Present address: Institute of Life Sciences, Ajinomoto Co., Inc., 1-1, Suzuki-cho, Kawasaki-ku, Kawasaki-shi, Kanagawa 210-8681, Japan
New model for assembly dynamics of bacterial tubulin in relation to the stages of DNA replication
Article first published online: 11 FEB 2009
© 2009 The Authors. Journal compilation © 2009 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd
Genes to Cells
Volume 14, Issue 3, pages 435–444, March 2009
How to Cite
Inoue, I., Ino, R. and Nishimura, A. (2009), New model for assembly dynamics of bacterial tubulin in relation to the stages of DNA replication. Genes to Cells, 14: 435–444. doi: 10.1111/j.1365-2443.2009.01280.x
Communicated by: Hiroji Aiba
- Issue published online: 24 FEB 2009
- Article first published online: 11 FEB 2009
- Received: 27 November 2008Accepted: 15 December 2008
How living cells receive their genome through cell division has been one of the important questions of biology. In prokaryotes, cell division starts with formation of a ring-shaped microtubule-like structure, FtsZ-ring, at the potential division site. All the previous models suggested that FtsZ-ring is formed coupling to termination or far after initiation of DNA replication. In contrast, we demonstrated that a close communication with DNA replication is maintained throughout the cell cycle. FtsZ starts to assemble to the cell center coupling to initiation of DNA replication, and stabilizes as FtsZ-ring at its termination, but does not constrict before separation of nucleoids. This combination of a positive and a negative control would guarantee that a successful replication event would inevitably induce one cell division such that each of the daughter cells would receive one and only one daughter nucleoid.
In all organisms, cell division is one of the most conspicuous and ubiquitous features of cellular life, and ‘when and where cells divide to split the genetic materials precisely into their descendants’ has been the prime concern. Bacteria are good model organism to answer these basic problems by its simplicity. Cell division might occurs coupling to DNA replication because synchronized initiation of chromosome replication goes hand in hand with synchronized cell division in a temperature sensitive initiation mutant of DNA replication, dnaC, of Escherichia coli (Yamazoe et al. 2005). Although, Jacob et al. first proposed the replicon model, which aimed to explain how a strict coordination of chromosome segregation events with the cell division might be achieved through the association of the nucleoid with a specific region of the cell membrane in E. coli (Jacob et al. 1963), the roles of membrane in chromosome positioning and segregation remain elusive.
Cell division of E. coli starts with assembly of FtsZ into a microtubule-like structure on the inner membrane of a potential division site in the initial stage of cell division, forming a ring structure, FtsZ-ring, which contracts at the final step of cell division (Bi & Lutkenhaus 1991; Löwe & Amos 1998). FtsZ is a homologue of eukaryotic tubulin (Erickson 1997) and structurally and functionary conserved in probably all prokaryotes (Margolin 2000), archaea (Baumann & Jackson 1996; Margolin et al. 1996; Wang & Lutkenhaus 1996), chloroplasts and some mitochondria (Osteryoung et al. 1998; Strepp et al. 1998), and is a key protein to have a pivotal role for cytokinesis in these organisms (Margolin 2000).
In the current ‘nucleoid occlusion model’ (Woldringh et al. 1991; Zaritsky & Woldringh 2003) or ‘replisome-masking model’ (Harry 2001), FtsZ-ring formation is supposed to start around termination or far after initiation of DNA replication cycle when the two nucleoids (Woldringh et al. 1991; Zaritsky & Woldringh 2003) (or replisomes (Harry 2001) or replication origins (Lau et al. 2003)) move away from midcell, and consequently a midcell structure that defines FtsZ assembly is unmasked.
