Transcription Profiling of the Model Cyanobacterium Synechococcus sp. Strain PCC 7002 by Next-Gen (SOLiD™) Sequencing of cDNA.

The genome of the unicellular, euryhaline cyanobacterium Synechococcus sp. PCC 7002 encodes about 3200 proteins. Transcripts were detected for nearly all annotated open reading frames by a global transcriptomic analysis by Next-Generation (SOLiD™) sequencing of cDNA. In the cDNA samples sequenced, ∼90% of the mapped sequences were derived from the 16S and 23S ribosomal RNAs and ∼10% of the sequences were derived from mRNAs. In cells grown photoautotrophically under standard conditions [38°C, 1% (v/v) CO(2) in air, 250 μmol photons m(-2) s(-1)], the highest transcript levels (up to 2% of the total mRNA for the most abundantly transcribed genes; e.g., cpcAB, psbA, psaA) were generally derived from genes encoding structural components of the photosynthetic apparatus. High-light exposure for 1 h caused changes in transcript levels for genes encoding proteins of the photosynthetic apparatus, Type-1 NADH dehydrogenase complex and ATP synthase, whereas dark incubation for 1 h resulted in a global decrease in transcript levels for photosynthesis-related genes and an increase in transcript levels for genes involved in carbohydrate degradation. Transcript levels for pyruvate kinase and the pyruvate dehydrogenase complex decreased sharply in cells incubated in the dark. Under dark anoxic (fermentative) conditions, transcript changes indicated a global decrease in transcripts for respiratory proteins and suggested that cells employ an alternative phosphoenolpyruvate degradation pathway via phosphoenolpyruvate synthase (ppsA) and the pyruvate:ferredoxin oxidoreductase (nifJ). Finally, the data suggested that an apparent operon involved in tetrapyrrole biosynthesis and fatty acid desaturation, acsF2-ho2-hemN2-desF, may be regulated by oxygen concentration.

Introduction

n class="Species">Synechococcus sp. strain PCC 7002 (hereafter pan> class="Species">Synechococcus 7002) is a euryhaline, unicellular cyanobacterium, which tolerates extremely high-light intensities and grows over a wide range of NaCl concentrations (Batterton Jr. and van Baalen, 1971; Sakamoto and Bryant, 2002; Nomura et al., 2006b). It has the fastest reported doubling time for any cyanobacterium: ∼2.6 h when grown on urea or ammonia and ∼4.0 h when grown on nitrate under optimal conditions [38°C, 1% (v/v) CO2 in air at a saturating light intensity of ∼250 μmol photons m−2 s−1]. The organism is easily transformable genetically (Stevens and Porter, 1980), the complete genome sequence is available (see http://www.ncbi.nlm.nih.gov/), and a versatile system for genetic complementation and gene overexpression is available (Xu et al., 2011). Collectively, these traits make Synechococcus 7002 a robust and promising platform for biotechnological applications, including the production of biofuels.

Using energy provided by sunlight, cyanobacteria produce the reducing equivalents required for n class="Chemical">CO2 reductionpan> anpan>d cellular metabolism from the oxidationpan> of pan> class="Chemical">water molecules via two photosystems, denoted photosystem II (PS II) and photosystem I (PS I; Bryant, 1994). The resulting reducing equivalents are mostly used for CO2 fixation and subsequent generation of carbohydrates and other metabolite building blocks, from which other cellular constituents are made. In addition to reducing equivalents, the photosynthetic apparatus generates a proton gradient across the thylakoid membrane, which is used for ATP synthesis. Because light is not continuously available, all phototrophic organisms must switch to another metabolic mode at night to produce the maintenance energy to support basic cell functions. Cyanobacteria have a complete respiratory electron transport chain that allows respiration with oxygen as terminal electron acceptor (Peschek et al., 2004; Bernroitner et al., 2008). However, in many natural environments, oxygen is rapidly consumed in the dark by cyanobacteria or other organisms (Stal, 1995; Steunou et al., 2008), and thus the local environmental conditions may quickly become anoxic. Under these conditions, most cyanobacteria can perform fermentation of stored carbohydrates to produce maintenance energy (Stal and Moezelaar, 1997; McNeely et al., 2010a,b). Thus, cyanobacteria must rapidly adjust to diurnal light availability, changing light intensities, and the availability of inorganic nutrients (N, P, S, Fe, etc.). A consequence of these fluctuating conditions is that cells experience rapid changes in cellular redox states due to changing photosynthetic and respiratory electron transfer processes, as well as oxygen levels, over the course of a day.

In order to maintain cellular levels of reducing equivalents generated by the photosystems within acceptable boundaries, cyanobacteria adjust their cellular contents of PS I, PS II, and light-harvesting phycobilisomes to maximize both n class="Chemical">ATP anpan>d pan> class="Chemical">NADPH production (Fujita et al., 1994). However, the availability of CO2, which is the major sink for the electrons produced, is an equally important factor. Thus, it is critical that cells balance reductant generation and CO2 fixation rates to avoid over-reduction of electron carriers, which can lead to the production of reactive oxygen and nitrogen species that can cause potentially lethal photooxidative damage to cells. Such adjustments can be performed at many different stages of gene expression, including mRNA synthesis (transcription); protein biosynthesis (translation); protein maturation, assembly, and stability; post-translational modification and allostery.

Because the complete genome of n class="Species">Synechococcus 7002 is available anpan>d the complete set of its predicted genpan>es is knpan>ownpan>, genpan>e expressionpan> inpan> this pan> class="Species">cyanobacterium can be studied at different levels. This study focused on the transcriptional level and was performed to characterize the expression of all predicted mRNA-encoding open reading frames (ORFs) of Synechococcus 7002 cells grown under selected physiological states. Several methods are available to obtain information about transcript levels. Microarrays have been extensively used to study global gene expression patterns in several cyanobacteria, including Synechocystis sp. PCC 6803, Nostoc sp. PCC 7120, Nostoc punctiforme, Synechococcus sp. PCC 7942, and Synechococcus sp. WH8102 (Hihara et al., 2001, 2003; Gill et al., 2002; Ehira et al., 2003; Postier et al., 2003; Singh et al., 2003; Sato et al., 2004; Kucho et al., 2005; Campbell et al., 2007, 2008; Foster et al., 2007; Nodop et al., 2008; Summerfield et al., 2008; Stuart et al., 2009; Tetu et al., 2009; Ostrowski et al., 2010; Rowland et al., 2010). However, microarrays suffer from a number of problems, including their relatively low sensitivity, accuracy, specificity, and reproducibility (see Draghici et al., 2006 for a review). Quantitative-RT-PCR allows a high degree of specificity and accuracy, but this method is not suitable for global analyses of transcription. Because Next-Generation (Next-Gen) sequencing can provide very large numbers of randomly distributed cDNA sequences over an entire cDNA sample (Cloonan et al., 2008), cDNA sequencing was employed to obtain global transcription information for Synechococcus 7002. In the studies described here, the global transcriptome of Synechococcus 7002 cells that had been grown under a well-defined, standard photolithoautotrophic conditions was determined. Additionally, the transcriptomes of cells that had been subjected to selected perturbations, including exposure to high-light intensity or darkness under oxic respiratory or anoxic fermentative conditions were compared to that of cells grown under standard conditions. Finally, the transcriptome of Synechococcus 7002 cells grown photolithoautotrophically under micro-oxic conditions was determined. The resulting datasets were compared to provide some initial insights into the patterns of gene expression that might be responsible for some of the physiological properties exhibited by Synechococcus 7002.

Materials and Methods

Sample preparation

n class="Species">Synechococcus 7002 was grownpan> inpan> 20-mm culture tubes conpan>tainpan>inpan>g medium A (25 mL) supplemenpan>ted with 1 mg of pan> class="Chemical">NaNO3 mL−1 as nitrogen source (designated as medium A+; Stevens and Porter, 1980). Medium A is a Tris-buffered (pH 8.2) medium containing 0.3 M NaCl and 20 mM magnesium-sulfate; the exact composition of medium A is 18 g L−1 NaCl, 0.6 g L−1 KCl, 1.0 g L−1 NaNO3, 5.0 g L−1 MgSO4·7H2O, 50 mg L−1 KH2PO4, 266 mg L−1 CaCl2, 30 mg L−1 Na2 EDTA·2H2O, 3.89 mg L−1 FeCl3·6H2O, 1 g L−1 Tris/HCl (pH 8.2), 1 mL L−1 P1 trace metal solution, 4 μg L−1 vitamin B12. P1 trace metal solution (1000×) contains the following substances: 34.26 g L−1 H3BO3, 4.32 g L−1 MnCl2·4H2O, 0.315 g L−1 ZnCl2, 0.03 g L−1 MoO3 (85%), 0.003 g L−1 CuSO4·5H2O, 0.01215 g L−1 CoCl2·6H2O. Unless otherwise specified, cultures were grown at 38°C with continuous illumination at 250 μmol photons m−2 s−1 and were sparged with 1% (v/v) CO2 in air (these optimal growth conditions are here defined as “standard conditions”). Pre-cultures were grown under these “standard conditions” under continuous illumination. Cultures for RNA analyses were inoculated at an OD730 nm between 0.05 and 0.1; and cells were subsequently grown under these conditions to OD730 nm = 0.7 (see Figure 1). For high-light-intensity treatment or for incubation in darkness, cells were grown to OD730 nm = 0.7 under the specified standard conditions and immediately prior to harvest were illuminated at ∼900 μmol photons m−2 s−1 or incubated in the dark for 1 h. For the latter treatment, cells were incubated in the dark while sparging with 1% (v/v) CO2 in N2 for 1 h. Photolithoautotrophic growth under micro-oxic conditions was performed by growing cells in the light while sparging with 1% (v/v) CO2 in N2 under otherwise standard conditions. To identify transcription changes during the growth of a batch culture, cells were harvested at OD730 nm = 0.4, 0.7, 1.0, 3.0, and 5.0 (see Figure 1). Cells were rapidly chilled and centrifuged (5 min, 5000 × g, 4°C), and the cell pellets were quickly frozen in liquid nitrogen and stored at −80°C until required for further processing.

Figure 1

Growth curve for . Growth curve under “standard” conditions in medium A+ indicating the OD730 nm values at which samples for transcription analyses were taken (arrows).

Growth curve for . Growth curve under “standard” conditions in medium A+ indicating the OD730 nm values at which sn class="Chemical">amples for transcription analyses were takenpan> (arrows).

RNA extraction

Frozen cells from 20 to 30 mL culture were resuspended in 10 mM Tris–HCl, pH 8.0 (400 μL) and lysozyme (40 μL of a 50-mg/mL stock) was added. The sn class="Chemical">ample (220 μL, each) was dispenpan>sed inpan>to two 1.5-mL reactionpan> tubes anpan>d inpan>cubated for 2 minpan> at room temperature. For proteinpan> denpan>aturationpan> anpan>d cell lysis, acidic pan> class="Chemical">phenol–chloroform solution [400 μL; 1:1 (v/v), pH 4.3] and BugBuster Protein Extraction Reagent (400 μL; Novagen) were added to each reaction tube, and the samples were vortexed two times for 1 min and stored on ice between treatments. The samples were centrifuged (2 min, 10,000 × g, 22°C), the aqueous phase was extracted once more with phenol–chloroform solution, and the combined extracts were finally extracted once with chloroform–isoamyl alcohol (24:1, v/v). The RNA was precipitated from the aqueous phase by adding one tenth volume of 3.0 M sodium acetate, pH 5.2 and 2.5 volumes of ethanol, and the precipitated RNA was washed twice with 70% (v/v) ethanol. After the resulting pellet was air-dried, the RNA was further processed using the High Pure RNA Isolation Kit (Roche) according to the recommendations of the manufacturer. Briefly, the RNA pellet was resuspended in DNase incubation buffer (460 μL), and RNasin RNase inhibitor (4 μL, Promega) and DNase I (60 μL) were added to the samples, which were incubated for 1 h at 22°C. The RNA was purified by size exclusion on the provided spin-columns, which also removed small RNAs (approximately <300 bp). When rRNA depletion was performed, the RNA was dissolved in TE buffer (10 mM Tris–HCl, pH 8.0, 1 mM EDTA) or twice-distilled H2O when it was used directly for cDNA library synthesis. Depletion of rRNA was performed using MICROBExpress Bacterial mRNA Purification Kit (Ambion) with cyanobacteria-specific capture oligonucleotides. RNA concentrations were determined using a NanoDrop ND-1000 Spectrophotometer (Thermo Scientific); a Qubit System (Invitrogen) was also used to determine RNA and DNA concentrations separately.

cDNA library construction and solid™ sequencing

Construction of cDn class="Chemical">NA libraries anpan>d SOLiD™ sequenpan>cinpan>g was performed inpan> the Genpan>omics Core Facility at The Penpan>nsylvanpan>ia State Unpan>iversity (Unpan>iversity Park, PA, USA). The cDNA libraries were constructed from 0.5 μg RNA sample using SOLiD™ Whole Transcriptome Analysis Kit (Applied Biosystems) and were barcoded by using the SOLiD™ Transcriptome Multiplexing Kit (Applied Biosystems). SOLiD™ ePCR Kit and SOLiD™ Bead Enrichment Kit (both Applied Biosystems) were used for processing the samples for sequencing, and either the SOLiD™ 3 or 3Plus protocol (Applied Biosystems) was used for sequencing.

The cDNA sequence data have been submitted to the NCBI Sequence Read Archive (SRA) under accession number SRP004049.

