INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

Two particularly important periodic biological events are circadian rhythms and the cell division cycle (CDC). The CDC operates in all growing organisms, and circadian rhythms are widely found in organisms from prokaryotic cyanobacteria to essentially all eukaryotes, including protista, fungi, plants, and animals up to human beings (7, 13, 20, 38, 41). Fundamental characteristics of circadian rhythms which define and distinguish them from other periodic phenomena in living organisms are (i) they are endogenous and genetically determined, (ii) the rhythms continue with a ~24-h period under constant conditions that is entrainable by environmental cycles, and (iii) the period length is compensated for changes in ambient temperature over a wide range of physiologically relevant temperatures (3, 12, 37, 42). Despite the fact that circadian rhythms display features that are not shared by the CDC oscillator (such as temperature compensation), some researchers have suggested that there might be a bidirectional interdependent linkage between these two oscillating systems (13, 14, 23). In the case of other circadian rhythms, there is evidence of such bidirectional linkage in that some outputs can feed back onto the central oscillator (36). For circadian rhythm-CDC coupling, other researchers have favored an alternative hypothesis that the circadian clock mechanism oscillates independently of the status of the CDC, but the CDC is dependent on the phase of the circadian clock such that cell division is gated to occur only in specific circadian phases (13, 14, 16, 35).

Cyanobacteria are the simplest organisms in which circadian rhythms have been clearly documented (20). Over the past 15 years, many circadian rhythms including rhythms of photosynthetic activity, nitrogen fixation, global gene expression, andmost relevant for this studycell division have been found in several cyanobacterial species. In fact, among photosynthetic organisms, our knowledge of clock components and interactions is most highly advanced in the unicellular cyanobacterium Synechococcus elongatus. The kai genes that are intimately involved in circadian timekeeping in Synechococcus have been cloned (18), and their homologs have been found in other cyanobacterial species as well as in archaea (6, 20, 21, 22, 29, 33). Deletion of any one of the kai genes does not affect viability (in single-strain cultures) but causes arhythmicity. As had been suggested for other model organisms such as mice, flies, and fungi (17, 28, 30, 43), Ishiura et al. (18) proposed for cyanobacteria that transcriptional and/or translational control of circadian clock genes by their own products (negative feedback regulation) is essential for circadian timekeeping. Biochemical and biophysical analyses of the processes by which the kai genes and their products are involved in circadian timekeeping are under way (19, 20).

We previously reported that cell division in S. elongatus is gated by a circadian oscillator (35). In light-dark (LD) cycles, division occurs only in the light phase. In constant light (LL), where growth is apparently continuous, the cells divide in the subjective day and late subjective night but are kept from dividing in the early subjective night by the circadian oscillator. In this study, we demonstrate that the same kai-dependent clock that regulates gene expression also controls cell division. The bacterial cell division gene ftsZ is expressed with a circadian pattern in Synechococcus. Most importantly, overexpression of FtsZ protein stops the cells from dividing while they continue to grow (resulting in filamentous cells) but does not affect circadian rhythms of gene expression. These results indicate that the circadian pacemaker that gates cell division and gene expression in Synechococcus oscillates stably and independently of feedback from the CDC.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Strains and culture conditions. S. elongatus PCC 7942 (wild type; also known as Anacystis nidulans R2 or Synechococcus sp. strain PCC 7942) and derivative strains were grown in modified BG-11 medium (15) at 30°C. Depending on the antibiotic resistance of each strain, spectinomycin (20 µg/ml), kanamycin (12.5 µg/ml), and/or chloramphenicol (10 µg/ml) was added to the medium. Continuous culturing and measurement of cell density of cultures were performed as previously described (35, 46).

