Chromodomain protein Tcd1 is required for macronuclear genome rearrangement and repair in Tetrahymena

The survival of an organism’s progeny depends on the maintenance of its genome. Programmed DNA rearrangement and repair in Tetrahymena occur during the differentiation of the developing somatic macronuclear genome from the germ line micronuclear genome. Tetrahymena chromodomain protein (Tcd1) exhibited dynamic localization from the parental to the developing macronuclei. In the developing macronuclei, Tcd1 colocalized with Pdd1 and H3K9me3. Furthermore, Tcd1 colocalized with Pdd1 in the conjusome and “donut structure” of DNA elimination heterochromatin region. During the growth and conjugation stages, TCD1 knockout cells appeared normal and similar to wild-type strains. In addition, these knockout cells proceeded to the 2MAC-1MIC stage. However, the progeny of the TCD1 knockout cells did not grow upon return to SPP medium and eventually died. The deletion of the internal elimination sequence R element was partially disrupted in the developing new macronuclei. Gamma H2A staining showed that Tcd1 loss induced the accumulation of DNA double-strand breaks and the failure of genome repair. These results suggest that the chromodomain protein Tcd1 is required for the rearrangement and repair of new macronuclear genome in Tetrahymena.

an epigenetically regulated RNAi-related mechanism 7 . During early conjugation stage, the micronuclear genome transcribes and produces double-stranded RNAs that are diced into 28 to 30 nucleotide scn-RNA by the Dicer-like protein Dcl1 8,9 . The scnRNAs then form complexes with the Argonaute protein Twi1 10 . The RNA helicase Ema1 facilitates interaction between the Twi1-scnRNA complex and the nascent ncRNAs in the parental MAC 11 . In the new MACs, the IES-specific scnRNA-Twi1 complexes target homologous sequences for chromatin modifications via the accumulation of H3K27me3 and H3K9me3. Furthermore, the chromodomain protein Pdd1 binds both H3K9me3 and H3K27me3, pdd3 preferentially recognizes H3K9me3 12 . The domesticated piggyBac transposase Tpb2p is responsible for the DNA cleavage during programmed DNA deletion 13 . The nonhomologous end-joining (NHEJ) core component TKu80 protein serves an end-protective role after DNA cleavage 14 . In the final step, Ligase IV and Xrcc4p are required for the covalent joining of the two broken ends 2 . A higher order heterochromatin structure is essential during this process to achieve effective genome rearrangement and NHEJ genome repair.
A chromatin organization modifier (chromo) domain is a 40 to 50 amino acid region that was originally identified as a conserved sequence motif between polycomb proteins and heterochromatin protein-1 (HP1) from Drosophila melanogaster 15,16 . Chromodomain-containing proteins can interact with diverse targets, including histones, DNA, and RNA 17 . The three "caging" aromatic residues of a chromodomain are necessary for binding trimethylated H3K9 and H3K27 18 . The tandem chromodomains of human CHD1 act cooperatively to specifically bind methylated H3K4 19 . The histone acetyltransferase MOF specifically binds through its chromodomain to roX2 RNA 20 . Chromodomains at the integrase C termini of chromoviruses direct the integration of retrotransposons to heterochromatin 21 . Recent studies have revealed that HP1 promotes early DNA damage response events, which affects the efficiency of DNA repair in mammals 22 . These reports suggest that chromodomains have evolved from a common ancestral fold to fulfill various functions in different molecular contexts and execute multiple functions depending on the developmental context 17,23 .
In the present study, we characterized a novel chromodomain protein Tcd1 (Tetrahymena Chromodomain protein 1). Tcd1 colocalized with H3K9me3 and Pdd1 in the developing macronuclei, specifically in the conjusome body and the IES elimination region. The progeny of the TCD1 germline knockout cells did not grow upon returning to SPP medium and eventually died. The deletion of the internal elimination sequence R element was disrupted in the newly developing new macronuclei and γ -H2A.X foci were maintained throughout the late conjugation stage in TCD1 knockout strains. These results suggest that Tcd1 plays roles in genome rearrangement and repair in Tetrahymena.

