Introduction

The carbon metabolism of cyanobacteria includes anabolic processes that build carbohydrates, like the Calvin-Benson-Bassham (CBB) cycle, as well as catabolic processes such as glycolytic pathways, in which these carbohydrates are consumed to provide intermediates, energy and reducing equivalents. In the day-night rhythm, carbohydrate fluxes alternate between anabolic and catabolic directions. In 2016, it was claimed that cyanobacteria and plants possess the Entner-Doudoroff (ED) pathway as a glycolytic route besides glycolysis (also known as Emden-Meyerhof-Parnas (EMP) pathway) and the oxidative pentose phosphate (OPP) pathway (Fig. 1A) (1). The ED pathway utilizes two enzymes that are unique to this route. The first enzyme is 6-phosphogluconate (6PG) dehydratase, abbreviated as EDD (ED dehydratase), which converts 6PG to 2-keto-3-deoxy-6-phosphogluconate (KDPG). The second enzyme is KDPG aldolase, which abbreviated as EDA (ED aldolase), which splits KDPG into pyruvate and glyceraldehyde 3-phosphate (GAP). This claim was mainly based on the biochemical characterization of purified KDPG aldolases (EDAs) from the cyanobacterium Synechocystis sp. PCC 6803 (hereafter Synechocystis) and the plant Hordeum vulgare, in combination with protein sequence alignments, growth experiments with selected cyanobacterial glycolytic deletion mutants and the detection of gluconate and KDPG in Synechocystis and H. vulgare, from which KDPG is a metabolite that is indicative of an active ED pathway (1). In the following years, Synechocystis mutants in which EDA was deleted either alone or in combination with other glycolytic enzymes, such as 6PG dehydrogenase (GND), which belongs to the OPP pathway, were characterized in detail and yielded contradictory results (3–7). These observations raised the question whether EDA could be either a promiscuous aldolase or an enzyme with moonlighting functions (3). Slr0452 from Synechocystis and its homologues in plants were considered to be EDDs (1). This consideration was based on protein sequence alignments, and in addition on a report of a promiscuous dihydroxy acid dehydratase (DHAD) with gluconate dehydratase (GAD) activity in Sulfolobus solfataricus also discussing similarities between the two dehydratases DHAD and EDD (1, 8). However, a recent study reported that the assumed EDDs in cyanobacteria and plants are instead in most cases exclusively DHADs (2). DHADs are specifically involved in the synthesis of the branched amino acids valine, leucine and isoleucine. Accordingly, raising further questions about the presence of the ED pathway in cyanobacteria and plants.

Figure 1:

The anabolic CBB cycle and the catabolic glycolytic EMP and OPP pathways share several enzymes, requiring a sophisticated regulation of metabolic routes. Under photomixotrophic conditions, organic carbon (e.g., glucose) is consumed from the medium and metabolized in parallel with CO₂ fixation, accelerating growth compared to photoautotrophic conditions (9). It was previously observed that growth of the Synechocystis mutant ΔpfkΔzwf in which the EMP and the OPP pathway were interrupted by deleting both phosphofructokinases (pfk-A1 and pfk-A2) and glucose-6-phosphate dehydrogenase (zwf), was still able to enhance its growth in the presence of glucose (see Fig. 1) (1). This suggested the presence of an alternative route for carbohydrate degradation. When bioinformatic protein sequence analyses revealed the presence of both enzymes that are unique to the glycolytic ED pathway: EDD (Slr0452) and the KDPG aldolase EDA (Sll0107), this contradiction seemed to be resolved. EDA was purified and verified biochemically as KDPG aldolase (1). In addition, two putative glucose dehydrogenases (GDHs) (Sll1709 and Slr1608) were identified through bioinformatic studies. Since ZWF, which supplies the substrate 6PG for EDD, was deleted in ΔpfkΔzwf, it was assumed that a putative GDH, together with a gluconate kinase (GK), would convert glucose (bypassing hexokinase (HK) and ZWF) to gluconate and further to 6PG (Fig. 1). Growth experiments with the respective deletion mutants seemed to support these assumptions (1). Furthermore, gluconate and KDPG were detected in cells of the Synechocystis wild type (WT) and ΔpfkΔzwf via IC-ESI-MSMS (1). Hence, it was concluded that the accelerated growth of the ΔpfkΔzwf mutant was made possible by the existence of a previously overlooked glycolytic ED pathway in combination with a GDH/GK bypass (1). Notably, at that time, it was not known whether the CBB cycle and glycolytic routes are spatially separated in the cytoplasm or rather intertwined with sophisticated regulation to enable a switch between anabolic and catabolic reactions. Subsequently, flux analyses with labelled glucose under photomixotrophic conditions with WT and Δeda yielded interesting, but also contradictory results, as will be outlined in detail in the following (3). It was observed that glycolytic routes and the CBB cycle are highly intertwined in that sense that catabolic glycolytic routes are shortened to shunts that feed carbohydrates as anaplerotic reactions into the anabolic CBB cycle and thereby accelerate CO₂ fixation (3, 4). Two glycolytic shunts were identified: the OPP shunt, comprising the two first reactions of the OPP pathway catalyzed by ZWF and GND, and the PGI shunt, which includes only the first enzyme of the EMP pathway, phosphoglucose isomerase (PGI) (see Fig. 1) (3, 4). Whereas 27.6% of the labeled glucose was directed to the glycogen pool, the majority, 63.1%, was metabolized via the PGI shunt into the regenerative CBB cycle, and 9.3% were channeled via the OPP shunt to the CBB cycle (3). Contrary to our expectations, no flux via the ED pathway was detected, and no KDPG was measurable, despite the use of ultrasensitive LC–MS/MS.

The earlier assumption that the accelerated growth on glucose of the glycolytic ΔpfkΔzwf mutant requires the GDH/GK bypass and the ED pathway, had been proven wrong by the described flux analyses, as this growth behavior can easily be explained by the existence of the PGI shunt which is responsible for the main flux of glucose into the CBB cycle and is still intact in the ΔpfkΔzwf mutant (Fig. 1) (3). However, although there was no detectable flux via the ED pathway in the WT, carbohydrate fluxes were still altered in Δeda. The flux via the PGI shunt was not modified, but a higher portion of labeled glucose was directed to the glycogen pool, and the flux through the OPP shunt was reduced (3, 4). Further studies yielded similarly conflicting results concerning the existence of the ED pathway and particularly concerning the phenotypes of Δeda and the double mutant ΔedaΔgnd. Both ΔedaΔgnd and Δeda were impaired in their ability to recover from nitrogen starvation, Δeda had a reduced ability to respond to CO₂ shifts, the reactivation of the dark arrested CBB cycle was delayed and growth of mutants in which either eda was deleted alone or in combination with other glycolytic enzymes was not understandable on the mere interruption of glycolytic fluxes (4, 5, 7). Therefore, suggesting that alterations in Δeda and ΔedaΔgnd might rather be based on regulatory aspects such as a moonlighting functions or substrate promiscuity of EDA instead of missing fluxes (3). Enzymes with substrate promiscuity accept a broad range of substrates because their active sites can accommodate molecules with different shapes or chemical groups. Therefore, they catalyze several reactions in addition to the primary, most efficient one (10). Moonlighting enzymes, on the other hand, perform additional functions that are independent of the catalytic activity and active site of an enzyme, such as protein-protein interactions, regulation of gene expression, or binding to RNA or DNA as transcription factors (11).

The most puzzling contradiction that remained was the observation that the intermediate 6-P gluconate(6PG) accumulates only in ΔedaΔgnd (where both 6PG effluxes should be interrupted in the case of an existing ED pathway) but neither in Δeda nor in Δgnd (see Fig. 1A) (4).

In order to understand previous results in detail against the background of contradictory data and the new bioinformatic information that EDD is absent in most cyanobacteria (2, 12), we went back to old data, purified and tested the putative EDD (Slr0452), checked the existence of the previously presumes GDH/GK bypass, performed a series of control measurements with different mutants and resolved the accumulation pattern of 6PG in deletion strains. A detailed biochemical characterization of the KDPG aldolase EDA (Sll0107) reveals its promiscuity and provides initial evidence for its possible role in the absence of EDD in the cyanobacterium Synechocystis.