Data analyses

The sequencer datasets were first converted into fastq format using the PERL script supplied with the BWA software package (Li and Durbin, 2009). The resulting fastq sequences were mapped in colorspace against the n class="Species">Synechococcus 7002 genpan>ome usinpan>g the Burrows–Wheeler algorithm, allowinpan>g four mismatches (>90% sequenpan>ce idenpan>tity). Prior to further anpan>alyses, the sequenpan>ces mappinpan>g to rRNA-coding regions (the major portion of all datasets) and non-uniquely mapping reads were removed from the output files generated by BWA. All protein-coding regions were analyzed for cDNA sequences mapping entirely within or partially covering the respective ORF (by at least one nucleotide), and the resulting hits for each ORF were counted. The relative transcript abundances for all ORFs were calculated as the number of sequences mapping in a given ORF divided by the total number of sequences mapping within any protein-coding region. To monitor differences in transcript level between two conditions, the relative transcript abundances under the two conditions were compared for all ORFs. The ratio is given as the relative transcript abundance under condition 2 divided by the relative transcript abundance under condition 1. For the “standard” growth conditions defined above, cDNA sequence data for three independent biological replicates were obtained. Based on these three “standard” datasets the mean and standard deviation for the relative transcript abundance was calculated for each ORF. When “standard” conditions served as the basis for a comparison, the probability for equal transcription was calculated for each ORF using the z-test (Ott and Longnecker, 2000). When conditions were compared, for which only single datasets were available, the chi-square test was applied to determine the probability level for equal transcription for each ORF (Monaghan et al., 2009).

To compare the mRNA levels for difn class="Chemical">ferenpan>t genpan>es withinpan> onpan>e span> class="Chemical">ample, the number of aligned sequences for a given ORF was normalized by the length of the ORF, and the results are reported as aligned sequences (“hits”) per kilobase. For comparisons of the same gene but in different samples, the number of aligned sequences was normalized relative to the total number of mRNA counts, because the gene length is constant but the total number of mRNA counts was variable and was also dependent upon the sequencing depth.

Results and Discussion

Analyzing the transcriptome of Synechococcus 7002 by solid™ sequencing of cDNA

Under the experimental design and conditions defined as “standard” (see Materials and Methods), cultures of n class="Species">Synechococcus 7002 grow exponpan>enpan>tially up to anpan> optical denpan>sity at 730 nm (OD730 nm) of ∼0.7 (see Figure 1; anpan>d Sakamoto anpan>d Bryanpan>t, 1998). Growth slows at higher OD730 nm values because of light limitationpan> imposed by self-shadinpan>g. Thus, cultures were harvested at OD730 nm = 0.7 to produce the maximal yield of cells inpan> exponpan>enpan>tial phase. RNA was extracted as described and used for cDNA sequencing. Depending on how many samples were pooled and barcoded for one sequencing run, and the version of the sequencing chemistry employed (SOLiD™ 3 or 3Plus), between ∼7 and 30 million mapped sequence reads were obtained with SOLiD™ sequencing (Table 1). In the absence of any depletion of rRNA, 89.9–94% of the sequences mapped to the 16S and 23S rRNAs. Because of the presence of some small, multigene families (e.g., psbA, psbD) that have similar sequences, the percentage of uniquely mapped sequences ranged from 94.9 to 98.3%. However, cDNA sequencing allows unique mapping even if only one base is different, and transcription differences can therefore be detected even for very similar genes. This is not possible when using microarrays, at least not with such a high accuracy, because that method is based on hybridization (Hihara et al., 2001).

Table 1

Number of sequences obtained by SOLiD™ sequencing for the samples analyzed in this study.

Sample	Mapped reads	Mapped in rDNA regions	Percent rDNA	Remaining mapped reads	Uniquely mapped reads	Percent unique reads
Standard 1	18,238,746	13,193,499	72.31	5,045,247	4,886,185	96.8
Standard 2	29,450,401	27,109,346	92.1	2,341,055	2,270,856	97.0
Standard 3	25,082,458	22,558,727	89.9	2,523,731	2,455,018	97.3
Dark oxic	7,375,663	6,406,311	86.91	969,352	949,679	98.0
Dark anoxic	16,252,102	13,828,205	85.11	2,423,897	2,382,528	98.3
High light	8,112,975	7,249,329	89.41	863,646	819,321	94.9
Low O2	6,834,732	6,251,835	91.5	582,897	569,435	97.7
OD 0.4	11,832,808	11,055,228	93.4	777,580	752,042	96.7
OD 1.0	16,964,973	15,545,651	91.6	1,419,322	1,381,923	97.4
OD 3.0	11,375,065	10,530,750	92.6	844,315	821,743	97.3
OD 5.0	9,375,497	8,813,140	94.0	562,357	548,269	97.5

The number of reads obtained for the different samples, number of mapped reads, number of reads mapping within the rDNA regions and outside rDNA regions and the number of reads mapping uniquely (outside rDNA regions) are given for the individual samples.

1These samples were treated to deplete rRNAs.

Number of sequences obtained by SOLiD™ sequencing for the sn class="Chemical">amples analyzed inpan> this study.

The number of reads obtained for the difn class="Chemical">ferenpan>t span> class="Chemical">amples, number of mapped reads, number of reads mapping within the rDNA regions and outside rDNA regions and the number of reads mapping uniquely (outside rDNA regions) are given for the individual samples.

1These sn class="Chemical">amples were treated to deplete rRn class="Chemical">NAs.

In some early experiments, Rn class="Chemical">NA span> class="Chemical">amples were processed in attempts to remove some of the rRNA sequences (see Materials and Methods for details). In one sample from cells grown under standard conditions, rRNA was depleted to a level about 20% lower than the average value for untreated samples (Table 1), but on average, the depletion resulted in a much smaller reduction in rRNA, generally only ∼5%. Because the depletion protocol employed was relatively ineffective, rRNA depletion was not employed prior to cDNA synthesis for most samples.

In the “standard 1” sn class="Chemical">ample, which had the highest number of nonpan>-rRNA sequences, transcripts were detected for nearly all of the 3235 predicted ORFs in the Synechococcus 7002 genome. Table 2 presents the 15 genes that had the highest number of mapped cDNA sequences in cells grown under standard conditions. As expected, most of these genes encoded structural components of the phycobilisomes and the two photosystems. For example, cpcA (SYNPCC7002_A2210), encoding the alpha subunit of phycocyanin (de Lorimier et al., 1984), accounted for 2.2% of the mapped cDNA sequences and psbA (SYNPCC7002_A1418) accounted for 2.0% of the mapped cDNA sequences. Due to the very high number of total sequences provided by SOLiD™ sequencing, transcripts could be detected even for genes transcribed at very low levels; there were only a 12 ORFs (SYNPCC7002_A2187, A2712, A2746, SYNPCC7002_D0029, D0030, SYNPCC7002_F0015, F0016, SYNPCC7002_G0039, G0040, G0050, G0051, and G0052) with no mapped cDNA sequences under standard growth conditions; for two of those (SYNPCC7002_A2187 and A2746) transcripts were found under at least one of the other conditions tested here, but at very low level (one or two mapped sequences only). These ORFs are either annotated as hypothetical proteins or integrases and resolvases, which suggests that they might be wrongly annotated or do not represent functional genes. The transcript levels for all genes and all conditions described in this study are provided in the Table S1 in Supplementary Material.

Table 2

The 15 most abundant mRNA species in the “standard 1” sample.

Locus tag	Counts	Counts/total counts	Counts/kb	Gene name	Gene product
SYNPCC7002_A0957	19392	0.00610	208516.1	psbT	Photosystem II reaction center, PsbT protein
SYNPCC7002_A2210	70907	0.02231	145899.2	cpcA	Phycocyanin, alpha subunit
SYNPCC7002_A2209	52347	0.01647	101447.7	cpcB	Phycocyanin, beta subunit
SYNPCC7002_A1929	43223	0.01360	89488.6	apcB	Allophycocyanin, beta subunit
SYNPCC7002_A2579	28125	0.00885	78781.5	–	Hypothetical protein
SYNPCC7002_C0011	13311	0.00419	70428.6	–	Hypothetical protein
SYNPCC7002_A1930	32380	0.01019	67039.3	apcA	Allophycocyanin alpha subunit
SYNPCC7002_A1589	15162	0.00477	62395.1	psaC	Photosystem I iron-sulfur center subunit VII
SYNPCC7002_A1418	63774	0.02006	59050.0	psbA	Photosystem q(b) protein
SYNPCC7002_A2804	33226	0.01045	58291.2	–	General secretion pathway protein
SYNPCC7002_A1928	9781	0.00308	48661.7	apcC	allophycocyanin-associated phycobilisome 7.8-kDa core-linker polypeptide
SYNPCC7002_A0272	12032	0.00379	47746.0	–	RNA-binding protein
SYNPCC7002_A2326	12655	0.00398	43488.0	petF	Ferredoxin I (2Fe-2S)
SYNPCC7002_A1008	21979	0.00691	43351.1	psaF	Photosystem I reaction center subunit III, PsaF
SYNPCC7002_A0167	14787	0.00465	42128.2	petJ	Cytochrome c6 precursor (cytochrome c553)

The 15 most abundant mRn class="Chemical">NA species inpan> the “standard 1” sn class="Chemical">ample.

Some cDNA samples were also analyzed by pyrosequencing (data not shown). Although similar results were obtained for the most abundant transcripts, because of the much smaller number of sequences obtained and the much greater cost per sequence, pyrosequencing was not pursued further because it was not a cost-effective method for global transcription profiling.

To assess the reproducibility of the cDNA sequencing approach, RNA sn class="Chemical">amples isolated from three inpan>depenpan>denpan>t cultures grownpan> under “stanpan>dard conpan>ditionpan>s” were inpan>depenpan>denpan>tly conpan>verted to cDNA anpan>d subjected to SOLiD™ sequenpan>cinpan>g. Two of these span> class="Chemical">amples were processed without rRNA depletion, while as described above, one sample (standard 1) was treated to deplete the rRNA sequences in the sample. Figure 2A shows a scatter plot comparing the transcript abundances calculated for the two non-depleted samples (standard 2 and standard 3). The majority of the data points occurred on a line with a slope of ∼1.0, which indicated that the mapped sequence values for each gene obtained from the independent samples were very similar. Very few data points fell outside the lines indicating a two-fold difference (Figure 2A), and most of those points are associated with genes that are transcribed at low levels, for which stochastic processes would produce the largest variation. These data suggested that, much like microarrays, differences reflecting greater than two-fold changes can be considered to be significant. Moreover, much smaller differences (∼25%) might be statistically significant for highly transcribed genes (note that p-values are reported for all comparisons in Table S2 in Supplementary Material).

Figure 2

Variation in the relative transcript abundance between standard samples. These scatter plots show the variation in the relative transcript abundance between (A) standard conditions sample 3 compared to standard conditions sample 2; (B) standard conditions sample 2 compared to standard conditions sample 1; and (C) standard conditions sample 3 compared to standard conditions sample 1. The gray lines indicate a two-fold change in either direction.

Variation in the relative transcript abundance between standard sn class="Chemical">amples. These scatter plots show the variationpan> inpan> the relative tranpan>script abundanpan>ce betweenpan> (A) stanpan>dard conpan>ditionpan>s span> class="Chemical">ample 3 compared to standard conditions sample 2; (B) standard conditions sample 2 compared to standard conditions sample 1; and (C) standard conditions sample 3 compared to standard conditions sample 1. The gray lines indicate a two-fold change in either direction.

Figures 2B,C show scatter plot comparisons of the results obtained for the sn class="Chemical">ample subjected to rRNA depletion plotted against the results for the two samples which were not depleted of rRNA. These two scatter plots are very similar to one another, but differ from that shown in Figure 2A by showing lightly greater variance. A larger number of genes showed transcript levels that differed by a factor of 2. This result suggested that the rRNA depletion process slightly altered the levels of some mRNAs. For this reason, and because the depletion method was variable and incomplete, rRNA depletion was abandoned after some initial trials. Because the values obtained for all three datasets were highly comparable, the transcript levels from the three “standard conditions” samples were averaged to produce a list of average transcript abundances for each gene for cells grown under standard conditions. These averages were subsequently used as the basis for comparisons of transcript abundances for other conditions (see Table 3; and Table S2 in Supplementary Material). Further, the high similarity of the datasets obtained for three “standard conditions” samples shows that reliable comparisons can be made even with a single dataset for a specific condition.

Table 3

Changes in the transcript level for genes of selected metabolic pathways.