Results
Identification of TCD1. To understand the function of chromodomain-containing proteins in genome organization, we identified 14 chromodomain-containing genes (including PDD1, PDD3, and HHP1) in the Tetrahymena genome database (http://www.ciliate.org) via a BLAST search with the conserved chromodomain sequence. Among these genes, TTHERM_01337400 (henceforth named TCD1) was not expressed in growing and starved cells; however, the mRNA expression of this gene was initiated at 2 h of mating and significantly increased between 8 h and 12 h of mating (Figs. 1b,1c). RT-PCR, 5′ -RACE, and 3′ -RACE were performed to obtain the complete mRNA sequence of TCD1. The open reading frame, 5′ UTR, and 3′ UTR of this gene comprise 2172, 37, and 165 bases, respectively. A comparison of cDNA and genomic DNA indicated that the cDNA contained four exons and encoded a putative 723 amino acid protein. Primary sequence analysis showed that Tcd1 has two chromodomains and one chromoshadow domain (labeled CD1, CD2, and CSD in Fig. 1a). CD1 resembles the canonical chromodomains in its primary sequence, whereas CD2 shares 38% identity with CD1. The expression pattern of TCD1 is consistent with the microarray data of TCD1 in the Tetrahymena Functional Gene Database (http:// tfgd.ihb.ac.cn) 24 . The conjugation-specific expression patterns suggest that Tcd1 has important functions during the conjugation stage.

Tcd1 Is Essential for Viable Progeny.
To investigate the function of Tcd1, we first disrupted the TCD1 gene in the MAC by homology-directed gene replacement. Similar to wild-type (WT) strains, The TCD1 somatic knockout strains grew normally during vegetative stage. The development of mating TCD1 somatic knockout strains is also normal and generated viable progeny during conjugation stage. Then we disrupted the TCD1 gene in the MAC and MIC. The complete replacement of TCD1 in the homozygous homokaryon Δ TCD1-6 and Δ TCD1-8 strains (lacking TCD1 in both the somatic MAC and the germline MIC) was confirmed by Southern blotting analysis (Fig. 2b). Similar to WT strains, the Δ TCD1-6 and Δ TCD1-8 strains exhibited normal vegetative growth. Considering that TCD1 was not expressed during the growth and starvation stages, we speculated that TCD1 is dispensable at the vegetative stage.
To elucidate the function of Tcd1 during sexual development, the mating progress of Δ TCD1-6 and Δ TCD1-8 cells was evaluated, and the development stages were analyzed. Although Δ TCD1 strains mating was delayed and maximum mating was approximately 10% lower than that in WT cells, their sexual development stage was normal and finally stay at the 2MAC-1MIC stage ( Supplementary Fig.  S1). The Δ TCD1 mating pairs then were cloned into SPP medium drops and their progeny survival was assessed. The progeny of the Δ TCD1 cells did not grow upon return to nutrient SPP and eventually died (Figs. 2c,2d). The conjugation-mediated transfer of proteins has been previously reported between two mating cells in Tetrahymena 25 . The Δ TCD1 cells were then mated with WT cells. Progeny viability was rescued when Δ TCD1 cells were crossed with WT cells (Fig. 2c). The results show that the Δ TCD1 phenotype was triggered by the absence of TCD1. Therefore, Tcd1 is essential for the formation of viable conjugation progeny.

