Molecular Microbiology (2012) 83(4), 856–865 䊏
doi:10.1111/j.1365-2958.2012.07971.x
First published online 18 January 2012
Light-dependent and asynchronous replication of
cyanobacterial multi-copy chromosomes
mmi_7971 856..865
Satoru Watanabe,1 Ryudo Ohbayashi,1 Yuh Shiwa,2
Aska Noda,1 Yu Kanesaki,2 Taku Chibazakura1 and
Hirofumi Yoshikawa1,2*
1
Department of Bioscience and 2Genome Research
Center, Tokyo University of Agriculture, Tokyo 156-8502,
Japan.
Summary
While bacteria such as Escherichia coli and Bacillus
subtilis harbour a single circular chromosome, some
freshwater cyanobacteria have multiple chromosomes per cell. The detailed mechanism(s) of cyanobacterial replication remains unclear. To elucidate the
replication origin (ori ), form and synchrony of the
multi-copy genome in freshwater cyanobacteria Synechococcus elongatus PCC 7942 we constructed
strain S. 7942TK that can incorporate 5-bromo-2’deoxyuridine (BrdU) into genomic DNA and analysed
its de novo DNA synthesis. The uptake of BrdU was
blocked under dark and resumed after transfer of the
culture to light conditions. Mapping analysis of
nascent DNA fragments using a next-generation
sequencer indicated that replication starts bidirectionally from a single ori, which locates in the upstream
region of the dnaN gene. Quantitative analysis of
BrdU-labelled DNA and whole-genome sequence
analysis indicated that the peak timing of replication
precedes that of cell division and that replication is
initiated asynchronously not only among cell populations but also among the multi-copy chromosomes.
Our findings suggest that replication initiation is regulated less stringently in S. 7942 than in E. coli and
B. subtilis.
Introduction
Replication is the most fundamental and essential
process in the cell cycle of all organisms. In Escherichia
coli and Bacillus subtilis, DNA replication starts from a
single origin (oriC) in their single circular chromosome
(Bramhill and Kornberg, 1988). The oriC region consists
Accepted 31 December, 2011. *For correspondence. E-mail
[email protected]; Tel. (+81) 3 5477 2758; Fax (+81) 3 5477
2668.
© 2012 Blackwell Publishing Ltd
of multiple repeated sequences that contain consensus
elements termed the DnaA box (Fuller et al., 1984). Replication proceeds bidirectionally around the chromosome
and terminates at a region on the opposite side of oriC
(terC). These replication processes are thought to be
tightly coupled to cell division in the bacterial cell cycle
(Cooper and Helmstetter, 1968; Wang and Levin, 2009).
Cyanobacteria are prokaryotic microorganisms; they
manifest an oxygen-producing photosynthetic system
similar to that of chloroplasts of higher plants. There is
increased interest in cyanobacteria because these photosynthetic organisms convert solar energy to biomass and
thus they may be useful for the production of biofuels.
However, the exploitation of cyanobacteria for bioengineering requires a thorough understanding of their proliferation mechanism(s).
The freshwater cyanobacteria Synechococcus elongatus PCC 7942 and Synechocystis sp. PCC 6803 (S. 7942
and S. 6803 respectively) have been used as model
organisms for phototrophs because their transformation
efficiency and growth rate are superior to those of marine
cyanobacteria and their complete genome sequences are
published. Several species of freshwater cyanobacteria
are oligo- and polyploid organisms harbouring multiple
genomic copies per cell (Mann and Carr, 1974; Binder
and Chisholm, 1995; Mori et al., 1996; Griese et al.,
2011). Among them, it has been reported that S. 7942 is
oligoploid, because it carries three to four genome copies
per cell (Mori et al., 1996; Griese et al., 2011). Chloroplasts of plants, which are thought to derive from cyanobacteria, also contain multi-copy chromosomes (Bendich,
1987; Kuroiwa, 1991). Similar polyploidy are observed in
other bacterial species such as Deinococcus radiodurans
(Minton, 1994) and Thermus thermophilus (Ohtani et al.,
2010).
The genome sequence of freshwater cyanobacteria is
also unique in terms of its GC skew, a plot of the normalized excess of the guanine (G) over the cytosine (C)
content in a subgenomic region with sliding windows
along the entire genome sequence (Lobry, 1996). In many
eubacteria including E. coli and B. subtilis, the GC skew
plot divides the genome into a region with an excess of G
over C and a region with an excess of C over G (the
leading and lagging strand respectively). The shift points
of the GC skew plot have been reported to correlate with
Light-dependent DNA replication in cyanobacteria 857
the loci of ori and ter (Frank and Lobry, 1999). On the
other hand, the GC skews of freshwater cyanobacteria
(S. 6301, S. 7942 and S. 6803), chloroplasts and
D. radiodurans, all of which carry multi-copy chromosomes, are distinct from those of other bacteria including
the marine cyanobacterium Synechococcus sp. WH 8102
(Fig. S1). Since these asymmetrical genomes have many
shift points between high-G and high-C regions, the location of oriC and terC cannot be predicted from GC skew
information alone. In S. 7942, a cluster of dnaA boxes has
been identified in the upstream region of dnaN (Liu and
Tsinoremas, 1996), located on the border between the
high- and low-GC regions in the GC skew analysis
(Fig. S1), and therefore, this region has been predicted as
a replication origin of the S. 7942 genome. Although there
is an informatics method to identify the oriC regions
(Zhang and Zhang, 2005), these predictions have not
been confirmed experimentally.
Here we document the location of the replication origin,
the replication form, and the replication synchrony of
S. 7942. Our data provide significant information for a
thorough understanding of not only the cyanobacterial cell
cycle but also the replication mechanism(s) of organisms
containing multi-copy and asymmetrical genomes.
