Biochemical elucidation of citrate accumulation in Synechocystis sp. PCC 6803 via kinetic analysis of aconitase

A unicellular cyanobacterium Synechocystis sp. PCC 6803 possesses a unique tricarboxylic acid (TCA) cycle, wherein the intracellular citrate levels are approximately 1.5–10 times higher than the levels of other TCA cycle metabolite. Aconitase catalyses the reversible isomerisation of citrate and isocitrate. Herein, we biochemically analysed Synechocystis sp. PCC 6803 aconitase (SyAcnB), using citrate and isocitrate as the substrates. We observed that the activity of SyAcnB for citrate was highest at pH 7.7 and 45 °C and for isocitrate at pH 8.0 and 53 °C. The Km value of SyAcnB for citrate was higher than that for isocitrate under the same conditions. The Km value of SyAcnB for isocitrate was 3.6-fold higher than the reported Km values of isocitrate dehydrogenase for isocitrate. Therefore, we suggest that citrate accumulation depends on the enzyme kinetics of SyAcnB, and 2-oxoglutarate production depends on the chemical equilibrium in this cyanobacterium.

Kinetic parameters of SyAcnB. The activity of SyAcnB for citrate was the highest at pH 7.7 and temperature 45-55 °C (Fig. 2a), and that for isocitrate as the substrate was the highest at pH 8.0 and a temperature of 53 °C (Fig. 2b). Thereafter, the activities of SyAcnB for citrate were measured at pH 7.7 and 45 °C and for isocitrate at pH 8.0 and 53 °C except where indicated.
The kinetic parameters of SyAcnB, using citrate and isocitrate as the substrates, were estimated from the saturation curves (Fig. 3a,b). The V max , k cat , and k cat /K m values of the activity of SyAcnB for citrate were 4.58 ± 0.07 unit/mg, 9.12 ± 0.14 s −1 , and 8.11 ± 0.23 s −1 mM −1 , respectively ( Table 2). The V max , k cat , and k cat /K m values of the activity of SyAcnB for isocitrate were 8.36 ± 0.17 unit/mg, 16.67 ± 0.34 s −1 , and 10.88 ± 0.93 s −1 mM −1 , respectively ( Table 2). The K m values of SyAcnB for citrate and isocitrate were 1.13 ± 0.04 and 1.54 ± 0.17 mM, respectively (Table 3).
Kinetic parameters were calculated under optimal conditions for both substrates. Therefore, we calculated the parameters by unifying the measurement conditions and plotting a substrate saturation curve at 30 °C, which is the optimal temperature for the growth of Synechocystis 6803 23 (Fig. 4a,b). In the presence of Tris-HCl (pH 7.0) at 30 °C, the V max and k cat /K m values of SyAcnB for citrate were 3.65 ± 0.19 unit/mg and 10.68 ± 0.76 s −1 mM −1 , respectively ( Table 2). The V max and k cat /K m values for isocitrate were 3.19 ± 0.18 unit/ mg and 30.17 ± 1.51 s −1 mM −1 respectively and 0.87-and 2.8-fold higher than those for citrate, respectively ( Table 2). In the presence of Tris-HCl (pH 8.0) at 30 °C, the V max and k cat /K m values of SyAcnB for citrate were 3.51 ± 0.21 unit/mg and 8.79 ± 0.27 s −1 mM −1 , respectively ( Table 2). The V max and k cat /K m values for isocitrate were 4.52 ± 0.31 unit/mg and 19.13 ± 1.12 s −1 mM −1 respectively and 1.3-and 2.2-fold higher than those for citrate, respectively (Table 2). Finally, in the presence of Tris-HCl (pH 9.0) at 30 °C, the V max and k cat /K m values of SyAcnB for citrate were 2.22 ± 0.09 unit/mg and 1.95 ± 0.22 s −1 mM −1 , respectively ( Table 2). The V max and k cat /K m values for isocitrate were 2.61 ± 0.11 unit/mg and 3.18 ± 0.47 s −1 mM −1 respectively and 1.2-and 1.6-fold higher than those for citrate, respectively ( Table 2). The K m values of the activity of SyAcnB for citrate at 30 °C were 0.68 ± 0.02 mM, 0.80 ± 0.03 mM, and 2.28 ± 0.16 mM at pH 7.0, 8.0, and 9.0, respectively, and the values for citrate were 3.2-, 1.7-, and 1.4-fold higher than those calculated for isocitrate at pH 7.0, 8.0, and 9.0, respectively (Table 3). Since there were some points where the correlation coefficient (R 2 value) was low at 30 °C, the same measurement was performed at 45 °C (Fig. 