Dynamic Interplay between Nucleoid Segregation and Genome Integrity in Chlamydomonas Chloroplasts1[OPEN]

Homologs of RecA recombinase and DNA gyrase are required for the proper nucleoid segregation and maintenance of genome integrity in Chlamydomonas chloroplasts. The chloroplast (cp) genome is organized as nucleoids that are dispersed throughout the cp stroma. Previously, a cp homolog of bacterial recombinase RecA (cpRECA) was shown to be involved in the maintenance of cp genome integrity by repairing damaged chloroplast DNA and by suppressing aberrant recombination between short dispersed repeats in the moss Physcomitrella patens. Here, overexpression and knockdown analysis of cpRECA in the green alga Chlamydomonas reinhardtii revealed that cpRECA was involved in cp nucleoid dynamics as well as having a role in maintaining cp genome integrity. Overexpression of cpRECA tagged with yellow fluorescent protein or hemagglutinin resulted in the formation of giant filamentous structures that colocalized exclusively to chloroplast DNA and cpRECA localized to cp nucleoids in a heterogenous manner. Knockdown of cpRECA led to a significant reduction in cp nucleoid number that was accompanied by nucleoid enlargement. This phenotype resembled those of gyrase inhibitor-treated cells and monokaryotic chloroplast mutant cells and suggested that cpRECA was involved in organizing cp nucleoid dynamics. The cp genome also was destabilized by induced recombination between short dispersed repeats in cpRECA-knockdown cells and gyrase inhibitor-treated cells. Taken together, these results suggest that cpRECA and gyrase are both involved in nucleoid dynamics and the maintenance of genome integrity and that the mechanisms underlying these processes may be intimately related in C. reinhardtii cps.