Locus tag	Ratio high light/std	Ratio dark oxic/std	Ratio dark anoxic/std	Ratio dark anoxic/dark oxic	Ratio micro-oxic/std	Gene name	Gene product
GLYCOLYSIS
SYNPCC7002_A2438	1.31	1.51	0.29	0.19	0.93	glk	Glucokinase
SYNPCC7002_A0964*	0.53	3.27	1.62	0.50	0.83	pgi	Glucose-6-phosphate isomerase
SYNPCC7002_A0162*	0.39	2.95	0.80	0.27	0.46	pfkA	6-Phosphofructokinase PfkA
SYNPCC7002_A0886	0.06	1.48	0.08	0.05	0.75	–	PfkB family of fructokinase
SYNPCC7002_A0329*	0.43	0.70	0.57	0.82	0.82	fbp	Fructose-1,6-bisphosphatase
SYNPCC7002_A0010	0.85	4.41	1.05	0.24	0.71	fbaB	Fructose-bisphosphate aldolase class I
SYNPCC7002_A2697	0.07	5.23	0.63	0.12	0.71	gap	Glyceraldehyde-3-phosphate DH, type I
SYNPCC7002_A1585#	2.58	0.46	0.39	0.84	0.71	pgk	Phosphoglycerate kinase
SYNPCC7002_A2233	1.03	1.23	0.27	0.22	1.10	gpm	2,3-Bisphosphoglycerate-independent phosphoglycerate mutase
SYNPCC7002_A2560	1.66	0.34	0.26	0.76	1.16	gpmB	Phosphoglycerate mutase
SYNPCC7002_A0073	2.30	0.50	0.11	0.21	0.79	eno	2-Phosphopyruvate hydratase (enolase)
PYRUVATE METABOLISM
SYNPCC7002_A1658	1.56	0.36	0.11	0.31	1.04	pyk	Pyruvate kinase
SYNPCC7002_A0250	0.18	19.27	177.82	9.23	0.56	ppsA	Phosphoenolpyruvate synthase
SYNPCC7002_A0353	1.47	0.38	0.10	0.25	1.41	pdhA	Pyruvate dehydrogenase E1 alpha chain
SYNPCC7002_A0655	1.78	0.53	0.23	0.43	0.97	pdhB	Pyruvate dehydrogenase E1 beta chain
SYNPCC7002_A1126	1.23	0.59	0.47	0.80	0.99	ipdA	Dihydrolipoamide dehydrogenase
SYNPCC7002_A0110	1.52	0.40	0.11	0.27	0.89	–	Dihydrolipoamide S-acetyltransferase
SYNPCC7002_A1443	0.33	24.68	107.55	4.36	0.86	nifJ	Pyruvate:ferredoxin (flavodoxin) oxidoreductase
OXIDATIVE PENTOSE PHOSPHATE CYCLE
SYNPCC7002_A0964*	0.53	3.27	1.62	0.50	0.83	pgi	Glucose-6-phosphate isomerase
SYNPCC7002_A1459	0.25	3.08	2.64	0.85	0.86	zwf	Glucose-6-phosphate 1-dehydrogenase
SYNPCC7002_A0928	0.33	2.54	0.32	0.13	0.79	pgl	6-Phosphogluconolactonase
SYNPCC7002_A0221	0.39	3.55	3.46	0.97	0.65	gnd	6-Phosphogluconate DH, decarboxylating
SYNPCC7002_A1269$	2.45	0.09	0.06	0.66	1.39	rpiA	Ribose 5-phosphate isomerase A
SYNPCC7002_A0324$	2.99	0.47	0.22	0.47	0.99	rpe	Ribulose-phosphate 3-epimerase
SYNPCC7002_A1022$	1.31	0.40	0.55	1.36	0.90	tkt	Transketolase
SYNPCC7002_A0010*	0.85	4.41	1.05	0.24	0.71	fbaB	Fructose-bisphosphate aldolase class I
SYNPCC7002_A0329*	0.43	0.70	0.57	0.82	0.82	fbp	Fructose-1,6-bisphosphatase
SYNPCC7002_A0162*	0.39	2.95	0.80	0.27	0.46	pfkA	6-Phosphofructokinase PfkA
SYNPCC7002_A1460	0.31	3.38	5.53	1.64	0.79	tal	Transaldolase
SYNPCC7002_A2558	1.38	0.59	0.44	0.74	0.99	talC	Transaldolase
CBB CYCLE
SYNPCC7002_A1796	3.93	0.35	0.48	1.38	1.01	rbcS	RuBisCO, small subunit
SYNPCC7002_A1798	3.07	0.28	0.35	1.24	0.96	rbcL	RuBisCO, large subunit
SYNPCC7002_A1797	3.01	0.44	0.50	1.13	0.93	rbcX	RbcX protein, RuBisCO chaperone
SYNPCC7002_A1585#	2.58	0.46	0.39	0.84	0.71	pgk	Phosphoglycerate kinase
SYNPCC7002_A0106	1.94	0.25	0.04	0.17	1.00	gap	Glyceraldehyde-3-phosphate DH, type I
SYNPCC7002_A1352	2.64	0.53	0.20	0.37	1.08	fba	Fructose-bisphosphate aldolase, class II
SYNPCC7002_A0329	0.43	0.70	0.57	0.82	0.82	fbp	Fructose-1,6-bisphosphatase
SYNPCC7002_A1301	2.34	0.34	0.10	0.31	0.81	glpX	Bacterial fructose-1,6-bisphosphatase
SYNPCC7002_A1022$	1.31	0.40	0.55	1.36	0.90	tkt	Transketolase
SYNPCC7002_A0595	2.09	0.46	0.09	0.19	0.92	tpiA	Triosephosphate isomerase
SYNPCC7002_A1269$	2.45	0.09	0.06	0.66	1.39	rpiA	Ribose 5-phosphate isomerase A
SYNPCC7002_A0324$	2.99	0.47	0.22	0.47	0.99	rpe	Ribulose-phosphate 3-epimerase
SYNPCC7002_A2857	2.34	0.39	0.03	0.09	1.15	prk	Phosphoribulokinase
SYNPCC7002_A2665	1.36	1.10	0.40	0.36	1.05	prk	Phosphoribulokinase
CO2 CONCENTRATION
SYNPCC7002_A1805	0.63	0.89	0.30	0.33	1.17	ndhF4	NADH DH subunit F4
SYNPCC7002_A1806	0.53	0.60	0.14	0.23	0.94	ndhD4	NADH DH subunit D4
SYNPCC7002_A1807	0.54	0.57	0.13	0.22	1.01	cupB	CO2 hydration protein
SYNPCC7002_A0172	1.08	0.39	0.31	0.80	0.78	ndhF3	NADH DH (plastoquinone) chain 5
SYNPCC7002_A0173	1.76	0.27	0.35	1.30	1.03	ndhD3	NADH-quinone oxidoreductase (subunit 4)
SYNPCC7002_A0174	1.49	0.76	0.19	0.25	0.89	cupA	CO2 hydration protein
SYNPCC7002_A0175	2.39	0.79	0.26	0.33	1.48	cupS	Conserved hypothetical protein
SYNPCC7002_A0171	1.39	0.66	0.70	1.05	0.64	rbcR	Transcription regulator RbcR
SYNPCC7002_A2371	0.96	1.15	1.25	1.09	0.76	bicA	Bicarbonate transporter, BicA
SYNPCC7002_A0470	3.43	1.02	0.29	0.28	1.19	sbtA	Sodium-dependent bicarbonate transporter
SYNPCC7002_A2612	1.55	0.62	0.07	0.11	0.90	ccmK	CO2 concentrating mechanism protein
SYNPCC7002_A2613	1.71	0.61	0.51	0.84	0.59	ccmK	CO2 concentrating mechanism protein
SYNPCC7002_A1802	2.22	0.37	0.02	0.06	1.05	ccmK	CO2 concentrating mechanism protein
SYNPCC7002_A1803	3.56	0.46	0.03	0.07	1.33	ccmK	CO2 concentrating mechanism protein
SYNPCC7002_A2389	0.95	1.59	1.59	1.00	0.49	ccmk1	CO2 concentrating mechanism protein
SYNPCC7002_A1801	1.92	0.38	0.03	0.07	1.29	ccmL	CO2 concentrating mechanism protein
SYNPCC7002_A1800	2.69	0.31	0.04	0.14	0.80	ccmM	CO2 concentrating mechanism protein
SYNPCC7002_A1799	1.57	0.15	0.05	0.35	1.35	ccmN	CO2 concentrating mechanism protein
SYNPCC7002_A1997	2.27	0.60	0.15	0.26	0.96	icfA	Carbonic anhydrase
PHOTOSYSTEM II
SYNPCC7002_A1418	2.03	0.73	0.20	0.27	1.12	psbA	Photosystem q(b) protein
SYNPCC7002_A2164	0.71	0.57	0.62	1.08	0.86	psbA	Photosystem q(b) protein
SYNPCC7002_A0157	5.20	0.70	1.17	1.67	1.00	psbA-II	PS II D1 subunit PsbA-II (Qb protein)
SYNPCC7002_A1759	2.03	0.79	0.61	0.78	1.42	psbB	PS II protein
SYNPCC7002_A1559	1.66	0.51	0.40	0.79	1.33	psbC	PS II 44 kDa subunit reaction center protein
SYNPCC7002_A1560	1.81	0.50	0.22	0.45	1.49	psbD	PS II D2 protein (photosystem q(a) protein)
SYNPCC7002_A2199	1.62	0.60	0.12	0.19	1.00	psbD	PS II D2 protein
SYNPCC7002_A0230	1.25	0.57	0.13	0.24	0.68	psbE	Cytochrome b559, alpha subunit (PsbE)
SYNPCC7002_A0231	0.82	0.33	0.26	0.78	0.62	psbF	Cytochrome b559, beta subunit (PsbF)
SYNPCC7002_A0808	0.93	0.37	0.42	1.14	1.61	psbH	Phosphoprotein of PS II
SYNPCC7002_A0233	1.02	0.56	0.40	0.72	1.24	psbJ	PS II subunit PsbJ
SYNPCC7002_A2779	0.95	0.27	0.08	0.31	0.36	psbK	PS II 4 kDa reaction center component
SYNPCC7002_A0232	0.85	0.39	0.29	0.74	0.79	psbL	PS II subunit PsbL
SYNPCC7002_A2151	0.89	0.15	0.07	0.46	0.36	psbM	PS II reaction center M protein
SYNPCC7002_A0809	4.54	0.36	0.22	0.61	1.14	psbN	PS II reaction center N protein.-related protein
SYNPCC7002_A0269	1.72	0.17	0.17	1.03	2.25	psbO	PS II manganese stabilizing protein PsbO
SYNPCC7002_A1303	0.73	0.68	0.22	0.33	1.73	psbP	PsbP
SYNPCC7002_A0957	0.38	0.40	0.23	0.57	0.32	psbT	PS II reaction center, PsbT protein
SYNPCC7002_A0322	1.11	0.39	0.07	0.18	1.47	psbU	PS II 12 kDa extrinsic protein (PsbU)
SYNPCC7002_A0112	0.77	0.30	0.11	0.36	1.22	psbV	Cytochrome c-550 precursor (cytochrome c550)
SYNPCC7002_A1258	0.90	0.65	0.30	0.46	0.64	psbW	PS II reaction center W protein
SYNPCC7002_A1312	1.01	1.14	4.87	4.27	1.31	psbW2	PS II complex subunit
SYNPCC7002_A1347	1.95	0.44	0.19	0.43	1.97	psbY	PS II PsbY protein
SYNPCC7002_A0135	1.17	0.18	0.27	1.51	1.10	psbZ	PS II subunit PsbZ
SYNPCC7002_A2533	0.82	0.29	0.13	0.45	0.96	psbZ	PS II 11 kDa protein
PHOTOSYSTEM I
SYNPCC7002_A1961	0.42	0.52	0.56	1.08	0.93	psaA	PS I P700 chlorophyll A apoprotein A1
SYNPCC7002_A1962	0.56	0.39	1.21	3.08	1.50	psaB	PS I protein A2
SYNPCC7002_A1589	0.32	0.17	0.06	0.33	0.62	psaC	PS I iron-sulfur center subunit VII
SYNPCC7002_A0682	0.73	0.70	0.11	0.16	1.69	psaD	PS I subunit II
SYNPCC7002_A1393	0.69	0.32	0.04	0.12	0.90	psaE	PS I reaction center subunit IV
SYNPCC7002_A1008	1.31	0.49	0.12	0.25	2.14	psaF	PSI reaction center subunit III, PsaF
SYNPCC7002_A2621	0.35	0.32	0.08	0.24	1.31	psaI	PS I reaction center subunit VIII
SYNPCC7002_A1009	0.36	0.22	0.12	0.53	0.64	psaJ	PS I reaction center, subunit IX/PsaJ
SYNPCC7002_A2401	0.28	0.30	0.04	0.13	1.29	psaK	PS I reaction center subunit X
SYNPCC7002_A2620	0.35	0.44	0.02	0.04	1.12	psaL	PS I reaction center subunit XI
SYNPCC7002_A1834	0.27	0.32	0.41	1.28	0.28	psaM	PS I protein M
SYNPCC7002_A0975	1.92	0.38	0.19	0.49	0.92	btpA	PS I biogenesis protein btpA
PHOTOSYNTHETIC ELECTRON TRANSPORT
SYNPCC7002_A1910	2.05	0.17	0.06	0.37	1.07	petA	Apocytochrome f precursor
SYNPCC7002_A0842	2.27	0.68	0.13	0.19	1.31	petB	Cytochrome b6
SYNPCC7002_A1909	2.25	0.19	0.14	0.74	1.25	petC	Rieske FeS protein
SYNPCC7002_G0076	0.28	0.00	0.50	n.d.	0.00	petC-II	Rieske iron-sulfur protein paralog
SYNPCC7002_A0841	1.30	0.60	0.12	0.21	1.13	petD	cytb6/f complex subunit IV
SYNPCC7002_A1097	0.84	0.69	0.23	0.33	0.76	–	Cytochrome b6/f complex, alt. iron-sulfur subunit
SYNPCC7002_A0374	1.41	0.22	0.07	0.30	0.74	petG	Cytochrome b6-f complex subunit 5
SYNPCC7002_A1311	1.31	0.66	0.11	0.17	0.58	petM	Cytochrome b6-f complex subunit VII
SYNPCC7002_A2391	1.55	0.86	0.50	0.58	1.10	petJ	Cytochrome c6
SYNPCC7002_A0167	0.82	0.53	0.25	0.47	0.71	petJ	Cytochrome c6 precursor (cytochrome c553)
SYNPCC7002_A0624	0.77	0.55	1.31	2.36	0.75	petF	Ferredoxin 2Fe-2S II
SYNPCC7002_A2192	1.44	0.60	0.70	1.17	1.37	petF	Ferredoxin (2Fe-2S)
SYNPCC7002_A2325	4.08	0.69	0.14	0.20	1.15	petF	Ferredoxin
SYNPCC7002_A2326	2.81	0.40	0.11	0.27	1.53	petF	Ferredoxin I (2Fe-2S)
SYNPCC7002_A2548	0.66	1.38	1.01	0.73	0.36	petF2	Ferredoxin PetF2
SYNPCC7002_A0853	2.17	0.29	0.21	0.72	1.35	petH	Ferredoxin-NADP reductase
PHYCOBILISOMES, LYASES AND DEGRADATION
SYNPCC7002_A1930	0.67	0.20	0.05	0.23	1.55	apcA	Allophycocyanin alpha subunit
SYNPCC7002_A1929	1.01	0.20	0.05	0.23	1.99	apcB	Allophycocyanin, beta subunit
SYNPCC7002_A1928	0.49	0.26	0.07	0.26	0.78	apcC	Allophycocyanin-associated core-linker
SYNPCC7002_A2140	0.86	0.89	0.33	0.37	1.27	apcD	Allophycocyanin B alpha subunit
SYNPCC7002_A2009	0.51	0.51	0.27	0.52	1.41	apcE	Phycobilisome core-membrane linker ApcE
SYNPCC7002_A1631	0.92	0.75	0.21	0.28	0.96	apcF	Allophycocyanin beta-18 subunit
SYNPCC7002_A2210	0.53	0.25	0.10	0.40	2.46	cpcA	Phycocyanin, alpha subunit
SYNPCC7002_A2209	0.39	0.13	0.03	0.21	1.82	cpcB	Phycocyanin, beta subunit
SYNPCC7002_A2211	0.10	0.23	0.25	1.09	1.62	cpcC	Phycocyanin-associated rod linker protein
SYNPCC7002_A2212	0.23	0.20	0.02	0.09	1.35	cpcD	Phycocyanin-associated, rod-terminating linker
SYNPCC7002_A0811	0.45	0.42	0.12	0.29	1.22	cpcG1	Phycobilisome rod-core linker CpcG
SYNPCC7002_A0639	0.23	1.04	0.15	0.14	0.93	cpcG2	Phycocyanin-associated rod-core linker
SYNPCC7002_A2010	0.48	0.26	0.03	0.13	1.13	cpcM	Phycobiliprotein beta subunit methylase CpcM
SYNPCC7002_A2213	0.51	0.20	0.03	0.14	0.94	cpcE	Phycocyanin alpha subunit lyase, CpcE
SYNPCC7002_A2214	0.55	0.40	0.10	0.26	0.80	cpcF	Phycocyanin alpha subunit lyase, CpcF