4c,d). In the presence of Tris-HCl (pH 9.0) at 45 °C, the V max , K m and k cat /K m values of SyAcnB for citrate were 2.88 ± 0.17 unit/mg, 1.58 ± 0.27 mM and 3.69 ± 0.61 s −1 mM −1 , respectively, and the V max , K m and k cat /K m values for isocitrate were 4.64 ± 0.60 unit/mg, 3.79 ± 1.18 mM and 2.54 ± 0.50 s −1 mM −1 respectively (Table 4). All K m values are summarised in Table S2. The results of adding the peptide AcnSP (aconitase small protein) showed that the V max , K m , and k cat /K m values of SyAcnB for citrate were 4.29 ± 0.08 unit/mg, 0.74 ± 0.09 mM, and 11.65 ± 1.24 s −1 mM −1 , respectively, and the V max , K m , and k cat /K m values of SyAcnB for isocitrate were 6.59 ± 0.08 unit/mg, 0.91 ± 0.05 mM, and 14.50 ± 0.72 s −1 mM −1 , respectively (Fig. S1). For both citrate and isocitrate, the addition of peptide AcnSP decreased the V max and K m values and increased the k cat /K m value.
The activity of SyAcnB in the presence of other TCA metabolites and cations. We examined the effects of various metabolites on SyAcnB activity. The concentrations of the substrates used were the K m values determined for each substrate. In the presence of 5 mM pyruvate, 2-OG, and l-aspartate, the activity of SyAcnB for citrate decreased to 69%, 72%, and 84% of that of the control, respectively (Fig. 5a). Additionally, in the presence of 5 mM pyruvate, 2-OG, l-glutamine, l-glutamate, and l-aspartate, the activity of SyAcnB for isocitrate decreased to 78%, 74%, 81%, 85%, and 89% of that of the control, respectively (Fig. 5b). The kinetic parameters of SyAcnB in the absence ( Table 2, 3) and the presence of 2-OG under optimal conditions (Fig. S2, S3) were compared. When citrate was used as a substrate, the addition of 1 mM 2-OG did not change the V max , K m , and k cat /K m values (Fig. S2a), but the addition of 5 mM 2-OG increased the K m value and decreased the k cat /K m value (Fig. S3a). When isocitrate was used as a substrate, the addition of 1 mM 2-OG decreased the V max and K m values and increased k cat /K m values (Fig. S2b), but the addition of 5 mM 2-OG decreased only the V max value (Fig. S3b).  Table 2, respectively. There was no effect on the SyAcnB activity irrespective of 2-OG presence at three pH conditions (Fig. S4). Furthermore, we examined the effects of monovalent and divalent cations on SyAcnB activity. K + had little effect on the activity of SyAcnB for citrate, whereas the activity decreased to 59% and 14% in the presence of Table 1. BLAST search results for ATP-citrate lyase, citryl-CoA synthetase, and citryl-CoA lyase. BLAST search for ATP-citrate lyase, citryl-CoA synthetase, and citryl-CoA lyase was performed using the Kyoto Encyclopedia of Genes and Genomes database (https:// www. genome. jp/ kegg/ genome. html). The following sequences were used for the search: A: ATP-citrate lyase alpha-subunit from Chlorobium limicola (Clim_1231), B: ATP-citrate lyase beta-subunit from Chlorobium limicola (Clim_1232), C: citryl-CoA synthetase small subunit from Hydrogenobacter thermophilus (HTH_0201), D: citryl-CoA synthetase large subunit from Hydrogenobacter thermophilus (HTH_1737), E: citryl-CoA lyase from Hydrogenobacter thermophilus (HTH_0311). www.nature.com/scientificreports/   www.nature.com/scientificreports/ 1 mM and 5 mM Ca 2+ , respectively, and 58% and 9% in the presence of 1 mM and 5 mM Mg 2+ , respectively (Fig. 6a). Unlike the results obtained for citrate, 5 mM Mg 2+ decreased the activity of SyAcnB for isocitrate to 75%, and Ca 2+ had little effect on the activity of SyAcnB for isocitrate (Fig. 6b). The activity of SyAcnB for citrate decreased to 6% and 35% in the presence of 1 mM Zn 2+ and Mn 2+ , respectively, and 9% with 5 mM Zn 2+ (Fig. 6a). Similarly, the activity of SyAcnB for isocitrate decreased to 6% and 23% in the presence of 1 mM Zn 2+ and Mn 2+ , respectively, and 3% and 7% in the presence of 5 mM Zn 2+ and Mn 2+ , respectively (Fig. 6b).