In most current studies, the cellular dynamics of FtsZ has been investigated using FtsZ-GFP (green fluorescent protein) fusion protein expressed from a plasmid-born ftsZ-gfp (Ma et al. 1996). However, the timing of FtsZ assembly may be normally controlled in part by its cellular concentration, and sometimes induces mini cell formation and cell filamentation depending on its level in a cell (Woldringh et al. 1991; Ward & Lutkenhaus 1985). For the case of conventional plasmid born ftsZ-gfp, timing of FtsZ assembly changes depending on its expression level, in spite of some trials for low-level expression (Sun & Margolin 1998). Moreover, transcription of the ftsZ gene appears to be tightly regulated by at least six promoters (Addinal & Holland 2002), and it is suggested that, although the total amount of FtsZ in the cell may not change dramatically throughout the cell cycle (Rueda et al. 2003), periodic expression of ftsZ replenishes the pool of active FtsZ and causes normal cell division, and constitutively expressed FtsZ from a plasmid-born ftsZ needs 40% more in a cell for normal cell division (Garrido et al. 1993; Sun & Margolin 1998). Furthermore, excess level of background may obscure the faint precursor FtsZ-ring of early stages, causing misread. Therefore, to demonstrate the native dynamics of FtsZ, it is very important to control the expression of ftsZ-gfp as possible as similar to that of native ftsZ.
Here, we demonstrated the several different stages of FtsZ-ring formation and their timing in relation to the stages of DNA replication in an ftsZ, ftsZ-gfp tandem configuration on the chromosome, a condition simpler and less artificial than in previous works. Based on our data, we propose a new model, which can account for a successful replication event inevitably inducing cell division so that the two daughter cells each receive one and only one nucleoid.
Constructed TZG cells grow faithfully natural
To ensure that the dynamics of FtsZ as observed by FtsZ-GFP faithfully reflected the dynamics under normal condition, we constructed a strain (TZG) in which the expression of FtsZ and FtsZ-GFP is under the control of the natural promoter of ftsZ, by inserting the ftsZ-gfp construct into just downstream of the chromosomal ftsZ (Fig. 1) in wild-type E. coli (MG1655) cells. We did not replace wild-type ftsZ by ftsZ-gfp, for the fear of GFP obstructing formation of the tube-like structure if all FtsZ were replaced by FtsZ-GFP. In addition, the ftsZ-gfp construct also encoded a sequence for a long, flexible 27-amino acid linker between the two protein moieties to avoid interference of FtsZ assembly by the GFP, whereas in previous works both were usually fused by only a short linker with several amino acids.
We did not detect any differences in generation time, cell density at saturation, or cell size between TZG cells and the parental MG1655 cells grown in minimal or rich medium. The relative amount of FtsZ-GFP per cell mass was constant during early log to late log stages of growth and amounted to about 10% of the total amount of FtsZ (Fig. 2). In addition, the insertion of ftsZ-gfp in tandem with the wild-type ftsZ gene did not affect expression and degradation of FtsZ (data not shown). We also examined the fraction of cells with an FtsZ-ring both by immunofluorescence microscopy, which detects both FtsZ and FtsZ-GFP, and phase-contrast fluorescence microscopy, which detects only FtsZ-GFP, and found that it was scarcely affected by the insertion of ftsZ-gfp. In contrast, we did find 1.4 times more FtsZ-ring under the conditions of FtsZ-GFP overproduction from plasmid (data not shown).