Results
Growth conditions for investigating DNA replication in
S. 7942
Since S. 7942 cells grow photoautotrophically, their DNA
replication is thought to be activated under light- and
arrested under dark conditions (Binder and Chisholm,
1990). In the stationary phase, the level of DNA synthesis
and the genome copy number are lower than in other
growth phases (Asato, 1979; Binder and Chisholm, 1990).
To investigate the replication initiation process of
S. 7942 we applied the following culture conditions. The
stationary-phase culture grown under continuous light for
10 days was diluted and incubated in the dark for 18 h and
then transferred to the light condition to restart cell growth
(Fig. S2A). The growth curve and the cell number after
transfer are shown in Fig. S2B and C. The cell mass, but
not the cell number, was increased 9 h post transfer. As
the S. 7942 cells showed exponential growth 24 h after
transfer and both their cell mass and cell number
increased significantly, we defined the periods around
these time points as the pre-log (up to approximately 15 h
post transfer) and log phases (15 h and longer post transfer) under this culture condition. During the log phase the
cells multiplied almost synchronously with a doubling time
of about 9 h (Fig. S2C).
Next we investigated the DNA content per cell in each
growth phase. Compared with dark- and briefly light-
exposed (1 h) cultures, about double the amount of DNA
was recovered from pre-log- (9 h) and log-phase (24 h)
cultures (Fig. S2D). When the dark culture was subjected
to flow cytometry (FACS), the cellular DNA profile exhibited a spike-shaped pattern, suggesting that a full round of
DNA replication was completed under the dark condition.
On the other hand, the FACS profile of pre-log-phase
cultures manifested a broader shape and the genome
copy number was significantly higher than in the dark
cultures (Fig. S2E). This suggests that replication of the
multi-copy chromosomes is initiated under the light condition during the pre-log phase.
Quantitative analysis of sequencing reads using a
next-generation sequencer
In rapidly growing bacterial cultures, e.g. E. coli and
B. subtilis, most cells contain two or more replication forks;
consequently they harbour two or more copies of the oriC
region and one copy of the terC region (Yoshikawa et al.,
1964). Sequencing of the whole genome on a nextgeneration sequencer showed that the read depth of oriC is
significantly greater than of terC in cells with multiple
replication forks (Srivatsan et al., 2008). To confirm this
finding we ascertained the location of the replication origin
in the B. subtilis genome (Yoshikawa and Ogasawara,
1991). We prepared genome libraries from log- and
stationary-phase B. subtilis cells. Using a next-generation
sequencer, we sequenced the libraries and mapped them
onto the B. subtilis 168 genome as a reference. Since the
replication frequency of log-phase is higher than of
stationary-phase B. subtilis cells, in log-phase cells the
ratio of sequence reads around the origin region was
approximately twofold higher than of the terC region,
resulting in a V-shaped distribution (Fig. 1A and B). Using
this technique we next tried to identify the replication origin
in the S. 7942 genome. Sequencing detected a circa 50 kb
deletion in the genome of our laboratory strain (indicated
by asterisks in Fig. 1C and D). This 50 kb deletion is from
711 254 to 759 931 of the S. 7942 genome (GenBank:
CP000100) and it is specific to our laboratory strain,
because the S. 7942 strains of Prof. Kondo’s laboratory
(Nagoya strain) do not contain the deletion. Since the
growth rates of our strain and the Nagoya strains containing the deleted region were equivalent, we concluded that
these genotypes have little effect on replication and used
our strain in our experiments. Compared with the distribution of sequence reads in B. subtilis, in S. 7942 there was
no dynamic change under dark and light (log phase) conditions (Fig. 1C and D, left). However, expanded plots
revealed that the read depth ratios around the dnaN gene
(Synpcc7942_0001) were slightly higher than the chromosome region opposite to dnaN in log-phase cells (Fig. 1C
and D, right). We mapped the sequence reads of S. 7942
© 2012 Blackwell Publishing Ltd, Molecular Microbiology, 83, 856–865
858 S. Watanabe et al. 䊏
Fig. 1. Identification of the replication origin by sequencing. B. subtilis and S. 7942 genomes were quantitatively sequenced on a
next-generation sequencer (GAII, Illumina) and the average read numbers per base in a 1 kb window (read depth) were mapped at the
respective genomic positions. The ratio of the read depth at each position to the sum of the read depths throughout the whole genome is
shown. B. subtilis genomic DNA prepared from stationary-phase cells (A), B. subtilis DNA from mid-log-phase cells (B), S. 7942 DNA from
cells grown under dark conditions for 18 h (C), S. 7942 DNA from cells grown under dark conditions for 18 h and under light conditions for
22 h (D). Expanded plots of (C) and (D) are shown on the right. The locations of oriC and terC in the B. subtilis genome and the dnaN gene in
S. 7942 are indicated by arrows. Asterisks: A sequence gap derived from the circa 50 kb genomic deletion in our S. 7942 strain.
onto the S. 6301 genome as a reference; S. 6301 and
S. 7942 are closely related except for the presence of a
188.6 kb inversion, and the location of the dnaN gene in the
S. 6301 genome is distinct from that in S. 7942 (Sugita
et al., 2007). A peak in the read depth ratio was observed
around the dnaN gene (syc1496_c, which locates at
1.62 Mbp position in the S. 6301 genome) (Fig. S3). Nevertheless, there was only approximately 10% difference in
the read depth ratio around the dnaN gene and the chromosome region opposite to dnaN (Figs 1D and S3B).