Chloroplasts (cps) are plant organelles that are responsible for photosynthesis and the supply of certain metabolites. Although cps contain their own genomic DNA, the genes encoding most of the functional cp proteins were transferred from chloroplast DNA (cpDNA) to nuclear DNA during the evolutionary process. As a result, cpDNA genomes are now small (generally 100-200 kb in size) compared with the genomes of ancestrally related cyanobacteria. The remaining cpDNA is densely packed and encodes genes essential for cp function and photosynthesis (Bungard, 2004;Green, 2011). In most plants, cpDNA is composed of four regions: two copies of a large inverted repeat, and large and small single-copy regions separated by the inverted repeats (Green, 2011). The cps usually possess 10 to 100 copies of the genomic DNA, some of which are organized into a nucleoid similar to that formed by bacterial chromosomal DNA (Kuroiwa, 1991). The cp nucleoids are composed mainly of cpDNA and constitutive nucleoid proteins (together termed the core nucleoid) and are associated with RNA and proteins involved in DNA and RNA metabolism. The nucleoid is thereby involved in the regulation of DNA replication and transcription (Powikrowska et al., 2014).
The cp nucleoids are found as small particles dispersed throughout the cp stroma in a diverse range of plant and algal taxa. However, the shape and distribution of cp nucleoids change dynamically according to cell cycle phase (Ehara et al., 1990), developmental stage (Kuroiwa et al., 1981), and metabolic status (Yehudai-Resheff et al., 2007). However, the mechanisms governing the dynamic micromorphology of cp nucleoids remain largely unknown. A few studies examined mutants defective in cp nucleoid segregation. The monokaryotic chloroplast (moc) mutants exhibiting impaired cp nucleoid dispersion were obtained by insertional mutagenesis in Chlamydomonas reinhardtii (Misumi et al., 1999). The moc mutants contained a single large cp nucleoid that was comparable in size to the nucleus, but the overall amount of cpDNA was slightly lower than in wild-type cells. The mutants exhibited aberrant cp nucleoid segregation, which resulted in cells with lower amounts of cpDNA. However, cpDNA copy number recovered quickly, probably as a result of accelerated cpDNA replication (Misumi et al., 1999). The gene responsible for moc phenotypes remains unidentified. Arabidopsis (Arabidopsis thaliana) GYRA and GYRB, which are homologs of bacterial-type topoisomerase II gyrase, modulate the topology of organelle DNA in Arabidopsis (Wall et al., 2004). Knockdown of GYRA or GYRB resulted in the aggregation of cp nucleoids into one or a few large particles in Nicotiana benthamiana (Cho et al., 2004). These studies indicate that there may be links between cp nucleoid micromorphology and cpDNA molecular structure that are related to cpDNA replication.
Distinct mechanisms maintain cpDNA integrity in plants that differ from the mechanisms underlying the establishment and maintenance of micromorphological cpDNA structure. Homologous recombination (HR) repair is a fundamental pathway that is critical in the recombination, replication, and repair of DNA defects such as double-strand breaks (DSBs) and stalled replication forks, thereby maintaining genome stability. RAD51 is a eukaryotic homolog (Shinohara et al., 1992) of RecA, a bacterial recombinase that plays an essential role in HR repair (Cox, 1999). In addition to RAD51, plants characteristically encode bacterial-type RecA homologs in their nuclear DNA (Lin et al., 2006). These plant RECAs are targeted to cps and/or mitochondria (Cerutti et al., 1992;Khazi et al., 2003;Odahara et al., 2007;Shedge et al., 2007;Inouye et al., 2008). Active HR occurs between large cp inverted repeats (Kolodner and Tewari, 1979), and RecA-related strand transfer activity is observed in cp extracts (Cerutti et al., 1993), suggesting that RecA-mediated HR mechanisms are active in cp. Expression of an Escherichia coli dominant-negative RecA protein in C. reinhardtii affected cp HR efficiency (Cerutti et al., 1995). Furthermore, transcription and translation of chloroplast-targeted RECA (cpRECA) is induced by treatment with DNA-damaging agents in various plants (Cerutti et al., 1992(Cerutti et al., , 1993Nakazato et al., 2003;Inouye et al., 2008). Taken together, these observations suggest that cpRECA is involved in the repair of damaged cpDNA.
Some mutant studies examined the role of cpRECA in the maintenance of cp genome integrity. In Physcomitrella patens, only modest growth defects were seen in knockout mutants of cpRECA (RECA2); however, the repair of damaged cpDNA was impaired, and mutants exhibited enhanced sensitivity to methyl methanesulfonate and UV light (Odahara et al., 2015). In addition, recombination between short dispersed repeats (SDRs; less than 100 bp) in cpDNA was higher in knockout mutants than in the wild type (Odahara et al., 2015). This suggests that cpRECA may suppress aberrant recombination between SDRs in cp and play a similar role in maintaining cp genome stability to that of mitochondrial RECA in mitochondrial genome stability (Odahara et al., 2009). After reproduction for several generations, Arabidopsis cpRECA mutants exhibited variegation in leaves and enhanced sensitivity to DSBand reactive oxygen species-inducing agents (Rowan et al., 2010;Jeon et al., 2013). In addition, compared with the wild type, mutants exhibit a reduction in cpDNA copy number, an increase in single-stranded cpDNA, and cpDNA structural alterations (Rowan et al., 2010). Recent next-generation sequencing approaches have revealed that Arabidopsis cpRECA acts to suppress U-turn like rearrangements mediated by microhomology (Zampini et al., 2015) In this study, we wished to explore the possible linkage between cp nucleoid micromorphology and the molecular structure and topology of cpDNA. In land plants, cp nucleoid morphology and molecular structure are complex and vary considerably depending on tissue type, developmental stage, age, and environment (Oldenburg and Bendich, 2015). Furthermore, as plant cells contain multiple cps that actively move around in response to environmental and other cues (Wada, 2013), monitoring the micromorphology of cp nucleoids at high resolution can be problematic (Terasawa and Sato, 2005). Therefore, we used the unicellular green alga Chlamydomonas reinhardtii as a simple model system. C. reinhardtii has a single cup-shaped cp containing approximately 80 copies of cpDNA (approximately 203 kb) organized into fiveto 10-cp nucleoids that can be readily observed using fluorescence microscopy. C. reinhardtii can be grown in homogenous culture, facilitating pharmacological analysis with various inhibitors, and abundant genome data (Merchant et al., 2007) and molecular tools are available.
In this research, we performed detailed functional analysis of cpRECA, with special focus on its impact on cp nucleoid dynamics and cp genome stability. We also analyzed the effect of gyrase inhibition on cp nucleoid dynamics and cp genome stability. The results demonstrate that cpRECA and gyrase are both involved in cp nucleoid dynamics and cp genome stability.

