Comparative Targeted Proteomics of the Central Metabolism and Photosystems in SigE Mutant Strains of Synechocystis sp. PCC 6803

A targeted proteome analysis was conducted to investigate the SigE dependent-regulation of central metabolism in Synechocystis sp. PCC 6803 by directly comparing the protein abundance profiles among the wild type, a sigE deletion mutant (ΔsigE), and a sigE over-expression (sigEox) strains. Expression levels of 112 target proteins, including the central metabolism related-enzymes and the subunits of the photosystems, were determined by quantifying the tryptic peptides in the multiple reaction monitoring (MRM) mode of liquid-chromatography–triple quadrupole mass spectrometry (LC–MS/MS). Comparison with gene-expression data showed that although the abundance of Gnd protein was closely correlated with that of gnd mRNA, there were poor correlations for GdhA/gdhA and glycogen degradation-related genes such as GlgX/glgX and GlgP/glgP pairs. These results suggested that the regulation of protein translation and degradation played a role in regulating protein abundance. The protein abundance profile suggested that SigE overexpression reduced the proteins involved in photosynthesis and increased GdhA abundance, which is involved in the nitrogen assimilation pathway using NADPH. The results obtained in this study successfully demonstrated that targeted proteome analysis enables direct comparison of the abundance of central metabolism- and photosystem-related proteins.


Introduction
As a photosynthetic organism, cyanobacteria can be used as a host organism for the direct production of bioenergy and chemicals from solar energy and CO 2 [1][2][3][4]. A model cyanobacteria, Synechocystis sp. PCC 6803, can adapt metabolically to environmental perturbations by modulating protein expression levels [5,6]. A sigma factor, SigE, is a global regulator of central carbon metabolism, whose expression is induced under nitrogen-starved conditions in an NtcA-dependent manner [7][8][9][10][11]. The expression of SigE is also induced at the end of day time or 12 h after light exposure, suggesting that SigE is responsible for a metabolic regulation required to survive in dark conditions [12].
For a detailed investigation of SigE functions, Synechocystis sp. PCC 6803 strains lacking (∆sigE) and over-expressing (sigEox) the sigE gene have been constructed in previous studies [10,13].  For the targeted proteome analysis, crude protein extracts were obtained from cells collected at mid-log phase (OD 730 = 0.4-0.7). Following reductive alkylation and digestion by trypsin, the tryptic peptide samples were served for the nano liquid-chromatography-triple quadrupole mass spectrometry (LC-MS/MS) analysis. Here, we employed a series of MRM assay methods with 3065 channels to analyze 686 tryptic peptides derived from 144 target proteins including enzyme and subunit proteins responsible for central metabolism and photosynthetic apparatus (Table S1). In order to compare protein levels precisely, fully 15 N-labeled tryptic peptide samples were prepared from WT, ∆sigE, and sigEox strains grown in modified BG11 medium containing 15 NH 4 Cl as the nitrogen source. The fully 15 N-labeled tryptic peptide samples were used as internal standards. The tryptic peptides, whose signals were commonly observed among 3 strains in the targeted proteomics data with the largest signal-to-noise ratios, were employed for protein quantification (Table S2). The MRM assays in this study successfully determined the levels of tryptic peptides derived from the 112 proteins (Table S2, Figures S1-S5). Because the signal change of multiple peptides constituting one protein were not significantly different, one of these peptides was selected as a representative for the quantification of protein ( Figures S1-S4). Figures 2 and 3 show the heat map representations of the fold change in the relative abundance of central metabolic enzymes ( Figure 2) and photosynthetic proteins ( Figure 3) in ∆sigE (∆sigE/WT) and sigEox (sigEox/WT) compared to those in WT. The enzyme abundance profiles were compared by volcano plots using thresholds with fold of change (FC) > 1.5, FC < 0.667 and p value < 0.05 ( Figure 4). The comparison of the protein abundance levels of 112 target proteins between WT and ∆sigE showed that the abundances of four proteins decreased in ∆sigE including Fda, Gnd, TalB, and GltA ( Figure 4A).
Although it was not statistically significant, the abundance of CbbA increased in ∆sigE (Figure 2). Although GltA is an enzyme for citrate synthase in the tricarboxylic acid (TCA) cycle, four proteins including CbbA, Fda, Gnd, and TalB were commonly responsible for the Calvin-Benson cycle and the OxPPP.
The protein abundance profile data showed that two enzymes in the pentose phosphate pathway (PPP), Gnd and TalB, were upregulated in sigEox, whereas abundances of these enzymes were decreased in ∆sigE (Figure 2, Figure 4). These results suggested that the expression of Gnd and TalB are positively regulated by SigE. Furthermore, the relative abundances of Fda (Slr0943, fructose-1,6-bisphosphate aldolase, class I) levels decreased in both the ∆sigE and sigEox strains while the expression of another aldolase, CbbA (Sll0018, fructose-1,6-bisphosphate aldolase class II), was negatively regulated by SigE (Figure 2, Figure 4). These results suggested that two aldolases, Fda and CbbA, had distinct roles in regulating the Calvin-Benson cycle. Indeed, it has been reported that sedoheptulose-1,7-bisphosphate (SBP) in addition to fructose-1,6-bisphosphate (FBP) could be a substrate of both Fda and CbbA. However, the SBP/FBP activity ratio of the CbbA was two times higher than that of the Fda [30]. The up-and down-regulation of proteins, however, did not affect the normal cell metabolic functions, as the three strains showed similar growth curves ( Figure 1).