Analysis of de novo DNA synthesis in S. 7942
De novo DNA synthesis can be monitored by detecting the
incorporation of BrdU, an analogue of thymidine (Dolbeare, 1995; Lewis and Errington, 1997; Hodson et al.,
2003). Since the S. 7942 wild-type strain lacks thymidine
kinase (TK) and therefore cannot incorporate BrdU into its
genomic DNA, we introduced the TK gene from herpes
simplex virus 2 into the S. 7942 genome (named
S. 7942TK). We introduced the TK gene, HA-tagged and
under the control of the Ptrc promoter, into the neutral site
of the S. 7942 genome with a spectinomycin resistance
gene (Fig. S4A). The growth of S. 7942TK was equivalent to
that of the wild-type strain. In S. 7942TK the TK protein was
expressed by a leaky activity of the Ptrc promoter without
IPTG; its expression was increased by transferring the
culture to the light condition (Fig. S4B). We monitored
BrdU uptake in a dark-synchronized S. 7942TK culture.
Under light- but not dark conditions, BrdU was incorporated into the S. 7942TK genome. Since BrdU incorporation
was sufficient without IPTG, we cultured this strain without
IPTG in subsequent experiments. After transfer of the
culture to the light condition BrdU uptake increased and
reached a plateau 9 h post transfer (Fig. 2A); it was clearly
inhibited by the addition of the replication inhibitor nalidixic
acid (Fig. 2B). To investigate the involvement of de novo
protein synthesis in replication initiation, we added
chloramphenicol, a translation inhibitor, at the time of transfer to the light condition. The uptake of BrdU was clearly
© 2012 Blackwell Publishing Ltd, Molecular Microbiology, 83, 856–865
Light-dependent DNA replication in cyanobacteria 859
Fig. 2. Detection of de novo DNA synthesis.
A–C. Immunoblot analyses of BrdU-labelled DNA. The growth
phase of S. 7942TK was synchronized by dark–light conditioning
and nascent DNA was labelled with BrdU during the indicated
periods. DNA samples (50 ng for A, 100 ng each for B and C)
extracted from the cells were blotted and analysed using anti-BrdU
antibody. Nalidixic acid (B, NDX, 3 mg ml-1) and chloramphenicol
(C, Cm, 100 mg ml-1) were added at the time of transfer to the light
condition. Ethanol was the solvent control.
D. Expression of the TK protein without (control) or with inhibitors
(NDX, Cm). RpoD1 was analysed as an internal control. Samples
(40 mg) were analysed by Western blotting using anti-HA and
anti-RpoD1 antibodies.
inhibited by chloramphenicol (Fig. 2C), although the TK
protein exists even after chloramphenicol addition
(Fig. 2D). These results suggest that replication initiation
depends on de novo protein synthesis.
Identification of the replication origin and form
The location of the replication origin can be identified by
quantitative sequencing of BrdU-labelled DNA (replication
sequencing, hereafter referred to as Repli-seq) (Hansen
et al., 2010). To identify the replication origin of S. 7942,
BrdU-labelled DNA, labelled for 1 h before (dark) and 0.5,
1 and 2 h after light exposure, was purified by immunoprecipitation with an anti-BrdU antibody (Fig. S5), and
genomic libraries prepared from the purified DNA were
analysed using a next-generation sequencer. The read
depth of each sample was normalized with a control prepared from immunoprecipitates using control mouse IgG.
Although there were no marked changes in distribution
between the dark cultures and cultures exposed for
30 min to light (Fig. 3A and B), we detected a clear peak
around the dnaN gene locus in the culture exposed to light
for 1 h (Fig. 3C). This peak was broadened in the 2 h light
culture (Fig. 3D). When we mapped the sequence reads
onto the S. 6301 genome as a reference (Fig. S6), we
observed a clear peak around the dnaN gene locus. This
finding was confirmed by quantitative real-time PCR
(qPCR) analysis using BrdU immunoprecipitated DNA.
We performed qPCR assays using primer sets specific for
several genomic regions (see Table S1). The qPCR signal
was first detected upstream of the dnaN gene; it spread
bidirectionally with time after transfer to the light condition
(Fig. S7). These findings were consistent with our Repliseq analysis.
To study the progression of DNA replication newly synthesized DNA was sequentially BrdU-labelled at 30 min
intervals before and after cell transfer to the light condition
and analysed by qPCR. The qPCR signal was first
detected in the dnaN upstream region, indicating that the
replication origin was near the dnaN gene (Fig. 4). In
addition, later on the peak was detected in equidistant
regions opposite the dnaN locus (Synpcc7942_2328
and Synpcc7942_0302). In the regions opposite to
dnaN (Synpcc7942_1001, Synpcc7942_1294 and Synpcc7942_1595), the peak signal was observed 3.5 h post
transfer to the light condition (Fig. 4). These results indicate that DNA replication is initiated as early as 0.5–1 h
after the transfer of cells to the light condition and proceeds
bidirectionally from around the dnaN gene.
Asynchronous DNA replication initiation in S. 7942
To investigate the synchrony of DNA replication among the
S. 7942 cells we examined BrdU-labelled cells using
immunofluorescence microscopy. As in E. coli and B. subtilis (Lewis and Errington, 1997; Adachi et al., 2005) we
observed one or two BrdU foci per S. 7942 cell (Fig. 5A).
Their formation was dependent on the duration of light
exposure and was clearly inhibited by nalidixic acid
(Fig. 5C). These findings indicate that the replication
machinery exists in a specific subcellular location although
multi-copy chromosomes are widely distributed inside the
S. 7942 cells (Fig. 5B), and that the formation of BrdU foci
is dependent on ongoing chromosomal replication. The
BrdU foci were randomly dispersed in the cells (Fig. S8)
although in E. coli they were located in the cell centre
(single focus) or at one-fourth and three-fourths points
along the long axis of the cells (two foci) that correspond
with potential division sites (Adachi et al., 2005).