Cloning and sequencing of ftsZ. Plasmid isolation, restriction digestion, ligation, transformation, and Southern blotting were performed essentially as described by Sambrook et al. (40). To make the ftsZ gene probe for screening genomic libraries, genomic PCR was performed with the primers FZTM1 (5'-CCT GAA TTC AAY ACN GAY GNC ARG C-3') and FZTM2 (5'-CCT GAA TTC GTN CCN GTN CCN CCN CCC CAT-3'). These primers were the same as those used by Zhang and coworkers for cloning ftsZ from Anabaena (11, 47). PCR mixtures of 100 µl contained 1 µmol of Tris-HCl (pH 9), 5 µmol of KCl, 150 nmol of MgCl₂, 1 µg of Triton X-100, 20 nmol of each of the deoxynucleoside triphosphates, 50 pmol of each primer, 75 ng of genomic DNA, and 2.5 U of Taq DNA polymerase (Promega, Madison, Wis.). PCR was performed on a Perkin-Elmer DNA thermal cycler (Applied Biosystems, Foster City, Calif.) with an initial hot start at 95°C followed by 30 cycles at 95°C (1 min), 55°C (1 min), and 74°C (2 min) and a final polymerization step of 74°C for 7 min. One major 220-bp DNA fragment was amplified by PCR. This PCR product was cleaved with EcoRI, cloned into the EcoRI site of pBluescript II vector (Stratagene, La Jolla, Calif.), and sequenced. The resulting sequence showed strong similarity to known ftsZ sequences, and the PCR product was used as a probe for genomic Southern analyses and for screening a cosmid library. To screen the cosmid library (provided by Susan Golden), hybridization was performed to a digoxigenin-labeled PCR probe (~20 ng/ml) in 5× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% N-lauroylsarcosine-0.2% sodium dodecyl sulfate-1% blocking reagent (Roche, Indianapolis, Ind.) at 68°C for 16 h, and hybridization of the probe on the filters was detected by the immuno-chemiluminescence method as recommended by the manufacturer (Roche). DNA fragments containing the ftsZ gene were isolated from identified cosmid clones, subcloned into a plasmid vector (pMT111 or pMT115), and sequenced. Plasmid templates for DNA sequencing were sequenced from the insert ends with T3, T7, M13 primers or with custom oligonucleotide primers. All double-stranded DNA templates were prepared with a Wizard Plasmid Miniprep kit (Promega) or by PCR and sequenced on an ABI377 DNA sequencer (Applied Biosystems). Database searches for similarity with other proteins were performed with BLASTP and TBLASTN (1) at the National Center for Biotechnology Information, National Institutes of Health, via the Internet.

RNA analyses. Cells were grown in 24-hour LD cycles (LD 12:12) and harvested. The culture was mixed with crushed ice to immediately chill the cells, and then the cell suspension was centrifuged at 15,000 × g at 4°C for 5 min. The cell pellets were frozen in liquid nitrogen and stored at 80°C. Total RNA was isolated by a modification of the method of Chomczynski and Sacchi (9). Five milliliters of TRI-Reagent (Molecular Research Center, Cincinnati, Ohis) was added to the frozen cell pellet in 13-ml centrifuge tubes, and the samples were vortexed with 2 g of glass beads for 5 min at room temperature. The homogenates were centrifuged at 10,000 × g at 4°C for 10 min; then the supernatants (~3.6 ml of each) were transferred into new tubes to which 360 µl of 1-bromo-3-chloropropane was added and vortexed vigorously for 15 s. The mixture was incubated at room temperature for 10 min and then spun at 10,000 × g for 15 min (4°C). About 2.2 ml of the aqueous phase was transferred to a new tube, and 2.2 ml of isopropyl alcohol was added and kept on ice for 10 min. Total RNA was collected by centrifugation at 12,000 × g at 4°C for 8 min, washed with 75% ethanol, dried, dissolved in water, and stored at 80°C.