Comparison with the Gene-Expression Data
The relative protein abundance levels determined by the targeted proteome analysis in this study were compared with the gene-expression data obtained by the transcriptome analysis

Comparison with the Gene-Expression Data
The relative protein abundance levels determined by the targeted proteome analysis in this study were compared with the gene-expression data obtained by the transcriptome analysis

Comparison with the Gene-Expression Data
The relative protein abundance levels determined by the targeted proteome analysis in this study were compared with the gene-expression data obtained by the transcriptome analysis performed in a previous study ( Figure 5). As shown in the above section, the abundances of Gnd protein in ∆sigE and sigEox strains were significantly lower and higher than that of WT, respectively. Essentially identical patterns were observed for gnd mRNA expression levels determined by the previous microarray analysis ( Figure 5). The similarities between protein abundance and gene-expression patterns were also observed for Zwf/zwf and TalB/talB in the PPP, and CbbA/cbbA, and Gap1/gap1 in glycolysis ( Figure 5). For instance, protein abundances of GlnA and GdhA in the nitrogen assimilation pathway were different from that of the corresponding genes ( Figure 5). Furthermore, the microarray analyses in the previous studies pointed out that the expression levels of glycogen degradation-related genes such as glgX (slr0237) and glgP (sll1356) were significantly changed by the regulation of sigE. However, this regulation was not observed in protein levels as the relative protein abundances of GlgP and GlgX were different from those of gene expressions ( Figure 5). Although this discrepancy may be derived from differences in culture conditions (shaking flask in this study vs. bubbling flask in the previous study), results also suggested that the regulation of protein translation and degradation played a role in regulating protein abundance. performed in a previous study ( Figure 5). As shown in the above section, the abundances of Gnd protein in ΔsigE and sigEox strains were significantly lower and higher than that of WT, respectively. Essentially identical patterns were observed for gnd mRNA expression levels determined by the previous microarray analysis ( Figure 5). The similarities between protein abundance and geneexpression patterns were also observed for Zwf/zwf and TalB/talB in the PPP, and CbbA/cbbA, and Gap1/gap1 in glycolysis ( Figure 5). For instance, protein abundances of GlnA and GdhA in the nitrogen assimilation pathway were different from that of the corresponding genes ( Figure 5). Furthermore, the microarray analyses in the previous studies pointed out that the expression levels of glycogen degradation-related genes such as glgX (slr0237) and glgP (sll1356) were significantly changed by the regulation of sigE. However, this regulation was not observed in protein levels as the relative protein abundances of GlgP and GlgX were different from those of gene expressions ( Figure  5). Although this discrepancy may be derived from differences in culture conditions (shaking flask in this study vs. bubbling flask in the previous study), results also suggested that the regulation of protein translation and degradation played a role in regulating protein abundance. and sigEox strains. The gene expression data obtained by the microarray analysis was obtained from a previous study [13]. Values are means ± standard deviation (SD) of three independent experiments. Asterisks indicate a significant difference assessed with two-sided Student's t-test with an α level of 0.05.

Effect of sigE Deletion and Overexpression on NADPH/NADP + Ratio
The targeted proteome analysis revealed that among the investigated proteins, the abundance of proteins in the PPP, such as Gnd and TalB, were most significantly and directly affected by SigE deletion and overexpression (Figure 2, Figure 4). The direct measurement of protein abundance confirmed that the activation and inactivation of the PPP via Gnd and TalB expression was a key . Comparison between gene expression and protein abundance in the wild-type (WT), ∆sigE, and sigEox strains. The gene expression data obtained by the microarray analysis was obtained from a previous study [13]. Values are means ± standard deviation (SD) of three independent experiments. Asterisks indicate a significant difference assessed with two-sided Student's t-test with an α level of 0.05.