Preferential Localization of cpRECA at the Periphery of cp Nucleoids
The intracellular localization pattern of cpRECA has been examined in tobacco (Nicotiana tabacum) leaves (Nakazato et al., 2003) but never in C. reinhardtii. Immunofluorescence microscopy using an antibody raised against C. reinhardtii cpRECA (Supplemental Fig. S1) confirmed that C. reinhardtii cpRECA localized to cp nucleoids (Fig. 1A). However, superimposition of the cpRECA signal and the DAPI-stained DNA signal indicated that cpRECA was not evenly distributed in cp nucleoids but instead was found at the cp nucleoid periphery, and only in a subset of cp nucleoids (Fig. 1A).
Yellow fluorescent protein (YFP) optimized for C. reinhardtii (Karcher et al., 2009) was used to confirm the cpRECA localization pattern in living cells. The cpRECA gene was fused with YFP and overexpressed under the control of the PSBD promoter in nuclear DNA in C. reinhardtii (Supplemental Fig. S2A). The cpRECA-YFP signal was again detected only in parts of the cp nucleoids, and the fluorescent signal indicated the formation of microfilament-like structures fringing the nucleoids (Fig.  1B). In some cells, however, the micromorphology of cp nucleoids was drastically deformed, and massive fibrous structures that traversed the cp were observed (Fig. 1C). Superresolution microscopy revealed that fibrous structures penetrated interspaces of the thylakoid membranes (Supplemental Fig. S2B; Supplemental Movies S1 and S2). The fibrous cp nucleoids were observed more frequently in cell lines with higher cpRECA expression levels. The variable cpRECA-YFP distribution patterns may be due to fluctuations in cpRECA-YFP expression levels within individual cells.
One possible explanation for the unusual fibrous structures was that aggregation between YFP molecules might have artificially caused the formation of fibrous cp nucleoids. To test this possibility, cpRECA fused to hemagglutinin (HA) instead of YFP expressed in wild-type cells. The HA tag was detected by immunofluorescence microscopy using anti-HA antibody. Again, similar cpRECA filamentous structures were observed that colocalized with cpDNA (Supplemental Fig. S2), excluding the possibility that the formation of the filamentous cpRECA-YFP structure was due to YFP aggregation.
The correspondence between cpRECA-YFP signals and cpDNA in the fibrous structures raises the possibility that large regions of cpDNA are present as singlestranded DNA (ssDNA), since RecA preferentially binds to ssDNA (Kowalczykowski et al., 1994). We tested this by comparing the mung bean (Vigna radiata) nuclease susceptibility of cpDNA in the wild type and the cpRECA-YFP-expressing strain, in which approximately 30% of cells showed fibrous cpRECA-YFP structures. Mung bean nuclease digests ssDNA and single-stranded RNA nonspecifically, but it does not digest fully paired duplexes. Total DNA was extracted, and equal amounts were treated with mung bean nuclease. Quantitative PCR (qPCR) analysis of two loci on cpDNA revealed that the sensitivity of cpDNA to mung bean nuclease treatment in the cpRECA-YFP-expressing strain was almost equivalent to that in the control strain (Supplemental Fig. S3).
To address whether cpRECA activity was required for the formation of the fibrous cp nucleoids, a K127R mutation was introduced into cpRECA-YFP that corresponded to a dominant-negative mutation (K72R) in the ATPase domain of E. coli RecA. RecA K72R is suggested to form filaments with normal RecA and inhibit RecA function depending on its ratio to normal RecA (Renzette and Sandler, 2008). Overexpression of cpRECA (K127R)-YFP resulted in the condensation of cp nucleoids, and neither fibrous nor wild-typelike cp nucleoids were observed (Fig. 1D), indicating that cpRECA was required for the formation of reinhardtii wild-type cells were immunolabeled using anti-cpRECA antibody. Fluorescence intensities of colocalized 49,6-diamino-phenylindole (DAPI) and cpRECA (indicated by the arrow) were determined (at right). B to D, Subcellular localization of cpRECA-YFP or cpRECA (K127R)-YFP. YFP fluorescence (B) and anti-YFP immunofluorescence (C and D) were observed in C. reinhardtii cells ectopically expressing cpRECA-YFP or cpRECA (K127R)-YFP. Cells were stained with DAPI. DIC, Differential interference contrast; N, nucleus. Bars = 5 mm.
fibrous cp nucleoids and the maintenance of wildtype-like cp nucleoid distribution.

Condensation of Chloroplast Nucleoids in cpRECA-Knockdown Cells
To further analyze the function of cpRECA, cpRECAknockdown (KD) strains were developed using an artificial microRNA system optimized for C. reinhardtii (Molnar et al., 2009). Two cpRECA-KD cell lines were generated by suppressing cpRECA transcription to 12% and 37% of wild-type levels (Supplemental Fig. S4A). The knockdown lines exhibited no significant growth defects under standard culture conditions (Supplemental Fig. S4B). However, the micromorphology of cp nucleoids was substantially altered in knockdown lines. Fluorescence microscopy using SYBR Green I, a double-stranded DNA-staining dye, showed remarkable differences in distribution, size, and number of cp nucleoids in cpRECA-KD cells compared with the wild type ( Fig. 2). The cpRECA-KD cells had considerably fewer cp nucleoids than the wild type: the modal number of nucleoids per cp in the wild type was 12, in contrast with the single nucleoid per cp in cpRECA-KD cells (Fig. 2B). The decrease in nucleoid number was associated with an increase in nucleoid size in cpRE-CA-KD cells (Supplemental Fig. S5), and the size of a single large cp nucleoid was frequently comparable to the size of the cell nucleus (Fig. 2Aa). The enlarged cp nucleoids seen in cpRECA-KD cells were generally round (Fig. 2Aa). However, the nucleoid structure was frequently distorted when it was associated with pyrenoids (subcellular microcompartments sequestering Rubisco). These visible phenotypes of cpRECA-KD cells are reminiscent of moc mutants (Misumi et al., 1999), which also contain only a single large cp nucleoid ( Fig. 2A, b and c).

