Intensive DNA Replication and Metabolism during the Lag Phase in Cyanobacteria

Unlike bacteria such as Escherichia coli and Bacillus subtilis, several species of freshwater cyanobacteria are known to contain multiple chromosomal copies per cell, at all stages of their cell cycle. We have characterized the replication of multi-copy chromosomes in the cyanobacterium Synechococcus elongatus PCC 7942 (hereafter Synechococcus 7942). In Synechococcus 7942, the replication of multi-copy chromosome is asynchronous, not only among cells but also among multi-copy chromosomes. This suggests that DNA replication is not tightly coupled to cell division in Synechococcus 7942. To address this hypothesis, we analysed the relationship between DNA replication and cell doubling at various growth phases of Synechococcus 7942 cell culture. Three distinct growth phases were characterised in Synechococcus 7942 batch culture: lag phase, exponential phase, and arithmetic (linear) phase. The chromosomal copy number was significantly higher during the lag phase than during the exponential and linear phases. Likewise, DNA replication activity was higher in the lag phase cells than in the exponential and linear phase cells, and the lag phase cells were more sensitive to nalidixic acid, a DNA gyrase inhibitor, than cells in other growth phases. To elucidate physiological differences in Synechococcus 7942 during the lag phase, we analysed the metabolome at each growth phase. In addition, we assessed the accumulation of central carbon metabolites, amino acids, and DNA precursors at each phase. The results of these analyses suggest that Synechococcus 7942 cells prepare for cell division during the lag phase by initiating intensive chromosomal DNA replication and accumulating metabolites necessary for the subsequent cell division and elongation steps that occur during the exponential growth and linear phases.


Introduction
Cell physiology during the exponential and stationary phases of bacterial growth has been studied extensively [1]. During exponential growth, cells undergo multiple rounds of DNA replication, accompanied by cell division, resulting in rapid cell proliferation. As a consequence of in modified BG11 medium [26] containing 35 mM NaNO 3 (double the amount in normal BG11). To synchronize cell growth, cells were cultured in BG11 medium for 10 days until they reached to OD 750 = 10 (stationary culture). They were then diluted to OD 750 = 0.2 in fresh BG11 medium and incubated for 18 hr in the dark. To reinitiate cell growth, cultures were reexposed to light conditions (time 0) and incubated in the presence of light for the remainder of the experiment. For the preparation of stationary culture of Synechococcus 7942 TK , the medium was supplemented with spectinomycin at a final concentration of 40 μg/ml.

Flow cytometry
Flow cytometry analyses were performed as previously described [23].

BrdU incorporation assay
BrdU incorporation was evaluated by counting the number of BrdU-positive cells, according to the method of Watanabe et al. [23].

Analysis of the composition of the BG11 growth medium
To analyse the composition of the BG11 growth medium, supernatant from Synechococcus 7942 culture grown in BG11 medium for 1 week was obtained by centrifugation and filtration. The composition (anions: nitrate, sulphate, chlorine, and phosphate; cations: sodium, potassium, calcium, magnesium, ferric and manganese) of fresh and spent BG11 media was analysed with the help of a commercial service (Yakult Central Institute, Tokyo, Japan).

Measurement of inorganic phosphate content of the supernatant from Synechococcus 7942 batch culture
The inorganic phosphate content of the batch culture was assayed as follows: after centrifugation, the supernatant obtained from Synechococcus 7942 culture at each time point was subjected to a malachite green assay (BioAssay Systems Co., CA, USA) and the amount of inorganic phosphate was measured according to the manufacturer's instructions.