Construction of the luciferase reporter strains and in vivo luminescence measurement. To construct ftsZ promoter (ftsZp)-bacterial luciferase transcriptional fusions, the vectors pAM1580 (2) and pAM1583I (2) (modified by M. Izumo [unpublished data]) were used for construction of the luxAB reporter strain. These vectors carry a promoterless luxAB gene set and transfer inserts to the neutral sites of the Synechococcus chromosome by homologous recombination. The upstream region of the ftsZ gene was isolated after restriction digestion of the ftsZ clone or amplified by PCR using Pfu DNA polymerase (Stratagene) and cloned into the unique StuI site upstream of the luxAB genes in the vectors. The constructs were linearized with NdeI to avoid homologous single recombination and then introduced into neutral site I (8) or neutral site II (2) of the wild-type Synechococcus chromosome by homologous double recombination (2, 15). Antibiotic-resistant colonies were selected, and correct transformations were confirmed by genomic PCR or Southern analysis. In vivo luminescence was measured as previously described by Kondo et al. (24, 26) and Mori et al. (35).

Overexpression of ftsZ in E. coli and cyanobacteria. Using primers TRCFZ1 (5'-TCG AGC TCA AGG AGG AAT AAC ATA TGA CCG ACC CTA TGC CGA TC-3') and TRCFZ2 (5'-TGG GAT CCC ATA TGC TAG GGT CGG TTT TGA ATT TTC CG-3') and the cosmid 8B5 as a template, we amplified a 1,215-bp SacI/BamHI fragment which contains the ftsZ gene (underlines denote created SacI and BamHI sites). An alternative ribosome binding site was introduced just upstream of the ftsZ open reading frame (ORF). This 1,215-bp SacI/BamHI fragment was inserted into the SacI-BamHI site downstream of the trc promoter of p322-Ptrc-NdeI, a derivative of p322-Ptrc (27), yielding plasmid pMT411. A 3.4-kb BglII segment of pMT411 that contains lacl^q, trcp::ftsZ, and an rrnB operon region (as a transcription terminator) was isolated and inserted into the BamHI site of either pAM1313 for targeting to neutral site I (2, 24) or pTS2K-NdeI, a derivative of pTS2K (27), for targeting to neutral site II. The constructs were introduced into either neutral site I or II on the chromosome of cyanobacterial strains by double crossover. The FtsZ protein was overexpressed in cyanobacterial strains in liquid or on solid (1.5% agar) BG-11 medium containing 0.5 or 1 mM isopropyl--D-thiogalactopyranoside (IPTG). The cells were observed under a light microscope, and the microscopic images (bright field with 20× or 40× objectives) were captured by a charge-coupled device camera.

Nucleotide sequence accession number. The nucleotide sequence of the S. elongatus ftsZ gene and flanking regions has been deposited in the DDBJ/EMBL/GenBank databases under accession number AF076530.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Circadian rhythm of cell division regulated by kaiC-dependent clock. In a previous study (35), we showed that S. elongatus cells exhibit a rhythm of cell division cycling in continuous cultures of rapidly growing cells. The period length of the cell division rhythm is altered in the same manner by mutations (25) that change the period length of the promoter activity rhythm of the psbAI gene (24, 35). The mutations in strains C22a and C27a (previously named SP22 and LP27, respectively), which altered period lengths of circadian rhythms of both cell division (35) and psbAI promoter activity (25), have been mapped to the circadian clock gene kaiC in the cluster of kai genes (18). We therefore examined the cell division phenotype of a kaiC-deficient strain that is arhythmic for the psbAI promoter rhythm. The wild-type (AMC149) and kaiC deletion (kaiC [46]) strains were grown in batch cultures. In both the wild-type and kaiC strains, cell division occurred in the day phases of LD 12:12 cycles (Fig. 1). After transfer to LL, the wild-type strain exhibited a stepwise growth curve; the cells divided in the subjective day and stopped dividing in the early subjective night (Fig. 1), as we reported previously (35). In contrast, in the kaiC strain, cell density increased without apparent rhythmicity until the culture reached stationary phase (Fig. 1). Lack of a cell division rhythm in the kaiC strain indicates that the same kaiC-dependent clock that regulates global gene expression regulates cell division and supports our prior finding using point mutations of the kaiC gene (35).