Decrease in Chloroplast Nucleoid Number in Gyrase Inhibitor-Treated Cells
The condensation of cp nucleoids in cpRECA-KD cells was reminiscent of a phenotype observed in Nicotiana benthamiana cells with repressed cp-targeted gyrase genes (Cho et al., 2004). Gyrase is a bacterial type II topoisomerase that uses ATP to introduce negative supercoils into DNA. Here, gyrase inhibitors (nalidixic acid and novobiocin) were used to investigate the possible contribution of gyrases to the regulation of cp nucleoid micromorphology in C. reinhardtii. Nalidixic acid and novobiocin have different modes of action. Nalidixic acid inhibits the dissociation of covalently bound gyrase from DNA, resulting in the production of DNA DSBs, whereas novobiocin inhibits ATPase activity and thereby inhibits the covalent binding of gyrase to DNA (Sugino et al., 1978;Chen et al., 1996). Treatment of C. reinhardtii wildtype cells with nalidixic acid or novobiocin inhibited growth (Supplemental Fig. S6A). Fluorescence microscopy of the inhibitor-treated cells revealed a dose-dependent decrease of cp nucleoid number, and most of the cells contained a single large cp nucleoid at the highest inhibitor concentration ( Fig.  3; Supplemental Fig. S6B). Corresponding with the decrease in cp nucleoid number, individual cp nucleoids were generally enlarged in inhibitor-treated cells. Most of the single cp nucleoids were round in shape and comparable in size to nucleoids in cpRECA-KD and moc cells (Fig. 3). These results suggest that the inhibition of gyrase function leads to a dramatic reduction in cp nucleoid number and an associated increase in nucleoid size.
Taken together, our data suggest that both cpRECA and gyrases are likely to be involved in the maintenance of cp nucleoid micromorphology through the regulation of cpDNA replication, recombination, and topology.

Destabilization of the Chloroplast Genome in cpRECA-KD Cells
Reductions in the levels of cpRECA and gyrase affected cp nucleoid number and size. Next, we examined whether the molecular structure of the cp genome is altered in cpRECA-KD lines and gyrase inhibitortreated cells. Previously, we showed that knockout of P. patens cpRECA2 resulted in cp genome instability as a consequence of increased recombination between SDRs (Odahara et al., 2015). Therefore, cp genome instability in C. reinhardtii cpRECA-KD cells was investigated with a particular focus on recombination between dispersed repeats. First, the C. reinhardtii cpDNA sequence was examined for repeat sequences using REPuter (Kurtz et al., 2001). Numerous repeats were identified that were consistent with previous analysis using MultiPipMaker (Maul et al., 2002). Approximately 2,600 pairs of repeats longer than 35 bp (0-bp mismatch) were identified in the C. reinhardtii cp genome (Supplemental Data Set S1).
Most of the identified repeats were unsuitable for use in the PCR analysis of recombination shown in Supplemental Figure S7, as they were palindromic or were found in large inverted repeats. Although several suitable loci were identified, recombination analysis was problematic in many of these for two main reasons. The design of primers specific to recombination products was hampered, first, by the presence of abundant repeats and, second, by low copy numbers or no copies of recombination products. However, through trial and error, suitable recombination products were identified from CD5 (77 bp, 1-bp mismatch), CI12 (94 bp, 1-bp mismatch), CD15 (59 bp, 0-bp mismatch), and CD20 (71 bp, 1-bp mismatch) repeats ( Fig. 4F; Supplemental Table S1). Intramolecular recombination between these repeats results in deletion and inversion of the flanking region when they are oriented as direct repeats and inverted repeats, respectively (Supplemental Fig. S7; Supplemental Table S1). Copies of CD5 and CD20 were found at additional loci with similarity to their surrounding sequences (CD59 and CD209; Fig. 4F; Supplemental Table S1); therefore, the PCR analysis detected recombination products from both CD5 and CD59 and products from CD20 and CD209 simultaneously. Recombination products from CD5 and CD20 were significantly more abundant in the cpRECA-KD strains than in the wild type, whereas recombination products from CI12 or CD15 were not significantly affected (Fig. 4, A-D). The increase in CD5 and CD20 products was more prominent in the knockdown line with the lowest cpRECA transcript levels (Fig. 4, A and  D). Analysis of cpDNA copy number relative to nuclear DNA using qPCR of the chloroplastic psbB locus and the nuclear cblp locus showed that cpDNA copy number in cpRECA-KD cells was 50% to 65% of that in the wild type (Fig. 4E). These results suggest that knockdown of cpRECA causes cp genomic instability by inducing aberrant recombination between SDRs and that the increase in recombination is accompanied by a reduction in cpDNA copy number.