Metabolomic analyses
Synchronized Synechococcus 7942 wild-type cells were harvested with fast filtration and immediately suspended in the extraction solution, as described previously [29]. Metabolite concentrations were analysed with capillary electrophoresis time-of-flight mass spectrometry (CE-TOFMS), using an Agilent CE capillary electrophoresis system equipped with an Agilent 6210 time-of-flight mass spectrometer, an Agilent 1100 isocratic HPLC pump, an Agilent G1603A CE-MS adapter kit, and an Agilent G1607A CE-ESI-MS sprayer kit with an attached platinum needle (Agilent Technologies, Waldbronn, Germany). Cationic metabolites were separated with a fused silica capillary, using 1 M formic acid as the electrolyte. Anionic metabolites and nucleotides were separated with a COSMO (+) capillary (Nacalai Tesque, Kyoto, Japan), using 50 mM ammonium acetate (pH 8.5) as the electrolyte. Details of the CE-TOFMS conditions and the methods used for data analysis were described previously [30]. Absolute quantifications were performed using metabolite standards and calculated as the amount of metabolite per 10 7 cells (i.e. pmol per 1 × 10 7 cells) in each sample (S2 Table). We were unable to separate the TOF-MS peaks of 3-phosphoglycerate (3-PG) and 2-phosphoglycerate (2-PG). Consequently, they are depicted as the combined total amounts of 3-PG and 2-PG.

Characterization of growth phases in Synechococcus 7942 batch cultures
In previous studies, we observed significant changes in the chromosomal copy number of Synechococcus 7942 in growth-synchronized cultures [23]. While the copy number decreased significantly during stationary phase, it increased when the cells were transferred to fresh BG11 medium and illuminated. To elucidate this phenomenon, we first studied the process of copy number change during long-term culture. A stationary culture of Synechococcus 7942 TK , which contains an HA-tagged thymidine kinase (HA-TK) gene with which to assess the frequency of DNA replication, was diluted, incubated in the dark for 18 h, and then transferred to the light condition to restart cell growth. As depicted in Fig 1A, Synechococcus 7942 TK cells entered exponential growth approximately 18 hr after exposure to light. Growth gradually slowed after 48 hr and nearly arrested after 120 hr. We therefore defined the exponential phase as hours 18-48, and beginning of the linear phase as 48 hours post-light exposure.
The DNA content of the cells at each growth phase was analysed with flow cytometry. The starting culture contained 2-4 chromosomal copies per cell and exhibited a spike-shaped pattern ( Fig 1B, Panel 0), consistent with cells that have completed a full round of DNA replication. Conversely, the DNA profile at 18 hr after transferring to the light condition ( Fig 1B,  Panel 18) manifested a broader shape, and the chromosomal copy number was significantly higher (4-10 copies per cell) than that in the starting culture. The profile gradually returned to a spiky shape at 48 hr post-transfer, and the copy number remained at approximately 2-4 copies per cell throughout the linear phase ( Fig 1B, Panels 72-168). The reduction in chromosomal copy number continued during the late stage of linear phase, eventually returning to 2-4 copies per cell by 96 hr (Fig 1B, Panels 96-168).
We next analysed the expression patterns of HA-TK and two essential proteins, RpoD1, the principal sigma factor [31,32], and DnaK2, an Hsp70 family chaperone protein involved in the stress response [33,34]. The expression of RpoD1 and HA-TK was markedly increased in the exponential phase (Fig 1C, 8-24 hr after inoculation), compared to expression in the starting culture. Levels then decreased during the linear phase (48-192 hr), indicating that RpoD1 is primarily involved in the regulation of genes necessary during the exponential growth of Synechococcus 7942. Conversely, DnaK2 was expressed at similar levels at each time point (Fig 1C). The Hsp70/DnaK chaperone family is required to prevent aggregation and maintain proper folding of many proteins. Therefore, it is likely that consistent expression of DnaK2 proteins is essential during all growth phases.
We quantified the major components of the liquid BG11 medium before and after 1 week of culturing Synechococcus 7942. Compared to the fresh BG11 medium, phosphate and ferric ions were almost lost in the supernatant of 1 week-old culture (S1 Table). In addition, potassium ion content of the 1-week-old culture was found to be less than half of that of fresh BG11. Since phosphate is a primary substrate for DNA synthesis, we additionally assayed the amount of inorganic phosphate under batch culture conditions. The phosphate content gradually decreased with the culturing time and reached an almost undetectable level 96 hr post-transfer (S1 Fig). To assess the cellular response of Synechococcus 7942 to phosphate-limitation in batch culture conditions, we investigated the expression of the phosphate binding proteins (sphX: Synpcc7942_2445 and pstS: Synpcc7942_2444), which are subunits of the ABC-type phosphate transporter and induced in Synechocystis sp. PCC 6803 under phosphate-limiting conditions [35,36]. The levels of both sphX and pstS transcripts transiently increased during the lag phase (9 hr post-transfer), and were dramatically upregulated in the late linear phase (72 hr and 120 hr post-transfer) (S1 Fig). This indicates that the expression of the phosphate transporter was strongly induced when the amount of phosphate in the BG11 growth medium was decreased to less than 15% of that in the fresh medium (S1 Fig, 72 hr post-transfer). At the beginning of the linear phase (48 hr post-transfer), the induction of phosphate-binding subunits was not observed, suggesting that the phosphate level at the time (more than 70% of that of fresh BG11 medium) was sufficient for cell survival and growth (S1 Fig, 48 hr post-transfer). These results suggest that phosphate limitation is not a major factor in triggering the switching from exponential phase to linear phase.