HA-Tcd1
Can Replace the Function of Endogenous Tcd1. A sequence that encodes two HA epitopes was inserted into the 5′ end of the TCD1 open reading frame to analyze the expression and localization of Tcd1 (Fig. 3a). The cassette Neo3-HA-TCD1 was introduced into the WT strain CU428 and B2086. Endogenous TCD1 gene was completely replaced with HA-TCD1, as indicated by Southern hybridization (Fig. 3b). The mating progress in HA-tagged TCD1 cells was normal and similar to that in WT cells. Furthermore, the survival of Δ TCD1 progeny can be rescued with HA-TCD1 cells (data not shown). The results indicated that HA-Tcd1 can perform the essential functions of Tcd1. Subsequently, the expression pattern of HA-Tcd1 was examined by Western blot analysis with anti-HA antibodies. A single band of approximately 85 kD was observed, which is consistent with the predicted molecular weight of HA-Tcd1 (Fig. 3c). The expression level of HA-Tcd1 was higher at 12 h to 14 h after mating, which indicates that Tcd1 accumulated at the late conjugation stage. Therefore, Tcd1 could have important functions during the late conjugation stage.   12,26,27 . Thus, we speculated that HA-Tcd1 can also colocalizes with Pdd1 and display partially functional redundancy.

HA-Tcd1 Dynamically
Aside from macronulear and anlagen localization, transient localization of HA-Tcd1 was observed in the cytoplasm; in addition, the cytoplasmic signal was close to the anterior of the mating pair, that is, the conjusome ( Supplementary Fig. S2). The nonmembrane-bound organelles were composed of a coarse reticulum and transiently appeared when developing MACs had initially emerged 28 . Pdd1 is the first component to be found in the conjusome 29 . Given its associating with Pdd1 and IES chromatin, Lia1 is also targeted to this structure 30 . Furthermore, HA-Tcd1 colocalized with Pdd1 in the conjusome ( Supplementary Fig. S3). The localization of HA-Tcd1 in the conjusome and its colocalization with Pdd1 in the structure strengthened their connection: these events could be involved in the DNA rearrangement machinery.

Tcd1
Partially Colocalizes with H3K9me3 in Developing MACs. In Tetrahymena, H3K9me3 is found only in anlagen and associated with chromatin that contains sequences targeted for elimination 31 .
To determine the possible localization of Tcd1 to regions of condensed chromatin in anlagen and its involvement in DNA elimination, we colocalized HA-Tcd1 with H3K9me3 during the anlagen stage. HA-Tcd1 dynamically colocalized with H3K9me3 in most regions during the anlagen stage. Previous studies have shown that H3K9me3 and H3k27me3 are required for heterochromatin formation and IES elimination 12,26,27,32 . Canonical chromodomain of Pdd1p interacts strongly with both H3K27me3 and K9me3, while the chromodomain of Pdd3p preferentially recognizes H3K9me3 12 . In the present study, Tcd1 colocalized with H3K9me3 at the periphery of anlagen (Figs. 5b,5c). Furthermore, the sequence alignment of the canonical chromodomain from Tcd1 and other well characterized readers like HP1, Pdd1, and Pdd3, showed the conserved three aromatic caging residues ( Supplementary Fig.  S4). The three aromatic residues are predicted to form a "cage" enclosing the methylammonium group of H3K9me3 or/and H3k27me3. The results suggest that Tcd1, similar to Pdd1 and pdd3, could bind to the heterochromatin marker H3K9me3 or H3K27me3 during the late conjugation stage in Tetrahymena.

Tcd1 and Pdd1 Colocalize to the DNA Elimination Structures. Previous data suggested that Pdd1
exists in a complex with eliminated DNA in a characteristic "doughnut-like" structure 27,33 . To understand whether or not Tcd1 also occurs in the "doughnut-like" structure, we prepared optical sections at the appropriate anlagen stage via confocal microscopy. Colocalization of Pdd1 and Tcd1 was detected with rabbit anti-Pdd1 and FITC-conjugated secondary antibodies (column II) and mouse anti-HA and rhodamine-conjugated secondary antibodies (column III), respectively. Progressive optical sections demonstrated that Tcd1 and Pdd1 colocalized in the doughnut structure (Fig. 6). A previous study reported that the Pdd1-derived doughnut structure correlates with DNA elimination in Tetrahymena 26 . Therefore, Tcd1 could also be involved in DNA elimination.