SYNPCC7002_A1822	1.07	0.74	0.13	0.18	1.03	cpcS	Phycobiliprotein lyase
SYNPCC7002_A2053	0.68	1.19	7.95	6.66	0.76	cpcU	Phycobiliprotein lyase
SYNPCC7002_A2095	1.67	0.55	0.16	0.28	1.80	cpcT	Bilin lyase
SYNPCC7002_A2772	1.67	0.36	0.39	1.06	0.96	cpcV	Similar to phycobiliprotein lyase CpcS and CpcU
SYNPCC7002_A1821	2.17	4.77	0.75	0.16	1.21	nblA	Phycobilisome degradation protein
SYNPCC7002_A0348	0.79	2.68	6.62	2.47	1.02	nblB	Phycobiliprotein lyase related protein
PORPHYRIN/CHLOROPHYLL BIOSYNTHESIS
SYNPCC7002_A1302	0.66	0.47	0.15	0.32	1.45	hemA	Glutamyl-tRNA reductase
SYNPCC7002_A2206	1.69	0.47	0.09	0.19	0.69	hemL	Glutamate-1-semialdehyde-2,1-aminomutase
SYNPCC7002_A1753	0.44	0.36	0.24	0.67	0.81	hemB	Delta-aminolevulinic acid dehydratase
SYNPCC7002_G0127	0.92	0.00	0.17	n.d.	0.14	hemB-II	Delta-aminolevulinic acid dehydratase
SYNPCC7002_A1610	1.07	0.91	0.10	0.11	0.81	hemC	Porphobilinogen deaminase
SYNPCC7002_A1192	0.36	1.56	2.34	1.50	0.36	cysG/hemD	Uroporphyrin-III synthase/methyltransferase
SYNPCC7002_A0823	0.91	0.40	0.30	0.75	0.56	hemE	Uroporphyrinogen decarboxylase
SYNPCC7002_A1828	0.40	0.66	0.09	0.13	1.17	hemF	Coproporphyrinogen III oxidase, aerobic
SYNPCC7002_A1990	1.02	1.08	1.20	1.12	16.51	hemN2	O2-independent coproporphyrinogen III oxidase
SYNPCC7002_A2831	1.44	1.47	1.85	1.26	1.88	hemN	O2-independent coproporphyrinogen III oxidase, putative
SYNPCC7002_A0644	0.56	1.67	0.37	0.22	0.69	hemJ	Protoporphyrinogen oxidase
SYNPCC7002_A2476	0.43	1.05	0.23	0.22	0.92	chlP	Geranylgeranyl reductase
SYNPCC7002_A0596	1.90	0.41	0.09	0.22	1.15	chlD	Mg-protoporphyrin IX chelatase subunit D
SYNPCC7002_A2256	1.71	0.32	0.11	0.33	0.80	chlI	Mg-chelatase ATPase subunit I
SYNPCC7002_A1000	1.04	1.45	3.65	2.52	0.61	chlH	Mg-chelatase, subunit H
SYNPCC7002_A1018	0.68	0.78	0.52	0.67	1.01	chlH	mG-protoporphyrin IX chelatase, subunit H
SYNPCC7002_A0908	1.92	0.49	0.46	0.95	0.91	chlM	Magnesium-protoporphyrin O-methyltransferase
SYNPCC7002_A0707	0.82	0.58	0.08	0.14	1.19	acsF	Magnesium-protoporphyrin IX monomethyl ester aerobic oxidative cyclase (oxygen-dependent)
SYNPCC7002_A1992	1.49	0.31	11.09	36.18	18.65	acsF2	Mg-protoporphyrin IX monomethyl ester cyclase
SYNPCC7002_A0210	0.41	0.34	0.10	0.31	0.91	por	Light-dependent protochlorophyllide reductase
SYNPCC7002_A1652	0.08	3.88	5.79	1.49	0.75	chlB	Light-independent protochlorophyllide reductase, B subunit
SYNPCC7002_A2345	0.16	4.76	0.78	0.16	1.40	chlN	Light-independent protochlorophyllide reductase, N subunit
SYNPCC7002_A2347	0.15	8.20	0.41	0.05	0.90	chlL	Light-independent protochlorophyllide reductase, iron-sulfur ATP-binding protein
SYNPCC7002_A0548	2.59	0.59	0.19	0.32	1.14	chlG	Chlorophyll synthase, ChlG
SYNPCC7002_A2589	1.17	0.68	0.40	0.59	1.01	hemH	Ferrochelatase
SYNPCC7002_A2508	0.53	0.31	0.05	0.16	0.63	hox1	Heme oxygenase (decyclizing)
SYNPCC7002_A1991	0.59	0.14	17.46	129.16	14.87	ho2	Heme oxygenase 2
SYNPCC7002_A2228	0.55	0.78	0.26	0.33	1.33	–	Phycocyanobilin:ferredoxin oxidoreductase
ATP SYNTHASE
SYNPCC7002_A0734	3.01	0.21	0.29	1.40	0.75	atpA	ATP synthase F1, alpha subunit
SYNPCC7002_A0739	1.60	0.07	0.02	0.23	1.65	atpB	ATP synthase F0, A subunit
SYNPCC7002_A0750	2.80	0.28	0.06	0.23	0.60	atpC	ATP synthase F1, epsilon subunit
SYNPCC7002_A0749	2.52	0.33	0.51	1.55	1.22	atpD	ATP synthase beta chain
SYNPCC7002_A0738	4.47	0.23	0.21	0.90	0.71	atpE	ATP synthase C chain (Lipid-binding protein)
SYNPCC7002_A0736	2.01	0.13	0.08	0.63	0.83	atpF	ATP synthase B chain (Subunit I)
SYNPCC7002_A0733	2.18	0.23	0.23	1.00	0.79	atpG	ATP synthase F1, gamma subunit
SYNPCC7002_A0737	1.45	0.13	0.09	0.73	1.07	atpG	ATP synthase B chain (Subunit II)
SYNPCC7002_A0735	2.03	0.10	0.16	1.51	0.88	atpH	ATP synthase F1, delta subunit
SYNPCC7002_A0740	1.65	0.14	0.04	0.31	0.62	atp1	ATP synthase subunit I
SYNPCC7002_G0151	0.21	1.58	0.33	0.21	1.34	atpA-II	ATP synthase F1, alpha subunit
SYNPCC7002_G0148	0.15	1.26	0.55	0.44	0.98	atpB-II	ATP synthase F0, A subunit
SYNPCC7002_G0145	0.43	3.39	0.35	0.10	1.28	atpC-II	ATP synthase epsilon subunit
SYNPCC7002_G0144	0.31	1.88	0.88	0.47	0.98	atpD-II	ATP synthase F1, beta subunit
SYNPCC7002_G0146	0.25	2.86	0.35	0.12	1.07	–	F0F1-ATPase subunit, putative
SYNPCC7002_G0147	0.08	1.1	0.17	0.16	1.2	–	Conserved hypothetical protein
SYNPCC7002_G0150	0.30	2.68	0.80	0.30	1.44	atpF-II	ATP synthase b subunit
SYNPCC7002_G0152	0.11	0.98	0.16	0.17	1.51	atpG-II	ATP synthase F1, gamma subunit
SYNPCC7002_G0149	0.34	2.64	0.93	0.35	0.79	atpH-II	ATP synthase c subunit
NADH DEHYDROGENASE
SYNPCC7002_A0926	1.19	0.64	0.26	0.42	1.34	ndhA	NADH dehydrogenase subunit A
SYNPCC7002_A2547	1.83	1.16	0.46	0.39	0.83	ndhB	NADH dehydrogenase subunit B
SYNPCC7002_A2748	2.23	0.90	1.85	2.06	1.29	ndhC	NADH dehydrogenase subunit C
SYNPCC7002_A2327	1.68	0.30	0.08	0.27	0.94	ndhD	NADH dehydrogenase subunit 4
SYNPCC7002_A2000	0.22	1.16	0.63	0.54	1.04	ndhD1	NADH dehydrogenase subunit D1
SYNPCC7002_A1973	15.62	0.84	0.20	0.24	1.48	ndhD2	NADH dehydrogenase subunit D2
SYNPCC7002_A0923	1.16	0.61	0.07	0.11	1.04	ndhE	NADH dehydrogenase subunit E
SYNPCC7002_A0854	1.61	0.85	0.83	0.97	1.05	ndhF	NADH dehydrogenase subunit 5
SYNPCC7002_A0924	1.21	0.85	0.16	0.18	1.31	ndhG	NADH dehydrogenase subunit G
SYNPCC7002_A2541	1.39	1.00	0.33	0.34	0.80	ndhH	NADH dehydrogenase subunit H
SYNPCC7002_A0925	1.11	0.77	0.10	0.13	1.10	ndhI	NADH-plastoquinone oxidoreductase, I subunit
SYNPCC7002_A2750	1.98	1.08	1.03	0.95	0.99	ndhJ	NADH dehydrogenase subunit J
SYNPCC7002_A2749	1.89	0.85	1.86	2.18	1.07	ndhK	NADH dehydrogenase subunit K
SYNPCC7002_A0560	2.22	0.78	0.63	0.81	1.30	ndhL	NADH DH subunit L(inorganic carbon transporter)
SYNPCC7002_A0569	1.71	1.10	0.44	0.40	1.77	ndhM	Conserved hypothetical protein
SYNPCC7002_A1143	1.17	1.49	1.02	0.69	0.80	ndhN	Conserved hypothetical protein
SYNPCC7002_A2094	1.19	0.84	1.23	1.47	0.83	ndhO	Conserved hypothetical protein
CYTOCHROME OXIDASE
SYNPCC7002_A1162	0.39	1.09	0.09	0.09	0.77	ctaCI	Cytochrome oxidase subunit II
SYNPCC7002_A1163	0.50	1.57	0.34	0.22	1.06	ctaDI	Cytochrome oxidase large subunit (subunit I)
SYNPCC7002_A1164	0.33	1.15	0.45	0.39	1.00	ctaEI	Cytochrome oxidase small subunit (subunit III)
SYNPCC7002_A0727	0.62	0.45	0.11	0.24	1.12	ctaCII	Cytochrome C oxidase subunit II
SYNPCC7002_A0726	1.13	0.86	0.20	0.23	0.96	ctaDII	Cytochrome oxidase II large subunit
SYNPCC7002_A0725	1.09	0.81	0.29	0.35	1.26	ctaEII	Cytochrome oxidase II small subunit
TRANSHYDROGENASE
SYNPCC7002_A0986	0.08	3.80	1.31	0.34	0.85	pntA	Transhydrogenase, chain alpha, part 1
SYNPCC7002_A0984	0.13	6.80	0.60	0.09	0.83	pntB	Transhydrogenase, subunit beta
SYNPCC7002_A0985	0.15	7.36	0.84	0.11	1.26	pntC	Transhydrogenase, chain alpha, part 2
LACTATE DEHYDROGENASE, ALCOHOL DEHYDROGENASE AND HYDROGENASE
SYNPCC7002_G0164	0.48	0.24	0.14	0.61	0.63	ldhA	Fermentative lactate dehydrogenase
SYNPCC7002_A1838	0.64	1.29	1.07	0.83	0.65	acsA	Acetate-CoA ligase
SYNPCC7002_A2015	0.77	2.37	0.46	0.19	0.94	acs	Acetyl-CoA synthetase
SYNPCC7002_A0868	0.59	1.10	0.41	0.38	0.90	–	Alcohol DH, Zn-binding family
SYNPCC7002_A2590	0.88	0.60	0.11	0.19	0.72	–	Short-chain alcohol dehydrogenase family
SYNPCC7002_A0195	0.30	6.62	16.39	2.47	1.24	hoxE	Hydrogenase subunit E
SYNPCC7002_A0196	0.26	7.23	4.78	0.66	1.06	hoxF	Hydrogenase large diaphorase subunit F
SYNPCC7002_A0197	0.33	4.88	2.15	0.44	0.76	hoxU	Hydrogenase small diaphorase subunit U
SYNPCC7002_A0198	0.37	4.32	0.99	0.23	1.05	hoxY	Hydrogenase small subunit Y
SYNPCC7002_A0200	0.31	3.11	0.40	0.13	0.53	hoxH	Hydrogenase large subunit H
FATTY ACID DESATURASES
SYNPCC7002_A2756	2.92	0.14	0.05	0.39	1.82	desA	Phosphatidylcholine desaturase
SYNPCC7002_A0159	9.49	0.00	0.05	n.d.	1.20	desB	Omega-3 acyl-lipid desaturase
SYNPCC7002_A2198	0.81	0.25	0.07	0.28	1.02	desC	Delta-9 acyl-lipid desaturase
SYNPCC7002_A2833	1.56	1.40	0.39	0.28	1.55	desE	Fatty acid desaturase
SYNPCC7002_A1989	0.59	0.27	1.40	5.19	20.88	desF	syn-2, delta 9 acyl-lipid fatty acid desaturase
IRON UPTAKE
SYNPCC7002_A2507	1.01	0.61	0.16	0.26	5.39	sufA	Iron transport protein
SYNPCC7002_G0079	1.05	0.97	2.51	2.59	6.23	–	ABC transporter, permease, FecCD family
SYNPCC7002_G0080	0.31	1.44	3.66	2.55	5.05	–	Iron ABC transporter, ATP-binding protein
SYNPCC7002_G0081	0.36	0.33	0.93	2.83	5.01	–	Outer membrane TonB-dependent receptor
SYNPCC7002_G0083	0.50	1.84	1.52	0.82	7.54	–	ABC transporter, iron-binding lipoprotein
SYNPCC7002_G0086	1.13	0.50	0.41	0.82	3.32	–	ATP-binding protein of ABC transporter for iron
SYNPCC7002_G0087	0.52	0.48	0.36	0.75	3.02	fecD	Iron compound ABC transporter (FecCD)
SYNPCC7002_G0088	0.67	0.65	0.11	0.16	3.96	fecC	FecCD transport (permease) family
SYNPCC7002_G0090	0.85	0.93	0.17	0.18	4.57	–	TonB family C-terminal domain protein
SYNPCC7002_G0091	0.29	0.54	1.02	1.89	1.27	fecB	Iron(III) dicitrate-binding periplasmic protein
SYNPCC7002_G0092	0.63	0.63	1.08	1.70	0.90	–	Iron(III) dicitrate-binding periplasmic protein
SYNPCC7002_G0093	0.00	0.99	1.64	1.65	2.91	–	TonB-dependent receptor
SYNPCC7002_G0095	0.97	0.45	0.42	0.94	2.36	–	Ferric aerobactin receptor
SYNPCC7002_G0102	1.16	0.61	1.52	2.47	0.90	–	Periplasmic binding protein; for iron siderophore
SYNPCC7002_G0138	1.13	0.55	0.15	0.27	11.93	–	TonB-dependent siderophore receptor
SYNPCC7002_G0139	0.52	0.00	0.11	n.d.	7.82	–	Periplasmic binding protein; iron transporter
OTHERS
SYNPCC7002_A0263	0.21	9.00	36.21	4.02	0.41	lrtA	Light-repressed protein LrtA
SYNPCC7002_A1827	0.80	0.32	0.06	0.20	0.98	nirA	Nitrite reductase
SYNPCC7002_A1314	1.12	0.17	0.13	0.77	0.98	narB	Nitrate reductase
SYNPCC7002_A1292	0.22	0.41	0.19	0.47	13.60	isiA	Photosystem I chlorophyll-binding protein
SYNPCC7002_A1291	1.37	0.00	0.30	n.d.	32.59	isiB	Flavodoxin
SYNPCC7002_A1442	0.84	31.80	49.22	1.55	0.86	–	Dihydroorotate dehydrogenase
SYNPCC7002_A2195	0.88	0.76	0.20	0.26	0.78	pyrD	Dihydroorotate dehydrogenase
SYNPCC7002_A2286	2.29	0.33	0.12	0.36	1.01	pstA	Phosphate ABC transporter, permease protein
SYNPCC7002_A1895	1.66	0.76	0.33	0.44	0.97	pstB	Phosphate import ATP-binding protein
SYNPCC7002_A2285	2.17	0.66	0.25	0.37	0.75	pstC	Phosphate ABC transporter, permease protein
SYNPCC7002_A2284	5.87	0.43	0.27	0.64	0.45	pstS	Phosphate transporter; substrate-binding protein
SYNPCC7002_A1321	1.00	2.77	3.12	1.13	0.65	flv1	Flavoprotein
SYNPCC7002_A1743	0.62	1.79	0.70	0.39	0.68	flv2	Flavin reductase like domain protein
SYNPCC7002_A0147	3.07	0.37	0.46	1.25	0.65	groELII	Chaperonin GroEL-II (Cpn 60)
SYNPCC7002_A2457	6.8	0.16	0.39	2.38	0.67	groES	Chaperonin, 10 kDa protein
SYNPCC7002_A2458	4.07	0.12	0.23	1.86	0.67	groEL	Chaperonin, 60 kDa protein