FtsZ-GFP configurations are classified into five groups
Exponentially growing TZG cells in minimal medium, that prevented overlap of plural generations, were observed by fluorescence microscopy. Generation time of TZG cells in this condition was 103 min (incidentally, that of parent, MG1655, was 99 min). The pattern of FtsZ-GFP assembly showed five types of cells (Fig. 3A); in most TZG small cells, FtsZ-GFP was distributed throughout the cytoplasm (stage a), but in a fraction of small cells FtsZ-GFP was localized at one of their poles (stage e), and in cells of medium length it formed faintly visible FtsZ-ring (stage b). Distinction between (a) and (b) was clear. Type (b) cells showed fair amount of variety of FtsZ assembly pattern as if the FtsZ assembly is on the half way. In most cells of approximately double size it formed a clearly defined ring (stage c), but in a fraction of double sized cells the ring has contracted together with cell membrane (stage d). In contrast, the artificially high expression of plasmid-born ftsZ-gfp (MG1655/pHFGgs50 cells) resulted in the failure to reveal distinction between the two types, (b) and (c), and it was impossible to detect type (e), probably because of the higher background caused by higher expression level of ftsZ-gfp (data not shown) as the previous works. The number and length of cells in each stage are measured and treated statistically (Fig. 3B). Percentage of each type was 27.4% (a), 38.2% (b), 22.8% (c), and 11.6% (d) of total cells, respectively.
FtsZ-GFP dynamics in a single generation of single TZG cells
The timing of FtsZ assembly in the cell cycle has usually been analysed with reference to cell length in the random population of an exponentially growing culture. To confirm the age dependence of cell length and FtsZ-GFP localization, we next followed FtsZ-GFP localization in single TZG cells using the on-chip microculture system (Wakamoto et al. 2001) as in Fig. 4. In this way, single, exponentially growing cells can be observed in real time by phase-contrasted microscopy. Just after cell division, FtsZ-GFP was localized at one pole (105X, Y of Fig. 4). Subsequently FtsZ-GFP scattered throughout the cell (5 min), subsequently FtsZ-GFP gradually started to assemble in the cell center forming a faint ring (20–35 min). It is noticeable that FtsZ-GFP started to assemble in the cell center at more early stage in a cell cycle than the stage predicted from the previous works. This changed to a clearly defined ring between 45 and 75 min, and this state was maintained until 95 min. At this stage, FtsZ-GFP was all concentrated in the ring and was scarcely detected elsewhere in the cytoplasm. The ring began to constrict at around 100 min, and the cell divided into two daughter cells at approximately 105 min, but FtsZ-GFP remained at the divided side pole of each daughter cell for another 5 min, after which FtsZ-GFP was again scattered throughout the cytoplasm. Thus, we have been able to confirm that the temporal sequence of FtsZ-GFP dynamics is stage (a), (b), (c), (d) and (e) in the Fig. 3 experiment.
Furthermore, the plot of cell length against cell age is exponential (Fig. 4B) as mentioned in a previous report (Cooper 1988) by mathematical formulation; If we assume the transition time from (a) to (b), (b) to (c) or (c) to (d) to be 25, 65 or 100 min after cell division, we can predict from Fig. 4B that in a random exponentially growing culture, percentage of cells in stage (a), (b), (c), or (d) will be approximately 30, 38, 27 or 5 %, respectively, using the theoretical age distribution, 2−t/T × (log 2)/2 (T: generation time) (Powell 1956) subject to the simplifying assumption that the generation time of all cells is constant and equal to 103 min. The predicted values are in good agreement with the percentage of cells of each stage in Fig. 3B.
DNA replication stages are classified into four groups
We next analysed the stages of DNA replication with reference to cell length in the random population of an exponentially growing culture to investigate the correlation between stages of FtsZ dynamics and DNA replication stages.