Results

In vitro characterization of the predicted EDD/DHAD in Synechocystis

Based on the biochemical characterization of the KDPG aldolase EDA in Synechocystis, we had previously assumed that Slr0452 encodes for an enzyme that is involved in both amino acid synthesis as dihydroxy acid dehydratase (DHAD) and the ED pathway as EDD, as it was shown that the bifunctional DHAD (DHAD, IlvD/EDD superfamily) in the (hyper)thermophilic archaeon Sulfolobus solfataricus, also accepts D-gluconate - and to a lesser extent, galactonate - as alternative substrates in addition to its canonical role in branched-chain amino acid biosynthesis (8). Similarly, DHAD from Rhizobium leguminosarum bv. trifolii exhibits activity toward several C5 and C6 sugar acids, including gluconate, and DHAD from Fontimonas thermophila has been engineered to increase its gluconate-dehydration activity by an order of magnitude. Together, these observations underscore the intrinsic substrate promiscuity of IlvD enzymes in the IlvD/EDD superfamily.

However, although at least some DHADs exhibit activity toward gluconate (sugar acids), they have not been shown to accept 6-phosphogluconate (6PG)—the physiological substrate of EDD in the canonical ED pathway—from any organism. One possible explanation for this limited overlap in substrate specificity was proposed by Evans et al.(2), who suggested that gluconate, being smaller and unphosphorylated, may fit within the DHAD substrate-binding pocket normally occupied by dihydroxy-isovalerate. However, the lack of crystal structures of DHAD bound to dihydroxy-isovalerate or of EDD bound to 6PG precludes direct structural comparisons, leaving the precise geometric and electrostatic determinants of substrate discrimination between these enzymes unresolved. Of note, archaea like S. solfataricus utilizing modified, branched ED pathway versions involving the dehydration of gluconate to KDG, do not employ a DHAD but a dedicated gluconate dehydratase (GAD) (phylogenetically grouped within the enolase superfamily, specifically a subgroup related to the mandelate racemase/muconate lactonizing enzyme (MR/MLE)) to catalyse this conversion. This indicates that DHAD side activity alone might be insufficient to sustain physiologically relevant flux through the modified ED pathway. To date, there is no evidence that DHAD can functionally replace a bona fide GAD from the enolase superfamily in these organisms in vivo.

A recent study supposes that Slr0452 from Synechocystis which was assumed to be an EDD is an exclusive DHAD based on the biochemical characterization of its homolog in soybean (Glycine max) (2). Notably, organisms harboring both DHAD and EDD activity typically encode an additional IlvD/EDD superfamily gene, likely originating from gene duplication. Thus, the presence of only a single DHAD candidate in Synechocystis argues against an EDD function. DHAD activity was indeed shown for Slr0452 in another study, however, its potential activity toward 6PG as a substrate was not evaluated, leaving unresolved whether Slr0452 encodes a strict DHAD or a bifunctional DHAD/EDD enzyme (13). To address this question, the slr0452 gene from Synechocystis was heterologously expressed in E. coli and recombinant protein was purified (Fig. S1). For comparison, a DHAD from the cyanobacterium Synechococcus elongatus was purified and assayed under identical conditions as a positive control (12) (Fig. S2). Enzymatic assays confirmed that Slr0452 catalyzes the dehydration of 2R-dihydroxyisovalerate (2R-DHIV) to 2-ketoisovalerate (KIV), consistent with DHAD function (Fig. S3A). Slr0452 exhibited a specific activity of 2.0 U mg⁻¹, whereas DHAD from S. elongatus showed 0.8 U mg⁻¹ with 2R-DHIV as substrate (Fig. S3A).

To test for potential dual DHAD/EDD activity, Slr0452 was subsequently assayed using 6PG as substrate. A biochemically characterized EDD from Caulobacter crescentus (14) served as a positive control. Despite screening a range of divalent metal cofactors (Mg²⁺, Mn²⁺, Zn²⁺, Co²⁺, Cu²⁺, and Fe²⁺), no measurable activity toward 6PG was detected for Slr0452 under any tested condition (Fig. S3B). In contrast, EDD from C. crescentus displayed robust 6PG dehydratase activity as expected (Fig. S2, Fig. S3B). To evaluate the possibility of a modified ED pathway in Synechocystis, Slr0452 and EDA (Sll0107) activities were also assayed using gluconate and KDG as substrates, respectively. S. solfataricus DHAD with GAD activity (SsoDHAD) and Sulfolobus acidocaldarius KD(P)G aldolase (SaciKD(P)GA), previously shown to catalyze reactions with gluconate and KDG, respectively (15, 16), were used as positive controls (Fig. S2, Fig. S4). Slr0452 and EDA from Synechocystis showed no activity on gluconate or KDG indicating that they are not involved in a modified ED pathway (Fig. S4).

Collectively, these results demonstrate that slr0452 encodes a monofunctional, metal-dependent DHAD rather than an EDD or a bifunctional DHAD/EDD enzyme. This confirms that Slr0452 is exclusively involved in amino acid synthesis and does not catalyze the first reaction of the ED pathway as previously suggested by others (2, 13).

Above all, these results show that Synechocystis lacks a canonical 6-phosphogluconate (6PG) or gluconate dehydratase and therefore does not contain either the canonical or modified glycolytic ED pathway.

Testing for the existence of the previously presumed GDH/GK bypass

We furthermore tested for the existence of the previously presumed glucose dehydrogenase/gluconate kinase (GDH/GK) bypass as an alternative route to hexokinase and ZWF for gluconate and 6PG production. Two putative GDHs (Sll1709, Slr1608) were previously identified in the genome based on protein sequence alignments (1). Crude cell extracts were analyzed for GDH and GK activity; however, analysis of crude extracts from Synechocystis WT, Δzwf, as well as a putative GDH1 (Sll1709) overexpression mutant sll1709:oe revealed no detectable GDH activity under either photoautotrophic no photomixotrophic conditions (Fig. S5, Fig. S6, Fig. 2A-B).

Figure 2:

Unfortunately, we were not able to overexpress putative GDH2 (Slr1608). Δzwf was tested on the assumption that the GDH/GK route, if existent, might be upregulated in this mutant. The sll1709:oe overexpression mutant was tested to ensure sufficient enzyme amounts. Cell crude extracts of Δzwf and purified putative GDH1 (Sll1709) were furthermore tested for GDH activity in an in vitro enzyme assay (Fig. S6, Fig. 2B). In this analysis, the two quinones 2,3-dimethoxy-5-methyl-1,5-benzoquinone (DMB) and duroquinone (DQ) were also tested as electron acceptors in addition to NADP⁺ and NAD⁺ since some glucose dehydrogenases are known to reduce quinones instead of NAD(P)⁺. No significant amount of GDH activity was detected for any tested electron acceptor in the GDH eluate, as well as in a cell extract of photomixotrophically grown Δzwf (Fig. 2B). Taken together, these results indicate that GDH activity, which converts glucose to gluconate, is absent in Synechocystis in vivo. However, the inability to detect enzymatic activity is never a completely safe proof for its absence as the conditions might not have been chosen correctly, the enzyme activity might be below detection limits, or the purified enzyme might have lost its activity during purification. Furthermore, we were not able to purify and test the second putative GDH2 (Slr1608). Therefore, the second enzyme of the putative GDH/GK route was sought. The search for homologues of GK in the Synechocystis genome, using GKs from Escherichia coli and Pseudomonas aeruginosa as baits, was not successful. In line with the in silico analyses, no gluconate kinase activity was detectable in either crude cell extracts from WT or Δzwf (Fig. 2C). These results furthermore strengthen the assumption that a GDH/GK bypass is absent in Synechocystis.

