol. 19, 5631–5641.

Cell 710

Chau, M.-F., and Orias, E. (1996). Developmentally programmed DNA rearrangements in Tetrahymena thermophila: isolation and sequence characterization of three new alternative deletion systems. Biol. Cell 86, 111–120.

(2001). Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins. Nature 410, 116–120.

Coyne, R.S., Chalker, D.L., and Yao, M.C. (1996). Genome downsizing during ciliate development: nuclear division of labor through chromosome restructuring. Annu. Rev. Genet. 30, 557–578.

Lin, R., Cook, R.G., and Allis, C.D. (1991). Proteolytic removal of core histone amino termini and dephosphorylation of histone H1 correlate with the formation of condensed chromatin and transcriptional silencing during Tetrahymena macronuclear development. Genes Dev. 5, 1601–1610.

Coyne, R.S., Nikiforov, M.A., Smothers, J.F., Allis, C.D., and Yao, M.C. (1999). Parental expression of the chromodomain protein Pdd1p is required for completion of programmed DNA elimination and nuclear differentiation. Mol. Cell 4, 865–872. Czermin, B., Schotta, G., Hulsmann, B.B., Brehm, A., Becker, P.B., Reuter, G., and Imhof, A. (2001). Physical and functional association of SU(VAR)3–9 and HDAC1 in Drosophila. EMBO Rep 2, 915–919. Dedon, P.C., Soults, J.A., Allis, C.D., and Gorovsky, M.A. (1991). Formaldehyde cross-linking and immunoprecipitation demonstrate developmental changes in H1 association with transcriptionally active genes. Mol. Cell. Biol. 11, 1729–1733. Duharcourt, S., and Yao, M.-C. (2002). Role of histone deacetylation in developmentally programmed DNA rearrangements in Tetrahymena thermophila. Eukaryotic Cell 1, 293–303. Gaertig, J., and Kapler, G. (2000). Transient and stable DNA transformation of Tetrahymena thermophila by electroporation. Methods Cell Biol. 62, 485–500. Godiska, R., and Yao, M.C. (1990). A programmed site-specific DNA rearrangement in Tetrahymena thermophila requires flanking polypurine tracts. Cell 61, 1237–1246. Godiska, R., James, C., and Yao, M.C. (1993). A distant 10-bp sequence specifies the boundaries of a programmed DNA deletion in Tetrahymena. Genes Dev. 7, 2357–2365.

Laemmli, U.K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685.

Litt, M.D., Simpson, M., Gaszner, M., Allis, C.D., and Felsenfeld, G. (2001). Correlation between histone lysine methylation and developmental changes at the chicken ␤-globin locus. Science 293, 2453– 2455. Madireddi, M.T., Davis, M.C., and Allis, C.D. (1994). Identification of a novel polypeptide involved in the formation of DNA-containing vesicles during macronuclear development in Tetrahymena. Dev. Biol. 165, 418–431. Madireddi, M.T., Coyne, R.S., Smothers, J.F., Mickey, K.M., Yao, M.C., and Allis, C.D. (1996). Pdd1p, a novel chromodomain-containing protein, links heterochromatin assembly and DNA elimination in Tetrahymena. Cell 87, 75–84. Matzke, M., Matzke, A.J., and Kooter, J.M. (2001). RNA: guiding gene silencing. Science 293, 1080–1083. Meyer, E., and Garnier, O. (2002). Non-Mendelian inheritance and homology-dependent effects in ciliates. Adv. Genet. 46, 305–337. Mochizuki, K., Fine, N.A., Fujisawa, T., and Gorovsky, M.A. (2002). Analysis of a piwi-related gene implicates small RNAs in genome rearrangement in Tetrahymena. Cell 110, this issue, 689–699. Muller, J. (1995). Transcriptional silencing by the Polycomb protein in Drosophila embryos. EMBO J. 14, 1209–1220.