Comparison of DNA replication frequency in the lag and exponential phases
We next compared the frequency of DNA replication in each phase. Because the cell number in Synechococcus 7942 TK cultures exhibited stepwise doubling rather than a linear increase after transfer to the light condition, the timing of cell division was synchronized under this condition (Fig 2A).
To assess DNA replication, Synechococcus 7942 TK cultures were treated at various time points (Fig 2B) with 5-bromo-2 0 -deoxyuridine (BrdU) for 30 min and then examined with immunofluorescence microscopy. As shown in Fig 2A, the cell number doubled between 18 and 24 hr post-transfer. The number of BrdU-positive cells steadily increased upon exposure to light, reaching a maximum 6-9 hr after transfer ( Fig 2B). Interestingly, the peak replication time (i.e. the point at which the proportion of BrdU-positive cells was highest) preceded the onset of cell division. Although thymidine kinase was clearly expressed until 120 hr post-transfer ( Fig 1C, HA-TK), the number of BrdU-positive cells decreased steadily after the peak replication time. The DNA profile analysis supported this observation. The spike-shaped profile gradually broadened between 0 and 24 hr post-transfer, and the chromosomal copy number increased relative to that in the starting culture (Fig 2C, Panels 0-24 hr post-transfer). After the 24 hr time point, which corresponded to the period just after the onset of cell division, the profile began to exhibit a spiked pattern and eventually regained the spike-shaped pattern at 72 hr post-transfer. We hereafter define the period between the culture transfer and the onset of cell division (~20 hr) as the lag phase. In this phase, intensive DNA replication occurred.

Accumulation of chromosomal copies in the lag phase and Synechococcus 7942 proliferation
We previously observed that treatment with the DNA gyrase inhibitor nalidixic acid (NDX) inhibited DNA replication in Synechococcus 7942 [22,23]. In order to establish that intensive DNA replication in the lag phase is essential, we compared the effects of NDX. Five sets of synchronized batch cultures were prepared and treated with NDX at 0, 6, 12, and 24 hr after transferring cells to illuminated conditions. While the control culture (without NDX) doubled 12-18 hr post-transfer (Fig 3A, closed circles), the culture treated with NDX before transfer Intensive DNA Replication during the Lag Phase in Cyanobacteria exhibited severe defects in cell growth and cell doubling (Fig 3A, open triangles). However, the effects of NDX treatment gradually decreased as the incubation time increased. Although treatment with NDX at 6 hr post-transfer abrogated cell doubling, the growth rate of the culture was only slightly slower than that of the control (Fig 3A, closed triangles). The addition of NDX at 12 hr post-transfer (open squares) did not affect the growth rate; however, there was a slight defect in cell doubling beginning at 30 hr after transfer. Lastly, the addition of NDX at 24 hr post-transfer (closed squares) had no effect on growth rate or doubling. We compared the morphology of the control cells and the cells treated with NDX at 12 hr post-transfer (Fig 3A,  open squares). Cells were harvested, stained with SYTOX Green, and visualized with fluorescence microscopy. Compared to cells in the control culture, the NDX-treated cells were larger, with an aberrant DNA distribution (Fig 3B). These results indicate that the lag phase can be clearly defined by the differences in the effects of NDX on cell proliferation and morphology. Furthermore, they suggest that DNA replication during the lag phase is an essential precursor to proliferation steps. Thus, there is likely a chromosomal copy number threshold that must be attained in order to promote efficient cell division.

