Microbiology (2014), 160, 2538–2550
DOI 10.1099/mic.0.081695-0
Deletion of the Synechocystis sp. PCC 6803
kaiAB1C1 gene cluster causes impaired cell
growth under light–dark conditions
Anja K. Dörrich,1 Jan Mitschke,2 Olga Siadat2 and Annegret Wilde2
Correspondence
1
Annegret Wilde
[email protected]
2
freiburg.de
Received 18 June 2014
Accepted 14 August 2014
Institute for Microbiology and Molecular Biology, Justus-Liebig-University, Heinrich-Buff-Ring 26,
35392 Giessen, Germany
Institute of Biology III, University of Freiburg, Schänzlestrasse 1, 79104 Freiburg, Germany
In contrast to Synechococcus elongatus PCC 7942, few data exist on the timing mechanism of
the widely used cyanobacterium Synechocystis sp. PCC 6803. The standard kaiAB1C1 operon
present in this organism was shown to encode a functional KaiC protein that interacted with KaiA,
similar to the S. elongatus PCC 7942 clock. Inactivation of this operon in Synechocystis sp. PCC
6803 resulted in a mutant with a strong growth defect when grown under light–dark cycles, which
was even more pronounced when glucose was added to the growth medium. In addition, mutants
showed a bleaching phenotype. No effects were detected in mutant cells grown under constant
light. Microarray experiments performed with cells grown for 1 day under a light–dark cycle
revealed many differentially regulated genes with known functions in the DkaiABC mutant in
comparison with the WT. We identified the genes encoding the cyanobacterial phytochrome
Cph1 and the light-repressed protein LrtA as well as several hypothetical ORFs with a complete
inverse behaviour in the light cycle. These transcripts showed a stronger accumulation in the light
but a weaker accumulation in the dark in DkaiABC cells in comparison with the WT. In general, we
found a considerable overlap with microarray data obtained for hik31 and sigE mutants. These
genes are known to be important regulators of cell metabolism in the dark. Strikingly, deletion of
the DkaiABC operon led to a much stronger phenotype under light–dark cycles in Synechocystis
sp. PCC 6803 than in Synechococcus sp. PCC 7942.
INTRODUCTION
Cyanobacteria are obligate phototrophs performing oxygenic photosynthesis. As such, they are highly dependent
on predictable daily changes in light availability. In order
to regulate their metabolism in accordance with daily
environmental changes, cyanobacteria have evolved a circadian timing mechanism. The underlying molecular mechanism was found to be fundamentally different from the
eukaryotic circadian clock, including the plant timing
mechanism. This is surprising because chloroplasts evolved
from cyanobacteria, and many plant genes in the
chloroplast as well as in the nucleus originate from a
cyanobacterial ancestor (Rujan & Martin, 2001). The
cyanobacterial circadian clock system consists basically of
three Kai proteins encoded by the kaiABC cluster forming
the central oscillator and several other components of the
Abbreviations: qRT, quantitative real-time; ZT, zeitgeber time.
The full microarray dataset is accessible from the GEO database with
the accession number GSE58572.
Two supplementary figures, two supplementary tables and two supplementary datasets are available with the online Supplementary Material.
2538
input and output signalling pathways. After 20 years of
work, the mechanism of this Kai-based circadian clock is
very well understood in the model system Synechococcus
elongatus PCC 7942 (hereafter Synechococcus 7942)
(recently reviewed by Pattanayak & Rust, 2014). Whilst
genomic data suggest that most cyanobacteria harbour
genes for the oscillator and subsets of the input and output
components (Shih et al., 2013), very little is known about
the timing mechanisms in other cyanobacteria. For some of
these, it was shown that a bona fide circadian clock system
exists in the cells. This was demonstrated for Cyanothece sp.
ATCC 51142 on the level of gene expression using
microarrays (Stöckel et al., 2008; Toepel et al., 2008) and
physiological analysis (Červený & Nedbal, 2009), as well as
for Thermosynechococcus elongatus using luciferase reporter
strains (Onai et al., 2004). Single reports of circadian
regulation have also been published for several other
cyanobacteria (for review, see Axmann et al. 2014). The
cyanobacterium Synechocystis sp. PCC 6803 (hereafter
Synechocystis 6803) is a model organism for the study of
photosynthesis, gene regulation and biotechnological
applications. The most important characteristics of this
strain are its natural competence for transformation
Downloaded from www.microbiologyresearch.org by
081695 G 2014 The Authors
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
Printed in Great Britain
Synechocystis 6803 DkaiABC mutant
(Grigorieva & Shestakov, 1982) and its ability to grow on
glucose as an energy and carbon source (Anderson &
McIntosh, 1991). Aoki et al. (1995) generated a luciferase
reporter strain for Synechocystis 6803 and showed a true
circadian expression of the dnaK gene by monitoring
bioluminescence. In a microarray analysis, Kucho et al.
(2005) demonstrated that not more than 10 % of the genes
in this organism exhibited a circadian rhythm of mRNA
accumulation. In contrast to Synechococcus 7942, the
genome of Synechocystis 6803 contains, in addition to the
standard kaiABC gene cluster (kaiAB1C1), an additional
kaiC2B2 cluster and two orphan kaiC3 and kaiB3 genes.
It was proposed that only the kaiAB1C1 cluster is important for circadian oscillations in Synechocystis 6803
(Aoki & Onai, 2009). To prove this hypothesis, we recently
evaluated the biochemical activities of the potential
components of a central oscillator (Wiegard et al., 2013).
All three Synechocystis 6803 KaiC proteins were analysed
regarding autokinase activities. The well-studied KaiC
protein from Synechococcus 7942 forms hexamers and
shows autokinase activity, which is stimulated by KaiA as
well as autophosphatase activity that is promoted indirectly
by KaiB. In a study by Wiegard et al. (2013), it was shown
that only the KaiC1 protein from Synechocystis 6803
interacts with KaiA and autophosphorylates actively when
KaiA is present in the assay. KaiC3 and KaiC2 are able to
autophosphorylate independently of KaiA. In addition,
purification of KaiC proteins from Synechocystis 6803 cells
suggested that there is no heterologous interaction between
the three different KaiC proteins. These analyses implied
that only the kaiAB1C1 gene cluster is able to drive a
timing mechanism in Synechocystis 6803, whereas the
additional kai genes might have other functions. Here,
we analysed the phenotype of a kaiAB1C1 deletion mutant
(DkaiABC) at the physiological and transcriptome levels.
overlapping BamHI and EcoRI restriction sites (underlined). Both
PCR products were fused in an overlap PCR using the kaiA upstream
forward primer and the kaiC1 downstream reverse primer. The
amplified overlap product was ligated into the pJET1.2/blunt cloning
vector (Thermo Scientific). A kanamycin resistance gene cassette
(KmR) from the pUC4K vector was then inserted into the introduced
BamHI restriction site. The orientation of KmR in the construct was
confirmed by sequencing. The kaiC3 (slr1942) deletion mutant
(DkaiC3) was constructed in a similar way using the primers kaiC3
upstream (forward 59-CCAGGTTAAGGCTGTGAAAG, reverse 59GGATCCCACGTACTTTCACTGCCCCATACTC) and kaiC3 downstream (forward 59-TACGTGGGATCCCAATTCTTCCCTTTGTAAACC; reverse 59-GGTTTCCGATAAGCGC). The upstream reverse
primer and the downstream forward primer contained overlapping
BamHI and BsaAI restriction sites (underlined). After ligation of the
overlap product into pJET1.2/blunt, a chloramphenicol resistance
gene cassette (CmR) from the pACYC184 vector was inserted into the
BsaAI restriction site.
Synechocystis WT cells were transformed with the resulting plasmids
pJET1.2-DkaiABC : : KmR or pJET1.2-DkaiC3 : : CmR. The deletion
mutants were maintained on BG11 agar plates with a stepwise
increase in antibiotics until a final concentration of 40 mg kanamycin
ml21 or 7 mg chloramphenicol ml21 was reached. Complete segregation of the mutant gene copies was confirmed by PCR or Southern
blot analysis following standard procedures.