Effect of sigE Deletion and Overexpression on NADPH/NADP + Ratio
The targeted proteome analysis revealed that among the investigated proteins, the abundance of proteins in the PPP, such as Gnd and TalB, were most significantly and directly affected by SigE deletion and overexpression (Figure 2, Figure 4). The direct measurement of protein abundance confirmed that the activation and inactivation of the PPP via Gnd and TalB expression was a key mechanism in the SigE-dependent metabolic regulation in Synechocystis sp. PCC 6803.
Although TalB (transaldolase) is a part of the Calvin-Benson cycle for carbon fixation, Gnd (gluconate-6-phosphate dehydratase) is responsible for NADPH regeneration in the OxPPP. Therefore, the SigE-dependent regulation of Gnd level could affect the NADPH regeneration rate. The NADPH/NADP + assay shown in Figure 6A showed that the NADPH/NADP + ratio in sigEox was 1.33 times larger than that of ∆sigE, and that the NADPH/NADP + ratios in ∆sigE and sigEox tended to decrease and increase compared to that of wild type, respectively. These results revealed that the abundance of Gnd in the OxPPP was significantly under the control of SigE and that changes in Gnd abundance should affect the NADPH balance via the OxPPP. It has been demonstrated that the metabolic flux level in the OxPPP is up-regulated by treatment with the PSII inhibitor, DCMU [3-(3,4-dichlorophenyl)-1,1-dimethylurea]. In photoheterotrophic conditions, cyanobacteria utilize glucose or glycogen as carbon sources and survive by reproducing NADPH through the OxPPP [31,32].
or glycogen as carbon sources and survive by reproducing NADPH through the OxPPP [31,32].

SigE-Dependent Regulation of the Nitrogen Assimilation Pathway
In cyanobacteria, ammonium is incorporated into 2-oxoglutarate (2-OG) via the glutamine synthetase (GS, GlnA) and glutamate synthase (GOGAT, GltB) cycle, known as the GS-GOGAT pathway [33]. The GS-GOGAT cycle is regulated by the global nitrogen assimilation regulator NtcA. Alternatively, ammonium can be incorporated directly into glutamate by NADP-dependent glutamate dehydrogenase (GdhA), in a less efficient but less energy-consuming manner [34]. The targeted proteome analysis performed in this study showed that the abundance of GdhA was significantly changed in ΔsigE strains (Figure 2, Figure 4). The increase in GdhA abundances in sigEox strains should associate with the regulation of the OxPPP as GdhA consumes NADPH for the nitrogen assimilation in Synechocystis sp. PCC 6803 [35]. However, increased expression of the gdhA gene was not found in the previous transcriptome analysis (Figure 5), suggesting that the regulation of GdhA abundance was sensitive to culture conditions and that protein translation and degradation play a role.
In order to confirm the effect of sigE deletion and over-expression on nitrogen assimilation, WT, sigEox, and ΔsigE strains were cultivated again under the same conditions to determine concentrations of remaining ammonium in the medium ( Figure 6B). The analysis showed that the ammonium concentration of ΔsigE at 48 h was 4.3 mM, which was 1.11 times larger than that of WT ( Figure 6B). On the other hand, the concentration of remaining ammonium in the sigEox strain was similar to that of WT. Similar results were also observed at 144 h ( Figure 6B). Those results suggested that deleting SigE could reduce nitrogen assimilation, probably by decreasing the GS-GOGAT pathway genes such as GdhA abundance and NADPH supply from the oxidative pentose phosphate pathway.