The ratio of relative transcript abundance under high light, dark oxic incubation, dark fermentative incubation, and low O *, #, and $, respectively.

n.d., not defined (due to division by 0); std, standard conditions.

The ratio of relative transcript abundance under high light, dark oxic incubation, dark n class="Chemical">fermenpan>tative inpan>cubation, anpan>d low O *, #, anpan>d $, respectively.

Transcription changes in cells subjected to high-light intensity

n class="Species">Synechococcus 7002 is knpan>ownpan> to be extremely toleranpan>t to high-light inpan>tenpan>sity (Sakamoto anpan>d Bryanpan>t, 2002; pan> class="Chemical">Nomura et al., 2006b), and sunlight intensity is one of the most important environmental factors in natural habitats of cyanobacteria. A light intensity of 250 μmol photons m−2 s−1, which is saturating for standard growth of dilute Synechococcus 7002 cultures, is often regarded as a high-light condition for Synechocystis sp. PCC 6803 (hereafter Synechocystis 6803; e.g., Hihara et al., 2001; Mizusawa et al., 2009). To induce high-light stress in Synechococcus 7002, a culture which was grown under standard conditions was exposed for 1 h to a light intensity of 900 μmol photons m−2 s−1. Table S2 in Supplementary Material provides comparisons for all genes under all conditions reported here, and it includes the p-values for the respective comparisons derived from the statistical analyses. Figure 3A shows a scatter plot of the relative transcript abundances for all ORFs in cells exposed to high-light intensity compared to the average value for standard conditions. Many genes show differences greater than two-fold, and the values for a few specific genes are indicated on the panel. The transcript levels of many genes involved in the Calvin–Benson–Bassham cycle were two- to four-fold higher in cells exposed to high light (Table 3). Transcripts for genes encoding the structural components of carboxysomes (ccmK, ccmL, ccmM, ccmN) and for carbonic anhydrase (icfA; Badger and Price, 2003; Yeates et al., 2008; Cannon et al., 2009) also increased 1.5- to 3-fold, and transcripts for sbtA, encoding the sodium-dependent bicarbonate transporter, showed an increase of 3.5-fold. Thus, a 1-h treatment with high light was sufficient to cause the well-known physiological response that high-light intensity is perceived by cyanobacterial cells as CO2 limitation (Badger and Price, 2003; Woodger et al., 2007). Interestingly, transcript levels for the genes encoding the so-called inducible CO2 uptake mechanism (ndhD3, ndhF3, cupA, and cupS; Ogawa and Mi, 2007) were either constant or only slightly higher (maximum two-fold increase); however, the transcript levels for the constitutive CO2-concentrating mechanism (ndhD4, ndhF4, cupB) decreased slightly (maximum decrease of approximately two-fold). Similar results were obtained in a microarray study in Synechocystis 6803; after a shift from low light to high-light intensity, the transcript levels of genes encoding ribulose 1,5-bisphosphate carboxylase/oxygenase (RuBisCO) and carboxysome components increased, whereas mRNA levels of genes encoding the inducible CO2 uptake mechanism increased only transiently or remained unchanged for the constitutive CO2 uptake mechanism (Hihara et al., 2001).

Figure 3

Changes in the relative transcript abundance after different treatments or growth. The scatter plots show the relative transcript abundances (A) after a 1-h high-light treatment, (B) after 1 h incubation in the dark (oxic conditions), and (C) after 1 h incubation in the dark under anoxic conditions compared to that for “standard conditions” (mean of three biological replicates). Scatter plot (D) shows the relative transcript abundances for a culture grown under micro-oxic conditions compared to a culture grown under “standard conditions”. The gray lines give two-fold changes in either direction. Selected genes are identified by name/locus tag number.

Changes in the relative transcript abundance after difn class="Chemical">ferent treatments or growth. The scatter plots show the relative transcript abundances (A) after a 1-h high-light treatment, (B) after 1 h incubation in the dark (oxic conditions), and (C) after 1 h incubation in the dark under anoxic conditions compared to that for “standard conditions” (mean of three biological replicates). Scatter plot (D) shows the relative transcript abundances for a culture grown under micro-oxic conditions compared to a culture grown under “standard conditions”. The gray lines give two-fold changes in either direction. Selected genes are identified by name/locus tag number.

The transcript levels for genes encoding the structural subunits of the Rn class="Chemical">NA polymerase anpan>d the ribosomal proteinpan>s did not chanpan>ge or inpan>creased slightly (two-fold at maximum; see Table S2 inpan> Supplemenpan>tary Material). Similar to results observed with pan> class="Species">Synechocystis 6803, transcript levels for the genes encoding the GroEL/GroES chaperone increased three- to seven-fold (Table 3; Figure 3A; Hihara et al., 2001).

Transcripts for ndhD2 were 15-fold higher in cells exposed to high light than in standard conditions (Table 3). However, transcripts for ndhD1, encoding a paralogous form of n class="Chemical">NdhD subunit for the Type-1 pan> class="Chemical">NADH dehydrogenase complex, were about four-fold lower in high-light grown cells. Cyanobacterial genomes encode paralogous NdhD and NdhF subunits (Ogawa and Mi, 2007). The NdhD1/NdhD2 and NdhF1/NdhF2 proteins are involved in formation of the Type-1 NADH dehydrogenase complex, which is required for respiration and cyclic electron flow, whereas the NdhD3/NdhD4 and NdhF3/NdhF4 proteins are subunits of specialized NADH dehydrogenase-like complexes required for CO2 concentration in cells (Battchikova and Aro, 2007; Ogawa and Mi, 2007). The higher ndhD2 mRNA level after high-light treatment might suggest a preferential involvement of NdhD2 in cyclic electron flow rather than in respiration. However, very little specific biochemical information is currently available for NADH dehydrogenase complexes containing NdhD2 (Battchikova and Aro, 2007). The transcript levels of the genes encoding other NADH dehydrogenase subunits were more or less constant, or increased slightly (less than or equal to two-fold); this might also indicate a slightly higher requirement for NADH dehydrogenase activity for cyclic electron flow at high-light intensity.

Interestingly, the transcript level for the desB gene, which encodes the Δ15 n class="Chemical">fatty acid desaturase, was 10-fold higher after the high-light treatmenpan>t (Table 3), anpan>d the mRNA level for desA (Δ12 desaturase) was three-fold higher. A similar observation was made for Synechocystis 6803 in a microarray analysis (Hihara et al., 2001). However, the transcript levels of the other fatty acid desaturases (desC, desE, desF) were unchanged. These results suggested that cells might increase the proportion of polyunsaturated fatty acids after high-light treatment. It has previously been shown that a desA mutant strain for Synechococcus 7002 exhibited a temperature-sensitive phenotype when cultures were grown at high-light intensity (Sakamoto and Bryant, 2002). Transcripts for pstS, encoding the substrate-binding protein of the phosphate transport system, were six-fold higher after high-light treatment (Table 3). However, the transcript levels for other genes for this ABC transport system (pstA, pstB, pstC) only increased approximately two-fold. Together with the higher transcript levels for genes involved in bicarbonate/CO2 uptake, this could point to an increased cell growth. However, the genes coding for components of nitrate and sulfate assimilation mechanisms do not show an increase in the mRNA levels (see Table S2 in Supplementary Material), and it has been reported that the growth rate of Synechococcus 7002 remains constant under moderately high-light conditions (Nomura et al., 2006a). Therefore, the up-regulation of the phosphate ABC transport system probably reflects an increased requirement for phosphate in cells exposed to high light, but this requirement probably is not directly linked to faster cell growth.

Genes showing much lower transcript levels after high-light treatment included glyceraldehyde-3-n class="Chemical">phosphate depan> class="Chemical">hydrogenase (gap, SYNPCC7002_A2697, ∼15-fold) and a bacterial/plant-type fructokinase (pfkB, SYNPCC7002_A0886, ∼15-fold; Table 3). Transcript levels for an alternative glyceraldehyde-3-phosphate dehydrogenase (gap, SYNPCC7002_A0106) were slightly higher (approximately two-fold) and for a second phosphofructokinase gene (pfkA, SYNPCC7002_A0162) were slightly lower (approximately two-fold). Interestingly, transcript levels for pntA, pntB, and pntC, encoding subunits of pyridine nucleotide transhydrogenase, were 7- to 13-fold lower after high-light treatment (Table 3). These results suggested that electron exchange between the NADH and NADPH pools might be less important in cells under high-light conditions.

The reducing equivalents required for cell growth are provided by the n class="Disease">photosystems in the light, anpan>d inpan> the absenpan>ce of photoinpan>hibitionpan> effects, high-light intensity should provide more reducing equivalents. On the other hand, when the light intensity is too high, the photosynthetic apparatus could produce excess reducing equivalents, which should cause cells to increase the transcript levels for genes involved in electron-consuming processes, e.g., the enzymes of the Calvin–Benson–Bassham cycle (see above). Phycobilisomes serve as the main antennae for photosynthesis in cyanobacteria, and they transfer excitation energy to both photosystems (Ashby and Mullineaux, 1999; Dong et al., 2009). Acclimation of this system directly affects the efficiency of excitation usage. Further, acclimative changes in the PS II to PS I ratio determines the ratio of linear electron transport (producing reductants (e.g., NADPH) from water oxidation) to cyclic electron transport (involving PS I and the cytochrome b6f complex), which generates proton motive force across the thylakoid membrane for ATP synthesis (Fujita et al., 1994).

