Central carbon metabolism in cyanobacteria comprises the Calvin-Benson-Bassham (CBB) cycle, glycolysis, the pentose phosphate (PP) pathway and the tricarboxylic acid (TCA) cycle. Redundancy in this complex metabolic network renders the rational engineering of cyanobacterial metabolism for the generation of biomass, biofuels and chemicals a challenge. Here we report the presence of a functional phosphoketolase pathway, which splits xylulose-5-phosphate (or fructose-6-phosphate) to acetate precursor acetyl phosphate, in an engineered strain of the model cyanobacterium Synechocystis (ΔglgC/xylAB), in which glycogen synthesis is blocked, and xylose catabolism enabled through the introduction of xylose isomerase and xylulokinase. We show that this mutant strain is able to metabolise xylose to acetate on nitrogen starvation. To see whether acetate production in the mutant is linked to the activity of phosphoketolase, we disrupted a putative phosphoketolase gene (slr0453) in the ΔglgC/xylAB strain, and monitored metabolic flux using (13)C labelling; acetate and 2-oxoglutarate production was reduced in the light. A metabolic flux analysis, based on isotopic data, suggests that the phosphoketolase pathway metabolises over 30% of the carbon consumed by ΔglgC/xylAB during photomixotrophic growth on xylose and CO2. Disruption of the putative phosphoketolase gene in wild-type Synechocystis also led to a deficiency in acetate production in the dark, indicative of a contribution of the phosphoketolase pathway to heterotrophic metabolism. We suggest that the phosphoketolase pathway, previously uncharacterized in photosynthetic organisms, confers flexibility in energy and carbon metabolism in cyanobacteria, and could be exploited to increase the efficiency of cyanobacterial carbon metabolism and photosynthetic productivity.

The oxygenic photosynthetic bacterium Synechocystis sp. PCC 6803 (S6803) is a model cyanobacterium widely used for fundamental research and biotechnology applications. Due to its polyploidy, existing methods for genome engineering of S6803 require multiple rounds of selection to modify all genome copies, which is time-consuming and inefficient. In this study, we engineered the Cas9 tool for one-step, segregation-free genome engineering. We further used our Cas9 tool to delete three of seven S6803 native plasmids. Our results show that all three small-size native plasmids, but not the large-size native plasmids, can be deleted with this tool. To further facilitate heterologous gene expression in S6803, a shuttle vector based on the native plasmid pCC5.2 was created. The shuttle vector can be introduced into Cas9-containing S6803 in one step without requiring segregation and can be stably maintained without antibiotic pressure for at least 30 days. Moreover, genes encoded on the shuttle vector remain functional after 30 days of continuous cultivation without selective pressure. Thus, this study provides a set of new tools for rapid modification of the S6803 genome and for stable expression of heterologous genes, potentially facilitating both fundamental research and biotechnology applications using S6803.