DNA Elimination is Defective in Conjugated ΔTCD1 cells. Pdd1 is required for scnRNA accumulation and IES elimination 7 . To understand whether or not Tcd1 affects scnRNA and IES elimination, the total scnRNA was analyzed during life cycle of Δ TCD1 cells. As shown in Fig. 7a, the scnRNAs generally accumulated and disappeared during the conjugation stage of Δ TCD1 cells. Subsequently, we used single-cell PCR to analyze the elimination of four IESs (M, R, Cal, and Tlr1 elements) in mating Δ TCD1 cells. The elimination of M, Cal, and Tlr1 elements in mating Δ TCD1 cells was similar to that in mating WT cells (data not shown). However, the deletion of R element was aborted in some mating cells (Fig. 7c). Previous results showed that R element is more sensitive to perturbation and more likely to show failed IES deletion than the other DNA deletion elements 4,34 . These data imply that Tcd1 is necessary for IES elimination and R element is a sensitive indicator of slight defects during DNA deletion in Tetrahymena. Tcd1 is Required for New Macronuclear Genome Repair. Partial R element elimination cannot explain the complete failure of development in Δ TCD1 progeny. Thus, we analyzed DNA repair in the newly developing macronuclear genome. γ -H2A.X signaling is an early cellular response to double-strand breaks. To determine whether or not the TCD1 knockout induces DNA damage accumulation, we checked the γ -H2A.X signal during Tetrahymena conjugation. Fig. 8 shows the immunofluorescence staining of mating WT cells and TCD1 knockout cells by using anti-γ -H2A.X antibody. γ -H2A.X staining was observed in mating WT and TCD1 knockout cells at the early meiotic and early anlagen stages. Surprisingly, γ -H2A.X staining was retained until the late anlagen stage of TCD1 knockout mating cells. By contrast, the signal disappeared at the late anlagen stage of WT mating cells. The results indicate that Tcd1 is required for genome repair during the development of new macronuclei.

Discussion
In this paper, we report a novel chromodomain-containing protein (Tcd1) from Tetrahymena. The following results suggest that Tcd1 is involved in DNA elimination and genome repair. First, TCD1 expression was undetected in growing or starved cells, but was observed in mating cells at 2 h after mating and increased significantly between 8 h and 12 h when DNA elimination occured; Tcd1 accumulated between 12 h and 14 h when genome repair had finished. Second, Tcd1 was dynamically localized from the parental MACs at the early crescent stage to the new MACs during anlagen formation. Similarly, Tcd1 colocalized with H3K9me3 and Pdd1 in the developing anlagen. Third, Tcd1 was specifically localized in the cytoplasmic conjusome and the nuclear doughnut structure, and interfered with the deletion of R element. Finally, γ -H2A.X staining was delayed to the late anlagen stage in TCD1 knockout mating cells.
A chromodomain, an evolutionarily conserved domain in a large class of chromatin-binding proteins, performs diverse functions 20,35,36 . At present, 14 chromodomain-containing proteins have been identified in Tetrahymena. Pdd1 is a well-studied member of this family and is required for DNA elimination during new MAC development 37 . Pdd3 containing one chromodomain physically associates with MIC-specific DNA sequences that are targeted for elimination 27 . The chromodomain of Pdd3 preferentially recognizes H3K9me3, the canonical chromodomain of Pdd1 interacts strongly with both H3K27me3 and H3K9me3 12,32 . The mutants disrupting H3K27me3 also affected Pdd1 localization 12 . The sequence alignment of the canonical chromodomains from different proteins showed the conserved three aromatic caging residues. The residues are predicted to form a "cage" enclosing the methylammonium group of H3K9me3 or/and H3k27me3. Similar to Pdd1 and Pdd3, Tcd1 also dynamically localized from the parental MACs to the new MACs and colocalized with H3K9me3 at the periphery of anlagen in the present study. Furthermore, Tcd1 and Pdd1 were colocalized in the doughnut structure. The similar staining pattern suggests that Tcd1 is a novel component of the DNA rearrangement machinery with newly evolved functions. The genome of