The transcript levels of the genes encoding PS I subunits, PS II subunits and other photosystem-related proteins changed upon high-light treatment. Transcript levels for PS II genes generally increased, whereas the transcript levels of PS I genes generally decreased (Figure 4). Similar observations were made for n class="Species">Synechocystis 6803 exposinpan>g cells to high light (300 μmol photonpan>s m−2 s−1 for 15 minpan> or 1 h; Hihara et al., 2001). Inpan> this study for pan> class="Species">Synechococcus 7002, transcripts for psbN and psbA-II (SYNPCC7002_A0157) increased approximately five-fold, and those for the other two psbA genes (SYNPCC7002_A1418 and SYNPCC7002_A2164) either increased slightly (approximately two-fold) or remained constant, respectively. Due to the high sequence similarity of SYNPCC7002_A0157 and SYNPCC7002_A1418, many psbA sequences did not map uniquely which made inferences difficult. However, transcripts for psbA gene SYNPCC7002_A2164 were present at a very low level compared to those for the other two psbA genes. Compared to the other two psbA paralogs (SYNPCC7002_A0157 and SYNPCC7002_A1418), which encode proteins of nearly identical amino acid sequence, the SYNPCC7002_A2164 PsbA paralog has many differences in the amino acid sequence. Multiple copies of psbA also occur in other cyanobacteria (e.g., Synechocystis 6803, Nostoc sp. PCC 7120, and Thermosynechococcus elongatus), and it has been reported that the transcription of these genes is modulated in response to changes in light intensity and O2 level (Schaefer and Golden, 1989; Summerfield et al., 2008; Sander et al., 2010). Our data suggest that transcripts of SYNPCC7002_A0157 increased five-fold at high-light intensities (Figure 3A), whereas the other tested conditions did not affect its transcription very much. Transcript levels for the other psbA genes were more or less constant under the conditions tested in this study or were slightly lower under dark fermentative conditions (in the case of SYNPCC7002_A1418). Considering the genes for all subunits of both photosystems, the data clearly showed that significant changes in the ratio of PS I and PS II begin with changes at the mRNA level (Figure 4). These changes are well correlated with the observation that the PS II to PS I ratio increases when cells are grown at higher light intensity (Fujita et al., 1994). When energy transfer from phycobilisomes to PS II is impaired in Synechococcus 7002, the PS II content of the cells increases (Zhao et al., 2001), whereas dark acclimation of wild type cells resulted in a state 2 transition (i.e., direct energy transfer from phycobilisomes to PS I; Huang et al., 2003; Dong et al., 2009). A recent microarray study in Synechocystis 6803 showed that direct excitation of either PS I or PS II caused significant changes in transcription. The genes with increased transcript levels were not only restricted to those genes of photosystem that was not being excited, but included genes for diverse metabolic processes as well (Singh et al., 2009).

Figure 4

Relative transcript levels of genes encoding components of photosystem I and photosystem II after a 1-h exposure to high light compared to standard conditions. The transcript level of genes encoding components of PS II (light gray) and PS I genes (dark gray) are displayed as the ratio of the relative transcript abundances for high-light-treated cells compared to the mean of three samples for standard conditions. A value of 1 represents an equal mRNA level under both conditions; values greater than 1 indicate an increase in relative transcript abundance; and values lower than 1 represent a decrease in relative transcript abundance after a 1-h exposure to high light.

Relative transcript levels of genes encoding components of photosystem I and n class="Disease">photosystem II after a 1-h exposure to high light compared to stanpan>dard conpan>ditionpan>s. The tranpan>script level of genpan>es enpan>codinpan>g componpan>enpan>ts of PS II (light gray) anpan>d PS I genpan>es (dark gray) are displayed as the ratio of the relative tranpan>script abundanpan>ces for high-light-treated cells compared to the meanpan> of three span> class="Chemical">amples for standard conditions. A value of 1 represents an equal mRNA level under both conditions; values greater than 1 indicate an increase in relative transcript abundance; and values lower than 1 represent a decrease in relative transcript abundance after a 1-h exposure to high light.

Among the components of the photosynthetic electron transport chain, the petH gene, encoding n class="Chemical">ferredoxinpan>:pan> class="Chemical">NADP+ oxidoreductase, showed approximately two-fold higher transcript levels in cells exposed to high light (Table 3). Transcript levels for genes for other components of electron transfer, such as the cytochrome b6f complex and cytochrome c also increased up to two-fold. The situation for petF, encoding the ferredoxin electron acceptor for PS I, was more complex, because several genes are annotated as ferredoxins. Some of these genes showed elevated transcript levels at high light (four-fold at maximum), whereas transcript levels for others remained constant. There was an overall decrease in transcript levels for phycocyanin and phycocyanin-associated linker proteins (2- to 10-fold; Figures 3A and 5), whereas the transcript level of allophycocyanin-associated genes decreased to a lesser extent (maximally approximately two-fold reduction; Figure 5). Transcripts for genes encoding the enzymes of heme and chlorophyll biosynthesis generally did not change very much after a 1-h exposure to high light with one major exception. Transcripts for chlB, chlL, and chlN genes, encoding the light-independent (dark-active) protochlorophyllide reductase, decreased ∼6- to 12-fold upon high-light treatment.

Figure 5

Changes in transcript levels of genes encoding phycobiliproteins, linker polypeptides, and enzymes specifically involved in their biogenesis/maturation. The transcript levels for genes encoding phycobiliproteins, linker polypeptides, and enzymes involved in phycobiliprotein biogenesis/maturation are displayed as the ratio of the relative transcript abundances for cells exposed to high light for 1 h (light gray) or incubated under dark oxic conditions for 1 h (dark gray) compared to the mean of abundances for three standard conditions samples.

Changes in transcript levels of genes encoding phycobiliproteins, linker polypeptides, and enzymes specifically involved in their biogenesis/maturation. The transcript levels for genes encoding phycobiliproteins, linker polypeptides, and enzymes involved in phycobiliprotein biogenesis/maturation are displayed as the ratio of the relative transcript abundances for cells exposed to high light for 1 h (light gray) or incubated under dark oxic conditions for 1 h (dark gray) compared to the mean of abundances for three standard conditions sn class="Chemical">amples.

The genes encoding the F0F1-type n class="Chemical">ATP synpan>thase showed higher tranpan>script levels (up to four-fold) after high-light treatmenpan>t thanpan> cells grownpan> under stanpan>dard conpan>ditionpan>s. This observationpan> suggested that the higher level of electronpan> tranpan>sport anpan>d protonpan> motive force produced under these conpan>ditionpan>s is probably used to enpan>hanpan>ce pan> class="Chemical">ATP synthesis. Finally, there are two gene clusters, ctaI and ctaII, that encode cytochrome oxidases in Synechococcus 7002 (Nomura et al., 2006a). Transcript levels for the ctaI genes were considerably higher compared to those for the ctaII genes (4- to 20-fold higher under standard conditions); this finding agrees with previous results showing that cytochrome oxidase I is the major terminal oxidase and is responsible for most oxygen uptake in the dark (Nomura et al., 2006b). Furthermore, the cytochrome oxidases are required to maintain cellular redox balance in the light. Previous studies had suggested that cytochrome oxidase II might play a role as a signal transducer to measure redox balance and trigger an oxidative stress response (Nomura et al., 2006b). Transcript levels for the ctaI genes decreased slightly (approximately two-fold) after high-light treatment, whereas the transcript levels for the ctaII genes were essentially unchanged (Table 3).

Flavoproteins related to SYn class="Chemical">NPCC7002_A1321 anpan>d SYpan> class="Chemical">NPCC7002_A1743 have previously been reported to act as oxygen photoreductases in Synechocystis sp. PCC 6803 (Helman et al., 2003; Hackenberg et al., 2009). Interestingly, high-light treatment did not increase the transcript levels for these two genes, and in fact, transcript levels for SYNPCC7002_A1743 might have decreased slightly (∼1.5-fold). This observation suggested that these flavoproteins might have functions in addition to their roles in the dissipation of excess electrons via the Mehler reaction.

Transcription changes in cells under dark oxic (respiratory) conditions

Light is obviously the key factor for photolithoautotrophic growth because it provides the reducing equivalents for n class="Chemical">CO2 reductionpan>. Light powers the electronpan> tranpan>sport reactionpan>s that genpan>erate protonpan>motive force for pan> class="Chemical">ATP synthesis, and thus directly fuels the biosynthesis of all cell components. For cyanobacteria in natural environments, cells are exposed to alternating periods of light and dark during a diel cycle, and light intensity may also increase or decrease depending on cloud cover and other factors. Changes in light availability will produce large changes in the supplies of ATP and reducing power in cyanobacterial cells. Therefore, gene expression patterns must be constantly readjusted to enable cells to cope with these changing circumstances.

To simulate the light-to-dark transition that would occur at dusk, a culture that had been grown under standard conditions was transn class="Chemical">ferred inpan>to darknpan>ess for 1 h while sparginpan>g with 1% (v/v) pan> class="Chemical">CO2 in air (i.e., under oxic conditions). Figure 3B shows a scatter plot comparing transcript levels in dark oxic conditions to those for standard conditions. The transcriptional changes that occurred upon dark oxic treatment were generally more severe compared to those associated with high-light treatment (Figure 3A). Both the number of genes showing a greater than two-fold change in transcript abundance, as well as the magnitude of the observed differential transcription changes, were higher. Transcripts for genes encoding components of the photosynthetic apparatus were still relatively abundant after a 1-h dark oxic incubation. Although transcript levels for a few genes remained unchanged, the relative transcript levels for genes encoding components of the photosynthetic apparatus were significantly lower (up to approximately six-fold) compared to their levels in cells under standard conditions (Table 3). These observations for transcript levels of photosystem-related genes are consistent with a comparable study in Synechocystis 6803, in which a culture was incubated in the dark and transcription changes were subsequently monitored as a function of time (Gill et al., 2002).

The transcript levels of genes encoding the enzymes of n class="Chemical">heme anpan>d pan> class="Chemical">chlorophyll biosynthesis were generally slightly lower upon dark incubation (approximately two- to three-fold for many of the genes); however, there was one exception: the three genes encoding the light-independent protochlorophyllide reductase (Kada et al., 2003; Bröcker et al., 2010; Muraki et al., 2010). The mRNA levels for the subunits of this enzyme (chlL, chlB, chlN) were four- to eight-fold higher in cells under dark oxic conditions. In the dark the transcript levels of genes encoding the structural subunits of the phycobilisomes and the bilin lyases were either approximately constant [e.g., apcD, cpcG2 (SYNPCC7002_A0639), apcF] or were lower by up to seven-fold (Figure 5). Although transcripts for both allophycocyanin and phycocyanin-related genes were lower in the dark, transcripts for phycocyanin and the peripheral rods were generally affected to a greater extent than genes encoding components of the phycobilisome core. Transcripts for nblA increased about five-fold upon dark incubation. Similar observations were made in a microarray study for the transcript levels for apcAB, cpcBA, and nblA genes in Synechocystis 6803 when a culture was incubated in the dark (Gill et al., 2002). Because NblA causes phycobiliproteins to become sensitive to degradation by Clp proteases (Baier et al., 2004; Karradt et al., 2008), this observation suggests that phycobilisomes and phycobiliproteins are actively degraded in the dark in order to reduce the cellular content of phycobiliproteins and to recycle the reduced carbon and nitrogen contained within the proteins of these antenna structures. Interestingly, the mRNA level for nblA also increased approximately two-fold after high-light treatment, which might be indicative of a similar reduction of light-harvesting components under high light.

After 1 h in the dark the mRn class="Chemical">NA levels for the Type-1 pan> class="Chemical">NADH dehydrogenase genes and for the cytochrome oxidase genes (both ctaI and ctaII) were similar to those in cells grown under standard conditions. However, transcript levels for most genes encoding the F0F1-type ATP synthase were significantly lower, 3- to 14-fold, after 1 h in the dark (Table 3). This correlates well with respiratory electron transport rates that are about 10-fold lower than the rate of oxygen evolution in the light (Nomura et al., 2006a). The Synechococcus 7002 genome encodes a second set of genes for an F0F1-type ATP synthase (annotated as ATPase II; SYNPCC7002_G0144 – SYNPCC7002_G0152). This plasmid-located (on plasmid pAQ7) gene cluster has been suggested to encode a Na+-translocating N-ATPase (Dibrova et al., 2010). The transcript levels for the genes in this particular gene cluster were very low under standard conditions (about 5% of the corresponding ATPase I levels) and were even lower in cells exposed to high light (Table 3). Transcript levels for the ATPase II genes, however, increased upon dark incubation (approximately three-fold). The transcription data thus showed that the genes encoding this putative N-ATPase are transcribed and are regulated; however, the biological function of this putative ATPase is currently unknown.

Genes involved in n class="Chemical">CO2 uptake, conpan>cenpan>trationpan> anpan>d fixationpan> showed lower tranpan>script levels after 1-h dark inpan>cubationpan> (Table 3); tranpan>scripts for the rbcL anpan>d rbcS genpan>es, enpan>codinpan>g RuBisCO, were about three-fold lower anpan>d tranpan>scripts for genpan>es enpan>codinpan>g componpan>enpan>ts of the carboxysome were up to six-fold lower. Tranpan>script levels for two genpan>es (ndhD3, ndhF3) enpan>codinpan>g subunits of the inpan>ducible pan> class="Chemical">CO2-concentrating complex were lower in the dark (approximately three-fold), and even the gene coding for a transcriptional regulator (rbcR/ccmR; Woodger et al., 2007) showed a slightly lower transcript level. However, the mRNA levels encoding subunits of the constitutive CO2-concentrating complex remained more or less constant. These data illustrate that genes involved in CO2 fixation, the major electron sink, are regulated at the transcriptional level and that this regulation includes not only the essential central components but also those for the more peripheral, inducible CO2 uptake system.