SigE-Dependent Regulation of the Nitrogen Assimilation Pathway
In cyanobacteria, ammonium is incorporated into 2-oxoglutarate (2-OG) via the glutamine synthetase (GS, GlnA) and glutamate synthase (GOGAT, GltB) cycle, known as the GS-GOGAT pathway [33]. The GS-GOGAT cycle is regulated by the global nitrogen assimilation regulator NtcA. Alternatively, ammonium can be incorporated directly into glutamate by NADP-dependent glutamate dehydrogenase (GdhA), in a less efficient but less energy-consuming manner [34]. The targeted proteome analysis performed in this study showed that the abundance of GdhA was significantly changed in ∆sigE strains (Figure 2, Figure 4). The increase in GdhA abundances in sigEox strains should associate with the regulation of the OxPPP as GdhA consumes NADPH for the nitrogen assimilation in Synechocystis sp. PCC 6803 [35]. However, increased expression of the gdhA gene was not found in the previous transcriptome analysis (Figure 5), suggesting that the regulation of GdhA abundance was sensitive to culture conditions and that protein translation and degradation play a role.
In order to confirm the effect of sigE deletion and over-expression on nitrogen assimilation, WT, sigEox, and ∆sigE strains were cultivated again under the same conditions to determine concentrations of remaining ammonium in the medium ( Figure 6B). The analysis showed that the ammonium concentration of ∆sigE at 48 h was 4.3 mM, which was 1.11 times larger than that of WT ( Figure 6B). On the other hand, the concentration of remaining ammonium in the sigEox strain was similar to that of WT. Similar results were also observed at 144 h ( Figure 6B). Those results suggested that deleting SigE could reduce nitrogen assimilation, probably by decreasing the GS-GOGAT pathway genes such as GdhA abundance and NADPH supply from the oxidative pentose phosphate pathway.

Protein Abundance Profiles of the Photosynthetic Apparatus
It was expected that the assimilated nitrogen was mainly used for the biosynthesis of proteins such as the photosynthetic apparatus. The volcano plot of the targeted proteome data (Figure 4) showed that the abundance of PsbO, RubA, AtpA and PetE was significantly decreased in the sigEox strains. PsbO is a manganese-stabilizing polypeptide in PSII that binds to a putative Mn-binding protein and keeps 2 of the 4 Mn-atoms. The absence of PsbO in Synechocystis affects the coordination of photosynthesis/respiration [36]. RubA is an iron-sulfur protein (rubredoxin) responsible for the assembly of PSI [37]. AtpA is the alpha subunit of ATP synthase. PetE is a plastocyanin that participates in electron transfer between P700 and the cytochrome b6-f complex in PSI [38]. The abundances of antenna proteins such as ApcA, ApcB and ApcC tended to decrease in the sigEox strain although this effect was not statistically significant.
The comparison of ultraviolet-visible (UV-VIS) spectra among the WT, sigEox, and ∆sigE strains revealed that the absorbance spectrum of ∆sigE was essentially similar to that of WT ( Figure 6C). The result was consistent with the targeted proteome data as the abundance of the photosystem-related proteins was not significantly changed in ∆sigE. On the other hand, the absorbance of the carotenoid, the phycocyanin and the chlorophyll in the photosynthetic apparatus (about 550 nm, 630 nm and 680 nm, respectively) was reduced in sigEox, suggesting that photosynthetic apparatus construction was disturbed by SigE overexpression. In particular, the significant decrease of the carotenoid is consistent with the previous studies in which photosynthesis activity was decreased under high light conditions and the activity of non-photochemical quenching (NPQ) was down-regulated in the sigEox strain [15,39]. The decrease of the carotenoid may be related to the fact that metabolic flow was directed to synthesis of the polyhydroxybutyrate (PHB) [15]. However, the photosynthesis activity was maintained under usual light [39]. Accordingly, NADPH/NADP + ratio in sigEox strain was not significantly different from WT strain ( Figure 6A). In sigEox strain, PPP and gdhA expression were much higher and photosystems expression were lower than in WT, but NADPH/ NADP + ratio, nitrogen assimilation and growth rate were not different. These results indicate that NADPH from PPP reduces the need from photosynthesis.
The down-regulation of photosynthetic apparatus could be explained in the context of the C/N balance [40]. As the photosynthetic apparatus are the most abundant protein complexes in cyanobacteria, it has been considered that antenna proteins could be a nitrogen sink in these species. For instance, phycobilisome and the associated linkers are degraded to supply nitrogen under conditions of N starvation [41]. It was reported that degradation of the nitrogen-rich phycobilisomes starts 12 h after nitrogen starvation [42], and also that reduced CO 2 fixation leads to the down-regulation of genes encoding proteins involved in nitrogen assimilation [43]. The targeted proteome analysis conducted in this study suggested that sigE globally regulated the metabolism to increase nitrogen assimilation in sigEox. These results suggested that the overexpression of sigE mimicked the nitrogen-limiting condition in Synechocystis cell. This is consistent with in the previous studies, in which sigE plays a role in recovery from nitrogen deprivation [44][45][46]. Decreases in the antenna protein would be compensated by improved light reaction efficiency in the photosystem.