Spot assays. Cell cultures of WT and mutants (DkaiABC and
DkaiC3) were grown under the conditions specified above. Cells were
METHODS
grown in medium lacking antibiotics to exclude effects from growth
on selective media. One day prior to the experiment, the cultures were
diluted to OD750 0.2 and were incubated further under the abovementioned conditions. For spot assays, cultures were split and diluted
to OD750 0.2 and 0.4, each followed by a dilution series from 100 to
1025. From each of the series, 5 ml was spotted in triplicates onto
BG11 agar plates supplemented with 0.3 % (w/v) Na2S2O3 and 0.2 %
(w/v) glucose if necessary. The plates were either incubated under
constant illumination or under a 12 h–12 h light–dark cycle. For
plates incubated under constant illumination, the dilution series
originating from the part of the culture that was diluted to OD750 0.2
was used. For plates incubated under the 12 h–12 h light–dark cycle,
the dilution series based on the part of the culture that was diluted to
OD750 0.4 was spotted. All plates were incubated until no further
colonies appeared in the highest dilution step.
Bacterial strains and growth conditions. The glucose-tolerant and
Absorbance spectra. Precultures of WT and mutants were grown
motile strain of Synechocystis 6803 used in this study was obtained
originally from S. Shestakov (Moscow State University, Russia) and
resequenced recently (Trautmann et al., 2012). Cell cultures of WT
and mutants were grown primarily under constant illumination with
white light at 50 mmol photons m22 s21 in BG11 medium (Rippka
et al., 1979) supplemented with 10 mM TES, pH 8 at 30 uC. For
the light–dark experiments performed in this study, cultures were
incubated under alternating (12 h–12 h light–dark cycle) illumination with the same intensity during the light period. Depending on
the experimental set-up, strains were either grown photoautotrophically or mixotrophically by the addition of 0.2 % (w/v) glucose.
under the same conditions as indicated for the spot assays. Three days
prior to the start of the experiment, the cultures were adjusted to
OD750 0.3. At time point 0 (t0), all cultures were diluted to OD750
0.15, partially supplemented with 0.2 % (w/v) glucose and incubated under a 12 h–12 h light–dark cycle. Absorbance spectra were
measured with a SPECORD 50 UV/Vis spectrophotometer (Analytik
Jena) at t0 (constant light) and after 6 days of incubation in the light–
dark cycle.
Mutant construction. For construction of DkaiABC, the upstream
region of kaiA (slr0756) and the downstream region of kaiC1
(slr0758) were amplified by PCR with specific primers and genomic
WT DNA as template. The primers used were kaiA upstream (forward
59-CAAAATATGGGCTTCATTAAAAGTACCG, reverse 59-GGATCCGAATTCAGAAAGGCACAAAAATCATCTAG) and kaiC1 downstream (forward 59-GAATTCGGATCCTGGTTGGGAAAGGCGGCATG; reverse 59-GTGACCCAGGATGGTCGG). The kaiA upstream
reverse primer and the kaiC1 downstream forward primer contained
http://mic.sgmjournals.org
RNA isolation. An aliquot of 10 ml cyanobacterial cell culture was
filtered through Supor-800 0.8 mm membrane filters (Pall). Each filter
was immediately put into a 15 ml reaction tube containing 1.5 ml
PGTX reagent (Pinto et al., 2009), fragmented by vortexing, frozen in
liquid nitrogen and stored at 220 uC. Samples were thawed for 7 min
at 95 uC in a water bath and vortexed several times in between to
further dissolve the membrane filter and to support cell lysis. After
incubation on ice for 5 min, 150 ml cold bromochloropropane was
added and the sample was mixed thoroughly by vortexing. Following
incubation at room temperature for 15 min, the phases were
separated by centrifugation at 6000 g for 15 min at 4 uC. The
aqueous phase was transferred to a 2 ml reaction tube, treated with an
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
2539
A. K. Dörrich and others
equal volume of phenol/chloroform/isoamylalcohol (25 : 24 : 1) and
centrifuged at 10000 g for 7 min at 4 uC. The RNA in the supernatant
was precipitated overnight at –20 uC by addition of an equal volume
of 2-propanol. The RNA was pelleted by centrifugation at 10000 g for
30 min at 4 uC, washed with 1 ml of 75 % (v/v) ethanol and pelleted
again (10000 g for 15 min at 4 uC). The RNA was air-dried for
10 min and resuspended in 50 ml diethylpyrocarbonate-treated
double-distilled H2O. The total extracted RNA was quantified using
a NanoDrop 1000 spectrophotometer (Thermo Scientific). Integrity
of the RNA was examined by electrophoretic separation on a 1.3 %
(w/v) formaldehyde agarose gel. Prior to the microarray analysis and
quantitative real-time (qRT)-PCRs, the RNA was treated with DNase
I (Invitrogen). Absence of genomic DNA in the samples was
confirmed by PCR analysis.
RNA as template. We performed non-template control reactions for
each primer pair and included negative controls omitting reverse
transcriptase to verify the absence of DNA. A melting curve analysis
confirmed the specificity of each primer pair. For each run, the
primer efficiency was calculated according to Pfaffl (2001). Raw data
were analysed using Bio-Rad CFX Manager software. Expression
levels were calculated within the linear range of all amplification
curves and normalized to the internal standard according to Pfaffl
(2001). Transcript accumulation was calculated as the mean of two
biological replicates, present in technical duplicates.
RESULTS
In order to investigate the physiological importance of
the kaiAB1C1 gene cluster in Synechocystis, we constructed
a mutant lacking the entire gene cluster (Fig. 1a). The
complete segregation of mutant gene copies was confirmed
by PCR (Fig. 1b). Using specific peptide-derived antisera
(Wiegard et al., 2013) against KaiC1 (Fig. 1c) in Western
blot analysis, no band corresponding to the respective
protein size was detected.
Microarray analysis. Labelling and hybridizations were performed
as described in Georg et al. (2009), with 3 mg RNA used for labelling
and 1.65 mg RNA for hybridization. Analysis was done in R (version
3.0.2) with the limma package (Smyth et al., 2005). Normalization was performed according to the process described in Georg
et al. (2009) with minor adjustments. Fold changes .0.9 were
considered significant. Microarray data were visualized in a wholegenome expression plot (Supplementary Dataset 1, available in the
online Supplementray Material). The full dataset is accessible from
the GEO database with the accession number GSE58572 and in
Supplementary Dataset 2.
Growth of DkaiABC is impaired under light–dark
conditions, but not under constant light
qRT-PCR. qRT-PCRs were performed in a one-step reaction using
Brilliant III Ultra-Fast SYBR Green QRT-PCR Master Mix (Agilent)
and a Bio-Rad CFX96 cycler. We used gene-specific primer pairs to
measure the accumulation of cph1 and dnaK2 transcripts normalized
to rnpB expression as internal standard. Primer sequences are listed in
Table S1. For each primer pair, a standard curve based on an RNA
dilution series was generated. The qRT-PCRs were performed in a
10 ml reaction volume using 0.5 mM of each primer and 0.5–4 ng total
To determine whether deletion of the kaiAB1C1 operon
results in a light-related phenotype, we performed spot
assays and analysed the viability of WT and DkaiABC
under constant light, as well as under alternating 12 h–12 h
light–dark conditions (Fig. 2). Investigation of a DkaiC3
mutant served as a control (for construction of this mutant
4354 bp
(a)
BamHI
kaiA-US-F
WT
US
B1
kaiA
kaiC1
DS
kaiC1-DS-R
BamHI
2651 bp
DkaiABC
kaiA-US-F
KmR
US
DS
kaiC1-DS-R
BamHI
ai
A
T
W
9
W
T
5
or
N
TC
4
Ve
ct
1
B
C
(c)
DkaiABC
Dk
(b)
BamHI
Anti-KaiC1
4354 bp
2651 bp
Anti-AtpB
Fig. 1. Construction of the DkaiABC mutant. (a) The kaiAB1C1 gene cluster was deleted and replaced by a kanamycin
resistance cassette (KmR). F, forward; R, reverse; US, upstream; DS, downstream. (b) Complete segregation of the mutant
gene copies was confirmed for four different clones (1, 4, 5 and 9) by PCR analysis. The primers used are indicated in (a). NTC,
non-template control. (c) The absence of the KaiC1 protein in DkaiABC was validated in a Western blot using a KaiC1-specific
antibody. Anti-AtpB served as a loading control.