Saccharomyces pombe contains at least nine chromodomain-containing proteins (e.g., Swi6, Chp1, Chp2, and Clr4). Swi6 is abundantly expressed and plays a dose-dependent role in forming a repressive structure through its self-association property 38,39 . The chromodomain of Chp1 binds to H3K9Me3 and is required for accurate chromosome segregation, whereas that of Chp2 associates with the histone deacetylase complex and is required for histone H3 lysine 14 deacetylation 40 . The Drosophila genome possesses at least five HP1 paralogs that have significantly different roles, ranging from the canonical heterochromatic function at the pericentric and telomeric regions to the exclusive localization and regulation of euchromatic genes 41 . In Tetrahymena, Pdd1 exhibits dynamic localization, apparently shuttling from the parental to the developing MACs through the conjusome 29 . As a reservoir of components from the old MACs, the conjusome stores or processes them for the newly developing MACs 29 . The recent detection of Lia1, Lia3, and Lia5 in the conjusome further supports the important  role of this structure in genome reorganization 28,29,42 . RNAi components are localized to the chromatoid body in male germ cells. Ago2 is the central protein component of the RNA-induced silencing complex and resides in cytoplasmic processing bodies (P-bodies) 43 . In consideration of the role of small RNAs in directing DNA rearrangement, the conjusome is evolutionarily equivalent to the P-bodies in ciliates 29 . Tcd1 is a novel conjusome component that could be involved in DNA rearrangement directed by an RNAi-related mechanism in Tetrahymena.
The development of new MACs in Tetrahymena is controlled by an epigenetically regulated, RNAi-related mechanism. Small RNAs are specifically expressed during conjugation do not accumulate in the absence of TWI1 or PDD1 expression 7 . By contrast, the global level of scnRNAs was not significantly affected in Δ TCD1 cells. These results argue that Tcd1 is not directly coupled with the generation and maintenance of scnRNAs in Tetrahymena. The protein could be a downstream factor of the process, and its expression profile supports this possibility.
The disruption of TCD1 did not produce an obvious phenotype during the vegetative growth and starvation stages. The expression of parental TCD1 is also not essential for sexual development. However, the expression of zygotic TCD1 is essential to produce viable conjugation progeny. Several protein components (Pdd1, Dcl1, Twi1, Ezl1, and Lia1) involved in IES elimination are required for the development of conjugation progeny 7,9,12,26,30 . All the mutant lines that abolished DNA rearrangement failed to complete development, arresting as exconjugants with two new MACs and two new MICs. However, Δ TCD1 cells completed conjugation and arrested at the exconjugant stage with two new MACs and one new MIC. Once returned to SPP medium, none of the progeny survived. Two possible reasons may explain these phenomena. One is that the elimination of R elements was inhibited in the progeny of Δ TCD1 strains. The partial deletion phenotype of IES was also previously observed in Δ EMA1, Δ TTN1/TTN2, and Δ HEN1 strains 11,34,44 . The other is that genome repair, an important process in progeny survival, cannot be completed in the developing MACs. The defect in ∆TKU80 cells involve the rejoining of the IES flanking regions that are retained in the MACs 14 . γ -H2A.X function as a docking site for protein complexes that bind to broken DNA ends and promote chromatin remodeling to join broken chromosomal DNA ends. The chromatin alterations may create structures that favor the recruitment or retention of some DNA repair proteins and thus favor a particular DNA repair pathway. HP1 proteins are involved in reorganization of high-order chromatin structure, which is essential for DNA repair 45 . Moreover, Tcd1 could probably recruit repair factors and remold the chromatin structure to fulfill genome repair at the late conjugation stage in Tetrahymena. Future experiments could further determine the different components of the Tcd1 complex and explore the mechanisms by which Tcd1 contributes to chromatin reorganization in genome repair.  