Chloroplast Genomic Instability in Gyrase Inhibitor-Treated Cells
The cp genome instability was observed in cpRE-CA-KD cells. Because inhibition of gyrase and knockdown of cpRECA had similar effects on cp nucleoid number and size, it was also possible that gyrase inhibition might similarly have effects on genome stability. To test this possibility, recombination between SDRs in cpDNA was analyzed in gyrase inhibitor-treated cells. First, qPCR analysis was performed for CD5, which exhibited elevated recombination in cpRECA-KD lines. The abundance of CD5 recombination products increased in wild-type cells treated with nalidixic acid or novobiocin, depending on inhibitor concentration (Fig.  5, A and B). Novobiocin-treated cells exhibited a more pronounced recombination effect than nalidixic acidtreated cells: the abundance of CD5 recombination products increased approximately 160-fold relative to the wild type at the highest novobiocin concentration, compared with approximately 9-fold at the highest nalidixic acid concentration (Fig. 5, A and B). Recombination products from CI12, CD15, and CD20 also increased in nalidixic acid-and novobiocin-treated cells compared with the wild type (Fig. 5, C-E). As with CD5, the effect was more pronounced with novobiocin than with nalidixic acid (Fig. 5, C-E).
The cpDNA copy number was determined using qPCR in wild-type and gyrase inhibitor-treated cells. The copy number in cells treated with either nalidixic acid or novobiocin decreased in a dose-dependent manner, with cpDNA copy numbers decreased to approximately 60% and approximately 20% of the wild Figure 3. Effects of gyrase inhibitors on cp nucleoids. Wild-type cells were cultivated in medium containing the gyrase inhibitor nalidixic acid or novobiocin. DNA was stained using SYBR Green I. SYBR Green I (green) and chlorophyll (red) fluorescence signals were observed simultaneously using fluorescence microscopy. N and cpN indicate nucleus and cp nucleoids, respectively. Bar = 5 mm. type in cells exposed to 1.6 mM nalidixic acid and 400 mM novobiocin, respectively (Fig. 5, F and G). Together, these results indicate that cp genome destabilization and cpDNA copy number reduction are associated with the impairment of cp nucleoid segregation in gyrase inhibitor-treated cells.

DISCUSSION
In this study, we showed that knockdown or overexpression of cpRECA, or treatment of cells with gyrase inhibitors, resulted in significant alterations to the micromorphology of cp nucleoids in C. reinhardtii. These alterations were accompanied by the impairment of cpDNA integrity and reductions in cpDNA copy number.
cpRECA is a homolog of bacterial RecA, a recombinase involved in homologous pairing and strand exchange in HR repair that has an important role in the repair of stalled and collapsed replication forks (Lusetti and Cox, 2002). Here, immunofluorescence staining ( Fig. 1) of cpRECA and the expression of cpRECA tagged with YFP or HA ( Fig. 1; Supplemental Fig. S2) confirmed that cpRECA localized to cp nucleoids. Notably, cpRECA did not associate with cp nucleoids homogenously but was found at the periphery of a small subset of cp nucleoids (Fig. 1). This nonuniform localization of cpRECA suggests that cpRECA is unlikely to be a constitutive core component of cp nucleoids and may indicate specific binding of cpRECA to damaged DNA.
In some cells, overexpression of cpRECA resulted in deformation of the dispersed cp nucleoid particles into a single thick filamentous structure. These structures were completely covered with cpRECA, in contrast with the nonhomogenous associations seen in cells with wildtype-like cp nucleoids (Fig. 1B), suggesting that cpRECA was the critical factor driving the formation of the fibrous cp nucleoid structures (Fig. 1D). Bacterial RecA binds preferentially to ssDNA to form a filamentous nucleoprotein structure (Kowalczykowski et al., 1994), suggesting that cpRECA may similarly bind to cpDNA in C. reinhardtii overexpressing cpRECA. However, mung bean nuclease/qPCR analysis revealed that the cpRECA-YFP overexpression line did not have more ssDNA regions in Figure 4. Effects of cpRECA knockdown on cpDNA. A to C, Quantification of CD5, CI12, and CD15 recombination products. qPCR analysis is shown for relative copy numbers of CD5 (A), CI12 (B), and CD15 (C) cpDNA recombination products. Copy numbers were normalized to cp psbB. Data represent averages of three replicates. D, PCR analysis of CD20 recombination products in cpRECA-KD lines using psbB as an amplified internal control. E, qPCR analysis of relative cpDNA psbB copy number, normalized to nuclear Cblp. F, Positions of CD5, CI12, CD15, and CD20 in C. reinhardtii cpDNA. Large inverted repeats are shown with thick black lines. SDRs are indicated with triangles. The cpDNA sequence corresponds to accession NC_005353 when taken clockwise from position 1. Data represent means of three replicates. *, P , 0.01 (compared with the wild type [WT]).
cpDNA than the control strain (Supplemental Fig.  S3). RecA forms nucleofilament also with doublestranded DNA, which is triggered by nucleation at ssDNA regions (Lusetti and Cox, 2002). The fibrous cp nucleoid/cpRECA structure might be largely composed of double-stranded DNA-cpRECA nucleofilaments, which would be insensitive to mung bean nuclease treatment.
Compared with the wild type, cpRECA-KD cells exhibited a decrease in cp nucleoid number and an increase in cp nucleoid size that, in some cases, produced a single large nucleoid comparable in size to the nucleus (Fig. 2). This cp nucleoid condensation phenotype was reminiscent of the moc mutant phenotype, the gene responsible for which has not yet been identified (Misumi et al., 1999). In bacteria, a recA mutation results in the production of anucleate cells, suggesting that RecA may be involved in bacterial chromosome segregation (Zyskind et al., 1992). In the bacterial recA mutant, stalled or collapsed replication forks may not be properly processed; therefore, subsequent chromosome segregation may be prevented. Defects in the repair of impaired replication forks in cpRECA-KD cps may similarly perturb cpDNA segregation. The association of cpRECA with only a subset of cp nucleoids suggests that cpRECA acts as an accessory protein at the nucleoids; however, it is surprising that overproduction or knockdown of a single accessory protein would have such a substantial impact on cp nucleoid micromorphology. Therefore, it may be of interest to investigate cpRECA functions in other processes, such as cp division, during which cp nucleoid shape and number undergo dynamic changes (Kuroiwa et al., 1981;Nakamura et al., 1986;Ehara et al., 1990).