2540
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
Microbiology 160
Synechocystis 6803 DkaiABC mutant
strain, see Fig. S1). All strains were grown in liquid cultures
under constant light, spotted at different dilutions onto
agar plates and incubated as indicated. Under photoautotrophic conditions (Fig. 2a) and continuous illumination,
all strains showed the same viability. When incubated
under light–dark conditions, WT and DkaiC3 cells were
not impaired in their growth compared with constant light
incubation. In contrast, DkaiABC showed a strongly
restricted viability and impaired growth behaviour. To
investigate whether the phenotype could have been caused
by a photosynthetic defect, we performed the same experiment under mixotrophic conditions by adding 0.2 % (w/v)
glucose to the agar plates (Fig. 2b). When incubated under
constant light, no difference in viability could be detected
between WT and DkaiC3, and when compared with
photoautotrophic conditions. In contrast, DkaiABC showed
a more yellowish pigmentation and slightly smaller colonies
than the other strains, and when compared with photoautotrophic growth. However, according to the number of
colonies in the different dilution steps, the viability of
mixotrophically grown cells was clearly not influenced under
constant light. On the contrary, when incubated under
alternating 12 h–12 h light–dark periods, the viability of
DkaiABC was markedly decreased compared with the WT
and the DkaiC3 control. The light–dark-related phenotype
was even stronger under mixotrophic than photoautotrophic conditions and therefore did not seem to be related
to a defect in photosynthesis function. In order to reduce
the probability of compensatory mutations, the DkaiABC
mutant cells were always kept under constant light
conditions and were subjected to light–dark cycles only for
a short experimental time.
Long-term light–dark incubation causes reduced
pigmentation and bleaching of DkaiABC
We measured whole-cell absorbance spectra of WT and
mutant cells grown in liquid cultures under constant light
and after 6 days of growth under a light–dark cycle (Fig. 3).
Whereas WT and mutant (DkaiABC and DkaiC3) cultures
grown under constant light showed similar absorbance
spectra (Fig. 3a), the DkaiABC mutant displayed an overall
reduced pigment content when incubated photoautotrophically under a light–dark cycle (Fig. 3b). The phycocyanin peak (628 nm) showed an even stronger decline when
mixotrophic conditions were applied (Fig. 3c). It is
noteworthy that the bleaching phenotype only occurred
after five or six light–dark periods and was not evident at
earlier time points.
Microarray analysis
In order to analyse changes on the transcriptome level in
the DkaiABC mutant under light–dark conditions compared
with the WT, we performed a microarray analysis. The
experimental setup of the RNA sampling strategy is shown
in Fig. 4. RNA samples used for the hybridizations are
marked in bold. After 3 days of growth in liquid cultures
http://mic.sgmjournals.org
(a)
Continuous light
Light–dark cycle
WT
DkaiABC
–Glucose
DkaiC3
100
10–5 100
Continuous light
(b)
10–5
Light–dark cycle
WT
DkaiABC
+Glucose
DkaiC3
100
10–5 100
10–5
Fig. 2. Growth and viability in light–dark cycles under photoautotrophic and photomixotrophic conditions. WT and mutant cells
were spotted in triplicate at different dilutions (100 to 10”5) onto
BG11 agar either (a) without or (b) with the addition of 0.2 % (w/v)
glucose. The plates were incubated under continuous or
alternating (12 h–12 h light–dark cycle) illumination of 50 mmol
photons m”2 s”1 until no further colonies appeared in the highest
dilution step. The observed effect was verified in three independent spot assays. One representative result is shown. The DkaiC3
mutant served as a control.
under continuous illumination, cells were shifted to light–
dark conditions. As DkaiABC mutant cells showed restricted growth under diurnal cycles, and to distinguish
between primary responses and secondary effects involved
in the bleaching process, we followed two independent
sampling approaches. In the first approach, cells were
harvested during the first period of alternating illumination
in the dark and light periods (early time points). In the
second and independent approach, cells were harvested in
the light period after 5 days of growth under a light–dark
cycle, when DkaiABC mutant cells showed a bleaching
phenotype (late time point). The early time points most
probably show the direct effects of the DkaiABC mutation
on gene transcription, whereas data resulting from the late
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
2541
A. K. Dörrich and others
Absorption
(a)
WT
DkaiABC
DkaiC3
0.08
0.06
Continuous light
–Glucose
0.04
0.02
0
400 450 500 550 600 650 700 750
(b)
Absorption
0.16
0.12
Light–dark cycle
–Glucose
0.08
0.04
0
400 450 500 550 600 650 700 750
In DkaiABC, various metabolic processes are
disarranged on transcriptomic level under
light–dark conditions
(c)
Absorption
0.12
0.09
Light–dark cycle
+Glucose
0.06
time point might reflect secondary effects due to growth
limitations. Every condition and time point is represented
by two replicates. The microarrays were processed and
normalized as described in Georg et al. (2009). If necessary,
probes with a P value .0.05 were omitted from the
calculation. From the remaining probes, a mean value was
calculated for each unique feature. For the expression
analysis, fold changes (log2) were calculated between the WT
and DkaiABC mutant. A list of top hits was established based
on these fold changes compared over the different time
points (Table 1). Table 1 includes the top 53 differentially
regulated transcripts in the first light–dark cycle of
annotated genes. A complete list of all microarray data is
provided in Supplementary Dataset 1. In addition, a
graphical representation of the fold change values for each
probe on the chromosome is shown in Supplementary
Dataset 2.
0.03
0
400 450 500 550 600 650 700 750
Wavelength (nm)
Fig. 3. Pigmentation in light–dark cycles under photoautotrophic
and mixotrophic conditions. Precultures were incubated under
continuous illumination for 3 days. (a) Prior to shifting to 12 h–
12 h light–dark conditions, absorbance spectra were measured.
(b, c) The precultures were then split and incubated under
alternating illumination for 6 days, either in the absence (b) or
presence (c) of 0.2 % (w/v) glucose. The DkaiC3 mutant served as
a control. Each spectrum represents the mean of three independent replicates.
During the first dark period, several transcripts accumulated to lower amounts in DkaiABC in comparison with the
WT (highlighted in blue, Table 1). Among them are two
ribosomal operons (rpl3–rpl15, consisting of 18 genes, and
the rpl21/27 operon), suggesting a downregulation of
protein synthesis in the dark in the DkaiABC mutant.
Only some representative genes of these operons are
listed in Table 1 due to our restrictive significance values.
However, the genome plot shown in Supplementary Dataset
1 suggests that all genes of the two ribosomal operons are
affected in their transcript abundance. Further more, genes
encoding proteins involved in electron transport, such as
petE, petD, petB as well as the respiratory electron transport
terminal oxidase (ctaC1D1E1 operon), accumulated to
lower amounts in the mutant in the dark [not all of the
genes belonging to these operons are listed in Table 1 due
to fold changes below the threshold used of ±1.0 fold
changes (log2) or to a weak P value]. However, whereas
Late
Early
L
L
L
Dilute cultures to
same OD750
0h
D
11.5 h
L
DL DL DL D
ZT17.5 ZT23.5 ZT0.5 ZT5.5
RNA samples
L
ZT5.5
D
ZT 17.5
RNA samples
Fig. 4. Timescale and sampling strategy of WT and DkaiABC cells grown under a 12 h–12 h light (L)–dark (D) cycle. The WT
and mutant were grown simultaneously in replicate liquid cultures. After 3 days of incubation under constant light, the cultures
were diluted to the same OD750 and conditions were shifted to alternating illumination. Time is given as zeitgeber time (ZT). At
the indicated time points (arrows), cells were harvested and total RNA was isolated. In a first approach, samples were taken
during the first diurnal period (early). In a second and independent approach, samples were taken after 6 days of incubation in
the light–dark cycle, when mutant cells started to bleach (late). The isolated RNA was used for microarray analysis (bold) and
qRT-PCR.
