Cyanobacterial multi-copy chromosomes and their replication

ABSTRACT

While the model bacteria Escherichia coli and Bacillus subtilis harbor single chromosomes, which is known as monoploidy, some freshwater cyanobacteria contain multiple chromosome copies per cell throughout their cell cycle, which is known as polyploidy. In the model cyanobacteria Synechococcus elongatus PCC 7942 and Synechocystis sp. PCC 6803, chromosome copy number (ploidy) is regulated in response to growth phase and environmental factors. In S. elongatus 7942, chromosome replication is asynchronous both among cells and chromosomes. Comparative analysis of S. elongatus 7942 and S. sp. 6803 revealed a variety of DNA replication mechanisms. In this review, the current knowledge of ploidy and DNA replication mechanisms in cyanobacteria is summarized together with information on the features common with plant chloroplasts. It is worth noting that the occurrence of polyploidy and its regulation are correlated with certain cyanobacterial lifestyles and are shared between some cyanobacteria and chloroplasts.

Abbreviations

NGS: next-generation sequencing; Repli-seq: replication sequencing; BrdU: 5-bromo-2′-deoxyuridine; TK: thymidine kinase; GCSI: GC skew index; PET: photosynthetic electron transport; RET: respiration electron transport; Cyt b₆f complex: cytochrome b₆f complex; PQ: plastoquinone; PC: plastocyanin.

GRAPHICAL ABSTRACT

Replication cycle of the cyanobacterium Synechococcus elongatus PCC 7942.

KEYWORDS:

• Cyanobacteria
• polyploidy
• DNA replication
• GC skew

Cyanobacteria are the predominant phototrophs in ocean and freshwater ecosystems and are one of the most widespread phylogenetic clades. Cyanobacteria have oxygen-producing photosynthetic capabilities, meaning that they can produce biomass by using solar energy, and they have recently gained attention for their potential as green cell factories for CO₂-neutral biosynthesis of various products. In particular, there is increasing interest in the production of biofuels and valuable chemicals using cyanobacteria [1–3]. Numerous aspects of cyanobacterial genetics and physiology have been intensively studied in recent years.

Some marine picocyanobacteria, such as the clade that includes Synechococcus sp. and Prochlorococcus sp., are monoploid [4,5] like Escherichia coli and Bacillus subtilis, whereas several freshwater cyanobacteria are polyploid, meaning that they harbor multiple chromosome copies per cell [4,6,7]. Similar polyploidy has been observed in some bacterial species, such as Deinococcus radiodurans [8], Thermus thermophilus [9], and in other symbiotic and parasitic bacteria [10–12], as well as in plant chloroplasts, which are evolutionarily derived from cyanobacteria [13,14]. Polyploidy has also been reported in several archaeal lineages [15–18].

The freshwater cyanobacteria Synechococcus elongatus PCC 7942 and Synechocystis sp. PCC 6803 (hereafter referred to as S. elongatus 7942 and S. sp. 6803, respectively) are frequently used as model phototrophs because their transformation efficiencies and growth rates are superior to those of other cyanobacteria. It has been reported that S. elongatus 7942 carries 3–4 chromosome copies (Figure 1) [19–21], whereas S. sp. 6803 contains more than 10 chromosome copies per cell during exponential growth phase [22–25]. Recently, Trichodesmium sp. was reported to be highly polyploid, containing as many as 100 chromosome copies per cell in the field, and 600 chromosome copies under laboratory conditions [26]. The chromosome copy numbers of other cyanobacteria were reported by Griese et al. [21]. Although an unusual chromosome copy number was reported for S. sp. 6803 in this report, a correction was made in a subsequent article by the same group [23].

Figure 1. Replication cycles of Escherichia coli and Synechococcus elongatus PCC 7942.