In order to further investigate if the GDH/GK bypass is absent from all cyanobacteria, we conducted a second blast (BlastP) analysis. Similarity thresholds were set at a query score > 30 %, an identity score > 30 % and an e-value < 1^-10. At these threshold settings we found 261 cyanobacteria that possess putative homologs of GKs from Escherichia coli and Pseudomonas aeruginosa (Table S1). We searched for homologs of GDHs in these 261 cyanobacteria next to Synechocystis. As baits the NAD⁺-dependent GDH 1GCO from Bacillus megaterium and the quinone-dependent GDH from Acinetobacter calcoaceticus were used (17, 18). While the results again showed a rather poor similarity for both putative GDHs (Sll1709 and Slr1608) from Synechocystis and thereby confirmed the lack of GDH activity in enzyme tests, good homologies were found for 244 (93 %) of the other 261 analyzed cyanobacteria (Fig. 2D, Table S1). Out of these 244 cyanobacteria that possess both a putative GDH and a putative GK homologue, we picked two strains, namely Spirulina sp. SIO3F2 and Lyngbya aestuarii and overexpressed and purified their putative GDHs (NEO87985.1 and WP_238987112) and the putative GK from Lyngbya aestuarii (WP_023068790.1) in E. coli (Fig. S7 to Fig. S9). Purification of the putative GK from Spirulina sp. SIO3F2 (NEO86388.1 MAG) was not successful. Enzyme activity assays showed a clear GDH activity for both the GDH from Spirulina and the GDH from Lyngbya (Table 1). Both NADP⁺ and NAD⁺ could function as cofactors, while higher activity was observed with NADP⁺ for both enzymes. The Spirulina and Lyngbya GDHs also used 2-Deoxy-D-glucose as substrate with 20-40% residual activity compared to glucose. Marginal activity with D-mannose was observed with the Lyngbya GDH (Table 1). All other tested substrates including fructose 6-phosphate, D-fucose, glucose 6-phosphate, lactose, D-glucosamine, D-lyxose, D-allose, D-arabinose, L-arabinose, D-xylose, L-xylose, and D-ribose were not used as substrates.

Table 1:

For both the GK from Lyngbya and Spirulina, expression in E. coli could be observed (Fig. S9), however, purification was only achieved for the GK from Lyngbya. For the Lyngbya GK a specific activity of 24.8 U/mg could be determined (Table 1). In combination with already shown GDH activity, this proves the presence of the GDH/GK bypass in this cyanobacterium.

Taken together, these results indicate that there is no GDH/GK bypass in Synechochstis, whereas it seems to be present in other cyanobacteria, as e.g. Lyngbya.

Based on these results, we currently have no explanation for the gluconate that was detected in previous IC-ESI-MSMS measurements in Synechocystis (1). Supplementation of the growth medium with gluconate did not enhance growth of the WT which furthermore supports the absence of a gluconate kinase or alternatively gluconate transport into the cells (Fig. 3A). An additional control was performed by testing photomixotrophic growth of a hexokinase mutant (Δhk, sll0539) assuming that a GDH/GK bypass if present might be upregulated in this mutant.

Figure 3:

Two enzymes (Sll0539 and Slr0329) were previously identified as glucose phosphorylating enzymes, with Sll0539 as the main contributor, while the deletion of Slr0329 did not affect on glucose uptake or photomixotrophic growth (19, 20). Therefore, only Δhk (Δsll0539) was studied (21).

Enzyme assays confirmed that the Δhk mutant lost its ability to phosphorylate glucose (Fig. 3B). It was furthermore unable to metabolize glucose from the medium and to accelerate its growth (Fig. 3C-D). The described phenotypes could be complemented in Δhk::hk. These data further support the assumption that a GDH/GK bypass is absent in Synechocystis, because it should allow glucose utilization in the absence of hexokinase (Hk, Sll0539) otherwise.

Investigation into the causes why 6PG accumulates in ΔedaΔgnd in the absence of the ED pathway, but not in Δeda or Δgnd

As our data indicate that the GDH/GK bypass is not present in Synechocystis, the only known source for 6PG would be the reaction catalyzed by ZWF (see Fig. 1). Based on this assumption, the accumulation of 6PG in ΔedaΔgnd in contrast to Δeda and Δgnd is unexpected in the absence of an existing ED pathway (see Fig. 1) (4). This pattern of 6PG enrichment had led to believe for many years, despite conflicting results, that there must be at least a small flux through the ED pathway in Synechocystis (4). So, to resolve this apparent contradiction, we attempted to resolve the origin of the 6PG accumulation pattern in ΔedaΔgnd, Δeda and Δgnd.

In order to test if ZWF is responsible for 6PG production in ΔedaΔgnd, this enzyme was deleted in addition resulting in the mutant ΔedaΔgndΔzwf. 6PG accumulation was abolished in ΔedaΔgndΔzwf as expected, indicating, that ZWF activity was responsible for 6PG production in ΔedaΔgnd (Fig. 4A).

Figure 4:

We subsequently tested ZWF activities in ΔedaΔgnd, Δeda, Δgnd and Δzwf as a control and found to our surprise that ZWF activity was completely lacking in Δgnd (Fig. 4B). Sequencing of the zwf gene in Δgnd, revealed an insertion of an adenine in position 399, which results in a premature stop condon, and obviously completely abolishes ZWF activity in the Δgnd mutant, whereas the mutants ΔedaΔgnd and Δeda lacked this insertion in their zwf genes (Fig. 4C). Immunoblots with specific antibodies against ZWF confirmed that ZWF is present in WT, Δeda and ΔedaΔgnd but is no longer expressed in Δgnd (Fig. 4D). Whereas ΔedaΔgnd showed ZWF activity that resembled or WT levels, ZWF activity was reduced to approximately 60% in Δeda (Fig. 4B). This observation is in line with the reduced photomixotrophic flux through the OPP shunt in this mutant that we had observed earlier (3). The construction of corresponding complemented mutants (Δeda::eda) is underway to test whether these differences are indeed due to the absence of EDA.

In summary, these data demonstrate that the accumulation pattern of 6PG in ΔedaΔgnd, Δeda and Δgnd is not caused by an active ED pathway, as previously assumed. Instead, the pattern can be explained by a secondary mutation in zwf in the Δgnd mutant which abolishes 6PG production in the Δgnd mutant.

In vitro characterization of EDA (Sll0107)

Taken together, all data so far support the conclusion that the ED pathway is absent in the cyanobacterium Synechocystis. One remaining puzzling observation is the reported detection of KDPG in Synechocystis in one study using IC-ESI-MS/MS, which can be indicative of an active ED pathway, whereas another study using LC–MS/MS failed to detect this metabolite (1, 3). Analytical artifacts can never be fully excluded. To clarify whether EDA (Sll0107) might catalyze the reversible aldol condensation of pyruvate and GAP to KDPG, which could explain the detection of KDPG, in the absence of EDD and to test for potential promiscuity of EDA with several substrates, the sll0107 gene (KEGG annotation) was expressed in E. coli. Although SDS–PAGE and Immunoblot blot analysis confirmed the expression of a ∼23 kDa protein, it could not be eluted from the Ni–TED resin, and crude extracts lacked detectable KDPG aldolase activity. These observations suggested expression of a non-functional or truncated protein. Genomic inspection of sll0107 (eda) revealed an alternative start codon (genome position 2,972,167) 72 nucleotides upstream of the annotated start codon (genome position 2,972,095) (Fig. 5). Comparative sequence and AlphaFold 3 analyses (22) using E. coli and C. crescentus KDPG aldolases (23) as a reference, showed that only the upstream-initiated longer sequence retained the N-terminal α-helix characteristic of class I KDPG aldolases (Fig. 5A).

Figure 5.

Therefore, the extended Sll0107 version was expressed in E coli, purified and resulted in active protein (Fig. S10-S12). SDS–PAGE showed a subunit mass of ∼25 kDa, consistent with the calculated 23 kDa, and size-exclusion chromatography indicated a native mass of ∼64 kDa, corresponding to a homotrimeric assembly typical of bacterial class I KDPG aldolases 13) (25, 26).

Biochemical enzyme characterization of Sll0107 (EDA)

Recombinant Sll0107 exhibited robust activity toward its canonical substrate KDPG, catalyzing its cleavage into pyruvate and GAP with a catalytic efficiency of 24.4 s^-1 mM^-1 (Fig. 6, Table 2). Inspired by recent reports that metal-independent class I KDPG aldolases from Synechococcus elongatus also accept oxaloacetate (12), we tested this substrate and found that EDA (Sll0107) catalyzed its cleavage to pyruvate and CO₂ with a catalytic efficiency of 0.58 s^-1 mM^-1 (Fig. 6, Table 2). Since elevated proline levels were previously reported in Δeda, we next examined activity with 2-keto-4-hydroxyglutarate (KHG), an intermediate of the proline degradation pathway as substrate (5, 27). EDA efficiently cleaved KHG to pyruvate and glyoxylate with a catalytic efficiency of 0.18 s^-1 mM^-1 (Fig. 6, Table 2).

Figure 6.