We first analysed each stage of DNA replication as follows; exponentially growing TZG cells were treated with rifampicin, cephalexin, and chrolamphenicol (Fig. 5). This is important to distinguish between a non-replicating nucleoid and a nucleoid that has just begun to replicate. The de novo synthesis of all proteins including DnaA and those necessary for chromosome partition is inhibited by rifampicin and chrolamphenicol (L¢bner-Olsen et al. 1989). As the result, further rounds of replication, separation of sister chromosomes, and increase of cytoplasm are all inhibited immediately after addition of the drugs, but on-going round of replication is completed using pre-existing proteins. By this treatment, a cell with a replicating chromosome results in a cell with a double sized nucleoid, but a cell with non-replicating chromosome results in a cell with a single sized nucleoid. Further, addition of chrolamphenicol causes condensation of nucleoid (Zusman et al. 1973), which would not otherwise be clearly visible in the DAPI stain. Cell division and cell elongation are inhibited by cephalexin, and thus it makes it possible to fix the cell length at the time of drug addition. The cells that have just initiated DNA replication end up as the smallest cells with a double sized nucleoid, whereas those that have just terminated DNA replication end up as the largest cells with a double sized nucleoid. Phase contrast fluorescent microscopy of these cells stained for nucleoid by DAPI showed four types of cells (Fig. 5A); small cells showing a small signal of fluorescence for nucleoid, which should be at the stage before initiation of replication (stage a), medium length cells showing a double sized signal without any gap at the center of signal: they should derive from cells in which DNA was replicating (stage b), similar cells but with a fine gap at the center, which should have terminated DNA replication and have begun separation of the two daughter nucleoids but still retain physical connection between them (stage c), and double length cells showing two separated small signals, which should be at the stage after completion of sister nucleoid separation (stage d). The stage (c) above has not been observed before. The number and size of cells on each stage are measured and treated statistically (Fig. 5B). Percentage of each stage was 27.6% (a), 39.2% (b), 25.2% (c), and 8.0% (d) of total cells, respectively.
Comparison between Figures 3B and 5B shows that the FtsZ-protein assembly cycle exactly corresponds to the stages of DNA replication cycle. The relative size of corresponding peaks in both figures is in good agreement. If we take the cross over points of curves as the measure of transition between phases, the start of FtsZ-GFP assembly towards cell center (relative cell size is 1.29) coincided with the relative cell size that is associated with the initiation of DNA replication (1.27) as predicted from the time sequence study on a single cell. The relative cell size (1.57) at the time when FtsZ-ring is changing from a faint to a defined clear ring coincided with the relative cell size (1.55) that is associated with termination of DNA replication. The two nucleoids, which have terminated replication, did not separate immediately. The ring began to constrict at a cell size (1.78) that is associated with nucleoid separation (1.80).
We confirmed the timing of initiation and termination of DNA replication by flow cytometry. Wild-type E. coli, MG1655, growing exponentially with a generation time of 105 min in a minimal medium, was fixed by ethanol, stained by the highly fluorescent SybrGreen and measured by flow cytometry with high resolution, and the intensity of fluorescence was plotted against the cell mass (Fig. 6). Cells with one nucleoid should be represented by dots around a horizontal line for a uniform intensity of fluorescence at small cell sizes (line I), but cells with two nucleoids should be represented by dots around another horizontal line for an intensity of fluorescence about twice that of cells with a single nucleoid at larger cell sizes (line II), and dots for cells with nucleoids undergoing replication should come around an oblique line between the two horizontal lines (line III). The cell mass at the cross-point between lines I and III (or lines II and III) should be the cell mass at the initiation (or termination) of DNA replication, respectively. The relative cell mass measured here at the initiation (1.24) and termination (1.59) agreed with the respective cell sizes of those in Figure 4 (1.27 and 1.55).
We, therefore, conclude that the FtsZ-ring formation begins as soon as DNA replication has started and that preparation for FtsZ-ring contraction starts as soon as DNA replication has terminated, but the ring does not constrict before separation of the two daughter nuclei.
Our picture of the coordination of DNA replication cycle and FtsZ tubulin assembly cycle is summarized in Figure 7.
All the previous models on the coordination of DNA replication and cell division have supposed the connection to begin at or around termination of DNA replication. But our data presented here shows that a close communication is maintained throughout the cell cycle.