The transcript levels for many genes involved in n class="Chemical">carbohydrate degradationpan> (e.g., glycolysis) inpan>creased inpan> cells after a 1-h dark treatmenpan>t (Table 3). Tranpan>scripts inpan>creased three-fold for pan> class="Chemical">pgi (glucose-6-phosphate isomerase) and pfkA (6-phosphofructokinase), four-fold for fbaB (fructose-bisphosphate aldolase class I), and five-fold for gap (SYNPCC7002_A2697; glyceraldehyde-3-phosphate dehydrogenase, type I). Some of the reactions of glycolysis or the oxidative pentose phosphate cycle and the Calvin–Benson–Bassham cycle are common to more than one of these pathways. Cyanobacteria usually have one enzyme that is used for the oxidative pentose phosphate cycle and another enzyme that is used for the Calvin–Benson–Bassham cycle (Knowles and Plaxton, 2003). The transcription data for genes encoding glyceraldehyde-3-phosphate dehydrogenase (gap), fructose-bisphosphate aldolase (fba), and fructose-1,6-bisphosphatase (fbp) reflect this phenomenon. Transcription of one set of gap and fba/fbp genes (gap/SYNPCC7002_A0106, fba/SYNPCC7002_A1352, glpX/SYNPCC7002_A1301) was regulated in the same way as the RuBisCO genes (i.e., increased transcript levels in high-light-treated cells and lower transcript levels in cells under dark oxic conditions). Transcript levels for the other set of genes was regulated oppositely: i.e., transcript levels for these genes (gap/SYNPCC7002_A2697, fbaB/SYNPCC7002_A0010, fbp/SYNPCC7002_A0329) were higher in cells after dark oxic treatment (or at least at about the same level in the case of SYNPCC7002_A0329) and lower in cells exposed to high light. Thus, although these biochemical pathways share certain reactions, the transcription data suggest that there are probably distinctive enzymes, which are differentially expressed, that change in response to light and/or the availability of reducing equivalents provided through the action of light.

The so-called light-repressed transcript (lrtA), which encodes the “light-repressed protein (Singer and Doolittle, 1974; Tan et al., 1994),” became extremely abundant after dark incubation. The lrtA transcripts increased nine-fold after dark, oxic treatment and increased to an even greater level under dark, n class="Chemical">fermenpan>tative conditions (Table 3). Conversely, the lrtA tranpan>scripts were about five-fold lower inpan> cells exposed to high light for 1 h. It has previously beenpan> demonstrated that lrtA tranpan>scripts are actively degraded after cells are exposed to light (Samartzidou anpan>d Widger, 1998).

Two genes in n class="Disease">pyruvate metabolism were amonpan>g those genpan>es for which tranpan>script levels inpan>creased the most uponpan> dark inpan>cubationpan> (see Figures 3B anpan>d 6). Tranpan>scripts for ppsA, enpan>codinpan>g pan> class="Chemical">phosphoenolpyruvate synthase increased 20-fold, and those for nifJ, encoding pyruvate:ferredoxin oxidoreductase increased 25-fold (Figure 6). Conversely, high-light treatment for 1 h caused transcripts for these two genes to decrease about three- and five-fold respectively. After cells were exposed to dark oxic conditions for 1 h, transcript levels for pyruvate kinase (pyk) and pyruvate dehydrogenase (pdhA, pdhB, ipdA, and SYNPCC7002_A0110) decreased approximately two-fold. However, the transcript levels for the genes encoding pyruvate kinase and pyruvate dehydrogenase did not change much (1.5-fold higher) after cells were exposed to high light for 1 h. Micro-oxic growth conditions (see below) likewise did not alter the transcript levels for genes involved in the pyruvate metabolism. The possible implications of these differences for pyruvate metabolism will be discussed below when describing transcription changes that occur during fermentative (dark anoxic) conditions.

Figure 6

Changes in the transcript levels for genes coding for proteins involved in pyruvate metabolism. Relative transcript levels for genes encoding proteins/enzymes involved in pyruvate metabolism [pyruvate (Pyr) kinase (pyk), phosphoenolpyruvate (PEP) synthase (ppsA), pyruvate dehydrogenase (Pyr-DH) complex (E1 alpha, E1 beta, E2 and E3 proteins; encoded by pdhA, pdhB, SYNPCC7002_A0110, ipdA), and pyruvate:ferredoxin oxidoreductase (Pyr:Fd OR; nifJ)] are shown. The proposed reaction pathways from PEP to acetyl-CoA with the respective enzymes are shown in (A,C); and the change in the mRNA level for the respective genes upon dark incubation is indicated by arrows. (B) The ratio of the transcript levels for cells after a 1-h dark oxic treatment (dark gray), a 1-h dark anoxic treatment (fermentative conditions; black), a 1-h high-light treatment (light gray), and photolithoautotrophic growth under micro-oxic conditions (medium gray) are compared to transcript levels in cells under standard conditions.

Changes in the transcript levels for genes coding for proteins involved in n class="Disease">pyruvate metabolism. Relative tranpan>script levels for genpan>es enpan>codinpan>g proteinpan>s/enpan>zymes inpan>volved inpan> pan> class="Disease">pyruvate metabolism [pyruvate (Pyr) kinase (pyk), phosphoenolpyruvate (PEP) synthase (ppsA), pyruvate dehydrogenase (Pyr-DH) complex (E1 alpha, E1 beta, E2 and E3 proteins; encoded by pdhA, pdhB, SYNPCC7002_A0110, ipdA), and pyruvate:ferredoxin oxidoreductase (Pyr:Fd OR; nifJ)] are shown. The proposed reaction pathways from PEP to acetyl-CoA with the respective enzymes are shown in (A,C); and the change in the mRNA level for the respective genes upon dark incubation is indicated by arrows. (B) The ratio of the transcript levels for cells after a 1-h dark oxic treatment (dark gray), a 1-h dark anoxic treatment (fermentative conditions; black), a 1-h high-light treatment (light gray), and photolithoautotrophic growth under micro-oxic conditions (medium gray) are compared to transcript levels in cells under standard conditions.

As noted above, high-light treatment caused transcript levels for pyridine nucleotide transn class="Chemical">hydrogenase subunits (pnpan>tA, pnpan>tB, pnpan>tC) to decrease sharply. Dark oxic inpan>cubationpan> had the opposite effect on the transcription of the transhydrogenase genes: mRNA levels increased approximately seven-fold for pntB and pntC and increased approximately four-fold for pntA. These transcriptional changes suggested that cells have a much higher requirement for NADH/NADPH electron exchange under dark respiratory conditions than in the light.

The mRNA level for SYNPCC7002_A1442, which encodes dihydroorotate den class="Chemical">hydrogenase, anpan> enpan>zyme inpan>volved inpan> pan> class="Chemical">pyrimidine metabolism, was 30-fold higher after a 1-h dark incubation; under dark fermentative conditions (see below) its transcript level was 50-fold higher than in standard conditions. The gene encoding this particular dihydroorotate dehydrogenase and nifJ are immediate neighbors, have the same orientation, and are separated by only 70 bp; these observations suggest that they maybe transcribed as an operon under a common regulatory mechanism. The transcript level for a second pyrD gene (SYNPCC7002_A2195) did not change much upon dark oxic treatment, but was about five-fold lower after 1 h under dark anoxic incubation when compared to cells grown under standard conditions. The transcript levels of both pyrD genes were similar to those of standard conditions when cells were grown under micro-oxic conditions (see below), suggesting that the regulation is not directly related to oxygen. The biochemical rationale for this exchange of PyrD proteins is not clear, but darkness is usually associated with lower oxygen levels for cyanobacteria. This could result in an indirect acclimation of pyrD transcription in response to changing oxygen levels or cellular redox potential.

Transcription of the genes coding for the two flavoproteins (SYn class="Chemical">NPCC7002_A1321 anpan>d SYpan> class="Chemical">NPCC7002_A1743) increased (three- and two-fold, respectively) after dark oxic incubation. A similar change was also observed upon dark anoxic incubation for SYNPCC7002_A1321 (three-fold higher), although no increase of the mRNA level for SYNPCC7002_A1743 was observed (see below). Growth under micro-oxic conditions (see below) resulted in a slight decrease (∼1.5-fold) in transcript levels for both genes, whereas almost no change was observed after high-light incubation. As noted above, these observations suggest that these flavoproteins might not exclusively function as catalysts to eliminate excess electrons in the light.

Transcriptional changes in cells under dark anoxic (fermentative) conditions

Because PS II cannot oxidize n class="Chemical">water anpan>d evolve pan> class="Chemical">O2 in the dark, O2 levels typically decrease in natural habitats of oxygenic photosynthetic organisms at night. The remaining O2 in these environments is often rapidly consumed by the respiratory activities of cyanobacteria and/or other microorganisms, which imposes dark anoxic (i.e., fermentative) conditions on the cyanobacteria. To simulate these conditions, cultures were placed in the dark and were sparged with 1% CO2 (v/v) in N2. Figure 3C shows a scatter plot in which the transcript level for each gene in cells exposed to dark anoxic conditions was plotted against the transcript level in cells grown under standard conditions. It is apparent that a larger number of genes exhibited greater than two-fold changes in transcription level than for dark oxic conditions (compare Figures 3B,C) and that many genes additionally showed transcript levels that were markedly higher or lower than this threshold. This is perhaps not surprising because two important environmental parameters, light and oxygen, were changed in this experiment. The transcript levels for genes encoding the two photosystems, electron transport proteins, phycobilisome components, and heme and chlorophyll biosynthesis were even lower in cells exposed to these fermentative conditions than in cells from dark oxic conditions (Table 3). A few exceptions, e.g., psbW2 and chlH (SYNPCC7002_A1000), had increased transcript levels under fermentative conditions compared to standard conditions. For each of these genes, there is a paralog (psbW/SYNPCC7002_A1258, chlH/SYNPCC7002_A1018) for which transcript levels decreased in cells exposed to fermentative conditions. Furthermore, transcripts for ho2 and acsF2 were much higher under dark anoxic conditions than under standard or dark oxic conditions (see further discussion concerning micro-oxic conditions below).

Transcripts for ppsA (n class="Chemical">phosphoenolpyruvate synpan>thetase) anpan>d nifJ (pan> class="Chemical">pyruvate:ferredoxin oxidoreductase) increased dramatically, 180- and 110-fold, respectively, under fermentative conditions. The latter value was in excellent agreement with results from quantitative-RT-PCR, which showed that nifJ transcripts increased 130 ± 23-fold when cells were incubated for 30-min under dark anoxic conditions (McNeely et al., 2010a; Xu, 2010). Conversely, transcripts for pyk (pyruvate kinase) and the pyruvate dehydrogenase complex (pdhA, pdhB, ipdA, and SYNPCC_A0110) were as much as 10-fold lower under dark anoxic conditions compared to standard conditions. Assuming that these very large transcriptional changes also reflect changes in enzyme activity levels in the cells, these results strongly suggested that the normal glycolytic pathway for phosphoenolpyruvate conversion to acetyl-coenzyme A (CoA) via pyruvate kinase and pyruvate dehydrogenase is replaced by an alternative pathway comprising phosphoenolpyruvate synthase and pyruvate:ferredoxin oxidoreductase. Transcription of nifJ has been reported under oxic conditions in several cyanobacteria (Schmitz et al., 2001); however, inactivation of pdhA in Synechococcus 7002 resulted in an acetate-requiring auxotrophic mutant strain (Xu, 2010). This result implies that pyruvate:ferredoxin oxidoreductase (NifJ) activity in cells grown under oxic conditions is insufficient to provide adequate acetyl-CoA to support autotrophic growth.

n class="Chemical">Phosphoenolpyruvate synpan>thase (PpsA) is the mainpan> enpan>zyme inpan>volved inpan> the conpan>versionpan> of pan> class="Chemical">phosphoenolpyruvate to pyruvate in a modified Embden–Meyerhof pathway that occurs in the archaeon Thermococcus kodakarensis, although both pyruvate kinase and phosphoenolpyruvate synthase are also present in this organism (Imanaka et al., 2006). The conversion of phosphoenolpyruvate into pyruvate by pyruvate kinase generates one ATP from ADP, whereas phosphoenolpyruvate synthase produces one ATP and pyruvate from phosphoenolpyruvate, AMP and phosphate (Imanaka et al., 2006). This means that phosphoenolpyruvate synthase conserves more energy than pyruvate kinase, which is very important under energy-limiting conditions such as fermentation. Pyruvate decarboxylation via the pyruvate dehydrogenase complex yields one equivalent each of NADH, acetyl-CoA and CO2. Pyruvate decarboxylation by the pyruvate:ferredoxin oxidoreductase also generates acetyl-CoA but instead of NADH produces two molecules of reduced ferredoxin, which have a much lower redox potential than the NADH/NAD+ couple (Tittmann, 2009). Reduced ferredoxin can be used directly to reduce nitrate, sulfite, and other substrates or can be used to produce NADPH via ferredoxin:NADP+ oxidoreductase. The availability of reducing equivalents at lower redox potentials should also benefit cells energetically when respiration is not possible. In summary, it appears quite likely that Synechococcus 7002 uses alternate pathways for the conversion of phosphoenolpyruvate into acetyl-CoA in the light and in the dark under fermentative conditions (Figure 6). The fermentative pathway might conserve more energy for cellular metabolism than the pathway that operates in the light or under dark oxic conditions.

Under dark anoxic conditions, no n class="Disease">oxidative phosphorylationpan> canpan> be performed because the terminpan>al electronpan> acceptor for the respiratory chainpan> (pan> class="Chemical">O2) is missing. These conditions caused dramatic changes in mRNA level for genes coding for the subunits of enzyme complexes involved in electron transport and ATP synthesis. The relative mRNA levels for the genes encoding the Type-1 NADH dehydrogenase complex, cytochrome oxidase and ATP synthase decreased more than 10-fold for some genes when compared to levels in cells under standard conditions (Table 3). Under fermentative conditions ATP synthesis for growth and cellular maintenance is produced by substrate-level phosphorylation. These reactions also typically produce reduced coenzymes or redox proteins, which must be regenerated by reduction of some substrate molecule. As for dark oxic conditions, transcripts for genes encoding components of the CO2 concentration and fixation pathways are lower than in cells grown under standard conditions. Reduction of nitrate and nitrite and excretion of ammonia could potentially provide another sink for electrons. However, transcript levels for narB (nitrate reductase) and nirA (nitrite reductase) were even lower in the dark fermentative sample (7- and 15-fold, respectively compared to levels in cells under standard conditions) than upon dark oxic incubation (seven- and three-fold, respectively, relative to levels in cells under standard conditions). Thus, nitrate reduction is probably not the major electron sink in absence of O2, although studies show that nitrate does compete with other oxidizing agents for electrons under these conditions (Gutthann et al., 2007; McNeely et al., 2010a,b).