Figure 5. Chloroplast genome instability in gyrase inhibitor-treated cells. A and B, qPCR analysis of relative CD5 copy numbers in wild-type cells treated with nalidixic acid (A) or novobiocin (B).
Copy numbers were normalized to cp psbB. C and D, qPCR analysis of CI12 (C) or CD15 (D) recombination products in cells treated with 1.6 mM nalidixic acid or 400 mM novobiocin. Copy numbers were normalized to psbB. E, PCR analysis of CD20 recombination products in cells treated with nalidixic acid or novobiocin, using psbB as an amplified internal control. WT, The wild type. F and G, qPCR analysis of relative cpDNA psbB copy number, normalized to nuclear Cblp, in wild-type cells treated with nalidixic acid (F) or novobiocin (G). The untreated wildtype ratio was set as 1. Data represent means of three replicates. *, P , 0.01 (compared with the untreated control).
The condensation of cp nucleoids also was observed in cells treated with the gyrase inhibitors novobiocin and nalidixic acid (Fig. 3). Novobiocin was more effective at inducing cp nucleoid condensation than was nalidixic acid, suggesting that condensation did not occur as a result of DSB formation but rather by the inhibition of ATPase activity and the resultant cpDNA structural and topological changes. Bacterial gyrase is a type II topoisomerase that can introduce negative supercoils. Gyrase-defective bacterial mutants exhibit impaired segregation of catenated chromosomes (Steck and Drlica, 1984). Similarly, cp chromosome segregation was impaired by knockdown of cp gyrase in N. benthamiana (Cho et al., 2004) and by treatment with nalidixic acid in the alga Cyanidioschyzon merolae (Itoh et al., 1997). Therefore, it is likely that gyrase has a common role in chromosome segregation in bacteria and cps. Our data suggest that both cpRECA and gyrases are involved in cpDNA segregation and the maintenance of cp nucleoid morphology via the control of cpDNA replication, recombination, and topology.
Recombination between cp SDRs was induced by knockdown of cpRECA. This suggests that cpRECA is involved in the maintenance of cp genome stability by the suppression of aberrant recombination in C. reinhardtii, as was shown previously in the land plant P. patens (Odahara et al., 2015). In contrast with the cp genomes of land plants, which have few repeats longer than 30 bp, the C. reinhardtii cp genome contains abundant repeats. However, despite this, recombination was induced in relatively few of the C. reinhardtii repeats in cpRECA-KD cells. This suggests that C. reinhardtii cpRECA may generally suppress recombination between short repeats, but additional suppression pathways that do not require cpRECA activity may be present in C. reinhardtii cp. In addition, hot spots or suppressed regions for recombination may be present in the cpDNA genome (Newman et al., 1992). Alternatively, the presence of abundant repeats might minimize the accumulation of recombination products from any two specific repeats.
SDR recombination products also accumulated in cells treated with gyrase inhibitors (Fig. 5), suggesting that cp genomes were destabilized as a result of induced recombination between SDRs in the inhibitor-treated cells. While the accumulation of CD5 recombination products was similar between inhibitor-treated cells and cpRECA-KD cells, CI12 and CD15 recombination products accumulated at higher levels in inhibitor-treated cells than in cpRECA-KD cells. These results suggest that recombination is induced in a wider range of repeats in inhibitor-treated cells than in cpRECA-KD cells. The difference in induced recombination may reflect differences in the roles of cpRECA and gyrase in the suppression of recombination.
In this study, we demonstrated that cp genome instability is accompanied by an impairment of nucleoid segregation in cpRECA-KD cells and cells treated with gyrase inhibitors. Our previous study suggested that the roles of mitochondrial RECA1 (Odahara et al., 2009) and chloroplastic RECA2 (Odahara et al., 2015) in the repair of stalled or collapsed replication forks in P. patens corresponded to the activities of bacterial RecA. RECA proteins may promote homologous recombination during the repair of stalled or collapsed replication forks, thereby preventing organelle genome instability. C. reinhardtii cpRECA may function similarly in the repair of stalled or collapsed replication forks, and cpRECA-independent recombination, which is suppressed to a low level in the wild type, may be induced during the repair of disordered replication forks in cells with the cpRECA-KD background. Consequently, the impairment of nucleoid segregation and genome instability in cpRECA-KD cells may be the consequence of defective replication repair processes (Supplemental Fig. S8).
Bacteria defective in gyrase experience replication fork arrest as a result of positive supercoil accumulations ahead of the fork. However, HR repair proteins, including RecA, are not required for the repair of these arrested forks (Grompone et al., 2003). This suggests that replication fork arrest induced by the accumulation of positive supercoiling is not subject to HR repair in E. coli. As in E. coli, gyrase in cps is involved in DNA supercoiling (Wall et al., 2004); therefore, it is likely that supercoil accumulation induces cp replication fork arrest in C. reinhardtii. Recombination between dispersed repeats in inhibitor-treated cp, which is probably independent of cpRECA, may occur when arrested replication forks are processed (Supplemental Fig. S8). This may correspond with recombination-dependent replication in novobiocin-treated C. reinhardtii cp, as proposed by Woelfle et al. (1993). Gyrase may maintain cp genome stability by reducing replication stresses that can lead to genome instability.
In summary, results from the induction of cp genome instability and the impairment of nucleoid segregation in two different cell types suggested that genome stability and nucleoid segregation in cp were intimately related. In addition, overexpression of cpRECA profoundly affected cp nucleoid structure, indicating its potential effects on nucleoid dynamics. Our findings provide further evidence toward an integrated understanding of nucleoid dynamics and genome integrity in cps.