A plausible evolutionary advantage of polyploidy is that the replicated chromosomes can compensate for damage to the other chromosomes. In cyanobacteria and chloroplasts, this chromosomal damage is highly likely because their DNA is always exposed to the oxidative stress that results from photosynthesis [27]. Another explanation for polyploidy is its correlation with cell volume, as chromosome copy number is proportional to cell size. Trichodesmium sp., which harbors 100–600 chromosome copies per cell, is >3 orders of magnitude larger by volume than monoploid picocyanobacteria, such as Prochlorococcus sp. and Synechococcus sp [26,28]. In addition, recent studies of S. elongatus 7942 revealed that the number of replicating chromosomes is restricted to maintain a stable gene copy number/cell volume ratio during growth [29,30].

In this review, the available knowledge of ploidy and the mechanisms of DNA replication in cyanobacteria, specifically S. elongatus 7942 and S. sp. 6803, is summarized together with information about their common features with chloroplasts. Notably, the occurrence of polyploidy and its regulation are correlated with certain cyanobacterial lifestyles and are shared between some cyanobacteria and plant chloroplasts.

Polyploidy changes

The degree of polyploidy in cyanobacteria is very flexible, and it depends on the growth conditions. In batch culture, cyanobacterial growth progresses from lag phase to exponential phase. This is typically followed by a period of linear growth that continues until the culture reaches a non-growing stationary phase. The degree of ploidy is significantly higher during lag phase than during the exponential and linear phases (Figure 2). Intensive DNA replication and increased ploidy have been observed during lag phase in both S. elongatus 7942 and S. sp. 6803. During lag phase, the chromosome copy number increases to 4–10 in S. elongatus 7942 [19,20] and to 10–20 in S. sp. 6803 [23,24,25]. This increase is necessary for the subsequent cell division and elongation steps that occur during the exponential and linear growth phases [19].

Figure 2. Growth phase and polyploidy in S. elongatus 7942.

In contrast, polyploidy is minimized during linear growth phase, which is associated with nutrient limitation in the medium as well as light limitation caused by cell self-shading [31–33]. In batch cultures of S. elongatus 7942 in BG11 medium, the phosphate is completely depleted at the start of linear growth phase (Figure 2). During linear growth phase, low-level growth was observed, even in the absence of an extracellular phosphate source along with a concomitant reduction in chromosome copy number to 2–3 copies per cell, suggesting that DNA can be used as a source of organic phosphate to support further growth (Figure 2) [19]. A similar phenomenon has been reported in S. sp. 6803. The chromosome number gradually decreased over time, and after 16 d of cultivation, the cells became monoploid [23]. This reduction in chromosome copy number during linear growth phase was suppressed by surplus phosphate [23]. Such ploidy changes are related to maintenance of the balance between the frequency of DNA replication and cell division, which is influenced by growth phase and environmental conditions and seems to be a common mechanism among polyploid organisms. In fact, such polyploidy changes have also been observed in haloarchaea [34,35].

S. elongatus 7942 has been widely used as a model for studying prokaryotic circadian rhythms [36]. Although cell division is controlled by the circadian rhythm [37–39], ploidy is stable even when the circadian rhythm of S. elongatus 7942 cells is synchronized [6], indicating that chromosome copy number regulation is independent of the circadian rhythm in this species. Under circadian rhythm-synchronized culture conditions, it is thought that only one or two chromosomes in the cell replicates in a light-dependent manner. Therefore, ploidy does not significantly change.

Cyanobacteria share certain characteristics related to ploidy changes with chloroplasts, which evolved from a freshwater lineage of cyanobacteria [40]. It has been reported that DNA replication in chloroplasts is not coupled to nuclear DNA replication or chloroplast division [41–43]. In green plants, such as wheat, barley, and pea, the chromosome number in chloroplasts varies according to the developmental stage [41,42,44–46]. Proplastids, which are undifferentiated chloroplasts, undergo intensive DNA amplification before cell division, which is a fundamental step toward subsequent proplastid division. In contrast, during leaf development and plant growth, the DNA content in chloroplasts is reduced to very low levels [47]. Mutants of chloroplast-specific DNA polymerase and RecA exhibited delays in plant growth and development and a reduction in chloroplast DNA content [48,49], suggesting that these proteins are important for chloroplast DNA replication. However, the mechanisms that control the chloroplast DNA copy number are still unclear.