2542
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
Microbiology 160
Synechocystis 6803 DkaiABC mutant
ctaC1D1E1 was downregulated also in the light, the petE,
petD and petB transcripts accumulated to higher amounts
in the mutant in the first light period in comparison with
the WT. This could be an indication of a primary
imbalance between photosynthetic and respiratory electron transport in the DkaiABC mutant. The gene cluster
slr0144–slr0152, which shows a similar expression profile
with a lower accumulation in the dark and a higher accumulation in the light in the mutant, is usually downregulated in cells experiencing oxidative stress (Singh et al.,
2004). The proteins encoded by this operon contain
binding domains for 2Fe–2S clusters and bilins, co-factors
of the cytochrome b6f complex and ferredoxins, and were
found to be associated with photosystem II pre-complexes
(Wegener et al., 2008). Furthermore, several unknown
genes, as well as genes for the protochlorophyllide reductase
subunit ChlL, RNase E and the water-soluble carotenoid
protein OCP, were downregulated in the dark in DkaiABC
compared with the WT. However, there were also transcripts accumulating to higher amounts in the mutant
compared with the WT in the dark (highlighted in yellow,
Table 1). Some of these transcripts, such as dnaJ, dnaK2,
hspA and clpB (slr1641), are regulated positively by
the histidine kinase Hik34 under stress conditions
(Paithoonrangsarid et al., 2004). Consequently, we detected
a higher transcript accumulation for hik34 in the dark in the
mutant. This was also true for the two photosystem I
transcripts psaAB and psaM as well as the slr1674/slr1675
operon, where slr1675 encoded a hydrogenase maturation
protein (Hoffmann et al., 2006). Although most genes
showed a differential expression only under one condition,
some of the transcripts, such as hypothetical genes slr0144–
slr0152 and slr1152, but most interestingly also the
transcripts for light-repressed protein lrtA and the cyanobacterial phytochrome cph1, accumulated in an inverse
manner in the DkaiABC mutant, i.e. lower accumulation
in the dark but higher accumulation in the light compared
with the WT. Finally, the transcript levels for sigE and norB
(encoding the cytochrome b subunit of nitric oxide
reductase) as well as for transaldolase talB, 6-phosphogluconate dehydrogenase gnd and several hypothetical proteins
were downregulated in the mutant during the first light
period, but also at the late time point in the light. Regarding
transcript accumulation after 5 days of light–dark cycles,
we detected downregulation of many (but not all) genes
known to be induced under low-carbon conditions, like
the inorganic carbon transporters sbtAB and genes for the
NDH1MS complex (sll1732–sll1734), as well as cmpB and
flv2. However, whereas other NADH-quinone oxidoreductase subunits do not show differential expression
under low-carbon conditions (Schwarz et al., 2013), ndhI,
ndhB and ndhJ were also downregulated in the DkaiABC
mutant. In addition, transcripts for psbA3 and psbA2 as
well as other transcripts encoding photosynthesis-related
proteins accumulated to lower levels, suggesting that the
cells were limited in their ability to grow optimally in the
light, and this may partly explain the bleaching phenotype
of the DkaiABC mutant. In contrast, most of the mRNAs
http://mic.sgmjournals.org
characterized by a higher expression level in mutant
cells in comparison with the WT at this late time point
coded for proteins classified into Gene Ontology categories such as cell envelope, fatty acid and amino acid
metabolism, DNA replication and transport functions,
suggesting that the cells may have also suffered from
nutrient limitations.
Accumulation of non-coding transcripts
Of the features on the microarray, 612 were considered to
be possible non-coding transcripts (Mitschke et al., 2011):
25 of them were located on the plasmids and 587 were
located on the chromosome. Of these, only 34 show a
strong fold change between the mutant and WT at
zeitgeber time (ZT) point ZT5.5 during the light phase
and only 16 (not including CRISPR and other plasmid
sequences) showed a strong differential regulation in the
mutant in the dark. Several of these differentially regulated
non-coding transcripts are most likely 59 or 39 UTRs of
adjacent annotated genes or internal transcripts. However,
putative non-coding transcripts NC-318 and NC-795 show
a different expression pattern in comparison with the
upstream and downstream coding regions. In addition, the
verified non-coding transcript SyR6, which has been shown
to be upregulated under increased light conditions (Georg
et al., 2009), is less strongly expressed in the DkaiABC
mutant compared with the WT. These non-coding transcripts, as well as antisense transcript showing significant
fold changes, are listed in Table S2. No functions for these
transcripts are known to date.
Validation of microarray data using qRT-PCR
In order to validate our microarray data, we chose two
genes, cph1 and dnaK2, which were inversely regulated in
the DkaiABC mutant compared with the WT (Table 1).
In our microarray analysis, we detected a decrease in
abundance of cph1-related transcripts in the dark (ZT17.5)
in DkaiABC compared with the WT, which became an
increase in the light (ZT5.5) (Fig. 5a). For dnaK2, we
detected an increase in transcript abundance in the dark
(ZT17.5) in DkaiABC compared with the WT, whilst
transcript levels of the WT and mutant showed no
differential expression in the light (ZT5.5) (Fig. 5b). To
validate these data, mRNA levels of cph1 and dnaK2 were
analysed by qRT-PCR (Fig. 6). In order to investigate
changes in the expression profiles of these genes at a higher
density, we used RNA samples from additional time points
as shown in Fig. 4. All samples used originated from the early
sampling approach (Fig. 4). On the whole, the qRT-PCR
data correspond to the results of the microarray experiment.
Whilst cph1 was downregulated in DkaiABC compared with
the WT in the dark and upregulated in the light, dnaK2
expression in the mutant was upregulated in the dark and
progressively fell over the light period (Fig. 6). Curiously, the
slight downregulation of dnaK2 in the light (ZT5.5) in
DkaiABC compared with the WT as shown by qRT-PCR
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
2543
A. K. Dörrich and others
Table 1. List of genes showing significant transcript accumulation in DkaiABC mutant cells relative to the WT under a light–dark
cycle (see Fig. 4)
The relative expression [fold changes (log2)] and adjusted P values are given. Changes of at least 1.0-fold and 21.0-fold and a P value of 0.05 were
regarded as significant. Downregulated transcript levels are marked in blue and upregulated levels in yellow. Genes are sorted by function and fold
change.
Locus tag
Gene name
Translation
slr1678
sll1802
sll1799
sll1807
Photosynthesis
slr0343
sll0199
slr1136
slr1835
smr0005
Metabolism
sll0329
slr1793
Cell envelope
ssl1533
ssl2501
sll1694
slr1841
Regulation
sll0359
slr1129
slr0473
sll1689
slr1594
slr1285
Others
slr1963
slr0749
sll0450
sll0947
sll0108
slr1675
DkaiABC/WT [fold changes (log2)]
Adjusted P value
Dark
Light
Light (late)
rpl21
50S ribosomal protein L21
rpl2
50S ribosomal protein L2
rpl3
50S ribosomal protein L3
rpl24
50S ribosomal protein L24
and respiration
petD
Cytochrome b6f complex subunit 4
petE
plastocyanin
ctaC
Cytochrome c oxidase subunit II
psaB
Photosystem I subunit PsaB
psaM
Photosystem I subunit PsaM
21.10
21.09
21.20
21.16
0.09
0.24
20.22
0.53
20.36
0.18
0.50
0.54
1.60610–3
5.63610–3
7.26610–3
3.20610–3
21.07
21.03
20.13
1.05
1.29
0.58
0.46
21.19
0.66
20.03
21.00
21.05
20.77
20.65
20.89
1.09610–3
5.04610–3
1.51610–5
6.22610–3
1.30610–4
gnd
talB
6-Phosphogluconate dehydrogenase
Transaldolase
20.48
20.18
21.31
21.11
20.96
21.05
6.53610–5
3.49610–5
ssl1533
ssl2501
pilA1
slr1841
Unknown protein
Unknown protein
Pilin polypeptide PilA1
Probable porin
22.67
22.37
21.85
21.02
21.07
20.07
0.03
0.46
21.29
21.62
20.86
0.16
4.57610–5
1.95610–5
1.55610–5
1.91610–2
cyabrB1
rne
cph1
sigE
slr1594
hik34
AbrB family transcriptional regulator
RNase E
Cyanobacterial phytochrome 1 Hik35
RNA polymerase sigma factor SigE
Two-component response regulator PatA subfamily
Two-component sensor histidine kinase
21.55
21.29
21.03
0.04
0.11
1.66
0.37
0.37
0.58
21.17
1.36
0.11
20.92
0.56
0.10
20.42
0.04
20.08
3.50610–5
2.50610–4
4.75610–2
1.80610–4
1.91610–4
2.68610–3
ocp
chlL
Water-soluble carotenoid protein
Light-independent protochlorophyllide reductase
subunit ChlL
Cytochrome b subunit of nitric oxide reductase
Light-repressed protein A
Ammonium methylammonium permease
Putative hydrogenase expression formation protein
HypA1
ClpB protein
CP12 polypeptide
DnaK protein
16.6 kDa small heat-shock protein
DnaJ protein
21.08
21.07
0.40
0.29
20.64
1.07
1.94610–3
1.76610–3
20.76
20.37
20.13
1.13
21.54
1.05
1.47
0.24
21.71
20.33
0.71
20.52
1.57610–5
1.46610–2
3.75610–5
5.41610–4
1.26
1.42
1.44
1.72
1.82
0.15
20.21
20.01
0.05
0.09
20.56
20.78
20.57
20.87
20.44
1.65610–5
1.40610–4
8.02610–4
2.18610–4
2.53610–5
AhpC/TSA family protein
Hypothetical ferredoxin
PetF-like protein, ferredoxin
Hypothetical protein
Hypothetical protein
Hypothetical protein
Unknown protein
Hypothetical protein
21.40
21.34
21.30
21.13
21.12
21.12
21.12
21.08
0.61
0.44
0.41
0.25
1.63
0.61
1.20
0.29
20.55
20.83
21.03
21.33
0.15
0.06
0.38
21.31
1.11610–4
5.11610–4
6.45610–5
4.13610–4
3.28610–5
2.78610–3
3.03610–3
3.94610–4
norB
lrtA
amt1
hypA1
slr1641
clpB2
ssl3364
cp12
sll0170
dnaK2
sll1514
hspA
slr0093
dnaJ
Hypothetical or unknown
sll1621
sll1621
slr0148
slr0148
slr0150
slr0150
sll1201
sll1201
slr1583
slr1583
slr0146
slr0146
slr0145
slr0145
sll1774
sll1774
2544
Annotation
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
Microbiology 160
Synechocystis 6803 DkaiABC mutant
Table 1. cont.