Table 2:

EDA exhibited a strong preference for KDPG as its primary substrate but also catalyzed reactions with OAA and KHG, albeit with substantially lower catalytic efficiency. The effect of NADP⁺ on EDA activity was examined, as NADP⁺ has been reported to act as an activator of Synechococcus elongatus EDA (12). Under the assay conditions used in this study, the addition of NADP⁺ at concentrations up to 3 mM did not result in any measurable increase in EDA activity (Fig. 6E). These findings are consistent with curated enzyme data reported in BRENDA Enzyme Database (28), in which NADP⁺ is not listed as an activator for EDA.

To assess reaction reversibility, a hallmark of class I aldolases, reported for archaeal KD(P)G aldolases (29), we tested the enzyme in the condensation direction using (i) pyruvate + GAP for KDPG formation, (ii) pyruvate + CO₂ for oxaloacetate, and (iii) pyruvate + glyoxylate for KHG synthesis. Only a low level of reverse activity, 0.18 U/mg (∼2% of the cleavage activity) was detected for KDPG formation using a discontinuous TBA assay, whereas no reverse activity was observed for oxaloacetate or KHG synthesis (Fig. S12).

Given that a soybean KDPG aldolase was recently reported to accept fructose 1,6-bisphosphate (FBP) (2) we also tested FBP as a substrate. However, EDA (Sll0107) displayed no detectable activity with FBP (Fig. 6A, 6F, Table 2). Also, malate was not used as substrate (Table 2).

To confirm that catalysis proceeds via the conserved class I mechanism, the catalytic lysine in EDA was replaced by glycine through site-directed mutagenesis. The mutant enzyme EDAK147G was expressed and purified at comparable levels to the WT protein but exhibited no detectable activity with any tested substrate (Fig. 6F).

Taken together, these data show that the KDPG aldolase EDA (Sll0107) of Synechocystis is an aldolase with substrate promiscuity, which in vitro cleaves KDPG to GAP and pyruvate, OAA to CO₂ and pyruvate, and KHG to glyoxylate and pyruvate (Fig. 6, Table 2). In addition, it synthesizes KDPG from GAP and pyruvate with very low efficiency (∼2% of the cleavage activity) (Fig. S12).

Since 6PG dehydratase (EDD) seems to be absent in Synechocystis, the canonical substrate of EDA, 2-keto-3-deoxy-6-phosphogluconate (KDPG) is not produced, and no alternative enzymatic source of KDPG is known to the best of our knowledge. Our in vitro enzyme assays therefore suggest that EDA may instead participate in alternative metabolic reactions, potentially linked to the PEP-pyruvate-OAA node and/or proline catabolism, as illustrated in Figure 7. However, further studies are required to clarify the physiological role of EDA in vivo.

Figure 7:

Discussion

Our data demonstrate that Synechocystis lacks a functional ED pathway, due to the absence of both EDD and the proposed GDH/GK bypass, which is also supported by photomixotrophic flux analyses (3). Instead, it possesses a promiscuous EDA aldolase, whose physiological role remains to be elucidated. The dehydratases EDD and DHAD have a common origin, a high degree of amino acid and structural similarity, and the same catalytic mechanism, but differ in two domains that contribute to substrate selection, which is why EDD and DHAD are enzymes with high substrate specificity that cannot replace one another (2). Apart from four identified cyanobacteria identified as possessing potential EDDs (based on the described domains), most cyanobacteria lack EDD despite the presence of EDA, which likewise applies to algae, moss and plants (2, 27). Our data demonstrate that Slr0452 in Synechocystis, which we previously assumed as EDD (1), lacks 6PG dehydratase activity and instead dehydrates DIV exclusively, consistent with recent reports (13). The same experimental results were recently obtained for the cyanobacterium Synechococcus 7942 which also lacks EDD but contains EDA (12). This raises the question of what role EDA could play in the absence of EDD and the ED pathway in cyanobacteria, moss, algae and plants. In soybean (Glycine max), GmEDA was found to catalyze three reactions in vitro: firstly, the cleavage of KDPG to GAP and pyruvate, secondly the cleavage of fructose 1,6-bisphosphate (FBP) to dihydroxyacetone phosphate (DHAP) and GAP and thirdly the condensation of erythrose 4-phosphate (E4P) and GAP to sedoheptulose 1,7-bisphosphate (SBP) (2). Accordingly, the idea was put forward that GmEDA is a promiscuous aldolase which catalyzes aldolase reactions in the CBB cycle and glycolysis in vivo, which is well in line with its localization in chloroplasts (2). In diatoms the ED pathway was suggested to operate in mitochondria providing GAP and pyruvate to lower glycolysis and the TCA cycle, based on bioinformatic analyses, bacterial complementation experiments and measurements on the conversion of 6PG to pyruvate in soluble protein extracts (30–32). However, as neither the EDD nor EDA from diatoms were thoroughly characterized biochemically yet, further studies would be informative.

Based on our in vitro data Synechocystis EDA might be connected to the TCA cycle and the PEP-pyruvate-OAA node by decarboxylating oxaloacetate to pyruvate and in proline catabolism by splitting KHG to glyoxylate and pyruvate. The latter assumption is supported by earlier findings of elevated proline levels in Δeda mutants of Synechocystis (5). For EDA from Synechococcus 7942, it has also recently been demonstrated that, in addition to cleaving KDPG, it also decarboxylates oxaloacetate (12). The catalytic efficiency of EDA on oxaloacetate is rather low (k_cat/K_m (mM^-1 s^-1): 0.437) in Synechococcus, however, its activity could be enhanced by NADP⁺. The Synechocystis EDA showed a 42-fold higher catalytic efficiency with KDPG compared to oxaloacetate (KDPG 24.4 s^-1 mM^-1, OAA 0.58 s^-1 mM^-1) and no activation by NADP⁺ (1-3 mM) was observed. As the Synechococcus Δeda mutant was compromised in terms of photosynthesis and growth under diurnal day/night cycles, the authors hypothesized that Synechococcus EDA might play a role in balancing NADPH/NADP⁺ ratios and photosynthesis by modulating TCA cycle activities (12). For Synechocystis EDA, the catalytic efficiency for KDPG compared to KHG was 136-fold higher (KDPG 24.4 s^-1 mM^-1, KHG 0.18 s^-1 mM^-1). However, EDA displayed no detectable activity with FBP (Fig. 6), which is consistent with the fact that class I KDPG aldolases are members of a different enzyme class (EC 4.1.2.14). Likewise, class I and class II FBP aldolases (EC 4.1.2.13) are not known to exhibit activity toward KDPG (26, 33, 34). Interestingly, EDA was able to condensate GAP and pyruvate to KDPG in vitro with low efficiency. It remains to be tested whether this reaction could play a role in vivo, in which KDPG could act as a regulatory metabolite at low concentrations. This could explain why KDPG was detected in WT and ΔpfkΔzwf in one study (1), although this observation could not be replicated later (3). Based on the current data, it is not possible determine whether the detected KDPG was due to a measurement artifact or not. The accumulation pattern of 6PG in WT, Δeda, Δgnd and ΔedaΔgnd, which contradicted the flux analyses and had been interpreted as evidence of an active ED metabolic pathway in Synechocystis (3, 4), could be explained by a secondary mutation in the Δgnd mutant (Fig. 4C) and now agrees with the flux analyses.

Our systematic re-examination of data on the proposed ED pathway in Synechocystis demonstrates how complementary approaches can contribute to elucidating cyanobacterial metabolism. By combining detailed biochemical enzyme characterization, genetic deletion and complementation, careful mutant validation, and taking into account metabolic flux analyses (3), we were able to clarify that neither EDD nor the GDH/GK bypass is present, and that the ED pathway is not functional in Synechocystis. These integrated studies further revealed the presence of a promiscuous KDPG aldolase whose broader metabolic involvement now opens exciting new research directions.

This study also highlights the importance of best practices to strengthen metabolic pathway reconstruction and provides the following lessons: Although protein sequence alignments provide valuable information, they cannot replace detailed biochemical enzyme tests, which are required to verify the predicted enzyme functions and to further test for substrate promiscuity. Furthermore, to demonstrate the presence of a metabolic pathway, it is advisable to characterize all enzymes of this pathway biochemically, rather than limiting analysis to individual key enzymes, and to also include flux analyses. Routine complementation of deletion mutants and sequencing of selected genes or the entire genome are effective means of identifying secondary mutations that can lead to misleading phenotypes. A final lesson is to exercise great caution when interpreting data and drawing conclusions, even if the data sets appear completely logical and conclusive at first glance (1). Over time, contradictory results have accumulated regarding the claim that there is an ED pathway in Synechocystis. In the present study, we have succeeded in clarifying a number of these contradictions. Taken together, these advances provide a solid foundation for future investigations into the in vivo role of the promiscuous aldolase EDA in cyanobacteria.