One of the new findings in this paper is that a remarkable change occurs as soon as DNA replication is initiated (stage b) whereas all previous works assumed that DNA replication influences FtsZ assembly state only at its termination or far from its initiation. Previous models such as the nucleoid occlusion model (Woldringh et al. 1991; Zaritsky & Woldringh 2003) and replisome-masking model (Harry 2001) are based on observations made under thymine starvation, under which condition, FtsZ-ring tends to localize to either side of the unreplicated nucleoid but not over the nucleoid, and FtsZ over-production, which precluded observation of FtsZ assembly at initiation of DNA replication. The mechanism of this early communication from DNA replication state to FtsZ assembly remains unknown, but is certainly a most interesting problem. Our hypothesis that the FtsZ-ring formation begins as soon as the initiation of DNA replication, is supported by the observation that in Caulobacter FtsZ-ring appears to assemble at the future division site during the early stages of DNA replication (Quardokus & Brurn 2002), and that during spore outgrowth in Bacillus subtilis a mid-cell FtsZ-ring is unable to form despite a clear DNA-free gap in this region if progression of the first round of chromosome replication is severely limited by thymine limitation (Regamey et al. 2000). We propose that FtsZ assembly rather requires the presence of the machinery for DNA replication, replisome (Kornberg 1980), instead of preclusion. When the replisome is absent under, for example, thymine starvation, or FtsZ is overproduced, FtsZ missing destination or surplus FtsZ would assemble at random abortively.
The period of on going DNA replication is represented in the division cycle by two mild gradients of FtsZ concentration with the cell center as the highest point. We suggest that these gradients are dynamic structure, and its biological significance might be monitoring the DNA replication process and coupling it to the division process. We call this period the ‘Replication monitoring period’. Recovery experiments after photobleaching in E. coli also suggest that FtsZ-ring is extremely dynamic structure, continually exchanging subunits with a cytoplasmic pool (Stricker et al. 2002; Anderson et al. 2004).
This dynamic state of FtsZ assembly is maintained while DNA replication is going on, but when DNA replication is terminated, a sharp band of FtsZ has been formed at the cell center, as had been observed previously, and the FtsZ molecules are in all probability in a polymerized state (stage c). After termination of replication, nucleoids remain at the same place for a while, but we assume that the replisome disappears and space might be made available for proteins that are necessary for stabilizing or constricting FtsZ-ring. We call this period the ‘Baton pass period’. Coupling molecules so far, such as Ap4A (Nishimura 1998), might work on this stage. At this stage the duplicated nucleoids begin to separate by pinching off at the center, but this process of separation does not proceed further and physical contact is retained between the two daughter nucleoids if treated with the drugs used in our experiments. We also observed this image for the first time. The FtsZ ring begins to constrict after the nucleoid separation has been completed, resulting in the cell division with each of the daughter cells receiving one nucleoid (stage d). Just after the cell has divided, FtsZ molecules are localized at the divided side pole of the daughter cells (stage e), but soon they are dispersed homogenously throughout the cell (stage a).
Such a close communication system maintained throughout the cell cycle could allow the cell to monitor the DNA replication processes and co-ordinate it with cell division. It couples initiation of DNA replication to initiation of the FtsZ assembly that is the first step of the processes leading to cell division, and prevents cell division during DNA replication and before daughter nucleoid separation. This combination of a positive and a negative control would guarantee that a successful replication event would inevitably induce one cell division such that each of the two daughter cells would receive one and only one daughter nucleoid.
Experiments in this article deals with only the simplest case where a next round of DNA replication does not start while the on going round is not terminated, but the case of growth in rich medium in which a second round is initiated while the first round is still going on requires special comments. This is evidence that information flow is from DNA replication system to FtsZ assembly system, as we have assumed all along. In a cell in such a state, Fts-Z-ring formed not only in the midcell, but second centers of FtsZ assembly could be formed in other positions, which could be carried over to the daughter cells at cell division. Thus there is no contradiction to our proposed model in the case of growth in rich medium.