Gorovsky, M.A., Yao, M.C., Keevert, J.B., and Pleger, G.L. (1975). Isolation of micro- and macronuclei of Tetrahymena pyriformis. Methods Cell Biol. 9, 311–327.

Nakayama, J., Klar, A.J., and Grewal, S.I. (2000). A chromodomain protein, Swi6, performs imprinting functions in fission yeast during mitosis and meiosis. Cell 101, 307–317.

Guttman, S.D., Glover, C.V., Allis, C.D., and Gorovsky, M.A. (1980). Heat shock, deciliation and release from anoxia induce the synthesis of the same set of polypeptides in starved T. pyriformis. Cell 22, 299–307.

Nakayama, J., Rice, J.C., Strahl, B.D., Allis, C.D., and Grewal, S.I. (2001). Role of histone H3 lysine 9 methylation in epigenetic control of heterochromatin assembly. Science 292, 110–113.

Heinonen, T.Y., and Pearlman, R.E. (1994). A germ line-specific sequence element in an intron in Tetrahymena thermophila. J. Biol. Chem. 269, 17428–17433.

Nielsen, P.R., Nietlispach, D., Mott, H.R., Callaghan, J., Bannister, A., Kouzarides, T., Murzin, A.G., Murzina, N.V., and Laue, E.D. (2002). Structure of the HP1 chromodomain bound to histone H3 methylated at lysine 9. Nature 416, 103–107.

Hwang, K.K., Eissenberg, J.C., and Worman, H.J. (2001). Transcriptional repression of euchromatic genes by Drosophila heterochromatin protein 1 and histone modifiers. Proc. Natl. Acad. Sci. USA 98, 11423–11427.

Nielsen, S.J., Schneider, R., Bauer, U.M., Bannister, A.J., Morrison, A., O’Carroll, D., Firestein, R., Cleary, M., Jenuwein, T., Herrera, R.E., and Kouzarides, T. (2001). Rb targets histone H3 methylation and HP1 to promoters. Nature 412, 561–565.

Jacobs, S.A., and Khorasanizadeh, S. (2002). Structure of HP1 chromodomain bound to a lysine 9-methylated histone H3 tail. Science 295, 2080–2083.

Nikiforov, M.A., Smothers, J.F., Gorovsky, M.A., and Allis, C.D. (1999). Excision of micronuclear-specific DNA requires parental expression of Pdd2p and occurs independently from DNA replication in Tetrahymena thermophila. Genes Dev. 13, 2852–2862.

Jacobs, S.A., Taverna, S.D., Zhang, Y., Briggs, S.D., Li, J., Eissenberg, J.C., Allis, C.D., and Khorasanizadeh, S. (2001). Specificity of the HP1 chromo domain for the methylated N-terminus of histone H3. EMBO J. 20, 5232–5241. James, T.C., Eissenberg, J.C., Craig, C., Dietrich, V., Hobson, A., and Elgin, S.C. (1989). Distribution patterns of HP1, a heterochromatinassociated nonhistone chromosomal protein of Drosophila. Eur. J. Cell Biol. 50, 170–180.

Nikiforov, M.A., Gorovsky, M.A., and Allis, C.D. (2000). A novel chromodomain protein, pdd3p, associates with internal eliminated sequences during macronuclear development in Tetrahymena thermophila. Mol. Cell. Biol. 20, 4128–4134. Noma, K., Allis, C.D., and Grewal, S.I. (2001). Transitions in distinct histone H3 methylation patterns at the heterochromatin domain boundaries. Science 293, 1150–1155.

Jenuwein, T., and Allis, C.D. (2001). Translating the histone code. Science 293, 1074–1080.

Platero, J.S., Hartnett, T., and Eissenberg, J.C. (1995). Functional analysis of the chromo domain of HP1. EMBO J. 14, 3977–3986.

Katoh, M., Hirono, M., Takemasa, T., Kimura, M., and Watanabe, Y. (1993). A micronucleus-specific sequence exists in the 5⬘-upstream region of calmodulin gene in Tetrahymena thermophila. Nucleic Acids Res. 21, 2409–2414.