Materials and Methods

Growth experiments

Synechocystis sp PCC 6803 WT and mutants (Table S2) were grown in 200 ml BG11 medium (pH 8) in glass tubes at 28°C under constant light (50 µmol m^-2 s^-1). Cultures were aerated with filter-sterilized air. Photomixotrophic cultures were supplemented with 10 mM D-glucose. Aliquots were taken under sterile conditions every 24 h to measure optical density photometrically at 750 nm.

Generation of mutants

Constructs for deletion mutants, complementation of Δhk and edaK147G were made by Gibson Assembly cloning, using the primers listed in Table S3. An antibiotic resistance cassette was fused to approximately 200 bp directly up- and downstream of the gene to be deleted and cloned into pBluescript. Synechocystis was transformed with these constructs to replace the gene with the antibiotic resistance cassette via homologous recombination. Mutants were checked via PCR for segregation.

Generation of a putative GDH1 (sll1709) His-tagged over-expression mutant (sll1709:oe) and overexpression of GDH homologues of Spirulina and Lyngbya

Two DNA fragments from the genome of Synechocystis were synthesized by Genescript: 1. a fragment containing 212 bp up- and 212 bp downstream of the sll1709 start codon, with a BamHI, XhoI and NdeI site in its middle and 20 bp of overlapping sequences with the pBluescript SK(+) vector at its ends (Fig. S5A). 2. a fragment containing a modified petE promotor, followed by His-tag, TEV cleavage recognition and linker encoding sequences, various restriction sites and 20 bp of overlapping sequences with the pBluescript SK(+) vector at its ends (Fig. S5B). Both fragments were separately cloned into the pBluescript SK(+) vector by Gibson Assembly cloning. A kanamycin antibiotic resistance cassette was inserted into the EcoRV site of the plasmid containing the modified petE promotor. Both plasmids were then digested with BamHI and NdeI and the promoter cassette was ligated into the alkaline phosphatase treated GDH1 plasmid to yield the final mutagenesis vector. This plasmid was transformed into Synechocystis and mutant segregation was confirmed by PCR analysis. GDH homologues of Spirulina and Lyngbya were overexpressed in E.coli (Fig. S7-8).

P3-His-GDH1 (sll1709) over-expression and His-tag purification

For the purification of GDH1 from Synechocystis, a 6 L photoautotrophic culture of the P3-His-GDH1 overexpression strain was grown to an OD₇₅₀ of about 1. Cells were harvested by centrifugation at 4000 rpm in a JLA-8.1000 rotor for 20 min at 4°C. Initially, His-GDH1 over-expression in the 6 L culture was assessed by SDS PAGE analysis followed by immunoblotting with a His-tag specific antibody (Fig S6). A specific band could be detected in the over-expression mutant, confirming expression and stable accumulation of the over-expressed and N-terminally His-tagged GDH1 protein. A small-scale purification (using a 50-mL aliquot of the large culture) demonstrated that the protein could be purified (Fig S6C). For the large-scale purification the remaining cells were resuspended in lysis buffer (50 mM NaPO₄ pH=7.0; 250 mM NaCl; 1 tablet complete protease inhibitor EDTA free (Roche) per 50 mL) and broken by passing them through a French Press cell at 1250 p.s.i. twice. Unbroken cells and membranes were removed by centrifugation in a Beckman ultracentrifuge using a 70 Ti rotor at 30.000 rpm for 45 min at 4°C. The decanted soluble extract was adjusted to a volume of 45 mL with lysis buffer and incubated with 5 mL TALON cobalt resin (Takara) for 1 h at 4°C. The resin was then washed extensively with 200 mL lysis buffer and subsequently with 100 mL lysis buffer containing 5 mM imidazole. Bound proteins were eluted with 10 mL elution buffer (50 mM NaPO₄ pH=7.0; 250 mM NaCl; 250 mM imidazole) and after dialysis overnight against the reaction buffer 50 mM Tris pH=7.5; 20 mM MgCl, the protein was concentrated in a Vivaspin Turbo 4 Ultrafiltration Unit (5 kDa MWCO). Commercially available Glucose dehydrogenase from Pseudomonas sp. (19359, Sigma) was resuspended in the reaction buffer for control measurements. Finally, protein concentrations were determined by Bradford assay using the Roti-Quant reagent (Carl Roth) and adjusted to 0.25 mg/mL. Samples taken during the purification procedure were assessed by SDS PAGE analysis followed by immunoblotting with antibodies specific against the His-tag.

Preparation of crude cell extracts

Enzyme activity was measured in crude cells extracts. For the generation of crude cell extracts of Synechocystis, an OD volume of 30-50 OD*mL sample culture was pelleted by centrifugation (5 min, 9000 g, 4 °C) and the pellet resuspended in 0.5 mL 100 mM Tris buffer (pH 7.6). The suspension was then transferred to centrifugation tubes with screwed lids (max 500 µl/cup) and glass beads (0.17-0.18 mm diameter; Sartorius, Göttingen, Germany) were added so that 1-2 mm liquid remained above the dispersion. Samples were vortexed for 10 min at 4 °C. Afterwards, the glass beads were pelleted by centrifugation in two consecutive centrifugation steps (1 min, 1000 g, 4 °C and 10 min, 1600 g, 4 °C). The supernatant was transferred to new centrifugation cups in between centrifugations. In order to precipitate cell debris, the supernatant was centrifuged (15 min, 18000 g, 4 °C) one more time. The supernatant constituted the final cell extract and was measured for its protein concentration and kept on ice until further enzyme activity measurements.

Protein quantification

Protein concentrations were determined according to Bradford (35). 1 μL protein sample (cell extract) was mixed in triplicates with 799 μL water. Afterwards, 200 μL Bradford reagent, consisting of 100 mg/L Coomassie Serva Blue G-250 (Bio-Rad, Munich, Germany), 5 % (v/v) ethanol and 8.5 % (v/v) H₃PO₄ was added to all samples as well as to 800 µL bovine serum albumin (BSA) standards in concentrations of 2, 5, 10, 15 and 20 μg/mL. After an incubation at RT for 10 min, absorption at 595 nm was measured and sample protein concentration calculated by using the standard curve.

Glucose dehydrogenase activity measurements

For glucose dehydrogenase (GDH) activity measurements in Synechocystis, NADPH turnover was monitored photometrically at 340 nm. Measurements were carried out in 50 µl crude cell extract or purified GDH in 100 mM Tris/HCl buffer (pH 7.4) in the presence of 20 mM glucose, 5 mM NAD(P)⁺ and 10 mM MgCl₂. Reactions were started by addition of glucose. The kinetics of the reaction were monitored with a double-beam spectrophotometer for 5-10 minutes. In case quinones were tested as electron acceptors, the GDH reaction was enzymatically coupled to the gluconate kinase (GK) and 6PG dehydrogenase (GND) reaction, the latter providing the required NADP⁺ turnover. This was achieved by adding 5 mM ATP, 0.02 U gluconate kinase from E. coli (Megazyme, Wicklow, Ireland) and 0.005 U 6-phosphogluconate dehydrogenase from E. coli (Megazyme, Wicklow, Ireland). The two tested quinones 2,3-dimethoxy-5-methyl-1,5-benzoquinone (DMB) and duroquinone (DQ) were supplied in a concentration of 5 mM. As a positive control, 0.05 U glucose dehydrogenase from Pseudomonas sp. (Sigma Aldrich, Steinheim, Germany) was added.

The GDH activity of recombinant enzymes from Spirulina sp. SIO3F2 and Lyngbya aestuarii was determined after GTS-His tag removal by TEV protease cleavage and purification. Activity was determined spectrophotometrically by monitoring glucose dependent reduction of NAD(P)⁺ to NAD(P)H at 340 nm and 30°C (extinction coefficient of NAD(P)H=6.22 mM^-1cm^-1). The standard assay mixture (final volume of 500 µL) contained 0.1 mM Tris/HCl (pH 7.7 at 30°C), 10 mM MgCl₂, 5 mM NAD(P)⁺, 20 mM Glucose, and 2.5 – 5 µg of recombinant protein. To determine the substrate specifity glucose was substituted by other sugars (for details see text).