The number of BrdU-positive cells gradually increased
with the duration of light exposure (Fig. 5C) and reached
a maximum (c. 70%) 6.5 h after transfer to the light con-
© 2012 Blackwell Publishing Ltd, Molecular Microbiology, 83, 856–865
860 S. Watanabe et al. 䊏
Fig. 3. Ratio of read depth at each genomic
position analysed by Repli-seq. Libraries
made from purified BrdU-labelled DNA were
sequenced on a next-generation sequencer
(GAII, Illumina). The sequence reads were
mapped onto the S. 7942 genome as a
reference. Each of the mapping data was
normalized by control data using the library
prepared from DNA precipitated with control
mouse IgG. Cells were labelled with BrdU for
1 h under the dark condition (A), or 0.5 (B),
1 (C) and 2 h (D) after transfer to the light
condition. The location of the dnaN gene in
the S. 7942 genome is shown by arrows.
Asterisks indicate the 50 kb deletions.
dition, suggesting that replication started asynchronously
among the cell populations in culture. The point at which
the proportion of BrdU-positive cells reached its peak (i.e.
the peak replication time, 6–9 h post transfer) preceded
the point of cell division onset (15 h, Fig. S2C); this is
consistent with our findings presented in Fig. 2A and
Fig. S2D and E. We performed whole-genome sequencing at the peak replication time. Different from the
V-shaped distribution of the sequence reads in B. subtilis
log-phase cultures (Fig. 1B), the sequence reads in
S. 7942 at the peak replication time (9 h post transfer)
exhibited an almost flat-shaped distribution (Fig. 5D), as
in log-phase cultures (Fig. 1D). This observation indicates
that only a small number of the replication origins in the
multi-copy chromosomes were fired even when most of
the cells were undergoing DNA replication. Taken
together, these results suggest that replication is initiated
asynchronously not only among cell populations but also
among multi-copy chromosomes. Cell division, on the
other hand, is relatively synchronized (Fig. S2C).
Discussion
Our genome-wide analysis indicated that the DNA replication of S. 7942 is initiated from a single origin located
proximate to the dnaN gene (Figs 1B and C and 3). Since
a cluster of dnaA boxes has been found in the upstream
region of dnaN (Liu and Tsinoremas, 1996) on the border
between the high- and low-GC regions in the GC skew
plot (Fig. S1), the dnaN upstream region was predicted as
the replication origin of the S. 7942 genome. The evidence we provided here strongly supports this prediction.
In the dnaN upstream region, the first qPCR signal
appeared 1 h after the transfer of cultures to the light
condition. No signal was observed after 30 min light exposure (Fig. 4), a period during which the cells are thought to
prepare for growth (e.g. accumulation of energy, gene
expression). In fact, the expression of RpoD1 and TK
protein increased post transfer (Fig. S4B). Furthermore,
DNA synthesis was clearly blocked by chloramphenicol
(Fig. 2B), suggesting that de novo protein synthesis is
essential for the initiation of DNA replication.
Consistent with an earlier report (Yoshikawa and
Ogasawara, 1991) and the GC skew analysis (Fig. S1)
(Lobry, 1996), the replication origin of B. subtilis was
readily identified by sequencing the log-phase genome
library on a next-generation sequencer (Fig. 1A and B).
On the other hand, although the distribution of sequence
reads in S. 7942 did not manifest an obvious peak, the
reads ratio was slightly but significantly higher around the
dnaN gene than in regions on the opposite side of dnaN in
log-phase and peak replication-time cultures (Figs 1D and
5D). The qPCR signal in the dnaN upstream region did not
disappear even when the first signal arrived at the site
directly opposite of dnaN (Fig. 4). These results suggest
that DNA replication was initiated asynchronously. In addi© 2012 Blackwell Publishing Ltd, Molecular Microbiology, 83, 856–865
Light-dependent DNA replication in cyanobacteria 861
Fig. 4. Quantitative real-time PCR analysis of purified BrdU-labelled DNA. Before and after transfer to the light condition, S. 7942TK cells
were labelled with BrdU for 0.5 h before being harvested at the indicated times. BrdU-labelled DNA was purified by immunoprecipitation with
anti-BrdU antibody and used as a template for qPCR analysis with primer sets specific for the genomic regions shown in the centre. In each
diagram the actual number of amplified DNA molecules was calculated by subtracting the background number in the dark-labelled samples
(-0.5 to 0). The BrdU labelling times are: -0.5 to 0 h (Dark), 0 to 0.5 h (0.5), 0.5 to 1 h (1), 1 to 1.5 h (1.5), 2 to 2.5 h (2.5), 3 to 3.5 h (3.5), 4
to 4.5 h (4.5), 5 to 5.5 h (5.5), 6 to 6.5 h (6.5) and 7 to 7.5 h (7.5).
tion, the formation of BrdU foci exhibited asynchrony
among cell populations in the culture (Fig. 5C). We posit
that these asynchronies are reflected in the nearly flatshaped distribution of the read depth ratios shown in
Figs 1D and 5D. Since the cell number increased stepwise rather than linearly after transfer of the cultures to the
light condition, the timing of cell division was apparently
synchronized under this condition (Fig. S2C). We suggest
that the replication initiation of multi-copy chromosomes
occurs independently and asynchronously. We also found
that DNA replication progressed bidirectionally (Figs 3
and 4) as in other eubacteria with a circular genome. On
the side directly opposite of dnaN (Synpcc7942_1294),
the qPCR peak appeared 3.5 h after transfer to the light
condition (Fig. 4), suggesting that under our culture condition about 3 h are required for S. 7942 cells to complete
one round of genome replication. This time is significantly
shorter than the doubling time during the log phase (c.