Intracellular distribution of the multiple chromosome copies and their regulation

The distribution of the multi-copy chromosomes in S. elongatus 7942 cells has been investigated by microscopy. The replication origin (oriC) and a site opposite of oriC (the putative terC) in the chromosome were labeled, and their intracellular distributions were assayed. The results of these assays indicated that these regions align along the major axis of the cell and that chromosome number is correlated with cell length (Figure 1) [50–51]. Polyphosphate granules, which are a high-energy phosphate source, and carboxysomes, which are metabolic protein microcompartments where carbon fixation occurs, were observed to behave in a similar manner [51,52].

The distribution of the multi-copy chromosomes is regulated by ParA family proteins, which are chromosome-encoded and participate in chromosomal segregation in several bacteria, including B. subtilis [53,54], Pseudomonas aeruginosa [55], and Vibrio cholerae [56,57]. In S. elongatus 7942, ParA (recently named McdA) was shown to regulate the subcellular position of both carboxysomes and multi-copy chromosomes. ParA exhibits DNA binding activity and repeatedly moves from one end of the cell to the other under the control of McdB protein binding [58–60]. Deletion of ParA specifically disrupts the order of both carboxysomes and chromosomes [50,58]. Since disruption of ParA affects both carboxysomes and multi-copy chromosomes in S. elongatus 7942, these findings indicate that ParA is involved in cellular compartmentalization (Figure 1). In addition, a parA deletion mutant was shown to be more sensitive to UV-C than the wild-type strain. It has also been shown that ParA interacts with SMC_N family proteins, which are highly conserved among cyanobacteria [50], and a SMC_N homologue is involved in chromosome compaction and segregation [61]. Therefore, like ParA, SMC_N family proteins may be involved in chromosome distribution and the UV stress response in S. elongatus 7942.

Genome structure

The genome sequences of several cyanobacteria are unique in terms of their GC skew, as indicated by plots of the normalized guanine (G) over cytosine (C) content in a subgenomic region with sliding windows along the entire genome [62]. In many monoploid organisms, including E. coli and B. subtilis, the GC skew plot divides the genome into a region two regions, one with an excess of G over C and one with an excess of C over G (the leading and lagging strands, respectively). In genomes with this symmetrical GC skew, the shift points of the GC skew plot are correlated with the replication loci oriC and terC (Figure 3) [63]. Thus, such plots have enabled the prediction of the locations of oriC and terC in the chromosomes of monoploid bacteria, including the picocyanobacteria Synechococcus sp. WH 8102 and Prochlorococcus marinus str. CCMP1375. In these monoploid bacteria, the location of oriC has been predicted with high precision using both GC skew and the binding sequence of DnaA, the initiator of DNA replication in bacteria [64].

Figure 3. Comparison of the GC skews in mono- and polyploid organisms.

The GC skews of polyploid organisms, such as S. elongatus 7942, S. sp. 6803, Gloeobacter violaceus PCC 7421, and D. radiodurans, are distinct from those of monoploid bacteria (Figure 3). The GC skew in polyploid organisms is asymmetrical and contains many high-G/high-C shift points. A similar pattern was also observed in the chloroplasts of Arabidopsis thaliana (Figure 3). The strength of the GC skew in a given bacterial chromosome can be calculated mathematically as the GC skew index (GCSI), which is useful for estimating the confidence levels for the prediction of oriC and terC based on GC skew [65–67] and are easily obtained from G-language genome analysis [68]. The GCSI scores of the genomes in monoploid organisms, including picocyanobacteria, are high, suggesting clear GC skews, while the GCSI scores of polyploid organisms are generally lower (Figure 3). A high GCSI score suggests a strong mutation or selection pressure induced by bi-directional replication machinery starting from a single origin [65–67]. Picocyanobacteria are likely to be susceptible to mutation or selection pressure due to their low chromosome copy number, whereas polyploid cyanobacteria are likely to be comparatively less susceptible. It has been suggested that a low GCSI score can also result from long doubling times [66]. However, in cyanobacteria, a low GCSI score seems to be related to ploidy rather than doubling time, because an inverse correlation between GCSI score and ploidy was observed in cyanobacteria (Figure 4). However, further study is needed to determine the relationship between genome structure and ploidy.