We examined the effects of Mg 2+ and Ca 2+ on the kinetic parameters of the activity of SyAcnB for citrate. The inhibitory effects of Mg 2+ and Ca 2+ on the activity of SyAcnB for citrate were concentration-dependent (1-5 mM) (Fig. 6c). In the presence of 1 mM Mg 2+ , the V max of the activity of SyAcnB for citrate was 5.66 ± 0.20 unit/mg, and its K m value increased to 3.01 ± 0.04 mM, whereas its k cat /K m value decreased to 3.76 ± 0.12 s −1 mM −1 (Fig. 6d). Similarly, in the presence of 1 mM Ca 2+ , the V max of the activity of SyAcnB for citrate was 5.26 ± 0.49 unit/mg, and its K m value increased to 2.61 ± 0.28 mM, whereas its k cat /K m value decreased to 4.01 ± 0.10 s −1 mM −1 (Fig. 6e).

Discussion
In this study, we demonstrated the biochemical properties of aconitase, which preferentially catalyses the reaction from isocitrate to citrate, in the unicellular cyanobacterium Synechocystis 6803 using citrate and isocitrate as the substrates.  www.nature.com/scientificreports/ We investigated the reactivation conditions for in vitro enzymatic reaction by altering the DTT concentration and incubation time. Previous studies have suggested the requirement of varying concentrations of DTT, such as 5 mM 22 or 1 mM 25,26 , for aconitase reactivation. We also revealed that the maximum activity of the enzyme varied with DTT concentration. Additionally, various incubation times have been suggested for aconitase reactivation, for example, 20 min at 25 °C 26 and 30-120 min on ice 22 . We demonstrated that aconitase was reactivated immediately after the addition of the reagents, and its maximum activity gradually decreased after 20 min. AcnB from E. coli is reactivated faster than AcnA, but the enzyme is unstable 10 . Thus, a long reactivation period for SyAcnB may degrade the Fe-S cluster and reduce its activity.
The optimal pH required for Corynebacterium glutamicum aconitase (for citrate) is 7.5-7.8 26 and that for Mycobacterium tuberculosis aconitase (for isocitrate) is 8.0 27 . These values are similar to the optimal pH values required for SyAcnB activity in the presence of citrate and isocitrate (Fig. 2a). The intracellular pH of Synechocystis 6803 in logarithmically growing cells has been reported to be approximately 7.5-7.7 under dark conditions 28 . This suggests that the optimal pH required for SyAcnB activity is suitable for the growth of Synechocystis 6803.