Poux, S., McCabe, D., and Pirrotta, V. (2001). Recruitment of components of Polycomb group chromatin complexes in Drosophila. Development 128, 75–85.

Klotz, I.M. (1997). Ligand-Receptor Energetics (New York: John Wiley and Sons). Kornberg, R.D., and Lorch, Y. (1999). Twenty-five years of the nucleosome, fundamental particle of the eukaryote chromosome. Cell 98, 285–294. Lachner, M., and Jenuwein, T. (2002). The many faces of histone lysine methylation. Curr. Opin. Cell Biol. 14, 286–298. Lachner, M., O’Carroll, D., Rea, S., Mechtler, K., and Jenuwein, T.

Rea, S., Eisenhaber, F., O’Carroll, D., Strahl, B.D., Sun, Z.W., Schmid, M., Opravil, S., Mechtler, K., Ponting, C.P., Allis, C.D., and Jenuwein, T. (2000). Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature 406, 593–599. Saveliev, S.V., and Cox, M.M. (1995). Transient DNA breaks associated with programmed genomic deletion events in conjugating cells of Tetrahymena thermophila. Genes Dev. 9, 248–255. Schultz, D.C., Ayyanathan, K., Negorev, D., Maul, G.G., and Rauscher, F.J., 3rd. (2002). SETDB1: a novel KAP-1-associated his-

Histone H3 Methylation Marks DNA for Elimination 711

tone H3, lysine 9-specific methyltransferase that contributes to HP1mediated silencing of euchromatic genes by KRAB zinc-finger proteins. Genes Dev. 16, 919–932. Smothers, J.F., Madireddi, M.T., Warner, F.D., and Allis, C.D. (1997a). Programmed DNA degradation and nucleolar biogenesis occur in distinct organelles during macronuclear development in Tetrahymena. J. Eukaryot. Microbiol. 44, 79–88. Smothers, J.F., Mizzen, C.A., Tubbert, M.M., Cook, R.G., and Allis, C.D. (1997b). Pdd1p associates with germline-restricted chromatin and a second novel anlagen-enriched protein in developmentally programmed DNA elimination structures. Development 124, 4537– 4545. Strahl, B.D., and Allis, C.D. (2000). The language of covalent histone modifications. Nature 403, 41–45. Strahl, B.D., Ohba, R., Cook, R.G., and Allis, C.D. (1999). Methylation of histone H3 at lysine 4 is highly conserved and correlates with transcriptionally active nuclei in Tetrahymena. Proc. Natl. Acad. Sci. USA 96, 14967–14972. van Holde, K.E. (1988). Histone modifications. In Chromatin, A. Rich, ed. (New York: Springer), pp. 111–148. Weiler, K.S., and Wakimoto, B.T. (1995). Heterochromatin and gene expression in Drosophila. Annu. Rev. Genet. 29, 577–605. Wolffe, A.P. (1998). Chromatin: Structure and Function (San Diego: Academic Press) Wolffe, A.P., and Hayes, J.J. (1999). Chromatin disruption and modification. Nucleic Acids Res. 27, 711–720. Yao, M.-C., and Yao, C.H. (1991). Transformation of Tetrahymena to cycloheximide resistance with a ribosomal protein gene through sequence replacement. Proc. Natl. Acad. Sci. USA 88, 9493–9497. Yao, M.-C., Duharcourt, S., and Chalker, D.L. (2002). Genome-wide rearrangements of DNA in ciliates. In Mobile DNA II, N. Craig, R. Craigie and M. Gellert, eds.(Washington D.C.: ASM Press), pp. 730–758. Yokoyama, R.W., and Yao, M.-C. (1982). Elimination of DNA sequences during macronuclear differentiation in Tetrahymena thermophila, as detected by in situ hybridization. Chromosoma 85, 11–22. Zamore, P.D. (2002). Ancient pathways programmed by small RNAs. Science 296, 1265–1269.