Figure 4. Correlation between chromosome copy number and GC skew index score in cyanobacteria.

Replication origin (oriC)

In S. elongatus 7942, oriC and the mechanism of replication have been experimentally identified by replication-sequencing (Repli-seq) using next-generation sequencing (NGS) technology [20]. Since S. elongatus 7942 lacks the thymidine kinase (TK) gene, which is necessary for the incorporation of 5-bromo-2′-deoxyuridine (BrdU, a thymidine analog) in DNA, a TK gene was introduced into S. elongatus 7942. The resulting TK-introduced strain was used for Repli-seq analysis and evaluation of DNA replication activity. BrdU-labeled S. elongatus 7942 DNA was purified and quantitatively analyzed by NGS. A clear peak, indicating BrdU-labeled DNA, was observed upstream of the dnaN gene, which contains the dnaA-box cluster (the binding site of the replication initiator DnaA) [69], on the border between the high- and low-GC regions in the GC skew analysis (Figure 3). The peak broadened as the BrdU labeling time increased, indicating bi-directional DNA replication in S. elongatus 7942, similar to that in monoploid organisms such as E. coli and B. subtilis.

The oriC-like region in the filamentous cyanobacterium Anabaena sp. PCC 7120 has also been shown to be located in between the dnaA and dnaN genes [70]. The oriC-like region of A. sp. 7120 supported autonomous plasmid replication in S. sp. 6803, indicating that S. sp. 6803 has the ability to utilize the identified oriC-like sequence. Notably, the integration of additional oriC-like regions into the chromosome of A. sp. 7120 led to the appearance of multiple, contiguous proheterocysts in the absence of nitrogen, suggesting the involvement of DNA replication in heterocyst development in A. sp. 7120 [70].

Unlike in S. elongatus 7942, which has a genome with a relatively clear GC skew compared to other cyanobacteria, in S. sp. 6803, the replication origin could not be identified [71]. No clear phenotypes, including growth, morphology, and DNA replication activity, were observed in S. sp. 6803 strains lacking the predicted oriC region or dnaA gene. In a Repli-seq analysis of S. sp. 6803, there were no peaks observed in either the predicted oriC region or any other genomic regions, suggesting that there are multiple replication origins that fire asynchronously in S. sp. 6803. This species possesses a genome with strand asymmetry; thus, the replication origin cannot be predicted from GC skew alone, as there are many shift points (Figure 3). This genomic composition suggests the existence of multiple replication origins in some prokaryotes, similar to that in eukaryotic nuclear chromosomes [72,73]. Therefore, further studies are needed to identify the replication origins in S. sp. 6803.

The mechanism by which chloroplast DNA is replicated is also still a mystery. It appears that chloroplast DNA can be replicated by more than one mechanism [74], including recombination-dependent replication [47,48,75], a double D-loop mechanism [76,77], and rolling circle replication [78]. Studying DNA replication in polyploid cyanobacteria will provide useful information for understanding the mechanism of chloroplast DNA replication.

Asynchronous DNA replication

Analyses based on NGS and microscopy have revealed asynchronous DNA replication in S. elongatus 7942. In rapidly growing bacterial cultures, e.g., E. coli and B. subtilis, DNA replication occurs in almost all cells, and consequently, these bacteria often harbor two or more copies of oriC and one copy of terC [79,80]. Whole-genome NGS showed that the copy number of oriC is significantly greater than that of terC in bacterial cells with multiple replication forks [81,82]. However, in S. elongatus 7942, there is little difference in copy numbers of oriC and terC, even at the peak of DNA replication activity, indicating that the multi-copy chromosomes in this species do not replicate at the same time [20]. Consistent with this observation, immunofluorescence microscopy using BrdU revealed that S. elongatus 7942 cells harbored one or two BrdU foci, indicating a replicated chromosome, even though its multi-copy chromosomes are widely distributed in the cell (Figure 1) [20]. A similar observation was made in S. sp. 6803 [29], suggesting that asynchronous DNA replication is a common feature of polyploid cyanobacteria.