View larger version (31K): [in this window] [in a new window]   FIG. 1.   Circadian rhythm of cell division in batch cultures of S. elongatus. Cell number data for the wild-type strain () and a clock-null kaiC strain (). The wild-type and kaiC (46) strains were grown in LD 12:12 and transferred into LL (45 µE/m2/s) at time zero. The last two LD cycles preceding LL are illustrated by the bars on the upper abscissa (white, light; black, darkness; gray, subjective night phases of LL). The left ordinate is for the wild-type strain, and the right ordinate is for the kaiC strain.

Cloning the ftsZ gene from Synechococcus. Although the circadian gating of cell division in cyanobacteria has been clearly demonstrated (35) (Fig. 1), the molecular bases for gating the CDC are still largely unknown. The protein FtsZ is ubiquitous in bacteria and is also found in chloroplasts. In bacteria, a cytoskeletal element called the Z ring is formed in the middle of the cell (5, 39), and septation occurs through the action of the Z ring. The FtsZ proteins assemble into the Z ring and are known to be essential for cell division (4, 10). Our previous study indicated that DNA replication occurs continuously and randomly throughout the circadian cycle within a population of rapidly growing cells but that cytokinesis is gated by the circadian clock such that division is forbidden in the early subjective night (35). Consequently, we hypothesized that genes related to cytokinesis (or septation) might play a role in the circadian gating of cell division; therefore, we cloned the ftsZ gene from Synechococcus and investigated its temporal expression patterns.

View larger version (34K): [in this window] [in a new window]   FIG. 2.   Primary structure and expression of the S. elongatus ftsZ gene. (A) Physical map of the ftsZ gene and its flanking regions. Segments used for constructing luxAB reporter strains are indicated by horizontal bars with the names of the transformed strains. P, PstI; Nc, NcoI; H, HindIII; S, SalI; B, BssHI; No, NotI; X, XhoI. (B) Northern blot analysis of the ftsZ gene. To show equivalent loading of the two lanes in the gel, ethidium bromide staining of the 16S rRNA band is shown. Cells were grown in LD 12:12 cycles and harvested in the day and night. In LD cycles, cyanobacteria divide only in the day (Fig. 1). (C) Determination of putative transcriptional start sites of the ftsZ gene by primer extension. The transcription start site was confirmed by S1 mapping (not shown). (D) Sequence of the promoter region of ftsZ. Putative transcriptional (right-angle arrows) and translational (straight arrows) start sites are indicated.

Circadian rhythms of ftsZp activity. To monitor the expression patterns of ftsZ, we constructed transcriptional fusion strains with the ftsZ promoter linked to a bacterial luciferase gene set. The upstream regions of the ftsZ gene were isolated and transcriptionally fused to the luxAB gene set (the DNA segments fused to luxAB are indicated in Fig. 2A), and the reporter constructs were introduced into neutral sites of the wild-type Synechococcus chromosome. Rhythmic fluctuations of in vivo luminescence from the luxAB reporter strain MTC508 are shown in Fig. 3B and D. The rhythms of luminescence peaked at the end of the subjective day (or the beginning of subjective night), with troughs near the subjective dawn both in slowly growing batch cultures (doubling time [DT] > 24 h [Fig. 3B]) and in continuously diluted cultures of rapidly dividing cells (DT ~ 12 h [Fig. 3C and D]). The phasing of this promoter rhythm was similar to that expressed by the kaiBC promoter, a class I gene (18, 31, 32). In batch cultures of the other reporter strains (MTC164, MTC188, MTC189, MTC190, and MTC191) in which shorter pieces of the 5' region of ftsZ were fused to luxAB (Fig. 2A), luminescence rhythms were as observed from strain MTC508 (data not shown). Therefore, the activity of the ftsZ promoter appears to be under the control of the circadian clock in Synechococcus.