Medium and chemicals
Minimal medium consisred of 42 mm Na2HPO4·12H2O, 17 mm KH2PO4, 9 mm NaCl, 19 mm NH4Cl, 1 mm MgSO4·7H2O, 0.1 mm CaCl2, 0.4% Glucose, 0.02% Casamino amino acids, and 40 µg/mL Thymine. Tymine, Gulucose, minimal salts, cephalexin, rifampicin, and 4′, 6-diamidino-2-phenyl-indole (DAPI) were purchased from Nakarai, casamino amino acids from Difco, chrolamphenicol from Sankyo, SybreGreen I from Invitrogen.
Construction of TZG
To create a linker between ftsZ and gfp encoding 50 amino acids, two single stranded DNAs, 5′-CATGGAGCTGCCGCTGCT GCCGCCGCTGCTGCCGCCGCTGCTGCCGCCGCTGC TGCCGCCGCTGCTGCCGCCGCTGCTGCCGCCGCTGC TGCCGCCGCTGCTGCCGCCGCTGCTGCCGCCGCTG CTGCCGCCGCTGCTGCCGCCAGATC-3′ and 5′-GATC GATCTGGCGGCAGCAGCGGCGGCAGCAGCGGCGGCA GCAGCGGCGGCAGCAGCGGCGGCAGCAGCGGCGGC AGCAGCGGCGGCAGCAGCGGCGGCAGCAGCGGCGG CAGCAGCGGCGGCAGCAGCGGCGGCAGCAGCGGCAGC TC-3′ were synthesized, annealed, and ligated with BamHI-NcoI fragment of pEGFP (Clontech, USA) which carries gfp, yielding pEGFPgs50. Next, to introduce cat, HindIII-EcoRI fragments of both pEGFPgs50 and pHSG576 (Takeshita et al. 1987) were ligated, yielding pHSGFPgs50. Then, to introduce ftsZ together with its 20bp upstream region, the region was amplified by PCR using genomic DNA of MG1655 as template and 5′-GACAAGCTTCAGGCACAAATCG-3′ and 5′-TGAGTCGACTCAGCTTGCTTA-3′ as primers, digested with HindIII-SalI, and introduced into the HindIII-SalI site of pHSGFPgs50, yielding pHFGgs50. DNA sequence of PCR-products around ftsZ-gfp in TZG showed that the linker length shortened corresponded to 27 amino acids.
TZG (tandem insertion of ftsZ-gfp) was established as follows; the ftsZ gene with its 20bp upstream region and cat was amplified by PCR using pHFGgs50 as template and 5′-TATCTGGATA TCCCAGCATTCCTGCGTAAGCAAGCTGATTAACAGG CACAAATCGG AGAGAAAC-3′ and 5′-GTTTAGCACAAA GAGCCTCGAAACCCAAATTCCAGTCAATTCAATGGCA CTACAGGCGCCG-3′ as primers, and integrated into just downstream of ftsZ on the chromosomal DNA in the wild-type strain, MG1655, by Wanner's method (Datsenko & Wanner 2000). Constructed ftsZ-gfp was finally transferred to MG1655 by P1-transduction.
Western blot analysis of FtsZ and FtsZ-GFP
Collected cells were lysed in the SDS-PAGE sample buffer at the concentration of 1 OD600/100 µl followed by brief sonication. Equal amounts of whole cell lysates were boiled for 3 min, separated by 15–25% SDS-PAGE using PAG min “daiichi” 15/25 (Daiichi Pure Chemicals Co. LTD) and electrotransferred onto Sequi-BlotTM PVDF membrane (Bio-Rad laboratories). Membranes were blocked with TBS-T containing 5% non-fat dry milk at room temperature for 1 h and subsequently incubated with polyclonal anti-FtsZ antibody, followed by incubation with Protein A-HRP conjugate (Bio-Rad laboratories). After washing with TBS-T, the membrane were developed using an ECL Western blotting system (Amersham).