9 h), suggesting that there is a gap between replication
and cell division or that multiple rounds of replication
occur within a cell division cycle. Indeed, the point of peak
replication (6–9 h post transfer, Figs 2A, 5C and S2D and
E) was significantly earlier than the onset of cell division
(15 h, Fig. S2C), an observation consistent with an earlier
report (Asato, 1979). Based on our findings we hypothesize that cell division is initiated when the multi-copy
© 2012 Blackwell Publishing Ltd, Molecular Microbiology, 83, 856–865
862 S. Watanabe et al. 䊏
Fig. 5. Asynchronous DNA replication in S. 7942 cells.
A. Localization of BrdU-labelled DNA in S. 7942TK cells. The cells were cultured in the dark for 18 h and 6 h after their transfer to the light
condition they were BrdU-labelled for 0.5 h and harvested. Fixed cells were examined by immunofluorescence microscopy using anti-BrdU
antibody. The bright-field and immunofluorescence images of BrdU are merged. Scale bar: 2 mm.
B. Localization of genomic DNA in S. 7942TK cells. After fixation as described in Experimental procedures S. 7942TK cells were stained with
1 mg ml-1 DAPI. A bright-field image (left) and DAPI-stained cells (right) are shown. Scale bar: 2 mm.
C. The proportion of BrdU-positive cells and the number of BrdU foci before and after transfer to the light condition. S. 7942TK cells were
labelled with BrdU for 0.5 h before being harvested at the indicated times and examined by anti-BrdU immunofluorescence microscopy.
8.5+NDX: 3 mg ml-1 nalidixic acid was added at the time of transfer to the light condition and BrdU-labelling was started from 8 h post transfer.
The proportion of BrdU-positive cells among 300 examined cells is shown in different colours based on the focus number per cell: blue = one
focus, red = 2 foci, green = 3 or more foci per cell.
D. Plots of the sequencing read depth (presented as in Fig. 1) prepared from S. 7942TK cells grown in the dark for 18 h and under the light
condition for 9 h. The location of the dnaN gene is shown by an arrow. Asterisk: the 50 kb deletion.
chromosomes reach a threshold number that is regulated
by checkpoint(s) responding to environmental and/or
internal conditions (see below).
Immunofluorescence microscopy revealed that like
E. coli and B. subtilis, S. 7942 cells harboured one or two
BrdU foci (Fig. 5A). In the former, subunits of DNA polymerase III colocalized with the BrdU foci (Lemon and
Grossman, 1998; Kongsuwan et al., 2002; Onogi et al.,
2002). We posit that the replication machinery is locoregional and colocalizes with newly synthesized DNA in the
S. 7942 cell, although its multi-copy chromosomes are
widely distributed inside the cell (Fig. 5B). Since the position of BrdU foci corresponds to potential division sites
(Adachi et al., 2005) that colocalize with the FtsZ division
rings, the replication process in E. coli is thought to be
highly co-ordinated with cell division (Wang and Levin,
2009). On the other hand, in S. 7942 cells, BrdU foci were
randomly dispersed (Fig. S8) although the FtsZ ring is
located in the centre of the cell as in E. coli (Miyagishima
et al., 2005). In S. 6803, multiple nucleoids segregate just
before the complete closing of the division septum,
leading to the suggestion that the co-ordination between
chromosome segregation and cell division in S. 6803 is
much less stringent than in B. subtilis (Schneider et al.,
2007). This is consistent with our observation that the
replication process of freshwater cyanobacteria containing multi-copy chromosomes is not tightly coupled to cell
division. Others (Dong and Golden, 2008; Mori, 2009;
Yang et al., 2010) proposed that cell division is under the
control of the circadian rhythm, which measures daily time
and adjusts to the predictable light–dark alteration. Its
involvement in DNA replication remains unclear, although
© 2012 Blackwell Publishing Ltd, Molecular Microbiology, 83, 856–865
Light-dependent DNA replication in cyanobacteria 863
the rate of DNA synthesis was constant even when the
circadian rhythm of S. 7942 cells was synchronized (Mori
et al., 1996).
Why do freshwater cyanobacteria harbour multiple
chromosomes and an asynchronous replication system?
We posit a correlation with the genome structure represented as an asymmetrical GC skew. We suspect that this
phenomenon is present in a range of organisms with
multi-copy and asymmetrical genomes. To elucidate the
process by which freshwater cyanobacteria replicate, the
replication machinery and replication regulators in cyanobacteria require further study.
Experimental procedures
Culture conditions and synchronization of growth
Unless otherwise indicated, wild-type S. 7942 and its derivatives were grown photoautotrophically at 30°C under continuous illumination (40 mE m-2 s-1) in BG-11 medium with 2%
CO2 bubbling. When appropriate, spectinomycin was added
at a final concentration of 40 mg ml-1. To synchronize cell
growth we cultured the cells in BG-11 medium until they
reached the stationary phase. They were then diluted to
OD750 = 0.2 with fresh BG-11 medium. After 18 h in the dark
the cultures were transferred to the light condition (time 0) to
restart cell growth.
Flow cytometry
For flow cytometry, 1 ml of cell culture was fixed in 0.005%
Tween 20 and 1% glutaraldehyde for 30 min at 4°C, washed
with 1 ml of PBS and stored at -30°C. After freeze–thawing
with liquid N2 the cells were treated with 100 mg ml-1 RNase A
in PBS at 37°C for 1 h, washed once with PBS and resuspended in 50 mM trisodium citrate (pH 8.0). Chromosomal
DNA was stained with 10 mM Sytox Green (Invitrogen, Carlsbad, CA, USA) for 3 h and flow cytometry was performed on
a FACS Calibur instrument (Becton-Dickinson, Palo Alto, CA,
USA). The genome copy number of S. 7942 was estimated
using the chloramphenicol-treated E. coli culture as a
standard.