Electrons produced during glycolysis can also be used to reduce protons to n class="Chemical">hydrogen. Like manpan>y other cyanpan>obacteria, pan> class="Species">Synechococcus 7002 produces a so-called bidirectional, NAD(P)H-oxidizing [NiFe]-hydrogenase (McNeely et al., 2010a,b; Xu, 2010). When cells were incubated under dark oxic conditions, transcripts for the genes encoding the hydrogenase increased three- to seven-fold. Interestingly, under dark fermentative conditions, transcripts for the three genes that encode the diaphorase moiety of the enzyme (hoxE, hoxF, hoxU) increased significantly, 2- to 16-fold, but the transcript levels for hoxH and hoxY, which encode the hydrogenase portion of the bidirectional hydrogenase, did not increase. Nevertheless, the reduction of protons to produce hydrogen does occur under these conditions (McNeely et al., 2010a,b), again indicating that post-transcriptional regulatory processes may be important. The diaphorase portion of the bidirectional hydrogenase has sequence similarities to the missing diaphorase subunits of the cyanobacterial Type-1 NADH dehydrogenase complex and has been suggested to connect the bidirectional [NiFe]-hydrogenase to the NADH dehydrogenase complex (Appel and Schulz, 1996; Schmitz and Bothe, 1996). However, it is currently not clear whether such a connection actually occurs, and it is furthermore not clear what advantage such a connection would provide to cells under fermentative conditions.

Interestingly, transcript levels for ldhA (SYn class="Chemical">NPCC7002_G0164, pan> class="Chemical">d-lactate dehydrogenase), acsA (SYNPCC7002_A1838, acetyl-CoA ligase), acs (SYNPCC7002_A2015, acetyl-CoA synthetase), and two putative adh genes (alcohol dehydrogenases, SYNPCC7002_A0868 and SYNPCC7002_A2590) were similar under all conditions tested in this study. Transcript levels for ldhA and the adh genes actually decreased somewhat under dark fermentative conditions. Because it is known that d-lactate is the major fermentation product produced by Synechococcus 7002 (McNeely et al., 2010b), these observations further establish that post-transcriptional regulation processes are important in establishing the fermentative capabilities of Synechococcus 7002.

Transcription changes in cells under micro-oxic conditions

In order to distinguish if transcription changes were resulting from the efn class="Chemical">fects of light or directly from the pan> class="Chemical">O2 levels, transcription profiling was performed with cells grown photoautotrophically under micro-oxic conditions. Micro-oxic conditions were achieved by sparging the culture with 1% (v/v) CO2 in N2 to continuously remove the O2 produced by PS II. Interestingly, transcript levels for this sample were similar to those of the standard samples, which were grown at ambient O2 [1% (v/v) CO2 in air; Figure 3D]. However, a few genes showed strongly increased mRNA levels. Specifically, transcripts for acsF2, ho2, hemN2, and desF were 15- to 20-fold higher than in cells grown under standard conditions (Table 3). These four genes are clustered (Figure 7) and could potentially be transcribed as an operon. However, under dark fermentative conditions only acsF2 and ho2 showed a strong up-regulation, whereas hemN2 and desF transcript levels were similar to those for cells in standard conditions. This could either mean that multiple promoters exist that are differentially used, that transcription terminates at different places dependent upon growth conditions, or that transcript segments have different stabilities under some growth conditions. Three of these genes (hemN2, acsF2, ho2) code for enzymes involved in heme, chlorophyll and phycocyanobilin biosynthesis, respectively. The Synechococcus 7002 genome harbors multiple genes for each of these enzymes: acsF (SYNPCC7002_A0707) and acsF2 (SYNPCC7002_A1992), encoding magnesium-protoporphyrin IX monomethyl ester oxidative cyclases; hox1 (SYNPCC7002_A2508) and ho2 (SYNPCC7002_A1991), encoding heme oxygenases that synthesize biliverdin; and hemF (SYNPCC7002_A1828), hemN (SYNPCC7002_A2831), and hemN2 (SYNPCC7002_A1990) encoding coproporphyrinogen III oxidases. In contrast to acsF2, ho2, and hemN2, which exhibited much higher mRNA levels under micro-oxic conditions, the transcript levels of acsF, hox1, and hemF were essentially the same as for cells in standard conditions. Dark incubation, on the other hand, resulted in lower mRNA levels for acsF, hox1, and hemF (two- to three-fold lower under dark oxic conditions and 10- to 20-fold lower under dark anoxic conditions). The transcript levels of hemN, however, were somewhat higher under both micro-oxic (less than two-fold) and dark incubation (both oxic and anoxic; 1.5- to 2-fold) compared to levels in cells grown under standard conditions.

Figure 7

Organization of the . The acsF2 gene cluster comprises four genes (acsF2, ho2, hemN2, and desF). There is an ORF encoding a putative transcriptional regulator (SYNPCC7002_A1993) in the region upstream from acsF2 but encoded on the opposite strand.

n class="Species">Synechocystis 6803 also has anpan> acsF2–hpan> class="Chemical">o2–hemN2 gene cluster encoded by ORFs sll1874, sll1875, sll1876. The expression and/or the activity of the respective enzymes has been reported to depend on oxygen levels (Minamizaki et al., 2008; Yilmaz et al., 2009; Goto et al., 2010). It has further been shown that the oxygen-dependent coproporphyrinogen III oxidase (HemF) is the major enzyme under ambient O2 levels, whereas one of the oxygen-independent HemN (sll1876) proteins is most active under micro-oxic conditions (Goto et al., 2010). The role of the second hemN (sll1917) gene product in Synechocystis 6803, with highest similarity (50% identity) to SYNPCC7002_A2831 (HemN) in Synechococcus 7002, is still unclear. In Synechocystis 6803 transcript levels of the clustered ORFs sll1874 (acsF2), sll1875 (ho2), and sll1876 (hemN2) coordinately increase under micro-oxic conditions (Minamizaki et al., 2008). Our data indicate that in Synechococcus 7002 orthologs of these three genes are possibly transcribed together under micro-oxic conditions with a fourth gene, desF, which encodes a acyl-lipid/fatty acid desaturase, a di-iron containing enzyme that uses O2 as a substrate.

Interestingly, transcripts for isiA, which encodes the n class="Chemical">iron-stress-inpan>duced, PS I-associated, pan> class="Chemical">chlorophyll a-binding IsiA protein, and isiB, which encodes flavodoxin, increased 15- and 30-fold, respectively, under micro-oxic conditions. Transcript levels for isiA and isiB, which are cotranscribed as an operon, greatly increase under Fe-limiting growth conditions (Leonhardt and Straus, 1992). This observation can be explained by lower Fe3+ availability under more reducing conditions than at ambient O2 levels. This suggestion is additionally supported by the fact that transcript levels for many genes coding for iron uptake systems also increased (up to 12-fold) under micro-oxic conditions compared to standard conditions (Table 3). Under both dark oxic and anoxic conditions, the isiA and isiB mRNA levels were lower compared to standard conditions, but genes encoding various iron uptake systems were also transcribed at levels similar to standard conditions, which suggested that a sufficient iron supply is available under these conditions. The requirement for iron is much lower in the dark, because synthesis of the photosynthetic apparatus, which consumes the majority of the iron taken up by cyanobacteria, is down-regulated in the dark (see above).

Transcriptional changes in a culture undergoing batch growth under standard conditions

To monitor changes in transcription during the transition from exponential growth to almost stationary phase, cells were harvested at difn class="Chemical">ferenpan>t cell denpan>sities (OD730 nm = 0.4, 1.0, 3.0, anpan>d 5.0) durinpan>g batch growth (see Figure 1; anpan>d Sakamoto anpan>d Bryanpan>t, 1998). Whenpan> the tranpan>scriptionpan> profiles for these span> class="Chemical">amples were compared to those for standard conditions (OD730 nm = 0.7), it was found that transcript levels were relatively constant throughout the exponential growth phase (Figure 8). At OD730 nm values of 0.4 and 1.0, transcription changes differed by at most ∼10-fold compared to transcript levels at OD730 nm = 0.7. However, the reliability for many of these differences was low, because they were associated with ORFs whose expression levels were low. As the cell density increased beyond OD730 nm = 1.0, more differences in transcript levels were observed. At OD730 nm = 3.0 the maximal transcript-level differences increased to ∼17-fold, and at OD730 nm = 5.0, the maximal differences were ∼20-fold. Differences of these magnitudes are likely to be highly significant, but these differences are generally smaller than many of the changes in response to light and oxygen.

Figure 8

Changes in the relative transcript abundance throughout standard batch growth. The scatter plots show the relative transcript abundances (A) at OD730 nm = 0.4, (B) at OD730 nm = 1.0, (C) at OD730 nm = 3.0, and (D) at OD730 nm = 5.0 compared to those for standard conditions (at OD730 nm = 0.7). The gray lines indicate two-fold changes in either direction; selected genes are identified by name/locus tag number.

Transcripts for the pstS gene, encoding the substrate-binding protein of the n class="Chemical">phosphate tranpan>sport system, inpan>creased about 20-fold at OD730 nm = 5.0 (Table 4). Tranpan>scripts for other genpan>es associated with this pan> class="Chemical">phosphate uptake system (pstA, pstB, pstC) also increased two- to four-fold. Furthermore, transcript levels for a predicted phosphatase (SYNPCC7002_A0893) and alkaline phosphatase (SYNPCC7002_A2352) increased about 10-fold. Transcript levels for all of these genes were nearly constant at lower cell densities. Collectively, the results at higher cell densities resembled microarray data that were obtained from Synechococcus sp. WH8102 after phosphate limitation (Tetu et al., 2009; Ostrowski et al., 2010).

Table 4

Prominent changes in transcript levels at different culture densities of a batch culture.

Locus tag	Ratio OD 0.4/std	Ratio OD 1.0/std	Ratio OD 3.0/std	Ratio OD 5.0/std	Gene name	Gene product
SYNPCC7002_A2284	1.85	1.40	1.42	18.26	pstS	Phosphate transport system substrate-binding protein
SYNPCC7002_A2286	0.87	1.23	1.13	2.55	pstA	Phosphate ABC transporter, permease protein
SYNPCC7002_A1895	0.90	0.87	0.73	2.09	pstB	Phosphate import ATP-binding protein
SYNPCC7002_A2285	1.49	1.18	1.07	4.21	pstC	Phosphate ABC transporter, permease protein
SYNPCC7002_A0893	0.60	0.61	1.23	11.20	–	Predicted phosphatase
SYNPCC7002_A2352	1.14	0.55	0.78	10.75	–	Alkaline phosphatase
SYNPCC7002_A1821	0.93	2.03	1.90	5.20	nblA	Putative phycobilisome degradation protein

The ratio of relative transcript abundance at OD.

std, standard conditions.

Prominent changes in transcript levels at difn class="Chemical">ferenpan>t culture denpan>sities of a batch culture.

As the cell density increased in the batch culture, transcript levels for nblA also increased. For exn class="Chemical">ample, at OD730 nm = 1.0 anpan>d 3.0, nblA tranpan>scripts were about two-fold higher thanpan> inpan> cells under stanpan>dard conpan>ditionpan>s, anpan>d at OD730 nm = 5.0, nblA tranpan>scripts were five-fold higher. pan> class="Chemical">NblA expression is associated with “chlorosis,” which occurs when phycobiliproteins are degraded by cyanobacterial cells in response to nutrient deprivation, high-light intensity and oxidative stress (Collier and Grossman, 1992, 1994; Bienert et al., 2006; Karradt et al., 2008). Nitrogen, sulfur, and carbon limitation generally produce the greatest increases in nblA expression (Collier and Grossman, 1994), but phosphate limitation can also cause chlorosis. For example, previous studies have shown that phosphate starvation was accompanied by extensive degradation of phycobiliproteins in Synechococcus sp. PCC 6301 and PCC 7002 (Batterton and van Baalen, 1968; Stevens et al., 1981). Because transcript levels for both nblA and phosphate acquisition increased in parallel at higher cell densities during batch growth, it is likely that cells growing in medium A+ in batch culture become phosphate-limited at OD730 nm values above ∼3.0. Interestingly, the transcript levels of genes involved in photosynthesis are not subject to major changes with increasing culture density (Table S2 in Supplementary Material), which is different from observations that have been made in Synechocystis 6803 (Foster et al., 2007).

Conclusion

The data presented in this study, which were obtained by deep sequencing of cDn class="Chemical">NA via SOLiD™ pan> class="Chemical">Next-Gen sequencing, identified mRNAs for nearly all annotated genes in the genome of Synechococcus 7002. The reproducibility of the method appears to be as good or better than microarrays, and the sequencing depth can be adapted to achieve virtually any desired dynamic range. The data clearly showed that cells significantly altered their transcription patterns for major metabolic processes (photosynthesis, CO2 fixation, sugar degradation, respiration) in response to changes in light and oxygen, and the observed changes were consistent with previous studies in this cyanobacterium and others. Greatly increased transcript levels for the genes encoding an alternative pathway for the conversion of phosphoenolpyruvate into acetyl-CoA suggests that this process occurs by completely different routes in Synechococcus 7002 during light/oxic and dark/anoxic conditions. Transcription changes for a putative acsF2–ho2–hemN2–desF operon suggest that the transcription of these genes may be directly regulated by oxygen concentration. If this proves to be the case, the transcription of these genes can be monitored as a reporter for attempts to modulate the intracellular oxygen concentrations in cells. Lowering the intracellular oxygen concentrations in cyanobacteria is a very important consideration in attempts to engineer cells for light-driven hydrogen production.

Conflict of Interest Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Supplementary Material

The Supplementary Material for this article can be found online at http://www.frontiersin.org/microbial_physiology_and_metabolism/abstract/9540