The optimal temperature required for SyAcnB activity was estimated to be 45-55 °C (Fig. 2b), which is higher than the optimal growth temperature (30 °C) required for Synechocystis 6803, as reported in a previous study 23 . The optimal temperature required for the maximum activity of aconitase from C. glutamicum and the thermophilic archaea Sulfolobus acidocaldarius has been reported to be approximately 50 °C and 75 °C, respectively 26,29 . Additionally, the optimal temperature required for the maximum activity of aconitase from C. glutamicum is higher than its optimal growth temperature (30 °C) 30 . As the optimal temperature required for aconitase activity is known only for a few microorganisms, it remains unknown whether the optimal temperature for aconitase activity is usually higher than that required for the growth of microorganisms, as in this case. However, this pattern has been observed in some enzymes of the TCA cycle in Synechocystis 6803, such as fumarase (SyFum) (30 °C) 31 , wherein the optimal temperature required for enzyme activity corresponds with the optimal growth temperature of the bacterium; on the other hand, the optimal temperature required for the activity of other enzymes, such as citrate synthase (CS) from Synechocystis 6803 (SyCS) (37 °C) 32 and malate dehydrogenase (MDH) from Synechocystis 6803 (SyMDH) (45-50 °C) 33 , is higher than the optimal growth temperature of the bacterium. Enzymes are thermally denatured and inactivated at high temperatures; however, the reaction rate www.nature.com/scientificreports/ increases with the increase in temperature. Therefore, the optimal temperature required for the activity of some enzymes may be higher than the optimal growth temperature of the microorganisms. The affinity of SyAcnB for citrate has been reported to be higher than that of aconitases from other microorganisms, namely, E. coli, S. acidocaldarius, and Salmonella enterica (Table 3) 25,29,34 . On the contrary, the affinity of SyAcnB for isocitrate has been reported to be lower than that of aconitase from other microorganisms such as E. coli, C. glutamicum, S. acidocaldarius, and S. enterica (Table 3) 25,26,29,34 . The K m value of the activity of aconitase from C. glutamicum for citrate was slightly lower than that for isocitrate, which is consistent with the results obtained for SyAcnB, whereas the K m values of the activity of aconitase from E. coli, S. acidocaldarius, S. enterica, Rattus norvegicus (mitochondrial), and Zea mays (mitochondrial) for citrate are higher than those reported for isocitrate (Table 3) 25,26,29,[34][35][36] . The calculated values for K m (citrate)/K m (isocitrate) ratio are shown in Table 3; the ratio was estimated to be 0.73 at the optimum activity of SyAcnB, which is close to that for C. glutamicum aconitase (0.87). At pH 7.0, 8.0, and 9.0, the ratios were above 1 but were lower than those estimated for other organisms, except for C. glutamicum (Table 3). These results suggest that the aforementioned microorganisms tend to oxidise citrate to isocitrate. These values correspond with higher intracellular concentrations of citrate than isocitrate in Synechocystis 6803, which was estimated by the absolute quantification of metabolites 17 . The k cat /K m value of the activity of SyAcnB for isocitrate was slightly higher than that for citrate; this value is similar to that for aconitase from C. glutamicum (40.8 s −1 mM −1 for citrate and 52.4 s −1 mM −1 for isocitrate) and S. enterica (1.00 s −1 mM −1 for citrate and 1.22 s −1 mM −1 for isocitrate) 26,34 . Unlike heterotrophic bacteria, the TCA cycle flux in Synechocystis 6803 is always low under photoautotrophic, photomixotrophic, and heterotrophic conditions [37][38][39][40] , which may explain why the k cat /K m value of SyAcnB is lower than that of C. glutamicum aconitase.
The higher the pH, the lower the K m (citrate)/K m (isocitrate) ratio of SyAcnB (Table 3). The direction of the TCA cycle in Synechocystis 6803 is strongly affected by the pH, and the in vitro reconstruction of oxaloacetate metabolism displays a higher yield of citrate at higher pH 41 . At higher pH, higher concentrations of citrate, which is the substrate for SyAcnB, are formed, and the reaction is more likely to proceed in the oxidative direction at chemical equilibrium.