Construction of the strain carrying the HA-tagged
thymidine kinase gene
To introduce the HA-tagged thymidine kinase (TK) gene into
the chromosome, we constructed plasmid pNSHA based on
plasmid pNSE1 that harbours a spectinomycin-resistant gene
in the fragment derived from the neutral site (Kato et al.,
2008). The SacI–BamHI fragment containing the HA-tag
sequence from pCS2+HA (Turner and Weintraub, 1994) was
cloned between the SacI and BamHI sites of pNSE1. A TK
gene fragment of herpes simplex virus 2 was amplified by
PCR from plasmid pMH6 (Hayashi et al., 2007) using the
primers TK-Bam-f and TK-Bgl-r (Table S1). After digestion
with BamHI and BglII, the fragment was cloned into pNSHA
digested with BamHI. The resulting plasmid was used to
transform the wild-type S. 7942 to spectinomycin resistance.
After PCR confirmation that the TK gene was correctly introduced into the chromosomal neutral site, the resulting strain
was named S. 7942TK.
Immunoblot analysis of BrdU-labelled DNA
S. 7942TK cells were cultured in BG-11 medium in the presence of 1 mM BrdU and harvested at the appropriate times.
When appropriate, nalidixic acid (NDX) and chloramphenicol
(Cm) were added at the time of transfer to the light condition
at a final concentration of 3 mg ml-1 and 100 mg ml-1
respectively. DNA was extracted from each sample using the
DNeasy Plant Kit (QIAGEN GmbH, Hilden, Germany) and
spotted onto Hybond-N+ membranes (GE Healthcare). After
blocking with 5% skim milk in TNT buffer (blocking buffer), the
membranes were incubated with antibodies. For the detection of BrdU, an anti-BrdU monoclonal antibody (Invitrogen)
was used as the primary antibody; horseradish peroxidaseconjugated anti-mouse IgG antibody (GE) was the secondary
antibody. Signals were detected and visualized using the
ChemiDoc XRS+ system (Bio-Rad Laboratories, Hercules,
California, USA).
Immunoprecipitation of BrdU-labelled DNA
BrdU-labelled DNA was purified by immunoprecipitation as
described previously (Cimbora et al., 2000) with minor
modifications. BrdU-labelled DNA (20 mg) was sheared into
500 bp pieces using a Covaris S-2 sonicator (Covaris,
Woburn, MA, USA) and boiled in 500 ml of immunoprecipitation (IP) buffer [0.1 M sodium phosphate (pH 7.0), 0.14 M
NaCl, 0.05% Triton X-100]. After 10 min chilling on ice, biotinconjugated mouse anti-BrdU monoclonal antibody (2 mg,
Invitrogen) or control mouse IgG (Sigma) was added to each
sample. After 30 min incubation at room temperature (RT),
100 ml of streptavidin-agarose (GE Healthcare) was added
and the samples were incubated overnight at RT. DNA–
protein complexes with agarose beads were pelleted by
1 min microcentrifugation at 4°C (the supernatant is indicated
as [Sup.] in Fig. S5), the pellets were washed with 750 ml of
IP buffer, resuspended in 200 ml of digestion buffer [50 mM
Tris (pH 8.0), 10 mM EDTA, 0.5% SDS, 250 mg ml-1 proteinase K] and digested for 3 h at 37°C. After phenol/chloroform
treatment, the DNA was ethanol-precipitated, briefly dried
and resuspended in TE buffer (precipitate [Ppt.] in Fig. S5).
Purified BrdU-labelled DNA was analysed by immunoblotting
(Fig. S5) and qPCR, and sequenced.
Quantitative real-time PCR of BrdU-labelled DNA
BrdU-labelled DNA isolated by immunoprecipitation was
analysed using the 7500 Real-Time PCR System (Applied
Biosystems) with a KAPA SYBR FAST qPCR kit (KAPA Biosystems, Woburn, MA, USA) and a primer set for each locus
(Table S1, target-f and target-r). Standard DNA (500 bp) was
prepared using the corresponding standard-f and standard-r
primer set (Table S1). The actual number of amplified DNA
molecules for each locus was calculated by the amplification
rate of the corresponding standard DNA fragment.
© 2012 Blackwell Publishing Ltd, Molecular Microbiology, 83, 856–865
864 S. Watanabe et al. 䊏
Immunofluorescence microscopy of BrdU-labelled cells
TK
S. 7942 cells were fixed in chilled methanol containing 1%
(w/v) paraformaldehyde and 10% (v/v) dimethyl sulphoxide for
5 min at -80°C and washed two times with PBS. After 15 min
treatment with 0.05% Triton X-100 in PBS the cells were
permeabilized for 30 min at 37°C with 0.2 mg ml-1 lysozyme
dissolved in 25 mM Tris-HCl (pH 7.5) and 10 mM EDTA, and
then washed twice with PBS. For observation of the BrdU
signals, cells were treated with 4 M HCl for 1 h at 37°C and
then washed with PBS. After blocking with 5% bovine serum
albumin in PBS (blocking buffer), the cells were immunostained for 2 h with mouse monoclonal anti-BrdU antibody
(Invitrogen) diluted 1:20 in the blocking buffer, washed twice
with the blocking buffer and incubated for 1 h with Alexa Fluor
488-conjugated goat anti-mouse antibody (Invitrogen) at a
1:200 dilution. After washing twice with the blocking buffer the
cells were examined under a fluorescence microscope
equipped with an Olympus DP71 digital camera (Olympus,
Tokyo, Japan) and analysed with MetaMorph image analysis
software (Molecular Devices, Downingtown, PA).