Locus tag
sll0761
slr0151
sml0011
slr0144
ssr2062
slr1152
sll1898
slr0333
slr0226
slr0271
sll1022
slr1674
sll0846
Gene name
sll0761
slr0151
sml0011
slr0144
ssr2062
slr1152
sll1898
slr0333
slr0226
slr0271
sll1022
slr1674
sll0846
Annotation
Unknown protein
Unknown protein
Hypothetical protein
Hypothetical protein
Hypothetical protein
Hypothetical protein
Hypothetical protein
Unknown protein
Unknown protein
Unknown protein
Hypothetical protein
Hypothetical protein
Hypothetical protein
was not evident from the microarray data. However, the
qRT-PCR analysis could detect even very small changes
in transcript abundance and was far more sensitive than
a microarray experiment. Thus, good agreement was
obtained between the microarray data (Fig. 5) and qRTPCR (Fig. 6).
DISCUSSION
Deletion of the DkaiABC operon leads to a
stronger phenotype in Synechocystis 6803 in
comparison with Synechococcus 7942
Synechocystis 6803 is one of the most widely used
cyanobacteria for the study of photosynthetic functions,
gene regulation, stress responses, biotechnological applications and photoreceptor functions. The main reason for
the development of Synechocystis 6803 into a real model
organism was the release of its genome sequence (Kaneko
et al., 1996). However, in contrast to Synechococcus 7942, it
never became a model for circadian rhythms in cyanobacteria. Only very few data exist on a possible circadian
clock of Synechocystis 6803, demonstrating that at least
some genes show a circadian gene expression (Kucho et al.,
2005). Ito et al. (2009) reanalysed the data, and found even
fewer genes oscillating in a circadian manner and with
lower amplitude. In a very recent study by Beck et al.
(2014), a new detailed microarray analysis of Synechocystis
6803 grown under a light–dark cycle followed by constant
light or dark conditions was performed. The authors
showed that ~27 % of Synechocystis 6803 genes exhibited a
rhythmic behaviour under the light–dark cycle, but these
oscillations were rapidly dampened under constant conditions. Furthermore, measurements of circadian gene
expression using other methods, especially luciferase gene
reporter assays, were not as clear and reproducible as with
http://mic.sgmjournals.org
DkaiABC/WT [fold changes (log2)]
Dark
Light
Light (late)
21.08
21.06
21.03
20.98
20.59
20.53
0.05
0.06
0.08
1.14
1.21
2.22
2.75
0.27
0.20
20.05
1.54
23.38
2.54
21.35
21.24
1.10
0.02
0.12
0.34
0.24
20.87
21.12
21.58
0.09
23.01
1.36
20.59
20.90
20.51
0.23
20.08
21.22
20.69
Adjusted P value
1.07610–4
5.09610–5
1.15610–4
4.81610–2
1.77610–4
2.19610–4
2.34610–5
2.05610–3
2.58610–2
2.31610–5
7.56610–5
2.58610–4
1.52610–5
Synechococcus 7942 (Aoki & Onai, 2009). It was speculated
that the presence of more than one kaiC gene copy could
lead to a more complicated timing system, in which
different kaiC gene products exhibit complementary
functions (Aoki & Onai, 2009). In addition, Synechocystis
6803 is able to grow in the dark with glucose as an energy
source – a characteristic that might influence behaviour
under light–dark cycles. Nevertheless, the cells seem to
encode a KaiAB1C1-based oscillator, with similar biochemical characteristics to the model clock system (Wiegard et al.,
2013). In order to prove the physiological relevance of the
putative timing mechanism, we created deletion mutants of
all three kaiC genes. We were not able to generate a fully
segregated DkaiC2B2 mutant strain (data not shown). The
DkaiC3 mutant strain did not show any phenotypic effect
under the conditions tested. Only deletion of the kaiAB1C1
gene cluster led to cells that showed a growth defect under
light–dark cycles.
In contrast to the well-pronounced phenotype of the
Synechocystis 6803 DkaiABC mutant under diurnal
conditions, Synechococcus 7942 mutants lacking the
kaiABC cluster grew as well as the WT (Ishiura et al.,
1998). However, when WT cells were grown in co-culture
with an arrhythmic strain under light–dark cycles, the
mutant strain was rapidly outcompeted by the WT strain.
This suggests that an intact clock system enhances the
fitness of Synechococcus 7942 in a rhythmic environment,
but not under constant light conditions (Johnson et al.,
2008). This effect seems to be more pronounced in
Synechocystis 6803, as we detected strong growth differences in separate cultures. Another striking effect is the
reduced ability of the Synechocystis 6803 DkaiABC mutant
to grow with glucose in diurnal cycles. As Synechococcus
7942 does not grow chemoheterotrophically, it is difficult
to compare this phenotype between the two cyanobacterial strains.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
2545
A. K. Dörrich and others
6
(+) With primary 5′ ends
cph1, hik35
slr0473
2606250
2608750
2605000
sII0462
2607500
ssI0900
2610000
slI0461
proA
6
4
–13.3
2
0
ncl1300
log2(Read number) (454 sequencing)
0
2
9.3
ncr1340
ncr1350
–9.3
log2(Expression) (microarray)
12.3
4
15.3
(a)
(–) Total RNA
WT_ZT_5.5
MU_ZT_5.5
WT_ZT_17.5
MU_ZT_17.5
2313750
sII0171
gcvT
2316250
2315000
2312500
slI0170
dnaK2
2317500
slI0169
sll0168
sll0167
8
–17.3
6
4
–13.3
2
0
ncl1150
log2(Read number) (454 sequencing)
9.3
0
2
4
(+) With primary 5′ ends
–9.3
log2(Expression) (microarray)
12.3
(b)
(–) Total RNA
WT_ZT_5.5
MU_ZT_5.5
WT_ZT_17.5
MU_ZT_17.5
Fig. 5. Details of gene expression analysis in WT and DkaiABC (MU) at ZT5.5 and ZT17.5 (early time points). Genome sections
harbouring (a) cph1 and (b) dnak2 are shown. The whole-genome plot is shown in Supplementary Dataset 1. Locations of the
annotated genes are indicated by blue boxes, and non-coding and antisense RNAs by yellow and red boxes, respectively. The
scale for the microarray data is given on the left y-axis. RNA sequencing readings from previous transcriptome analyses under
standard conditions (Mitschke et al., 2011) are plotted on the right y-axis. Black arrows highlight the change in abundance of
cph1- and dnaK2-related transcripts.