Live-cell imaging has made it possible to visualize the movement of multi-copy chromosomes in S. elongatus 7942 [83,84]. Studies have shown that the distribution of multi-copy chromosomes is correlated with cell division, and multi-copy chromosomes replicate asynchronously in a linear-like fashion according to cell length. There is no specific spatial location where multi-copy chromosomes are replicated in the cell [20,83,84], whereas in monoploid organisms such as E. coli, chromosome replication is preferentially localized either at the center or the quartile points of the major axis [85,86]. After DNA replication, chromosome segregation into daughter cells in S. elongatus 7942 occurs in a seemingly nonrandom fashion, which may be the result of a cellular process that transiently organizes the chromosomes just before the completion of cell division.

Similar observations of DNA replication and chromosomal segregation have been reported in S. sp. 6803 [29,87], which lends further support to the notion that asynchronous DNA replication is a common feature of polyploid cyanobacteria.

The replication initiator protein DnaA

DNA replication is the most fundamental process in the cell cycle of all organisms. The main factors involved in DNA replication are highly conserved among cyanobacteria, and most of them are essential in S. elongatus 7942 (Table 1) [88]. In model bacteria, such as E. coli, DNA replication is initiated by the binding of DnaA to oriC, which is followed by the recruitment of the components of the replisome [89,90]. The formation of an appropriate DnaA–oriC complex is a critical step in replication control that occurs only once per cell division cycle. Dysfunction of DnaA regulation in E. coli leads to asynchronous DNA replication initiation and polyploidy [91–94], suggesting that regulation of the initiation of DNA replication is important for determining chromosome copy number. Like in most bacteria, DnaA is essential in S. elongatus 7942 [88], and it regulates DNA replication initiation by binding to oriC (Table 1) [29,71]. The binding of DnaA to oriC in S. elongatus 7942 depends on photosynthesis (see the following section). Notably, the lethal phenotype of a dnaA disruption mutant was suppressed by the integration of an episomal plasmid (pANL) into the chromosome, and in this strain, chromosomal replication was initiated via the plasmid replication initiation system [71].

Table 1. DNA replication factors in cyanobacteria.

Gene name	KEGG orthology	S. elongatus 7942	S. sp. 6803	S. sp. 6714	A. sp 7120	P. marinus 1375	G. violaceus 7421
dnaA	Chromosomal replication initiator	Synpcc7942_1100	sll0848	D082_00010	alr2009	Pro_0567	gvip205
dnaB	Replicative DNA helicase	Synpcc7942_1325	slr0833*	D082_07900*	all0578* all7274 (α)	Pro_1835	gll0034*
dnaG	DNA primase	Synpcc7942_1292	sll1868	D082_01160	all1323	Pro_0757	gvip547
dnaE	DNA polymerase III subunit α	–	–	–	–	Pro_0751	gvip538
dnaE1	DNA polymerase III subunit α (N-terminal)	Synpcc7942_2199*	slr0603*	D082_10620*	all3578*	–	–
dnaE2	DNA polymerase III subunit α (C-terminal)	Synpcc7942_0570*	sll1572*	D082_03500*	alr1054*	–	–
dnaX	DNA polymerase III subunit γ/τ	Synpcc7942_0912	sll1360*	D082_24430	alr4932 alr7570 (β)	Pro_1818	gvip228
holA	DNA polymerase III subunit δ	Synpcc7942_0405	sll0544	D082_24840	alr3444 alr7574 (β)	Pro_1807	gll0066
holB	DNA polymerase III subunit δ’	Synpcc7942_0094	slr0446	D082_06000	all4705 alr7575 (β)	Pro_0151	gvip200
dnaN	DNA polymerase III subunit β	Synpcc7942_0001	slr0965	D082_19380	alr2010 alr7569 (β)	Pro_0001	gvip446
priA	Primosomal protein N’	Synpcc7942_0651	sll0270	D082_22900	all4248	Pro_0497	gll0025
polA	DNA polymerase I	Synpcc7942_0194	syn:slr0707	D082_14220	alr1254	Pro_1234	gvip065
ssb1	ssDNA-binding protein	Synpcc7942_0301	slr0925	D082_20810	alr7579	Pro_1784	glr2487
ssb2	ssDNA-binding protein YCF41	Synpcc7942_0079	slr1034	D082_09250	all4779	Pro_0502	glr2488 gvip045
LigA	DNA ligase (NAD+)	Synpcc7942_0277	sll1209	D082_17450	all1717	Pro_1841	glr0727
rnhA	Ribonuclease HI	Synpcc7942_2183	slr0080	D082_26030	alr0142	Pro_0226	gvip491
rnhB	Ribonuclease HII	Synpcc7942_0879	slr1130
gyrA1	DNA gyrase subunit A	Synpcc7942_0254	slr0417	D082_15290	all0860	Pro_1137	gvip425
gyrA2	DNA gyrase subunit A	Synpcc7942_1694	sll1941	D082_25930	all0441	Pro_0005	gvip274
gyrB	DNA gyrase subunit B	Synpcc7942_2491	sll2005*	D082_01740	all5265	Pro_1796	gvip344
nrdA	Ribonucleotide reductase, class Ia	–	slr1164	D082_00790	–	–	–
nrdB	Ribonucleotide reductase, class Ia	–	slr0591	D082_27380 D082_27390	–	–	–
nrdJ	Ribonucleotide reductase, class II	Synpcc7942_1609*	–	–	all4035*	Pro_0815	gll3966*