Library preparation for genome sequencing
Genomic DNA of B. subtilis and S. 7942 was extracted with
the DNeasy Blood and Tissue kit (Qiagen) and the DNeasy
Plant kit (Qiagen) respectively. For sequencing each sample
was prepared according to Illumina protocols. Briefly,
genomic DNA (5 mg) was fragmented to an average length of
200 bp using the Covaris S2 system (Covaris). After repair of
the fragmented DNA and the ligation of ‘A’ to the 3′ end,
Illumina Index PE adapters were ligated to the fragments and
the samples were size-selected for a length of 300 bp using
E-Gel SizeSelect 2% (Invitrogen). The size-selected products
were PCR-amplified for 18 cycles with the primers InPE1.0,
InPE2.0, and an index primer containing 6 nt barcodes
(Illumina). The final products were validated using an Agilent
Bioanalyser 2100 (Agilent, CA, USA). For Repli-seq analysis
we used 20 mg of purified BrdU-labelled DNA for library
preparation in the same manner, except that PCR amplification was with 30 cycles.
Sequencing and data analysis
The barcoded libraries were used for cluster generation in a
multiplexed flow cell lane in the Illumina Genome Analyser II
system. After the sequencing reactions were completed, the
Illumina analysis pipeline (CASAVA 1.6.0) was used for
image analysis, base calling and quality score calibration.
Reads were sorted by the barcodes and exported in the
FASTQ format. The reads from each sample were aligned to
the published complete genomes of the B. subtilis 168
(GenBank: AL009126), S. 7942 (GenBank: CP000100) and
S. 6301 strain (GenBank: AP008231) by MAQ software (Ver.
0.7.1) (Li et al., 2008). The read depth was obtained by calculating the average number of reads per base in each 1 kb
window using the MAQ cns2win command and the ratio of
each read depth to the total read depth was calculated. For
Repli-seq analysis each read depth was normalized with
control data using the library prepared from DNA precipitated
with mouse IgG.
Acknowledgements
We thank H. Masukata (Osaka University) for providing
plasmid pMH6 which contains the thymidine kinase gene of
herpes simplex virus 2 and T. Kondo (Nagoya University) for
providing the S. 7942 Nagoya strains. All authors have no
conflict of interest to declare. This work was supported by the
Ministry of Education, Culture, Sports, Science and Technology [Grants-in-Aid for Scientific Research (S0801025)].
References
Adachi, S., Kohiyama, M., Onogi, T., and Hiraga, S. (2005)
Localization of replication forks in wild-type and mukB
mutant cells of Escherichia coli. Mol Genet Genomics 274:
264–271.
Asato, Y. (1979) Macromolecular synthesis in synchronized
cultures of Anacystis nidulans. J Bacteriol 140: 65–72.
Bendich, A.J. (1987) Why do chloroplasts and mitochondria
contain so many copies of their genome? Bioessays 6:
279–282.
Binder, B.J., and Chisholm, S.W. (1990) Relationship
between DNA cycle and growth rate in Synechococcus sp.
strain PCC 6301. J Bacteriol 172: 2313–2319.
Binder, B.J., and Chisholm, S.W. (1995) Cell cycle regulation
in marine Synechococcus sp. strains. Appl Environ Microbiol 61: 708–717.
Bramhill, D., and Kornberg, A. (1988) A model for initiation at
origins of DNA replication. Cell 54: 915–918.
Cimbora, D.M., Schubeler, D., Reik, A., Hamilton, J., Francastel, C., Epner, E.M., and Groudine, M. (2000) Longdistance control of origin choice and replication timing in
the human beta-globin locus are independent of the locus
control region. Mol Cell Biol 20: 5581–5591.
Cooper, S., and Helmstetter, C.E. (1968) Chromosome replication and the division cycle of Escherichia coli B/r. J Mol
Biol 31: 519–540.
Dolbeare, F. (1995) Bromodeoxyuridine: a diagnostic tool in
biology and medicine, Part I: Historical perspectives, histochemical methods and cell kinetics. Histochem J 27:
339–369.
Dong, G., and Golden, S.S. (2008) How a cyanobacterium
tells time. Curr Opin Microbiol 11: 541–546.
Frank, A.C., and Lobry, J.R. (1999) Asymmetric substitution
patterns: a review of possible underlying mutational or
selective mechanisms. Gene 238: 65–77.
Fuller, R.S., Funnell, B.E., and Kornberg, A. (1984) The dnaA
protein complex with the E. coli chromosomal replication
origin (oriC) and other DNA sites. Cell 38: 889–900.
Griese, M., Lange, C., and Soppa, J. (2011) Ploidy in
cyanobacteria. FEMS Microbiol Lett 323: 124–131.
Hansen, R.S., Thomas, S., Sandstrom, R., Canfield, T.K.,
Thurman, R.E., Weaver, M., et al. (2010) Sequencing
newly replicated DNA reveals widespread plasticity in
human replication timing. Proc Natl Acad Sci USA 107:
139–144.
Hayashi, M., Katou, Y., Itoh, T., Tazumi, A., Yamada, Y.,
Takahashi, T., et al. (2007) Genome-wide localization of
pre-RC sites and identification of replication origins in
fission yeast. EMBO J 26: 1327–1339.