The TCA cycle in Synechocystis 6803 is characterised by the citrate accumulation at high levels in the cells, although 2-OG is generated through the oxidative TCA cycle from citrate under normal phototrophic growth conditions 17 . In Synechocystis 6803, isocitrate dehydrogenase (ICD) can catalyse isocitrate to form 2-OG. The K m value of the activity of ICD from Synechocystis 6803 (SyICD) for isocitrate was estimated to be 5.7 × 10 -3 -5.9 × 10 -2 mM 42 . The K m values of the activity of SyAcnB for isocitrate were 3.6-fold higher than those www.nature.com/scientificreports/ of the activity of SyICD. Therefore, isocitrate is thought to be metabolised mainly by SyICD, rather than by SyAcnB, enabling the cells to produce 2-OG. The three lines of evidence, 1) the high level of citrate accumulation in Synechocystis 6803, 2) the lack of citrate-metabolising enzymes such as ACL and CCS/CCL [43][44][45] (Table 1), and 3) the lack of SyCS activity degrading a citrate 32 , suggest that SyAcnB enhances the reaction in the direction of citrate to isocitrate to produce 2-OG. In this way, the properties of two enzymes, SyAcnB and SyICD, facilitate citrate accumulation and 2-OG generation at the same time. The peptide AcnSP affects the kinetic parameters of SyAcnB 21 and we performed biochemical analysis using AcnSP (Fig. S1). In both cases, V max and K m values decreased as in previous studies using cis-aconitate as a substrate, suggesting that AcnSP does not a significant effect on the reaction direction but boosting the reaction between citrate and isocitrate catalysed by SyAcnB. SyAcnB activities in both directions were inhibited by 2-OG in our study (Fig. 5); however, 2-OG has not been reported as an inhibitor of aconitase thus far. Therefore, we tested the effects of 2-OG in detail by comparing the kinetic parameters when 1 or 5 mM 2-OG was added (Fig. S2, S3) with those when it was not added (Table 2). When citrate was used as a substrate, 5 mM 2-OG acts as an inhibitor (Fig. S3a), but not at 1 mM 2-OG deduced from k cat /K m values of SyAcnB (Fig. S2a). Whereas when isocitrate was used as a substrate, 1 mM 2-OG acts as an activator (Fig. S2b), but not at 5 mM 2-OG deduced from k cat /K m values of SyAcnB (Fig. S3b). The addition of 0.44 mM 2-OG, the intracellular concentration in Synechocystis cells 24 , did not decrease SyAcnB activities (Fig. S4), and hence, 2-OG could play a role in the inhibition of SyAcnB activities when too many reactions of the oxidative reaction of the TCA cycle have proceeded. We also found that the activity of aconitase from Z. mays (mitochondrial) is inhibited by succinate and malate 36 , whereas that of SyAcnB was not inhibited by these organic acids.
The activity of SyAcnB for citrate was strongly inhibited by Mg 2+ and Ca 2+ ions (Fig. 6a). Mg 2+ and Ca 2+ increased the K m value. The k cat /K m values for citrate in the presence of 1 mM Mg 2+ and Ca 2+ were estimated to be 46% and 49% of the control, respectively. As per a previous report, the citrate/isocitrate concentration ratio for aconitase from rat heart was altered by Mg 2+ and Ca 2+ ions, and the equilibrium leaned towards citrate 46 . Comparing the effects of Mg 2+ and Ca 2+ on the activities of enzymes in the TCA cycle from Synechocystis 6803 revealed that the activity of SyCS increased to 1463% and 1050% of the control in the presence of 100 mM Mg 2+ and Ca 2+ , respectively, and that the activity of SyMDH increased to 160% and 190% of the control in the presence of 1 mM and 10 mM Mg 2+ , respectively 32,33 . Additionally, SyICD requires Mg 2+ or Mn 2+ as a cofactor for its activity 42 . The concentration of free Mg 2+ ions in the stroma of spinach chloroplasts varies between dark and light conditions 47 . Thus, depending on culture conditions, the concentration of free Mg 2+ in Synechocystis 6803 cells may be altered 48 , which may affect the equilibrium of aconitase. Also, SyAcnB activities in both directions were strongly inhibited by Mn 2+ and Zn 2+ (Fig. 6a,b). The mitochondrial aconitase activity from rat AF5 cells decreased to 48% and 19% of the control in the presence of 2 mM and 5 mM Mn 2+ , respectively 49 . The activity of aconitase from rat prostate epithelial cells for citrate was inhibited by Zn 2+ , but this effect was not observed for isocitrate 50 . The activity of SyCS decreased to 37% of that in the control in the presence of 100 mM Mn 2+32 , and the activity of SyFum was inhibited by 10 mM Mn 2+ when l-malate was used as a substrate 31 . Moreover, the activity of SyFum was strongly inhibited by 1 mM Zn 2+31 . Presently, the understanding of the physiological significance of metal ions in Synechocystis 6803 is limited.