ORFs marked with asterisks contain cis- or trans-splicing inteins. Note: The DnaB homologue in S. sp. 6714 (D082_07900) possesses two cis-splicing inteins.

The essential genes in S. elongatus 7942, which were previously identified by a comprehensive analysis using transposon mutant library [88], are shown in bold.

A. sp 7120 contains additional copies of dnaB, dnaN, dnaX, holA, and holB in plasmids pCC7120alpha (α) and pCC7120beta (β).

Dependence on DnaA varies among cyanobacterial species. In S. sp. 6803 and A. sp. 7120, it has been shown that dnaA is not essential for DNA replication [71,95]. In the symbiotic cyanobacterium Nostoc azolla, the dnaA gene is inactivated by a transposon insertion [96]. dnaA is also not conserved in a diatom endosymbiont of cyanobacterial origin (termed a spheroid body) with a 2.8 Mb genome [97]. Most of the DNA replication enzymes in chloroplasts originated from cyanobacteria and other organisms [98] and are observable vestiges of their symbiotic evolution. However, the only dnaA ortholog is not conserved in the chloroplasts of any plant or alga with sequenced genomes [71]. Genomic analyses suggest that dependence on DnaA varies among cyanobacteria and was lost before the symbiotic relationship that led to the evolution of chloroplast was established.

DNA methylation

DNA methylation is a conserved epigenetic modification that is important for gene regulation and genome stability. Methylation by deoxyadenosine methylase (Dam) plays an important role in DNA replication, DNA repair, and the regulation of gene expression [99–102]. In E. coli, Dam functions as a maintenance methylase during a time window when the chromosome is hemi-methylated following the passage of a replication fork and the parental and daughter strands can be distinguished by their GATC methylation status. The GATC sequences that are hemi-methylated by Dam are involved in multiple aspects of DNA replication, including sequestration of oriC, transcriptional repression of DnaA, and cohesion of sister chromatids in the wake of the progressing replication fork [99,100,102]. In E. coli, although Dam is not essential, a genetic relationship between dam and genes involved in DNA repair has been reported [103–106]. Recently, a novel function for Dam has been reported, the control of aberrant DnaA–oriC-independent replication [107]. Dam-like methylase plays an essential role in V. cholerae, which contains a 1 Mbp secondary chromosome (chromosome II) [108]. Although replication of chromosome II is DnaA-independent, it strictly depends on methylation by Dam [108–111].