Hodson, J.A., Bailis, J.M., and Forsburg, S.L. (2003) Efficient
© 2012 Blackwell Publishing Ltd, Molecular Microbiology, 83, 856–865
Light-dependent DNA replication in cyanobacteria 865
labeling of fission yeast Schizosaccharomyces pombe with
thymidine and BUdR. Nucleic Acids Res 31: e134.
Kato, H., Chibazakura, T., and Yoshikawa, H. (2008) NblR is
a novel one-component response regulator in the cyanobacterium Synechococcus elongatus PCC 7942. Biosci
Biotechnol Biochem 72: 1072–1079.
Kongsuwan, K., Dalrymple, B., Wijffels, G., and Jennings,
P.A. (2002) Cellular localisation of the clamp protein during
DNA replication. FEMS Microbiol Lett 216: 255–262.
Kuroiwa, T. (1991) The replication, differentiation, and inheritance of plastids with emphasis on the concept of
organelle nuclei. Int Rev Cytol 128: 1–61.
Lemon, K.P., and Grossman, A.D. (1998) Localization of bacterial DNA polymerase: evidence for a factory model of
replication. Science 282: 1516–1519.
Lewis, P.J., and Errington, J. (1997) Direct evidence for active
segregation of oriC regions of the Bacillus subtilis chromosome and co-localization with the SpoOJ partitioning
protein. Mol Microbiol 25: 945–954.
Li, H., Ruan, J., and Durbin, R. (2008) Mapping short DNA
sequencing reads and calling variants using mapping
quality scores. Genome Res 18: 1851–1858.
Liu, Y., and Tsinoremas, N.F. (1996) An unusual gene
arrangement for the putative chromosome replication
origin and circadian expression of dnaN in Synechococcus
sp. strain PCC 7942. Gene 172: 105–109.
Lobry, J.R. (1996) Asymmetric substitution patterns in the
two DNA strands of bacteria. Mol Biol Evol 13: 660–665.
Mann, N., and Carr, N.G. (1974) Control of macromolecular
composition and cell division in the blue-green algae Anacystis nidulans. J Gen Microbiol 83: 399–405.
Minton, K.W. (1994) DNA repair in the extremely radioresistant bacterium Deinococcus radiodurans. Mol Microbiol 13:
9–15.
Miyagishima, S.Y., Wolk, C., and Osteryoung, K.W. (2005)
Identification of cyanobacterial cell division genes by comparative and mutational analyses. Mol Microbiol 56: 126–
143.
Mori, T. (2009) Cell division cycles and circadian rhythms. In
Bacterial Circadian Programs. Ditty, J., Mackey, S.R., and
Johnson, C.H. (eds). Berlin: Springer, pp. 183–204.
Mori, T., Binder, B., and Johnson, C.H. (1996) Circadian
gating of cell division in cyanobacteria growing with
average doubling times of less than 24 hours. Proc Natl
Acad Sci USA 93: 10183–10188.
Ohtani, N., Tomita, M., and Itaya, M. (2010) An extreme
thermophile, Thermus thermophilus, is a polyploid
bacterium. J Bacteriol 192: 5499–5505.
Onogi, T., Ohsumi, K., Katayama, T., and Hiraga, S. (2002)
Replication-dependent recruitment of the beta-subunit of
DNA polymerase III from cytosolic spaces to replication
forks in Escherichia coli. J Bacteriol 184: 867–870.
Schneider, D., Fuhrmann, E., Scholz, I., Hess, W.R., and
Graumann, P.L. (2007) Fluorescence staining of live
cyanobacterial cells suggest non-stringent chromosome
segregation and absence of a connection between cytoplasmic and thylakoid membranes. BMC Cell Biol 8: 39.
Srivatsan, A., Han, Y., Peng, J., Tehranchi, A.K., Gibbs, R.,
Wang, J.D., and Chen, R. (2008) High-precision, wholegenome sequencing of laboratory strains facilitates genetic
studies. PLoS Genet 4: e1000139.
Sugita, C., Ogata, K., Shikata, M., Jikuya, H., Takano, J.,
Furumichi, M., et al. (2007) Complete nucleotide sequence
of the freshwater unicellular cyanobacterium Synechococcus elongatus PCC 6301 chromosome: gene content and
organization. Photosynth Res 93: 55–67.
Turner, D.L., and Weintraub, H. (1994) Expression of achaetescute homolog 3 in Xenopus embryos converts ectodermal
cells to a neural fate. Genes Dev 8: 1434–1447.
Wang, J.D., and Levin, P.A. (2009) Metabolism, cell growth
and the bacterial cell cycle. Nat Rev Microbiol 7: 822–827.
Yang, Q., Pando, B.F., Dong, G., Golden, S.S., and van
Oudenaarden, A. (2010) Circadian gating of the cell cycle
revealed in single cyanobacterial cells. Science 327:
1522–1526.
Yoshikawa, H., and Ogasawara, N. (1991) Structure and
function of DnaA and the DnaA-box in eubacteria: evolutionary relationships of bacterial replication origins. Mol
Microbiol 5: 2589–2597.
Yoshikawa, H., O’Sullivan, A., and Sueoka, N. (1964)
Sequential replication of the Bacillus subtilis chromosome.
3. Regulation of initiation. Proc Natl Acad Sci USA 52:
973–980.
Zhang, R., and Zhang, C.T. (2005) Identification of replication
origins in archaeal genomes based on the Z-curve method.
Archaea 1: 335–346.
Supporting information
Additional supporting information may be found in the online
version of this article.
Please note: Wiley-Blackwell are not responsible for the
content or functionality of any supporting materials supplied
by the authors. Any queries (other than missing material)
should be directed to the corresponding author for the article.
© 2012 Blackwell Publishing Ltd, Molecular Microbiology, 83, 856–865