Gluconate kinase activity measurements

For gluconate kinase (GK) activity measurements, the GK reaction was enzymatically coupled to the gluconate kinase (GK) and 6-phosphogluconate dehydrogenase (GND) reaction, the latter providing NADP⁺ reduction, which was monitored photometrically at 340 nm. Measurements were carried out in 50 µl crude cell extract in 100 mM Tris buffer (pH 7.4) in the presence of 20 mM gluconate, 5 mM NADP⁺, 10 mM MgCl₂., 5 mM ATP and 0.005 U 6-phosphogluconate dehydrogenase. Reactions were started by addition of gluconate and kinetics were monitored with a double-beam spectrophotometer for 5-10 minutes. As a positive control, 0.05 U gluconate kinase from E. coli (Megazyme, Wicklow, Ireland) was added.

Recombinant GK activity was determined spectrophotometrically in a continuous coupled pyruvate kinase/lactate dehydrogenase (PK/LDH) assay by monitoring NADH oxidation at 340 nm (30 °C) (extinction coefficient of NAD(P)H=6.22 mM^-1cm^-1). Gluconate phosphorylation was followed indirectly via ADP formation, which was converted by PK/LDH in a NADH-dependent reaction. The standard assay (500 µl) contained 100 mM Tris/HCL (pH 7.7), 10 mM MgCl2, 0.2 mM NADH, 1 mM PEP, 2 mM ATP, 1-10 mM gluconate, auxiliary enzymes PK and LDH, and 0.7-15µg purified recombinant enzyme (after GST-tag cleavage).

Glucokinase/hexokinase activity measurements

For glucokinase/hexokinase (GLK/HK) activity measurements, the GLK/HK reaction was enzymatically coupled to the glucose-6-phosphate dehydrogenase (ZWF) reaction, whose NADP⁺ turnover was monitored photometrically at 340 nm. Measurements were carried out in 50 µl crude cell extract in 100 mM Tris buffer (pH 7.4) in the presence of 10 mM glucose or fructose, 5 mM NADP⁺, 10 mM MgCl₂., 5 mM ATP and 0.005 U 6-phosphogluconate dehydrogenase from E. coli (Megazyme, Wicklow, Ireland). Reactions were started by addition of glucose and kinetics were monitored with a double-beam spectrophotometer for 5-10 minutes.

Quantification of 6-phosphogluconate

Quantification of 6-phosphogluconate (6PG) was achieved photometrically by enzymatic conversion of sample 6PG to ribulose-5-phosphate and CO₂, which goes hand in hand with equimolar formation of NADPH. 5 to 50 OD*mL cell culture was pelleted by centrifugation (5 min, 9000 g, 4 °C) and cell lysis was achieved by solving the pellet in 1 mL 0.2 M HCl, followed by a 15 min incubation at 95 °C. The lysates were centrifuged (10 min, 18000 g, 4°C) and the supernatants transferred to new tubes where they were neutralized by addition of 1 mL 1 M Tris/HCl buffer (pH 8.0). Samples were then split into two (each sub-pool 900 μL), that were supplied with 90 μL of an 11.11 mM NADP⁺ solution. 10 μL of a 5 U/mL 6-PG dehydrogenase from (Megazyme, Wicklow, Ireland) solution was added to one of the samples, whereas 10 μL of water was added to the second sample as a control. All samples were then incubated at 37 °C for 30 min and the absorption at 340 nm was measured using a double-beam spectrophotometer. 6PG standards were prepared (0 mM, 0.05 mM, 0.1 mM, 0.2 mM, 0.4 mM and 0.8 mM) and measured in the same way as samples. Δabsorption values of standards were used to create a standard curve, which was then used to convert sample Δabsorption values into 6PG concentrations. The resulting 6PG concentration (mM) was converted into specific cellular 6PG content (μg/OD*mL) by multiplying it with the molar mass and dividing it by the respective OD volume.

Quantification of glucose

For quantification of medium glucose concentrations, 1 mL Synechocystis culture was pelleted by centrifugation (5 min, 10000 g, RT) and 50 μL of supernatant were transferred to a new cup. If medium glucose concentration was expected to exceed 5 mM, samples were diluted. 50 µL glucose standards with the following concentrations were prepared: 0 mM, 0.625 mM, 1.25 mM, 2.5 mM and 5 mM. Furthermore, a ‘NADP buffer’, containing 150 mM NADP⁺, 150 mM ATP, 300 mM MgCl₂ in 100 mM Tris-HCl (pH 7.4,) was prepared. 940 μL ‘NADP buffer’ were added to all samples and standards, which were then transferred to acryl-cuvettes and absorption at 340 nm was measured in a double-beam spectrophotometer (baseline absorption). After that, 0.17 U Hexokinase as well as 0.085 U glucose-6-phosphate dehydrogenase were added to all samples and standards. Samples and standards were then incubated at 37 °C for 60 min to ensure complete enzymatic conversion. Finally, absorption at 340 nm was measured again (endpoint absorption). Baseline absorbance values were subtracted from endpoint readings to calculate ΔA₃₄₀. A glucose standard curve was generated from the standards and used to determine glucose concentrations in the culture supernatants.

Blast analysis

In order to search for the existence of a gluconate kinase in the genome of Synechocystis and cyanobacteria in general, the fasta protein sequences of the well characterized gluconate kinases from Escherichia coli (KIG95398.1) and from Pseudomonas aeruginosa (WP134279162.1) were used as baits for the protein/protein blast tool at NCBI. Similarly, the fasta sequences of the NAD⁺ dependent glucose dehydrogenase 1GCO from Bacillus megaterium and the quinone dependent glucose dehydrogenase 1C9U from Acinetobacter calcoaceticus were used as baits for the search of gluconate dehydrogenases in cyanobacteria. Next to Synechocystis, 261 cyanobacterial species were analyzed. Homology thresholds were set at a query score > 30 %, an identity score > 30 % and an e-value < 1^-10.

SDS PAGE and immunoblotting analyses

To assess the level of ZWF protein in various strains, soluble protein extracts were prepared and analysed by SDS PAGE and immunoblotting. 5 µg of protein were loaded per lane on a 12.5% (w/v) polyacrylamide BisTris gel using MES running buffer (250 mM MES, 250 mM Tris, 5 mM EDTA, 0.5 % (w/v) SDS). Gels were Coomassie-stained or electroblotted onto nitrocellulose membrane. For immunoblotting, 5% (w/v) milk powder in 1x PBS-T was used as blocking solution and all washing steps were performed with 1x PBS-T. For detection a ZWF antibody (αZwf, Agrisera, Sweden, Phyto AB, PHY5113A) in a 1:5k dilution in 1x PBS-T and a horseradish peroxidase-conjugated secondary antibody (anti-rabbit,1:10k in 1x PBS-T Cytiva) were used. As a loading control an antibody against 30S ribosomal protein S1 (αRps, Agrisera, Sweden) was utilized.

Gene cloning and protein overexpression

Codon-optimized genes from Synechocystis (eda, sll0107, edd/dhad, slr0452) and Synechococcus elongatus PCC 7942 (dhad, syc0898c) were synthesized by BioCat GmbH (Germany). The dhad gene from Sulfolobus solfataricus (ss03107), encoding a DHAD with GAD activity was synthesized by Eurofins (Eurofins Genomics, Ebersberg, Germany). A longer sll0107 variant, extended by 24 amino acids relative to the KEGG annotation with an N-terminal 6×His tag was synthesized by BioCat GmbH (Germany) as well. The slr0452 and syc0898c genes were inserted into pET24a vectors encoding a C-terminal 6×His-tag. The saci_0225 gene from Sulfolobus acidocaldarius (encoding KDPG aldolase) was cloned into the pET15b vector using NdeI and BamHI restriction sites. Genes encoding eda (ccna_01562) and edd (ccna_02134) from Caulobacter crescentus were cloned into pET28b and pET15b using NdeI/BamHI and NdeI/HindIII, respectively. All constructs were verified by Sanger sequencing (LGC Genomics, Berlin, Germany). For protein expression, plasmids were transformed into E. coli Rosetta (DE3) (Agilent Technologies). Cultures were grown in Terrific Broth (TB; 22 g L⁻¹ yeast extract, 12 g L⁻¹ tryptone, 4 mL L⁻¹ glycerol) supplemented with the appropriate antibiotics (100 µg mL⁻¹ ampicillin or kanamycin, and 30 µg mL⁻¹ chloramphenicol) at 37 °C with shaking at 180 rpm. Protein expression was induced at an OD₆₀₀ of 0.6–0.8 by addition of 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG), followed by incubation at 18 °C for 18–22 h. Cells were harvested by centrifugation (15 min, 8630 × g, 4 °C) and stored at −70 °C until further use.