In S. sp. 6803, six DNA methylases (M.Ssp6803I–VI) have been identified [112,113]. Among them, M.Ssp6803III and M.Ssp6803IV are Dam-like methyltransferases, encoded by slr1803 and slr6050, respectively, that play crucial roles in S. sp., 6803, as these genes are essential under laboratory conditions [113]. Similar observations have been reported in S. elongatus 7942 and A. sp. 7120 [88,114], and the homologue of M.Ssp6803III in S. elongatus 7942 (Synpcc7942_1790) is also essential. The cytosine methyltransferase M.Ssp6803II, encoded by sll0729, in S. sp. 6803 is also important because it modifies at least 90% of the GG^m4CC recognition sequences, and a disruption mutant of sll0729 exhibited a severe growth defect, with a reduction in chlorophyll a content [113]. A suppressor of the sll0729 disruptant decreased the chromosome copy number and increased the levels of DNA topoisomerase 4 subunit A [25], implicating GG^m4CC methylation and DNA replication in DNA repair in S. sp. 6803. However, further experiments are needed to clarify the physiological role of methylation in cyanobacteria.

The occurrence of cytosine methylation in plastid DNA has been reported in several plants, including sycamore [115], tomato [116], and maize [117]. It has been suggested that cytosine methylation is involved in the transcriptional regulation and differentiation of plastids, although some observations have cast doubt on its physiological significance [118–121]. In Chlamydomonas reinhardtii, there are two mating types (mt+ and mt-), and the chloroplast DNA in mt+ gametes is methylated by a cytosine DNA methyltransferase [122,123]. Only chloroplast DNA from the mt+ parent remained in the mating cells, thus resulting in uniparental inheritance of the mt+ chloroplast DNA [124]. It has been suggested that the difference in DNA methylation could be responsible for the selection of chloroplast DNA by a specific endonuclease [125] and/or the frequency control of plastid DNA replication in mt- and mt+ cells [126].

Regulation of DNA replication

DNA replication activity, as evaluated by the uptake of BrdU, in three cyanobacterial species, S. elongatus 7942, S. sp. 6803, and A. sp. 7120, increased under light conditions and was inactivated by inhibitors of photosynthetic electron transport (PET) [20,127,128], indicating that activation of DNA replication through PET is conserved among cyanobacteria. In S. elongatus 7942, the binding of DnaA to oriC was promoted under light conditions and was inhibited in the dark or in the presence of PET inhibitors, even though DnaA levels remained constant [29,71]. These findings suggest that in S. elongatus 7942, DNA replication is regulated by the oriC binding activity of DnaA, and not the DnaA protein level, which is associated with the integrity of PET. In contrast, in S. sp. 6803 and A. sp. 7120, DNA replication is regulated by a DnaA-independent mechanism [71]. Therefore, although DNA replication is regulated by the electron transport system, the molecular mechanisms that regulate its initiation probably differ among cyanobacteria.

It has been shown that there are differences in DNA replication under dark conditions among these three aforementioned cyanobacteria [127]. In S. sp. 6803 and A. sp. 7120, DNA replication continued after the bacteria were transferred to the dark, whereas in S. elongatus 7942, DNA replication drastically decreased after transfer to the dark. In cyanobacteria, a portion of the electron transport chain, between plastoquinone (PQ) and plastocyanin (PC), is shared by photosynthesis and respiration. Under dark conditions, the respiratory electron transport (RET) system is activated and PET is inhibited. The respiration activity and ATP content in the dark were higher in S. sp. 6803 and A. sp. 7120 than in S. elongatus 7942, suggesting that the observed differences in metabolic activity are correlated with the distinct DNA replication activities that occur in the dark in cyanobacteria [127].

In the unicellular alga C. reinhardtii, which contains one chloroplast per cell, chloroplast DNA is replicated during the light phase but not during the dark phase, independent of the cell cycle and the timing of chloroplast division in photoautotrophic culture [43]. Inhibition of PET blocked chloroplast DNA replication. However, chloroplast DNA was replicated when the cells were grown heterotrophically in the dark, raising the possibility that chloroplast DNA replication is coupled with the reducing power of photosynthesis or respiration as was observed in cyanobacteria [127].

There are several observations suggesting a relationship between central carbon metabolism and DNA replication in E. coli and B. subtilis [129,130], although this has not yet been confirmed. In these bacteria, central metabolic activities directly affect RET activity. The observation of a relationship between PET or RET and DNA replication in cyanobacteria contributes to our understanding of the fundamental mechanisms of proliferation common among bacteria.