2546
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
Microbiology 160
Synechocystis 6803 DkaiABC mutant
Differential expression
[log2(fold change)]
2
cph1
1
0
–1
Differential expression
[log2(fold change)]
2
ZT
5.
5
ZT
0.
5
5
ZT
23
.
5
ZT
17
.
11
.5
h
–2
dnaK2
1
0
5
5.
ZT
5
0.
ZT
.5
23
ZT
.5
17
ZT
11
.5
h
–1
Fig. 6. Validation of microarray data by qRT-PCR. Example mRNA
levels of cph1 and dnaK2 were measured and normalized to rnpB
expression as internal standard. Results are shown as the relative
expression of mRNA in the DkaiABC mutant compared with the
WT. Values were calculated as the mean of two biological
replicates present in technical duplicates. RNA samples were
derived from the early time points as indicated in Fig. 4. Both
methods (microarray and qRT-PCR) showed good agreement.
DkaiABC shares important phenotype similarities
with other mutants in Synechocystis 6803
Interestingly, the phenotype of the DkaiABC mutant
shows important similarities to growth defects of a
previously described plasmid hik31 operon deletion mutant
(slr6039–slr6041, including two-component system genes)
(Nagarajan et al., 2014) and the group 2 sigma factor sigE
mutant (Osanai et al., 2005). Both mutants have been
shown to have defects when grown mixotrophically under
alternating light–dark cycles or during light-activated
heterotrophic growth. Strikingly, in addition to the growth
defect, the hik31 operon mutant also showed a pigmentation defect under light–dark cycles that deteriorated under
mixotrophic conditions (Nagarajan et al., 2014), as was
found here for the DkaiABC mutant. The authors suggested
that the phenotypic effect caused by glucose resulted from
insufficient metabolism after sugar entry. A microarray
experiment showed that deletion of the hik31 operon
caused reduced translation and a general deceleration of
key metabolic processes (Nagarajan et al., 2014). At least
http://mic.sgmjournals.org
two ribosomal operons are also downregulated in the dark
in DkaiABC mutant cells. Despite this notable phenotypic
overlap, hik31-related transcript levels were not changed in
DkaiABC under the conditions tested. However, as also
proposed for the hik31 operon mutant, downregulation of
sigE might at least partly cause the growth defect observed
here for the DkaiABC mutant. As in a sigE disruption
mutant (Osanai et al., 2005), two major genes of the
oxidative pentose phosphate pathway (gnd and tal) are
downregulated in DkaiABC at all three investigated time
points during the light–dark cycle. SigE, as a transcriptional
activator of sugar catabolic genes, could therefore contribute to the growth defect observed here for the DkaiABC
mutant by downregulation of its mRNA. Corroboratively,
the sigE disruption mutant was found to be unable to
proliferate under light-activated heterotrophic growth
conditions. This also holds true for the DkaiABC mutant
(Fig. S2). Interestingly, cells of a mutant where only the
kaiA gene was disrupted were also no longer able to grow
under these conditions (Aoki & Onai, 2009), indicating a
fundamentally important role of KaiA. The exhibition of
pleiotropic phenotypes caused by SigE overexpression was
investigated recently by Osanai et al. (2013). The authors
report effects on metabolism, photosynthesis, hydrogen
production and cell morphology. Most outstanding were
aberrant cell division and increased cell size in the sigE
overexpression strain. The impact of SigE on this multitude of cellular processes supports its significance for
Synechocystis 6803 cells.
Another strain showing a similar phenotype to DkaiABC is
a mutant of hik8, the sasA homologue of Synechocystis 6803
(Singh & Sherman, 2005). SasA is a component of the
output pathway of the circadian clock in Synechococcus
7942. It interacts with KaiC and phosphorylates the
response regulator RpaA (Iwasaki et al., 2000; Takai et al.,
2006). Both RpaA and the similar response regulator, RpaB
regulate expression of the kaiBC operon in Synechococcus
7942. Moreover, RpaA was recently shown to be a master
regulator of global circadian gene expression that directly
targets circadian clock components as well as circadian
effectors and sigma factors (Markson et al., 2013). Interestingly, sasA- and rpaA-deficient mutants of Synechococcus
7942 both grew more slowly than the WT and a DkaiABC
mutant strain under light–dark cycles, whereas no differences were detected under constant light (Takai et al.,
2006). This suggests that inactivation of the output
signalling pathway has a stronger effect than deletion of
the central oscillator in Synechococcus 7942. In contrast, both
DkaiABC and sasA mutant strains have similar phenotypes
in Synechocystis 6803, i.e. slower growth under light–dark
cycles under mixotrophic conditions and no growth under
heterotrophic conditions (Singh & Sherman, 2005; and this
work). The cognate response regulators, RpaA and RpaB
(regulator of phycobilisome-associated), also exist in
Synechocystis 6803 where they are suggested to regulate the
coupling of phycobilisomes to photosystems (Emlyn-Jones
et al., 1999). Majeed et al. (2012) demonstrated a role of
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
2547
A. K. Dörrich and others
RpaA in high-light acclimation in Synechocystis 6803,
whereas RpaB was shown to bind to high-light regulatory
sequences (HLR1), an upstream element of several photosynthesis-related genes, especially of photosystem I genes
(Seino et al., 2009; Takahashi et al., 2010). To date, it has not
been demonstrated that these two response regulators and
SasA are involved in timing mechanisms in Synechocystis
6803. Even though rpaB was not amongst the top 53
differentially regulated genes in DkaiABC (Table 1), it was
slightly downregulated at all time points in the mutant
[fold changes (log2) of 20.58 (dark) and 20.88 (light);
Supplementary Datasets 1 and 2]. However, it is not clear to
what extent this lower accumulation of rpaB contributes to
the phenotype of the DkaiABC mutant. The photosystem I
psaAB operon, a verified target of RpaB (Takahashi et al.,
2010), as well as psaM were upregulated significantly in the
dark in the mutant. In addition, several other genes
encoding different regulators with known and unknown
functions, such as cph1, cyabrB1, hik34 and the patA
response regulator slr1594, were transcribed differentially
in the DkaiABC mutant. It is known that cph1 transcription
is affected in sigE and hik8 mutants (Osanai et al., 2011;
Singh & Sherman, 2005). Although Garcı́a-Domı́nguez et al.
(2000) discussed a role of Cph1 in the control of processes
required for adaptation in light–dark transitions, no
phenotype of a Dcph1 strain related to diurnal rhythms
has been described. Nevertheless, secondary effects on
transcription due to altered accumulation of transcriptional
regulators are very likely.
secondary effect. However, Aoki et al. (1995) studied dnaK2
promoter activity using bioluminescence in Synechocystis
6803 and demonstrated true circadian oscillations for this
gene. Thus, dnaK2 might also be a primary target of a
putative timing mechanism in this organism.
Overall, the transcription of genes involved in a multitude
of different metabolic processes is affected in DkaiABC
under 12 h–12 h light–dark conditions. However, as we
detected a downregulation of RNase E and two non-coding
transcripts in the dark, regulation of the post-transcriptional level may also be affected in the mutant. Proteomic
approaches would be needed to reveal the reasons for the
growth defect of the DkaiABC mutant under a light–dark
cycle. For Synechococcus 7942, Guerreiro et al. (2014)
showed recently that in comparison with the high level of
cycling transcripts in this model strain, only a few proteins
cycle in a circadian manner, with the majority accumulating in the dark period. The authors speculate that posttranscriptional regulation might contribute to discrepancies between transcript and protein accumulation. For
Synechocystis 6803, many RNA regulators have been
predicted and there are several hints that these non-coding
transcripts fulfil important functions in light–dark acclimation (Georg et al., 2009; Mitschke et al., 2011).
However, the targeting of mRNAs by regulatory small
RNAs does not necessarily include a change in transcript
abundance. This could explain the relatively low number of
differentially regulated genes in the DkaiABC mutant.