View larger version (36K): [in this window] [in a new window]   FIG. 3.   Rhythms of luminescence in ftsZp::luxAB reporter strain MTC508. (A and B) In vivo luminescence from 3-ml batch liquid cultures of kaiBCp::luxAB (A) and ftsZp::luxAB (B) strains was measured as described by Kondo et al. (24, 26). (C and D). The ftsZp::luxAB cells were grown in LD 12:12 (125 µE/m2/s) and then released into LL, and continuous dilution of the culture was started. Cell density of the culture was monitored every hour, and a growth curve calculated from the actual cell density and the dilution rate (dilution rate = 43.3 ml of medium exchanged every hour in a total volume of 780 ml) is plotted in panel C. The average doubling time of this culture was 12.1 h. From the culture in panel C, 1 ml of cell suspension was withdrawn every 3 h and for measurement of luminescence (D) as described by Mori et al. (35).

Overexpression of FtsZ in cyanobacteria halts division. The FtsZ protein was overexpressed in trcp::ftsZ transformants in liquid BG-11 medium containing 0.5 mM IPTG. As was also found for E. coli (44), overexpression of FtsZ causes the cyanobacterial cells to become filamentous (Fig. 4A). When the FtsZ protein was overexpressed in trcp::ftsZ cells growing on solid (1.5% agar) BG-11 medium containing 1 mM IPTG, single cells formed a long filament (Fig. 4C), whereas colonies formed from single cells of the trcp::null strain on the same medium (Fig. 4B). Microscopic examination confirmed that the filamentous cells did not form septa. Figures 4A and C indicate that cell division, but not cell growth, was stopped when FtsZ was overexpressed from the trc promoter.

View larger version (91K): [in this window] [in a new window]   FIG. 4.   Cell division of growing cyanobacteria is stopped by overexpression of FtsZ, resulting in filamentous cells. (A) The trcp::ftsZ strain was grown for 101 h in liquid BG-11 medium supplemented with 0.5 mM IPTG. The insert in the bottom right corner of panel A shows trcp::null cells as a control under the same conditions. (B and C) The trcp::ftsZ and trcp::null cells were also grown on solid (1.5% agar) BG-11 medium supplemented with 1 mM IPTG for 48 h. A colony of trcp::null cells (B) and a filamentous trcp::ftsZ cell (C) are shown. Presumably both the colony in (panel B) and the filament in (panel C) were derived from a single initial cell.

Circadian rhythms of gene expression in nondividing cyanobacteria. To determine whether circadian rhythms of gene expression persist in nondividing cells, we transformed three different luxAB reporter strains with the trcp::ftsZ construct. Figure 5 shows luminescence patterns that report rhythms of promoter activities of the psbAI, kaiBC, and ftsZ genes in Synechococcus. Whether or not cell division was stopped by overexpression of FtsZ, the expression patterns of all three genes maintained robust circadian fluctuations for at least 4 to 5 days. Rather surprisingly, overexpression of FtsZ protein did not appear to affect the level of ftsZp activity, implying that there is little or no feedback of FtsZ abundance on the ftsZ promoter.

View larger version (37K): [in this window] [in a new window]   FIG. 5.   Luminescence rhythms in dividing and nondividing cyanobacteria in liquid cultures. In vivo luminescence was monitored as described by Kondo et al. (24) in the following reporter strains: (A) psbAIp::luxAB; (B) FtsZ overexpression in the psbAIp::luxAB strain; (C) kaiBCp::luxAB; (D) FtsZ overexpression in the kaiBCp::luxAB strain; (E) ftsZp::luxAB; (B) FtsZ overexpression in the ftsZp::luxAB strain. FtsZ protein was overexpressed continuously with 0.5 mM IPTG in panels B, D, and F, and filamentous morphology in those cultures was confirmed microscopically.