Splicing of DNA replication proteins

Protein splicing is a post-translational process involving a large family of proteins called inteins, which have been identified in all three domains of life [131,132]. Inteins are autocatalytic protein domains that excise protein precursors and ligate their flanking regions to a peptide bond. Most inteins are expressed within a single polypeptide chain (cis-splicing inteins), but some are split into two polypeptides, each containing one extein and one intein fragment (trans-splicing inteins). In archaea and mycobacteria, inteins have been found to be responsive to a range of stressors and environmental conditions, including temperature [133], DNA damage [134], salt [135], redox [135], and reactive oxygen species (ROS) [135]. These conditions are often either highly relevant to the environmental niche of the organism or related to the function of the intein-containing protein.

In cyanobacteria, both cis- and trans-splicing inteins have been identified in DNA replication proteins, except for in marine picocyanobacteria, including P. marinus 1375 (Table 1) [132]. In S. sp. 6803, three cis-splicing inteins have been identified in DNA helicase (DnaB) [136,137], the τ subunit of DNA polymerase III (DnaX) [138], and the DNA gyrase B subunit (GyrB) (Table 1) [139]. There are no inteins in DnaX and GyrB in Synechocystis sp. PCC 6714, which is a close relative of S. sp. 6803 [140], although S. sp. 6714 possesses two cis-splicing inteins in DnaB. Another cis-splicing intein has been identified in class II-type ribonucleotide reductase in S. elongatus 7942, A. sp. 7120, and G. violaceus 7421 (Table 1) [141]. These observations suggest that the acquisition of cis-splicing inteins occurred frequently during cyanobacterial evolution and might be mediated by horizontal gene transfer [132].

It is known that DnaE proteins that are split due to the presence of trans-splicing inteins and are conserved among most freshwater cyanobacteria, including S. elongatus 7942, S. sp. 6803, S. sp. 6714, and A. sp. 7120 (Table 1) [142,143]. Since a split DnaE is not present in picocyanobacteria or G. violaceus 7421, which are classified in different clades than freshwater cyanobacteria [141,143,144], the process of splitting DnaE with a trans-splicing intein likely occurred in a common ancestor of the freshwater cyanobacteria after the branching of picocyanobacteria and other cyanobacteria. The splicing point of DnaE is identical among cyanobacteria, indicating that the separation event of DnaE protein occurred once in the evolutional history of cyanobacteria. Measurement of in vivo splicing efficiencies and in vitro kinetics revealed that, in cyanobacteria, split DnaE inteins can catalyze protein trans-splicing in tens of seconds [145–147]. Because they splice efficiently and produce fully functional host proteins, inteins are used as biological tools in the fields of protein biology and synthetic biology [145,148–150].

Although there is no direct evidence answering the question of why inteins are present in DNA replication factors in cyanobacteria, it was recently proposed that the DnaB helicase intein functions as an oxidative stress sensor in Mycobacterium smegmatis [151]. There are two cis-splicing inteins within M. smegmatis DnaB helicase, DnaBi1, and DnaBi2 [151]. DnaBi1 splicing is reversibly inhibited by oxidative and nitrosative insults through the formation of an intramolecular disulfide bond in the catalytic cysteine of DnaBi1. Inhibition of cis-splicing in DnaB directly pauses replication and contributes to the maintenance of genome integrity. It is expected that inteins in cyanobacteria may have a similar function because the intracellular conditions in cyanobacteria induce the production of ROS [27]. Therefore, intein regulation might function as a safeguard against replication stress in the presence of excess ROS in cyanobacterial cells. Further studies are needed to characterize the inteins in cyanobacteria.

Acknowledgments

I sincerely thank Dr. Ryudo Ohbayashi and Dr. Kazuharu Arakawa for critical reading of the manuscript and constructive suggestions.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by grants-in-aid [grant numbers 25850056, 16K07675, and 17H05451] from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, and the Advanced Low Carbon Technology Research and Development Program of the Japan Science and Technology Agency [JST-ALCA, JPMJAL1608].