Microscopic observation of DAPI stained cells
Nucleoids were stained with DAPI solution using the method developed by S. Hiraga et al. (1989), and fluorescence signal was observed by phase-contrasted fluorescence microscope, OPTIPHOTO-2, (Nikon) connected to ORCA-ER-1394 Hamamatsu color chilled 3 CCD camera (Hamamatsu Photonics) with B exciting cubette, DM400 (Nikon).
On-chip microculture system
The on-chip microculture system was constructed as described previously (Wakamoto et al. 2001). The system consisted of a micro-chamber array plate, a cover chamber attached to the array plate separated by semi-permeable membrane, a phase-contrast/fluorescent microscope (Olympus IX-70) inverted microscope, and optical tweezers to trap and discard the unnecessary cells into the waste chamber through a narrow channel. All were maintained at 30 °C. Cells were grown exponential in the minimal medium. The micro-chambers were limited to a single cell each by using optical tweezers (1064 nm) to remove surplus cells into a waste chamber. We found that the generation time of cells growing in the micro-chamber (105 min) did not differ from that of randomly growing tube cultures (103 min), indicating that on-chip micro-culture system did not affect the growth of the TZG cells. The real time dynamics of FtsZ-GFP were observed by fluorescent microscopy periodically. Under the condition, at least three continuous cell cycles could be observed without bleaching. Fluorescent images were acquired simultaneously by using a charge-coupled-device (CCD) camera (Olympus, CS230). Intensities of both cell and FtsZ-GFP fluorescence were recorded along the apse line of the cell using NIH image.
Wild-type, MG1655, cells were grown exponential in the minimal medium at 29 °C, generation time was 105 min in this condition, and fixed by Ethanol and washed in 10 mm Tris (pH 7.4) as described previously (Skaratad et al. 1985), treated with RNase (1 mg/mL) and stained with the double stranded DNA-specific fluorescent dye, SybrGreen I (final dilution 10−4), for 30 min at room temperature. Measurements were carried out with flow cytometer, CyAn ADP (Beckman Coulter K.K.), and analysed by Win MDI.
This article would not have been possible to complete without the thoughtful comments (especially for verification of our result using the theoretical age distribution) and proofreading by Professor emeritus Yoshinobu Sugino. We are also grateful to Drs Koreaki Itoh, Sota Hiraga and Masaaki Wachi for their critical reading of this manuscript and helpful comments. We also thank Dr Kenji Yasuda for lending the device of on-chip microculture system and his technical support, and Dr Joe Lutkenhaus for kindly providing anti-FtsZ antibody.
- 2002) The tubulin ancester, FtsZ, draughtsman, designer and driving force for bacterial cytokinesis. J. Mol. Biol. 318, 219–236. & (
- 2004) Assembly dynamics of FtsZ rings in Bacillus subtilis and Escherichia coli and effects of FtsZ-regulating protein. J. Bacteriol. 186, 5775–5781. , & (
- 1996) An archaebacterial homologue of the essential cell division protein FtsZ. Proc. Natl. Acad. Sci. USA 93, 6726–6730. & (
- 1991) FtsZ ring structure associated with division in Escherichia coli. Nature 354, 161–164. , & (
- 1988) What is the bacterial growth law during the division cycle? J. Bacteriol. 170, 5001–5005. (
- 2000) One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97, 6640–6645. & (
- 1997) FtsZ, a tubulin homologue in prokaryote division. Trends Cell Biol. 7, 362–367. (
- 1993) Transcription of ftsZ oscillates during the cell cycle of Escherichia coli. EMBO J. 12, 3957–3965. , , , & (
- 2001) Bacterial cell division: regulating Z-ring formation. Mol. Microbiol. 40, 795–803. (