Identification of TCD1. TCD1 (TTHERM_01337400) was identified via a BLAST search of the Tetrahymena genome (http://www.ciliate.org) on the basis of its homology with the fission yeast S. pombe chromodomain protein Clr4. Additional homologous regions of the Tcd1 protein sequences were identified with Lasergene 7.0 (DNA star). Genomic DNA was prepared from B2086 cells. Total RNA was extracted with Trizol reagent (Invitrogen) from mating B2086 and CU428 cells at 8 h after mixing, the extracted RNA was treated with DNase I (Invitrogen). First-strand cDNA was synthesized using SuperScript III reverse transcriptase (Invitrogen). 3′ -RACE and 5′ -RACE were performed with the SMART RACE cDNA Amplification Kit (Clontech) to determine the transcription termination and start site, respectively. The individual constructs were introduced into conjugating CU428 and B2086 cells at 2.5 h by using the Biolistic PDS-1000/He particle-delivery system (Bio-Rad) as previously described 47 . Heterozygous micronuclear transformants were screened by progeny resistance to 100 μ g/mL paromomycin and 15 μ g/ mL 6-methylpurine. Homozygous micronuclear knockout heterokaryons were generated by crossing heterozygous micronuclear transformants with B*VI or B*VII star strains. Homozygous micronuclear knockout heterokaryons were identified by paromomycin/CdCl 2 sensitivity and verified through crosses with CU427, which produced progeny resistant to 100 μ g/mL paromomycin and 25 μ g/mL cycloheximide. Homozygous homokaryon TCD1 knockout cells (Δ TCD1-6 and Δ TCD1-8) were generated by crossing homozygous heterokaryons of compatible mating types 48 . 10 μ g DNA was digested with BglII, and the resulting fragments were separated in agarose gel followed by blotting onto a nylon membrane. The membrane was probed with the labeled short PCR product of the TCD1 5′ N-terminal fragment for Southern blot analysis. The localization of the probes was visualized by autoradiography.

Generation of TCD1
To generate HA tagged TCD1 cells, the 5′ and 3′ fragments of HA-TCD1 were amplified from genomic DNA with the primer pairs TCD1-HA2-FW1/TCD1-NHA-RV and TCD1-NHA-FW/TCD1-HA2-RV1, respectively. The full HA-TCD1 sequence was obtained from the 5′ and 3′ fragment of HA-TCD1 via overlapping PCR with the primer pair TCD1-HA2-in-FW2/TCD1-HA2-in-RV2. The Neo3 cassette was amplified from pBS-MnB-3 with the prime pair TCD1-NEO-5FW/ TCD1-NEO-3RV. The over-lapping amplified product was cloned into a plasmid vector, and the Neo3 cassette was introduced into the Hpa I site in the 5′ -nontranscribed sequence of TCD1. The constructs were introduced into CU428 and B2086 cells via the Biolistic PDS-1000/He particle-delivery system. The endogenous TCD1 was completely replaced with HA-TCD1 by phenotypic assortment and selection in increasing concentrations of paromomycin. Complete replacement was confirmed by Southern hybridization.
Western Blot Analysis. Whole-cell proteins from approximately 2.5 × 10 3 cells were separated by 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and then transferred onto PVDF membranes. Blots were incubated with 1: 2, 000 dilution of mouse anti-HA antibody (Covance, Berkeley, CA, USA) in blocking solution (5% milk, 0.1% Tween 20 in PBS) and visualized by incubation with a 1: 5, 000 dilution of HRP-conjugated anti-mouse IgG antibody (Zymed Labs Inc., South San Francisco, CA, USA) in blocking solution. Finally, the blots were reacted with Western Blot Chemiluminescence Reagent (NEN Life Science, Boston, MA, USA).
Analysis of Small RNA. Total RNA was extracted with TRIzol reagent, separated on 12% polyacrylamide-urea DNA sequencing gels, and then visualized as previously described 7 .
Analysis of IES Processing. Mating pairs were cloned into drops of 10 mM Tris-HCl (pH 7.4) at 12 h after mixing and then incubated for 24 h at 30 °C. One of the separated exconjugants was transferred to 1 μ L lysis buffer, incubated at 37 °C for 30 min, and then boiled for 3 min. The cell lysate was used for the first set of PCR (40 cycles) at a final volume of 25 μ L, and 1 μ L of the first reaction was used for the second set of PCR (40 cycles). The following primers were used: R5-1/R5-2; R3-1/R3-2 7 .