EDA (KDPG aldolase) mutation at the active center

Class I KDPG aldolases are metal-independent enzymes that utilize a catalytic lysine residue to form a Schiff base intermediate during catalysis. In Eda this lysine (Lys147, based on the extended sll0107 sequence) was identified as essential for activity toward multiple sugar acid substrates (36). To generate a catalytically inactive variant, Lys147 was substituted with glycine (K147G), which lacks both side-chain and net charge. Site-directed mutagenesis was performed as described (37). Q5® High-Fidelity DNA Polymerase and its recommended reaction buffer (NEB) were utilized for the amplification process. The primers used for mutagenesis are listed in Table S3, with the nucleotide substitutions introducing the K147G exchange underlined. As a template the codon-optimized eda gene from Synechocystis (sll0107) that was synthesized by BioCat GmbH (Germany), was utilized. Successful mutagenesis was confirmed by DNA sequencing (LGC genomics, Berlin, Germany). The resulting EDA variant was named EDAK147G.

Protein purification

Recombinant Synechocystis EDA (both versions), EDAK147G, and DHAD (Slr0452), recombinant Caulobacter crescentus CcEDA and CcEDD, Sulfolobus acidocaldarius SaciKD(P)GA, and Sulfolobus solfataricus SsoDHAD were purified by immobilized metal affinity chromatography (IMAC) followed by size-exclusion chromatography (SEC) for EDA and Slr0452. Frozen cell pellets were resuspended in 50 mM HEPES–NaOH (pH 7.8, 30 °C), 300 mM NaCl at a ratio of 1 g wet cells per 3 mL buffer. For EDD/DHADs purification buffer was supplemented with 10 mM MgCl₂ and 1 mM DTT. Cells were disrupted by sonication (3 × 5 min; amplitude 50; cycle 0.5) using a UP200S homogenizer (Hielscher Ultrasonics, Brandenburg, Germany). Cell debris was removed by centrifugation (45 min, 21,130 × g, 4 °C). His-tagged proteins, except Synechocystis DHAD (Slr0452), were purified from the supernatant using Ni–TED columns (Macherey-Nagel, Düren, Germany) according to the manufacturer’s protocol. Synechocystis DHAD (Slr0452) was purified on a self-packed Ni–NTA affinity column. Elution fractions containing the target protein were pooled and concentrated using Vivaspin 20 centrifugal concentrators (Sartorius Stedim Biotech; 30 kDa cutoff for EDD/DHAD and 3 kDa cutoff for EDA). Concentrated samples of Synechocystis EDA (6 mg protein) were further purified by SEC on a HiLoad 16/600 Superdex 200 prep grade column (Cytiva, Marlborough, MA, USA) equilibrated with 50 mM HEPES–NaOH (pH 7.8, 30 °C), 300 mM NaCl and additionally contained 10 mM MgCl₂ and 1 mM DTT for DHADs purification. Protein-containing fractions (4.5 mg after SEC) were analyzed by SDS–PAGE and enzymatic activity assays. Purified proteins were stored at −70 °C in 25% (v/v) glycerol. Protein concentrations were determined by the Bradford assay (38) using bovine serum albumin (Merck, Darmstadt, Germany) as the standard.

Determination of the native molecular mass of EDA

Following IMAC purification, 6 mg of purified EDA was subjected to SEC on a HiLoad 16/600 Superdex 200 prep-grade column (Cytiva, Marlborough, MA, USA). The column was calibrated under identical buffer conditions using standard proteins from the LMW and HMW gel-filtration calibration kits (Cytiva, Marlborough, MA, USA): Aprotinin (6.5 kDa), Ribonuclease A (13.7 kDa), Ovalbumin (44 kDa), Aldolase (158 kDa), and Ferritin (440 kDa), with blue dextran to determine the column void volume. The native molecular mass of EDA was estimated from the resulting calibration curve.

In vitro EDA enzyme characterization

The catalytic activities of EDA, EDAK147G, SaciKD(P)GA, and CcEDA in the KDPG cleavage direction were determined using a continuous, coupled spectrophotometric assay. EDA catalyzes the cleavage of 2-keto-3-deoxy-6-phosphogluconate (KDPG) into pyruvate and glyceraldehyde 3-phosphate (GAP). Pyruvate formation was monitored indirectly by its reduction to lactate using lactate dehydrogenase (LDH; rabbit muscle, Merck, Darmstadt, Germany), coupled to the oxidation of NADH to NAD⁺. The decrease in absorbance at 340 nm was recorded in 96-well plates (BRANDplates®, BRAND, Wertheim, Germany) using a Tecan Infinite M200 plate reader (Tecan Group AG, Männedorf, Switzerland) at 30 °C. NADH concentration was quantified against a standard curve ranging from 0–0.7 mM.

For KDPG cleavage assays (200 µL final volume), reactions contained 0.1 M HEPES–NaOH (pH 7.8, 30 °C), 0.7 mM NADH, 2 mM KDPG, 2 U LDH, and the indicated enzyme (typically 1 µg EDA, EDAK147G, SaciKD(P)GA, or 0.06 µg CcEDA). Substrate specificity toward oxaloacetate (OAA) and 2-keto-4-hydroxyglutarate (KHG) was assayed under the same conditions as KDPG assay, as they all generate pyruvate as one of the two product, however, for the KHG assay, 6 µg of EDA was used.

FBP cleavage activity forming GAP and DHAP was assayed using a coupled system containing glycerol-3-phosphate dehydrogenase (G3PDH; rabbit muscle, Merck, Darmstadt, Germany), which converts the DHAP product to glycerol 3-phosphate coupled to the oxidation of NADH to NAD⁺. Reaction mixtures (200 µL) contained 0.1 M HEPES–NaOH (pH 7.8, 30 °C), 0.7 mM NADH, 5 mM FBP and 2 U G3PDH. FBP aldolase (from Saccharomyces cervisiae; Merck, Darsmstadt, Germany) served as the positive control. Reactions were initiated by addition of enzyme, and all measurements were performed in triplicate. Negative controls omitted either enzyme or substrate. To obtain the kinetic parameters (Table 1), a classical Michaelis-Menten rate equation was fit to the complete dataset using NonlinearModelFit function in Wolfram Mathematica v14.2. EDA activity in the reverse (condensation) direction—forming KDPG from pyruvate and D-GAP was quantified using the colorimetric thio-barbituric acid (TBA) assay as described (39). Reactions (1 mL) contained each 20 mM pyruvate and 10 mM D-GAP in 0.1 M HEPES–NaOH (pH 7.8, 30 °C) with 15 µg EDA or 3 µg SaciKD(P)GA, SaciKD(P)GA was assayed at 70°C. At two-minute intervals, 50 µL aliquots were quenched with 5 µL of 12% trichloroacetic acid. Absorbance of the TBA–KDPG adduct was measured at 549 nm (ε = 67.8 mM⁻¹ cm⁻¹), and SaciKD(P)GA served as a positive control. EDA activity in reverse direction – forming OAA was assayed using malate dehydrogenase (MDH, S. cervisiae, Merck, Darmstadt, Germany) that forms malate from OAA by oxidation of NADPH to NADP⁺. EDA activity in reverse direction – forming KHG from pyruvate and glyoxylate was assayed by measuring the pyruvate consumption in a discontinuous assay. Pyruvate (5 mM) and glyoxylate (5 mM) were incubated with Eda (5 µg) in in 1 mL of 0.1 M HEPES–NaOH buffer (pH 7.8) at 30 °C. At defined time points (0, 2, 4, 6, 8, 10, 20, and 30 min), 50 µL aliquots were withdrawn and quenched by the addition of 5 µL of 12% (v/v) TCA. Samples were immediately placed on ice and centrifuged at 8,130 × g for 10 min to remove precipitated protein. Supernatants (50 µL) were diluted into 950 µL of 0.1 M HEPES–NaOH (pH 7.8). Pyruvate consumption was quantified by a lactate dehydrogenase (LDH)–coupled assay using 0.7 mM NADH and 1 U of LDH (rabbit muscle; Merck, Darmstadt, Germany). The decrease in NADH absorbance reflected the conversion of pyruvate to lactate by LDH and was used to compare pyruvate levels across time points. It was ensured that auxiliary enzymes were not rate limiting. One unit (1 U) of enzyme activity is defined as 1 µmol substrate consumed or product formed per minute.