Plant Materials and Culture Medium
Chlamydomonas reinhardtii wild-type strains CC-125 (mating type plus) and CC-124 (mating type minus; Harris, 1989), the cell wall-less mutant cw15 (Harris, 1989), and mutant strains moc A84 and moc G33 (Misumi et al., 1999) were used in this study. Cells were cultivated in Tris-acetate-phosphate (TAP) liquid medium (Harris, 1989) at 20°C under constant white light. For growth comparisons, cells were cultivated in TAP liquid medium and then spotted onto TAP agar medium after cell counting. For gyrase inhibitor treatment, cells were cultivated in TAP liquid medium containing novobiocin or nalidixic acid at the concentrations given in the figure legends.
Knockdown of cpRECA in C. reinhardtii An artificial microRNA system utilizing pChlami RNA3 (Molnar et al., 2009) was used to knockdown the C. reinhardtii cpRECA gene (GenBank accession no. AB048829). Artificial microRNA sequences for cpRECA (oligonucleotide DNA P1 and P2) were designed using WMD3 (http://wmd3.weigelworld.org/cgibin/webapp.cgi; Molnar et al., 2009). Oligonucleotides were annealed and then cloned into the SpeI site of pChlami RNA3 (a generous gift from Dr. David Baulcombe, University of Cambridge). The resulting plasmid was introduced into C. reinhardtii cc125 cells by electroporation (Shimogawara et al., 1998), and transformants were selected on TAP agar medium containing paromomycin. Knockdown of cpRECA was analyzed by quantitative reverse transcription-PCR (parameters are given below) with primers P3 and P4 for cpRECA and P5 and P6 for Cblp. Primer sequences are provided in Supplemental Table S2.

Extraction of Nucleic Acids
C. reinhardtii genomic DNA was extracted as follows. C. reinhardtii cells were resuspended in 2% SDS, 400 mM NaCl, 40 mM EDTA, and 100 mM Tris-HCl, pH 8. DNA was then extracted using phenol/chloroform and precipitated by adding CTAB. RNA was extracted from C. reinhardtii cells using Sepasol (Nacalai). After DNaseI treatment to remove residual genomic DNA, reverse transcription was performed using total RNA, random hexamers, and reverse transcriptase ReverTra Ace (Toyobo).