Metabolome profiles of Synechococcus 7942 during batch culture
To investigate the overall cellular physiology during the lag phase, we compared the metabolome at various time points after transferring cells to illuminated conditions. The timing of cell division in Synechococcus 7942 wild-type cells was synchronized, as described above; the cells were harvested at each point (Fig 4) and analysed with CE-TOFMS. The total amounts of 53 metabolites, normalized to the cell number (pmol per 10 7 cells), are shown in S2 Table,  The level of each metabolite varied throughout the growth phases. Because stationary cultures were diluted and used for the synchronization step, the metabolic state of the starting cultures was similar to those cultured under nutrient-and light-limited conditions. Consequently, upon transfer of these cultures to illuminated conditions, it was necessary for the cells to rapidly produce the various metabolites needed for proliferation. Central carbon metabolites, which are involved in glycolysis, the oxidative pentose phosphate (OPP) pathway, and the TCA cycle (glucose 6-phosphate (G6P), glucose 1-phosphate (G1P), fructose 6-phosphate (F6P), fructose 1,6-bisphosphate (FBP), 6-phosphogluconate (6-PG), and sedoheptulose-7-phosphate (S7P)), accumulated rapidly within 3 hr post-transfer (Fig 4 and S2 Fig). Likewise, the purine precursors, inosine 5 0 -monophosphate (IMP), adenosine, and guanosine, and the pyrimidine metabolites, cytidine, dTMP, and dTTP, increased transiently during this period (Fig 4 and S4 Fig). These metabolites are essential for energy production and DNA synthesis. These findings are consistent with the initiation of chromosomal replication during the lag phase (Fig 2B). The amounts of these metabolites decreased at 9 hr post-transfer, suggesting that central carbon metabolism and pyrimidine metabolism were obstructed between the 3 hr and 9 hr time periods. However, the accumulation of these molecules was apparently resolved by the activation of metabolic flow downstream of each metabolic pathway. Citrate was the only other molecule with a profile similar to those of the central carbon metabolites: levels increased at 3 hr and then decreased by 9 hr after transfer (Fig 4). In contrast, the levels of five metabolites (glutamate (Glu), proline (Pro), glutathione, NADP + , and SAM) transiently decreased at 3 hr but recovered by 9 hr post-transfer (S3 and S5 Figs). The initial reductions in these metabolites might be due to the overflow of carbon metabolites, as mentioned above. As a consequence of their usage, the pools of these molecules might decrease transiently. Several metabolites downstream of the central carbon metabolic pathway (3-and 2-phosphoglycerate (3PG and 2PG), phosphoenolpyruvate (PEP), acetyl CoA, TCA cycle metabolites, and amino   acids) as well as ATP and GTP accumulated at 9 hr post-transfer (Fig 4, S2-S4 Figs). The levels of the DNA precursors adenosine and CDP increased transiently at 18 hr post-transfer (Fig 4), which corresponded to the period at which DNA replication rates decreased (Fig 2B). These observations indicate that, in addition to the onset of DNA replication (Fig 2B), there is a dynamic metabolic change during the lag phase, and that activation of the metabolic pathways proceeds in an ordered manner during this period.
At the onset of cell division (18 hr post-transfer) and during the exponential phase (27 hr post-transfer), metabolite levels were depressed, suggesting that metabolic flow progressed smoothly, directly reflecting the rates of cell proliferation. At the beginning of the linear phase (48 hr post-transfer), growth rates began to decrease (Fig 1A) and several metabolites, especially amino acids, started to accumulate (S2 Table, Fig 4, and S3 Fig), indicating that the obstruction of these metabolic pathways had already begun. These responses may act as a trigger for transition from exponential to linear growth phase. A transient accumulation of F6P, FBP, G6P, ATP, and dTTP was observed at the linear phase (72 hr post-transfer) (Fig 4, S2 and S4 Figs). In contrast, the levels of S7P, amino acids, and glutathione continued to increase during the linear phase (Fig 4, S3 and S5 Figs). These metabolic responses may be due to the effects of both lightand partial nutrient-limitation during the linear phase. Because the amounts of these metabolites in the starting culture were low, the levels of these molecules were likely depleted during the dark incubation period after dilution in fresh BG11 medium. Notably, the profile of lysine (Lys) was distinct from those of the other amino acids (Fig 4). The level of Lys immediately decreased post-transfer and gradually increased thereafter, suggesting that the Lys degradation pathway is activated by light irradiation. However, such regulation has yet to be characterized.