Transcriptional changes in the DkaiABC mutant
CONCLUSION
Beck et al. (2014) showed that at least 27 % of Synechocystis
6803 genes oscillate under light–dark conditions. These
authors sorted all genes into 10 clusters by their maximal
accumulation during different phases. Transcripts upregulated in the early light phase are categorized into cluster 1,
whereas cluster 6 harbours genes accumulating in the late
phase. Clusters 7 and 8 contain night active genes, and
clusters 9 and 10 contain non-oscillating genes (Beck et al.,
2014). Interestingly, 83 % of genes are differentially
regulated in the DkaiABC mutant in our analysis group
into clusters 1–6, representing genes peaking at daytime
with a preference for genes whose accumulation peaks at
the end of the day. These data imply that although we used
only one time point for the light period, we also detected
by our microarray genes that peak at other time points
during the day (among them genes involved in metabolism
and regulation but also many hypothetical genes). None of
the genes overlap with clusters 7 and 8. In addition, 17 % of
the genes showing an altered expression pattern in the
DkaiABC mutant grouped into non-oscillating genes
belonging to cluster 9, according to Beck et al. (2014).
These mainly represent genes encoding chaperones, such as
DnaK2, DnaJ and HspA, as well as the ClpB2 protease.
Transcripts for these genes accumulated more strongly
during the first dark period in the DkaiABC mutant,
suggesting a stress response of the cells, most probably as a
Analysing the transcriptome of a Synechocystis 6803
DkaiAB1C1 mutant grown under a light–dark cycle
revealed only a small overlap with the microarray analyses
performed by Kucho et al. (2005), which aimed to identify
genes showing a circadian behaviour. When we compare our data to the new transcriptomic analysis of
Synechocystis 6803 grown under light–dark cycles (Beck
et al., 2014), we find a large overlap with transcripts
showing a rhythmic accumulation during the day.
Nevertheless, many of these transcripts are also downregulated in the dark in our analysis. Still, our work clearly
demonstrates a physiological role of the kaiAB1C1 gene
cluster next to the recently investigated biochemical
functionality of its gene products (Wiegard et al., 2013).
According to our transcript data, many processes, such as
translation, photosynthesis, respiration and metabolism,
seem to be affected when Synechocystis 6803 lacks the
kaiAB1C1 gene cluster. However, several regulators are also
differentially expressed in the mutant and it would
therefore be premature to speculate about primary targets
of the clock proteins on transcription in Synechocystis 6803.
More analyses on the accumulation of the respective
proteins are needed, as well as investigations of the role of
RpaA and RpaB homologues in a putative timing process
in Synechocystis 6803. Moreover, from the work done on
Synechococcus 7942 we know that circadian gene expression
2548
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
Microbiology 160
Synechocystis 6803 DkaiABC mutant
does not necessarily lead to circadian metabolic patterns
(Guerreiro et al., 2014). As an example, photosynthetic
oxygen evolution was shown to be rhythmic under
constant light after light–dark cycles in Synechocystis
6803, but not in Synechococcus 7942 (Yen et al., 2004).
Therefore, we also have to learn more about the timing of
metabolic processes in different cyanobacteria.
Synechocystis sp. PCC 6803. Additional homologues of hypA and hypB
are not active in hydrogenase maturation. FEBS J 273, 4516–4527.
Ishiura, M., Kutsuna, S., Aoki, S., Iwasaki, H., Andersson, C. R.,
Tanabe, A., Golden, S. S., Johnson, C. H. & Kondo, T. (1998).
Expression of a gene cluster kaiABC as a circadian feedback process in
cyanobacteria. Science 281, 1519–1523.
Ito, H., Mutsuda, M., Murayama, Y., Tomita, J., Hosokawa, N.,
Terauchi, K., Sugita, C., Sugita, M., Kondo, T. & Iwasaki, H. (2009).
Cyanobacterial daily life with Kai-based circadian and diurnal
genome-wide transcriptional control in Synechococcus elongatus.
Proc Natl Acad Sci U S A 106, 14168–14173.
ACKNOWLEDGEMENTS
We thank Ilka Axmann, Anika Wiegard and Christian Beck for very
helpful discussions, and Ulrike Ruppert, Anne Stinn and Gudrun
Krüger for excellent technical assistance. This work was supported by a
grant from the German Research Fund (DFG) (WI 2014/5-1) to A. W.
Iwasaki, H., Williams, S. B., Kitayama, Y., Ishiura, M., Golden, S. S. &
Kondo, T. (2000). A KaiC-interacting sensory histidine kinase, SasA,
necessary to sustain robust circadian oscillation in cyanobacteria. Cell
101, 223–233.
Johnson, C. H., Mori, T. & Xu, Y. (2008). A cyanobacterial circadian
clockwork. Curr Biol 18, R816–R825.
REFERENCES
Kaneko, T., Sato, S., Kotani, H., Tanaka, A., Asamizu, E., Nakamura,
Y., Miyajima, N., Hirosawa, M., Sugiura, M. & other authors (1996).
Anderson, S. L. & McIntosh, L. (1991). Light-activated heterotrophic
growth of the cyanobacterium Synechocystis sp. strain PCC 6803: a
blue-light-requiring process. J Bacteriol 173, 2761–2767.
Sequence analysis of the genome of the unicellular cyanobacterium
Synechocystis sp. strain PCC6803. II. Sequence determination of the
entire genome and assignment of potential protein-coding regions.
DNA Res 3, 109–136.
Aoki, S. & Onai, K. (2009). Circadian clocks of Synechocystis sp. strain
PCC 6803, Thermosynechococcus elongatus, Prochlorococcus spp.,
Trichodesmium spp. and other species. In Bacterial Circadian Programs, pp. 259–282. Edited by J. L. Ditty, S. R. Mackey &
C. H. Johnson. Berlin: Springer.
Kucho, K.-i., Okamoto, K., Tsuchiya, Y., Nomura, S., Nango, M.,
Kanehisa, M. & Ishiura, M. (2005). Global analysis of circadian
Aoki, S., Kondo, T. & Ishiura, M. (1995). Circadian expression of the
Majeed, W., Zhang, Y., Xue, Y., Ranade, S., Blue, R. N., Wang, Q. &
He, Q. (2012). RpaA regulates the accumulation of monomeric
dnaK gene in the cyanobacterium Synechocystis sp. strain PCC 6803.
J Bacteriol 177, 5606–5611.
Axmann, I. M., Hertel, S., Wiegard, A., Dörrich, A. K. & Wilde, A.
(2014). Diversity of KaiC-based timing systems in marine cyano-
bacteria. Mar Genomics 14, 3–16.
Beck, C., Hertel, S., Rediger, A., Lehmann, R., Wiegard, A., Kölsch, A.,
Heilmann, B., Georg, J., Hess, W. R. & Axmann, I. M. (2014). Daily
expression pattern of protein-coding genes and small noncoding
RNAs in Synechocystis sp. PCC 6803. Appl Environ Microbiol 80,
5195–5206.
Červený, J. & Nedbal, L. (2009). Metabolic rhythms of the
cyanobacterium Cyanothece sp. ATCC 51142 correlate with modeled
dynamics of circadian clock. J Biol Rhythms 24, 295–303.
Emlyn-Jones, D., Ashby, M. K. & Mullineaux, C. W. (1999). A gene
required for the regulation of photosynthetic light harvesting in
the cyanobacterium Synechocystis 6803. Mol Microbiol 33, 1050–1058.
Garcı́a-Domı́nguez, M., Muro-Pastor, M. I., Reyes, J. C. & Florencio,
F. J. (2000). Light-dependent regulation of cyanobacterial phyto-
chrome expression. J Bacteriol 182, 38–44.
Georg, J., Voss, B., Scholz, I., Mitschke, J., Wilde, A. & Hess, W. R.
(2009). Evidence for a major role of antisense RNAs in cyanobacterial
gene regulation. Mol Syst Biol 5, 305.
Grigorieva, G. & Shestakov, S. (1982). Transformation in the
cyanobacterium Synechocystis sp. 6803. FEMS Microbiol Lett 13, 367–
370.
Guerreiro, A. C., Benevento, M., Lehmann, R., van Breukelen, B.,
Post, H., Giansanti, P., Maarten Altelaar, A. F., Axmann, I. M. & Heck,
A. J. (2014). Daily rhythms in the cyanobacterium Synechococcus
elongatus probed by high-resolution mass spectrometry based
proteomics reveals a small-defined set of cyclic proteins. Mol Cell
Proteomics 13, 2042–2055.