PCR Analysis of cpDNA
qPCR analysis of C. reinhardtii cpDNA was performed using total genomic DNA and primers as follows: P7 and P8 for psbB, P9 and P10 for psbD, P11 and P12 for DNA recombined between CD5 repeats, P13 and P14 for CD15, P15 and P16 for CI12, and P17 and P18 for Cblp. qPCR was performed using the MX3000P QPCR System (Agilent) and FastStart Universal SYBR Green Master (Rox; Roche). PCR analysis was performed using C. reinhardtii total genomic DNA and primers P19 and P20 for CD20 and P21 and P22 for psbB.

Analysis of ssDNA Nuclease Sensitivity
Total genomic DNA was extracted from cells cultivated under standard culture conditions. For heat denaturation, the extracted genomic DNA was incubated at 98°C for 5 min and then chilled on ice. One microgram of genomic DNA was digested with 1 unit of mung bean (Vigna radiata) nuclease (New England Biolabs) at 30°C for 30 min. qPCR analysis was performed using the 7500 Fast Real-Time PCR System (Applied Biosystems) and primers P21 and P22 for psbB, P23 and P24 for psbD, and P17 and P18 for Cblp.

Fluorescence Microscopy
For YFP-tagging analysis, cpRECA was amplified from genomic DNA using primers P25 and P26, and YFP, which was a generous gift from Dr. Ralph Bock (Max-Planck-Institute), was amplified using primers P27 and P28. Fragments were inserted into NdeI-BamHI and XbaI-EcoRI sites of pGenD-C3HA (Chlamydomonas Resource Center), respectively. The expression level of cpRECA-YFP was analyzed by quantitative reverse transcription-PCR with primers P3 and P4 for cpRECA and P29 and P30 for MAA7. Cells were stained using a 1:1,000 dilution of SYBR Green or 1 mg mL 21 DAPI to stain cp nucleoids. DAPI staining was performed after fixation with 2.5% glutaraldehyde. Cells were observed with a BX51 fluorescence microscope (Olympus) for standard observations and with a TCS SP5 confocal laser scanning microscope (Leica Microsystems) for superresolution analysis.

Antibody Preparation
Primers P31 and P32 were used to amplify cpRECA cDNA. The resulting fragment was cloned into pQE80l (Qiagen), and the plasmid was transformed into Escherichia coli strain BL21. The E. coli culture was grown in Luria-Bertani medium at 37°C. Once the culture reached OD 600 0.7 to 1, isopropyl b-D-1-thiogalactopyranoside was added to a final concentration of 1 mM. Expressed protein was purified using Ni-NTA agarose (Qiagen) according to the manufacturer's instructions. Antibodies were raised by injecting the purified recombinant protein into mice every 2 weeks for a total of five times.

Immunoblot Analysis
Proteins extracted from C. reinhardtii cells were separated using 12.5% SDS-PAGE and transferred onto a polyvinylidene fluoride membrane. Detection was performed using diluted antibody (1:2,500) and the ECL Plus western-blot detection kit (GE Healthcare). Chemiluminescence was detected with a luminoimage analyzer (LAS4000; Fujifilm) and analyzed using Multi Gauge version 4.0 software (Fujifilm). Gels were also stained using Coomassie Brilliant Blue to assess loading.

Indirect Immunofluorescence Microscopy
Proteins were detected by immunofluorescence staining as described previously (Nishimura et al., 2002). Cells were harvested and fixed in 3.8% formaldehyde on ice for 10 min. After washing with Tris-buffered saline (TBS), the cells were gently permeabilized with 0.1% Triton X-100 on ice for 20 min. Cells were then resuspended in TBS buffer containing 5% bovine serum albumin and 0.25% Tween 20 and then kept on ice for 30 min. The cells were then incubated on ice for 1 h with the primary antibody (anti-YFP mouse IgG [Takara Bio] and anti-HA mouse IgG [Santa Cruz Biotechnology]) at a 1:1,000 dilution. After washing, the cells were resuspended in TBS buffer on ice for 30 min. The secondary antibody (Alexa Fluor 488 goat anti-mouse IgG) was added at a 1:1,000 dilution, and cells were incubated on ice for 1 h. Finally, cells were washed, stained with 1% DAPI, and observed by fluorescence microscopy under UV light (for DAPI) or blue light (for Alexa).

Supplemental Data
The following supplemental materials are available.
Supplemental Figure S2. Overexpression of cpRECA tagged with YFP or HA.
Supplemental Figure S4. Knockdown of the cpRECA gene in C. reinhardtii.
Supplemental Figure S6. Effect of gyrase inhibitors on C. reinhardtii growth and cp nucleoid number.
Supplemental Figure S7. qPCR assay for the amplification of recombination products from direct and inverted repeats.
Supplemental Figure S8. A model for genome instability and the defect in nucleoid segregation in cps.
Supplemental Table S1. C. reinhardtii cp repeats tested for recombination.
Supplemental Table S2. Oligonucleotide DNAs used in this study.