Bacterial Strains and Culture Conditions
Synechocystis sp. PCC 6803 GT strain, isolated by Williams [47], the ∆sigE strain [10] in which the sigE (sll1689) gene in the genome was disrupted, and the sigEox strain [13] which constitutively expresses the gene were used in this study. These strains were grown in modified BG11 medium, containing 5 mM NH 4 Cl as a nitrogen source. Cells were grown in 100 mL of medium in 500 mL Erlenmeyer flasks for batch culture under photoautotrophic conditions under continuous light (about 40 µmol m −2 s −1 ) and 34 • C, and cells in the linear growth phase (OD 730 = 0.4-0.7) [48] were collected for analysis. Kanamycin (10 µg mL −1 ) was added in the precultures of ∆sigE and sigEox.

Sample Preparation for Proteome Analysis
Total proteins were extracted as described in Picotti et al. [49]. Cell-culture medium containing cells in the logarithmic growth phase (100 mL, OD 730 = 0.4-0.7) was collected by centrifugation (5000× g, 4 • C, 5 min). Pellets were resuspended in 1 mL of lysis buffer [50 mM HEPES, 15% Glycerol, 15 mM DTT (Dithiothreitol), 100 mM KCl, 5 mM EDTA (Ethylenediamine-N,N,N ,N -tetraacetic acid, disodium salt, dihydrate), and one cOmplete protease inhibitors cocktail (Roche, Basel, Switzerland)]. The suspension was transferred to an Eppendorf tube containing zirconia beads (0.6 and 6 mm beads) and disrupted with a Beads Crusher µT-12 (TAITEC, Saitama, Japan) (3000 min −1 , 6 min). The resulting solution was centrifuged (15,000 rpm, 4 • C, 5 min), and the supernatant was transferred to a proteomics Eppendorf tube (protein low-adsorption tube) to obtain protein extraction samples. The protein concentration of the extracted sample was measured by the Bradford method, and the total protein amount was adjusted to 50 µg. Denatured buffer (500 mM Tris-HCl, 10 mM EDTA, 7 M Guanidine HCl) was added to the adjusted sample to a total volume of 220 µL.

Reduction and Alkylation/Methanol Chloroform Precipitation
One microliter of 50 mg mL −1 DTT was added and shaken at room temperature for 1 h using a tube mixer (CM-1000 Cute Mixer, EYELA, Tokyo, Japan), then 2.5 µL of 50 mg mL −1 iodoacetamide (IAA) was added and shaken for 1 h to reduce/alkylate the proteins. Next, the proteins were purified by methanol/chloroform precipitation. Proteins were purified as described in Wessel and Flügge [50]. Cold methanol (600 µL) was added to the sample solution and mixed by inversion, then 150 µL of cold chloroform was added and mixed by inversion. Cold milli-Q water (450 µL) was added and mixed by inversion, followed by centrifugation (15,000 rpm, 4 • C, 5 min). The upper layer was removed, and 450 mL of cold methanol was added and mixed gently by inversion. After centrifuging (15,000 rpm, 4 • C, 5 min) using a swing rotor, the supernatant was removed then additional centrifugation (15,000 rpm, 4 • C, 1 min) was performed to completely remove the supernatant.

Trypsin/LysC Digestion
Trypsin hydrolyzed the ester bonds on the carboxyl side of Arg and Lys and LysC on the carboxyl side of Lys. With trypsin alone, Lys which can not be leaved remains, but by combining LysC, the overall degradation efficiency and reproducibility are improved. Trypsin/LysC digestion was performed as described previously [26]. To the supernatant obtained from the above steps, 9 µL of 6 M urea was added and shaken for about 10 min at room temperature using a tube mixer. Thirty-six microliters of 0.1 M Tris-HCl (pH 8.5) was added and ultrasonic treatment and standing on ice were repeated twice for 30 s using an ultrasonic washer (Branson 2510, Danbury, CT, USA) to resuspend the protein precipitate. One microliter of 0.5 mg mL −1 LysC solution and 2.5 µL of 1% Protease Max solution ware added and mixed by tapping, then incubated at 25 • C for 3 h. One microliter of 0.5 mg mL −1 trypsin solution was added and mixed by tapping, then incubated at 37 • C for 16 h.