To test whether EDA exhibits activity toward 2-keto-3-deoxygluconate (KDG), an additional continuous coupled assay was employed. Reaction mixtures (200 µL final volume) contained 0.1 M HEPES–NaOH (pH 7.8, at 30 °C), 20 mM gluconate, 20 µg Sulfolobus solfataricus DHAD (SsoDHAD; SSO3107), 10 mM MgCl₂, and 1 mM DTT. The mixtures were incubated with 20 µg of either SsoDHAD at 50 °C or Slr0452 at 30 °C for 2 h. Reactions were initiated by the addition of EDA or Sulfolobus acidocaldarius KD(P)GA (SaciKD(P)GA; Saci_0225) as positive control, which converts the reaction product 2-keto-3-deoxygluconate (KDG) into pyruvate and glyceraldehyde. Pyruvate formation was detected as described above for the KDPG cleavage assay. All assays were performed in triplicate. Negative control reactions were performed in the absence of either the enzyme or the substrate.

In vitro EDD enzyme assay

EDD catalyzes the dehydration of 6PG to form KDPG in the classical Entner–Doudoroff pathway. Enzyme activity was monitored using a continuous, coupled spectrophotometric assay in which the KDPG product was subsequently cleaved by excess CcEDA to yield pyruvate and GAP. Pyruvate formation was detected through its reduction to lactate by lactate dehydrogenase (LDH; rabbit muscle, Merck, Darmstadt, Germany), coupled to the oxidation of NADH to NAD⁺. The decrease in absorbance at 340 nm was recorded in 96-well plates (BRANDplates®, BRAND, Wertheim, Germany) using a Tecan Infinite M200 plate reader (Tecan Group AG, Männedorf, Switzerland) at 30 °C. NADH concentrations were quantified against a standard calibration curve (0–0.7 mM). Reaction mixtures (200 µL final volume) contained 0.1 M HEPES–NaOH (pH 7.8, 30 °C), 0.7 mM NADH, 10 mM MgCl₂, 5 mM 6PG, 5 µg CcEDA (recombinantly produced and characterized in this study, 150 U mg⁻¹), 2 U LDH, and the respective amounts of Slr0452 (20 µg) as the test sample and CcEDD (5 µg) as positive control (40). Reactions were initiated by enzyme addition, and all assays were performed in triplicate. Negative controls omitted either the enzyme or substrate.

To test whether Slr0452 exhibits activity toward gluconate, as reported for the bifunctional EDD/DHAD from Sulfolobus solfactorius (15), an additional continuous coupled assay was employed. Reaction mixtures (200 µL final volume) contained 0.1 M HEPES–NaOH (pH 7.8, at 30 °C), 0.7 mM NADH, 10 mM MgCl₂, and 1 mM DTT. The mixtures were incubated with 20 µg of either S. solfataricus DHAD (SsoDHAD) at 50 °C or Slr0452 at 30 °C. Reactions were initiated by the addition of Sulfolobus acidocaldarius KD(P)GA aldolase, which converts the reaction product 2-keto-3-deoxygluconate (KDG) into pyruvate and glyceraldehyde (16). Pyruvate formation was detected as described above for the KDPG cleavage assay. All assays were performed at least in triplicate. Negative controls omitted either the enzyme or substrate. It was ensured that auxiliary enzymes were not rate limiting. One unit (1 U) of enzyme activity is defined as 1 µmol substrate consumed or product formed per minute.

In vitro DHAD enzyme assay

Dihydroxyacid dehydratase (DHAD; ilvD) catalyzes a key step in the biosynthesis of branched-chain amino acids (BCAAs: leucine, isoleucine, and valine). The activities of Slr0452 and S. elongatus DHAD (SeDHAD) were assayed using the physiological substrate 2R-dihydroxyisovalerate (DHIV; Merck, Darmstadt, Germany), which is dehydrated to yield 2-ketoisovalerate (KIV). Product formation was quantified using a discontinuous colorimetric assay as described previously (13).

Reaction mixtures (100 µL) contained 50 mM Tris–HCl (pH 7.8), 1 mM dithiothreitol (DTT), 10 mM MgCl₂, and 1 mM 2R-DHIV. Reactions were initiated by the addition of 0.25 µM enzyme, incubated aerobically at room temperature for 30 min, and terminated with 25 µL of 2 M HCl. The resulting KIV was derivatized by addition of 50 µL of saturated 2,4-dinitrophenylhydrazine (DNPH; Häberle, Germany) in 2 M HCl and incubated for 30 min at room temperature. Subsequently, 25 µL of 10 M NaOH were added, samples were mixed, and precipitates were removed by centrifugation (5,000 × g, 10 min). Supernatants (100 µL) were transferred to clear 96-well microplates, and absorbance of the KIV-DNPH adduct was measured at 540 nm. Quantification was based on a KIV-DNPH calibration curve (0–1 mM), which exhibited linear response across this range. All enzyme assays were performed in triplicate.

To test whether Slr0452 exhibits activity toward gluconate, as reported for the bifunctional DHAD from Sulfolobus solfataricus, (15) an additional continuous coupled assay was employed. Reaction mixtures (200 µL final volume) contained 0.1 M HEPES–NaOH (pH 7.8, at 30 °C), 20 mM gluconate, 10 mM MgCl₂, and 1 mM DTT. The mixtures were incubated with 20 µg of either SsoDHAD at 50 °C or Slr0452 at 30 °C. Reactions were initiated by the addition of SaciKD(P)GA, which converts the reaction product 2-keto-3-deoxygluconate (KDG) into pyruvate and glyceraldehyde (16). Pyruvate formation was detected as described above for the KDPG cleavage assay. All assays were performed in triplicate. Negative control reactions were performed in the absence of either the enzyme or the substrate. It was ensured that auxiliary enzymes were not rate limiting. One unit (1 U) of enzyme activity is defined as 1 µmol substrate consumed or product formed per minute.

Bioinformatic analysis

Crystal structures were obtained from the Protein Data Bank (PDB), and structural models were retrieved from the AlphaFold Protein Structure Database (version 3.0) (22). Structural analyses and comparative assessments were performed using UCSF ChimeraX (41). Amino acid sequences for alignment were retrieved from UniProt, and multiple sequence alignments were generated using the EMBL-EBI Job Dispatcher framework (24), followed by manual refinement and editing in BioEdit.

Data availability

All study data are included in the article and SI Appendix. In addition, the kinetic data and model simulations are available as Excel files and Mathematica notebooks on the FAIRDOMHub (https://fairdomhub.org/investigations/775) and will be made public upon publication of the manuscript.

Acknowledgements

This study was supported by grants from and the German Science Foundation (DFG Gu1522/2-2, Gu1522/5-1, SI 642/14-1, SI 642/14-2, FOR2816, GRK2749/1). We acknowledge financial assistance from the DSI/NRF in South Africa (grant NRF-SARCHI 82813 to J. L. S.). We thank Markus Brand, Frauke Caliebe, Friederike Hörsch, Carolin Kremer and Maria Patti for engaged discussions.

Additional information

Author’s contributions

RO, MT, RF, AM, MB, and CP, performed the experiments; MB, CB, JLS, BS, KG supervised, RO, MT, RF, AM, MB, CP, CB, JLS, MH, BS, and KG analyzed data; RO, AM, and KG wrote the original draft; all authors wrote, reviewed, and edited; BS, and KG conceptualized the study; BS and KG acquired funding.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work the author(s) used ChatGPT (Open AI 2025), ChatGPT Team (5.1) and DeepL in order to optimize language. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

Funding

Deutsche Forschungsgemeinschaft (DFG) (SI 642/14-2)

• Ravi Shakar Ojha

Deutsche Forschungsgemeinschaft (DFG) (Gu1522/2-2)

• Marius Theune

Deutsche Forschungsgemeinschaft (DFG) (GRK2749/1)

• Ruben Fritsche

Deutsche Forschungsgemeinschaft (DFG) (Gu1522/5-1)

• Alexander Makowka

Deutsche Forschungsgemeinschaft (DFG) (SI 642/14-1)

• Carmen Peraglie

DSI/NRF in South Africa (NRF-SARCHI 82813)

• Jacky L Snoep

Deutsche Forschungsgemeinschaft (DFG) (FOR2816)

• Martin Hagemann

Additional files

Supplementary information