Hoffmann, D., Gutekunst, K., Klissenbauer, M., Schulz-Friedrich, R.
& Appel, J. (2006). Mutagenesis of hydrogenase accessory genes of
http://mic.sgmjournals.org
expression in the cyanobacterium Synechocystis sp. strain PCC 6803.
J Bacteriol 187, 2190–2199.
photosystem I and PsbA under high light conditions in Synechocystis
sp. PCC 6803. PLoS ONE 7, e45139.
Markson, J. S., Piechura, J. R., Puszynska, A. M. & O’Shea, E. K.
(2013). Circadian control of global gene expression by the
cyanobacterial master regulator RpaA. Cell 155, 1396–1408.
Mitschke, J., Georg, J., Scholz, I., Sharma, C. M., Dienst, D.,
Bantscheff, J., Voss, B., Steglich, C., Wilde, A. & other authors
(2011). An experimentally anchored map of transcriptional start sites
in the model cyanobacterium Synechocystis sp. PCC6803. Proc Natl
Acad Sci U S A 108, 2124–2129.
Nagarajan, S., Srivastava, S. & Sherman, L. A. (2014). Essential role
of the plasmid hik31 operon in regulating central metabolism in the
dark in Synechocystis sp. PCC 6803. Mol Microbiol 91, 79–97.
Onai, K., Morishita, M., Itoh, S., Okamoto, K. & Ishiura, M. (2004).
Circadian rhythms in the thermophilic cyanobacterium
Thermosynechococcus elongatus: compensation of period length over
a wide temperature range. J Bacteriol 186, 4972–4977.
Osanai, T., Kanesaki, Y., Nakano, T., Takahashi, H., Asayama, M.,
Shirai, M., Kanehisa, M., Suzuki, I., Murata, N. & Tanaka, K. (2005).
Positive regulation of sugar catabolic pathways in the cyanobacterium
Synechocystis sp. PCC 6803 by the group 2 s factor sigE. J Biol Chem
280, 30653–30659.
Osanai, T., Oikawa, A., Azuma, M., Tanaka, K., Saito, K., Hirai, M. Y. &
Ikeuchi, M. (2011). Genetic engineering of group 2 s factor SigE
widely activates expressions of sugar catabolic genes in Synechocystis
species PCC 6803. J Biol Chem 286, 30962–30971.
Osanai, T., Kuwahara, A., Iijima, H., Toyooka, K., Sato, M., Tanaka, K.,
Ikeuchi, M., Saito, K. & Hirai, M. Y. (2013). Pleiotropic effect of sigE
over-expression on cell morphology, photosynthesis and hydrogen
production in Synechocystis sp. PCC 6803. Plant J 76, 456–465.
Paithoonrangsarid, K., Shoumskaya, M. A., Kanesaki, Y., Satoh, S.,
Tabata, S., Los, D. A., Zinchenko, V. V., Hayashi, H., Tanticharoen, M.
& other authors (2004). Five histidine kinases perceive osmotic stress
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
2549
A. K. Dörrich and others
and regulate distinct sets of genes in Synechocystis. J Biol Chem 279,
53078–53086.
Pattanayak, G. & Rust, M. J. (2014). The cyanobacterial clock and
metabolism. Curr Opin Microbiol 18, 90–95.
Pfaffl, M. W. (2001). A new mathematical model for relative
quantification in real-time RT-PCR. Nucleic Acids Res 29, e45.
Pinto, F. L., Thapper, A., Sontheim, W. & Lindblad, P. (2009). Analysis
of current and alternative phenol based RNA extraction methodologies for cyanobacteria. BMC Mol Biol 10, 79.
Rippka, R., Deruelles, J., Waterbury, J. B., Herdman, M. & Stanier,
R. Y. (1979). Generic assignments, strain histories and properties of
pure cultures of cyanobacteria. J Gen Microbiol 111, 1–61.
Rujan, T. & Martin, W. (2001). How many genes in Arabidopsis come
from cyanobacteria? An estimate from 386 protein phylogenies.
Trends Genet 17, 113–120.
Schwarz, D., Schubert, H., Georg, J., Hess, W. R. & Hagemann, M.
(2013). The gene sml0013 of Synechocystis species strain PCC 6803
encodes for a novel subunit of the NAD(P)H oxidoreductase or
complex I that is ubiquitously distributed among cyanobacteria. Plant
Physiol 163, 1191–1202.
Seino, Y., Takahashi, T. & Hihara, Y. (2009). The response regulator
RpaB binds to the upstream element of photosystem I genes to work
for positive regulation under low-light conditions in Synechocystis sp.
Strain PCC 6803. J Bacteriol 191, 1581–1586.
Shih, P. M., Wu, D., Latifi, A., Axen, S. D., Fewer, D. P., Talla, E.,
Calteau, A., Cai, F., Tandeau de Marsac, N. & other authors (2013).
Improving the coverage of the cyanobacterial phylum using diversitydriven genome sequencing. Proc Natl Acad Sci U S A 110, 1053–1058.
Singh, A. K. & Sherman, L. A. (2005). Pleiotropic effect of a histidine
kinase on carbohydrate metabolism in Synechocystis sp. strain PCC
6803 and its requirement for heterotrophic growth. J Bacteriol 187,
2368–2376.
Smyth, G. K., Michaud, J. & Scott, H. S. (2005). Use of within-array
replicate spots for assessing differential expression in microarray
experiments. Bioinformatics 21, 2067–2075.
Stöckel, J., Welsh, E. A., Liberton, M., Kunnvakkam, R., Aurora, R. &
Pakrasi, H. B. (2008). Global transcriptomic analysis of Cyanothece
51142 reveals robust diurnal oscillation of central metabolic
processes. Proc Natl Acad Sci U S A 105, 6156–6161.
Takahashi, T., Nakai, N., Muramatsu, M. & Hihara, Y. (2010). Role of
multiple HLR1 sequences in the regulation of the dual promoters of
the psaAB genes in Synechocystis sp. PCC 6803. J Bacteriol 192, 4031–
4036.
Takai, N., Nakajima, M., Oyama, T., Kito, R., Sugita, C., Sugita, M.,
Kondo, T. & Iwasaki, H. (2006). A KaiC-associating SasA–RpaA two-
component regulatory system as a major circadian timing mediator in
cyanobacteria. Proc Natl Acad Sci U S A 103, 12109–12114.
Toepel, J., Welsh, E., Summerfield, T. C., Pakrasi, H. B. & Sherman,
L. A. (2008). Differential transcriptional analysis of the cyanobacter-
ium Cyanothece sp. strain ATCC 51142 during light–dark and
continuous-light growth. J Bacteriol 190, 3904–3913.
Trautmann, D., Voss, B., Wilde, A., Al-Babili, S. & Hess, W. R. (2012).
Microevolution in cyanobacteria: re-sequencing a motile substrain of
Synechocystis sp. PCC 6803. DNA Res 19, 435–448.
Wegener, K. M., Welsh, E. A., Thornton, L. E., Keren, N., Jacobs, J. M.,
Hixson, K. K., Monroe, M. E., Camp, D. G., II, Smith, R. D. & Pakrasi,
H. B. (2008). High sensitivity proteomics assisted discovery of a novel
operon involved in the assembly of photosystem II, a membrane
protein complex. J Biol Chem 283, 27829–27837.
Wiegard, A., Dörrich, A. K., Deinzer, H. T., Beck, C., Wilde, A.,
Holtzendorff, J. & Axmann, I. M. (2013). Biochemical analysis of three
putative KaiC clock proteins from Synechocystis sp. PCC 6803 suggests
their functional divergence. Microbiology 159, 948–958.
Yen, U.-C., Huang, T.-C. & Yen, T.-C. (2004). Observation of the
Singh, A. K., Li, H. & Sherman, L. A. (2004). Microarray analysis and
circadian photosynthetic rhythm in cyanobacteria with a dissolvedoxygen meter. Plant Sci 166, 949–952.
redox control of gene expression in the cyanobacterium Synechocystis
sp. PCC 6803. Physiol Plant 120, 27–35.
Edited by: C.-C. Zhang
2550
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 31 Jul 2017 16:21:39
Microbiology 160