In this study, we determined the biochemical properties of SyAcnB and demonstrated that citrate accumulation depends on the enzyme kinetics of SyAcnB. The consumption of isocitrate by SyICD to produce 2-OG overcomes the kinetic barrier of the SyAcnB enzyme. Currently, the study is limited to biochemical analysis; further genetic manipulation of SyAcnB might reveal its importance in citrate metabolism in cyanobacteria.

Methods
Construction of cloning vector for the expression of recombinant SyAcnB. The nucleotide sequence of acnB (slr0665), obtained from the sequenced genome of Synechocystis 6803 at KEGG database (https:// www. genome. jp/ kegg/ kegg_ ja. html), was synthesised by Eurofins Genomics Japan (Tokyo, Japan). The synthesised fragment was inserted within the BamHI-XhoI site of the vector pGEX6P-1 (GE Healthcare Japan, Tokyo, Japan).
The cloned expression vector was transformed in competent E. coli DH5α cells (Takara Bio, Shiga, Japan), and the transformed E. coli cells were cultivated in 5 L of Luria-Bertani medium at 30 °C with shaking at 150 rpm. Recombinant protein expression was induced overnight by adding 0.01 mM isopropyl β-D-1thiogalactopyranoside (Wako Chemicals, Osaka, Japan) to the medium.  PO 4 , and 0.05% Tween 20) and lysed through sonication (model VC-750; EYELA, Tokyo, Japan). The procedure was repeated 10 times for 10 s at 20% intensity. The lysed cells were centrifuged at 13,000 × g for 15 min at 4 °C. The supernatant was transferred to a 50-mL tube, and 560 µL of Glutathione Sepharose 4 B resin (GE Healthcare Japan, Tokyo, Japan) was added. Thereafter, the mixture was gently shaken for 30 min on ice. To remove the supernatant, the mixture was centrifuged at 5,800 × g for 2 min at 4 °C. The resin was re-suspended in 700 µL of PBST and washed five times. After washing, the recombinant protein was eluted with 700 µL of glutathione-S-transferase (GST) elution buffer (50 mM Tris-HCl (pH 9.6) and 10 mM reduced glutathione) five times, and the protein was concentrated using a Vivaspin 500 MWCO 50,000 device (Sartorius, Göttingen, Germany). The protein concentration was measured using a Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, Rockford, IL, USA). To verify protein purification, sodium dodecyl sulphate-polyacrylamide gel electrophoresis was carried out, and the gel was stained using Instant Blue reagent (Expedeon Protein Solutions, San Diego, CA, USA). www.nature.com/scientificreports/ Enzyme assay. Before measuring the enzyme activity, purified 50 pmol SyAcnB was reactivated by adding 25 µL of a solution containing 5 mM dl-dithiothreitol (DTT), 100 µM Na 2 S, and 100 µM (NH 4 ) 2 Fe(SO 4 ) 2· 6H 2 O and incubating the mixture at 20 °C for 1 ~ 30 min. The activity of SyAcnB was measured by mixing 50 pmol holo-SyAcnB with 1 mL of the assay solution (100 mM Tris-HCl (pH 7.0-9.0) or MES-NaOH (pH 6.0-7.0) and 20 mM trisodium citrate dihydrate or 20 mM dl-isocitrate trisodium salt hydrate). The enzymatic reaction was initiated by adding reactivated SyAcnB. The formation of cis-aconitate was monitored by measuring the absorbance at 240 nm using a Hitachi U-3310 spectrophotometer (Hitachi High-Tech, Tokyo, Japan) 51 . One unit of SyAcnB activity was defined as the formation of 1 µmol cis-aconitate per minute. Unit/mg represents the value of one unit divided by the amount of purified protein (mg). The K m and V max values were calculated using curve fitting of Michaelis-Menten equation with the KaleidaGraph ver. 4.5 software and the k cat values were calculated from V max values. The 44 amino acid sequence of AcnSP from Synechocystis 6803 was synthesized by Eurofins Genomics Japan (Tokyo, Japan) with a purity of 91.6%.
Statistical analysis. Paired two-tailed Student's t-tests were performed to calculate the P-values using Microsoft Excel for Windows (Redmond, WA, USA). All experiments were independently carried out three times.

Data availability
All the materials and data are available by contacting the corresponding author.