- 1989) Chromosome partitioning in Escherichia coli: novel mutants producing anucleate cells. J. Bacteriol. 171, 1496–1505. , , , , , & (
- 1963) On the regulation of DNA replication in Bacteria. Cold. Spr. Harb. Symp. Quant. Biol. 28, 329–348. , & (
- 1980) DNA Replication. New York: Freeman. (
- 1989) The DnaA protein determines the initiation mass of Escherichia coli K-12. Cell 57, 881–889. , , , & (
- 2003) Spatial and temporal organization of replicating Escherichia coli chromosomes. Mol. Microbiol. 49, 731–743. , , & (
- 1998) Crystal structure of the bacterial cell-division protein FtsZ. Nature 391, 203–206. , & (
- 1996) Colocalization of cell division proteins FtsZ and FtsA to cytoskeltal structures in living Escherichia coli cells using green fluorescent protein. Proc. Natl. Acad. Sci. USA 93, 12998–13003. , & (
- 2000) Themes and variations in prokaryotic cell division. FEMS Microbiol. Rev. 24, 531–548. (
- 1996) Isolation of an ftsZ homolog from the archaebacterium Halobacterium salinarium: implications for the evolution of FtsZ and tubulin. J. Bacteriol. 178, 1320–1327. , & (
- 1998) Timing of cell division: Ap4A as the signal. Trends Biochem. Sci. 23, 157–159. (
- 1998) Chloroplast division in higher plants requires members of two functionally divergent gene famiies with homology to bacterial FtsZ. Plant Cell 10, 1991–2004. , , , & (
- 1956) Growth rate and generation time of bacteria, with special reference to continuous cultue. J. Gen. Microbiol. 15, 492–511. (
- 2002) DNA replication initiation is required for mid-cell positioning of FtsZ rings in Caulobacter crescentus. Mol. Microbiol. 45, 605–616. & (
- 2000) Mid-cell Z ring assembly in the absence of entry into the elongation phase of the round of replication in bacteria: coordinating chromosome replication with cell division. Mol. Microbiol. 38, 423–434. , & (
- 2003) Concentration and assembly of the division ring protein FtsZ, FtsA, ZipA during the Escherichia coli cell cycle. J. Bacteriol. 185, 3344–3351. , & (
- 1985) Escherichia coli DNA distributions measured by flow cytometry and compared with theoretical computer simulations. J. Bacteriol. 163, 661–668. , & (
- 1998) Plant nuclear gene knockout reveals a role in plastid division for the homolog of the bacterial cell division protein FtsZ, an ancestral tubulin. Proc. Natl. Acad. Sci. USA 95, 4368–4373. , , , & (
- 2002) Rapid assembly dynamics of Escherichia coli FtsZ-ring demonstrated by fluorescence recovery after photobleaching. Proc. Natl. Acad. Sci. USA 99, 3171–3175. , , & (
- 1998) FtsZ dynamics during the division cycle of live Escherichia coli cells. J. Bacteriol. 180, 2050–2056. & (
- 1987) High-copy-number and low-copy-number plasmid vectors for lacZa-complementation and chloramphenicol- or kanamycine-resistance selection. Gene 61, 63–74. , , , & (
- 2001) Analysis of single-cell differences by use of an on-chip microculture system and optical trapping. Fresenius J. Anal. Chem. 371, 276–281. , , & (
- 1996) FtsZ ring: the eubacterial division apparatus conserved in archaebacteria. Mol. Microbiol. 21, 313–319. & (
- 1985) Overproduction of FtsZ induces minicell formation in E. coli. Cell 42, 941–949. & (
- 1991) Toporegulation of bacterial division according to the nucleoid occusion model. Res. Microbiol. 142, 309–320. , , & (
- 2005) Sequential binding of SeqA protein to nascent DNA segments at replication forks in synchronized cultures of Escherichia coli. Mol. Microbiol. 55, 289–98. , , , & (
- 2003) Localizing cell division in spherical Escherichia coli by nucleoid occlusion. FEMS Microbiol. Lett. 226, 209–214. & (
- 1973) Nucleoid condensation and cell division in Escherichia coli MX74T2 ts52 after inhibition of protein synthesis. J. Bacteriol. 115, 1167–1178. , , & (