Desalting Samples
Sample desalination was performed as described previously [51][52][53]. Milli-Q water (7.5 µL) and 3 µL of 50% formic acid aqueous solution were added to the trypsin digestion product, and the mixture was stirred with a vortex mixer and centrifuged (15,000 rpm, 4 • C, 5 min). Fifty-eight microliters of the supernatant were obtained as a result. In order to carry out relative quantitative analysis, 15 N samples, obtained by culturing Synechocystis sp. PCC 6803 in BG11 medium whose nitrogen source was replaced with 15 NH 4 Cl, and 14 N samples were mixed so that the protein amount became 1:1 to prepare an analytical sample. The samples were diluted five times by Reagent A (5% acetonitrile, 0.1% formic acid). The diluted samples were desalted by handmade Stage-tip (3M Empore disk C18). The column was equilibrated with the same bed volume of Reagent B (80% acetonitrile, 0.1% formic acid). Then the column was washed with reagent A by centrifugation (4000× g, 10 min, RT). The samples corresponding to 5.8 µg were loaded onto the column, followed by centrifugation (4000× g, 10 min, RT). The column was washed by a bed volume of reagent A by centrifugation (4000× g, 10 min, RT) twice. The same volume of reagent B was added and centrifuged (4000× g, 10 min, RT), and the prepared sample was dried and solidified with a centrifugal concentrator, and then dissolved with 36 µL of 0.1% formic acid before nano LC-MS/MS analysis.

Design of Multiple Reaction Monitoring (MRM) Assay
First, the target 144 proteins related to the central metabolic pathway and photosynthetic apparatus were selected from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://www.genome.jp/kegg/kegg2.html). The amino acid sequences of target proteins were obtained from Cyanobase (http://genome.microbedb.jp/CyanoBase). The MRM method used to quantify these 144 proteins was created by the open software Skyline version 2.6 [54]. Each protein was subjected to a tryptic peptide filter of 8 to 25 residues and five y-fragments (y1 to y5) were selected for each peptide. Samples of Synechocystis sp. PCC 6803 were analyzed once by nano LC-MS/MS (LCMS-8060, Shimadzu, Kyoto, Japan) by the provisional MRM method. From the results obtained by the analysis, peak picking was performed based on the shape, coelution, and intensity of the peak, and the best transitions up to 5 were selected, respectively. For proteins with no suitable tryptic peptides and transitions, transitions were quantified for all y fragments and b fragments, from which the tryptic peptides suitable for quantitation were selected again to create the final MRM method.

Quantification of NH 3 in the Medium
The extracellular NH 3 assay was performed by F-kit (JK International) according to the manufacturer's protocol. Briefly, the medium was centrifuged (15,000 rpm, 4 • C, 1 min) and the supernatant was obtained, followed by heating at 80 • C for 20 min to inactivate remaining enzyme. Reagent Mix (200 µL) was reacted with 6.7 µL of the sample and incubated. A 340 was measured after 5 and 20 min.

Ultraviolet-Visible (UV-VIS) Spectrum
A DU 8000 was used to measure the UV-VIS spectrum. Absorption spectra of cell suspensions were measured according to the 'opal glass method,' with a translucent cuvette placed in front of the detector to minimize the effect of light scattering [56]. The results obtained were normalized to the absorbance at 730 nm of chlorophyll as 1.0.

Conclusions
Targeted proteome analysis was conducted in this study to directly compare the abundance of the central metabolism-and the photosystem-related proteins among WT, ∆sigE, and sigEox strains of Synechocystis sp. PCC 6803. The analysis showed the SigE-dependent regulation of central metabolism at the protein abundance level. Among the investigated proteins, the most direct or tight regulation via gene expression was observed for the proteins in the pentose phosphate pathway such as TalB and Gnd. Further investigations of protein abundance and their modification states such as phosphorylation by proteomic analysis should uncover the detailed regulatory mechanism of central carbon metabolism in cyanobacteria.

Supplementary Materials:
The following are available online, Figure S1: Selected reaction monitoring (SRM) spectra of the peptide EVTASLVGADAGK (ApcB), Figure S2: SRM spectra of the peptide SYFASGELR (ApcB), Figure S3: SRM spectra of the peptide AVLPQNLTQAQR (Gnd), Figure S4: SRM spectra of the peptide ELEPILTK (Gnd), Figure S5: SRM spectra of the peptide VPATIEEIAAR (Eno), Figure S6: Comparison with the relative abundances of mRNAs and proteins, Table S1: MRM assay method used in this study, Table S2: Targeted proteomics data; Table S3: All analyzed peptides.