Discussion
We demonstrated that the chromosomal copy number of Synechococcus 7942 varied depending on the growth phase. A graphic depiction of the growth phases of Synechococcus 7942 during batch culture is shown in Fig 6. In particular, ploidy level and metabolite production were found to increase significantly during the lag phase prior to the onset of cell division, indicating that DNA replication and cell division are not closely coupled in the Synechococcus cell cycle. These phenomena are related to regeneration of cells at stationary phase under nutrient-and light-limiting conditions (S1 Table,  We propose that the chromosomal copy number is determined by the balance between DNA replication and cell growth, including cellular metabolism and cell division. (1) In the lag phase, cells undergo rapid DNA replication before the onset of cell division, which is accompanied by a marked increase in the chromosomal copy number (Fig 2). NDX had a strong effect in the lag phase culture (Fig 3), indicating that the accumulation of chromosomal copies during this period is essential for the initiation of exponential growth. Likewise, a significant, orderly increase in metabolism occurs during the lag phase. However, phosphate content was not found to decrease during the lag phase (S1 Fig), suggesting that Synechococcus 7942 utilizes intracellular phosphate during the lag phase. Cells proceeded to the exponential growth phase after generating the necessary chromosomal copies and accumulating sufficient amounts of appropriate metabolites. Our results suggest there is a checkpoint that regulates the transition from the lag phase to the exponential phase. (2) The rate of DNA replication is slower in the exponential phase than in the lag phase, while metabolism proceeds smoothly until the linear phase, reflecting the rapid growth rate. As a result, the chromosome copy number begins to decrease during the exponential phase. In contrast, organisms with a single chromosomal copy, such as E. coli and B. subtilis, exhibit high rates of DNA replication during the exponential phase. (3) At the onset of the linear phase, which corresponds to the beginning of light and nutrient limitation, DNA replication and cellular metabolism are markedly reduced; however, cell growth and cell division continue. Hence, the linear phase cultures exhibit a spike-shaped DNA content profile, as shown in Figs 1B and 2D. It is noteworthy that the switch from exponential phase to linear phase seems to be triggered by light limitation rather than phosphate limitation. The principal sigma factor RpoD1 level began to decrease after 48 hr post-transfer (Fig 1C), indicating that the observed changes in the transcription profile reflect the cellular responses described above.
The polyploidy and intensive DNA replication observed in Synechococcus 7942 before the first cell division suggests that this species possesses an uncoupled DNA replication system. In Synechocystis 6803, multiple chromosome copies are segregated randomly into the two daughter cells during cell division, suggesting that the coordination between chromosome segregation and cell division in Synechocystis 6803 is much less stringent than that in B. subtilis [37].
Similar observations have been made in studies on plastids in plant cells, which arose from a freshwater cyanobacterial lineage [38]. It has been reported that DNA replication in plastids is not tightly linked to nuclear DNA replication or plastid division [39,40]. Plastids undergo intensive DNA amplification before cell division, which is a fundamental step towards subsequent plastid division. In green plants, such as wheat, barley and pea, the plastid DNA number varies during the developmental stages [39][40][41][42][43][44], suggesting that the frequency of DNA replication and plastid division changes at each stage. Additionally, in the unicellular algae Chlamydomonas reinhardtii, which contain one plastid per cell, the plastid DNA replicates multiples times before plastid division [45]. In addition, phosphate limitation strongly affects the copy number of plastid DNA in Chlamydomonas [46]. Our findings are expected to not only contribute to an improved mechanistic understanding of cyanobacterial growth physiology, but also to provide insights into the evolution of plastids.