Improved lipid production via fatty acid biosynthesis and free fatty acid recycling in engineered Synechocystis sp. PCC 6803

Abstract

Background: Cyanobacteria are potential sources for third generation biofuels. Their capacity for biofuel production has been widely improved using metabolically engineered strains. In this study, we employed metabolic engineering design with target genes involved in selected processes including the fatty acid synthesis (a cassette of accD, accA, accC and accB encoding acetyl-CoA carboxylase, ACC), phospholipid hydrolysis (lipA encoding lipase A), alkane synthesis (aar encoding acyl-ACP reductase, AAR), and recycling of free fatty acid (FFA) (aas encoding acyl-acyl carrier protein synthetase, AAS) in the unicellular cyanobacterium Synechocystis sp. PCC 6803.

Results: To enhance lipid production, engineered strains were successfully obtained including an aas-overexpressing strain (OXAas), an aas-overexpressing strain with aar knockout (OXAas/KOAar), and an accDACB-overexpressing strain with lipA knockout (OXAccDACB/KOLipA). All engineered strains grew slightly slower than wild-type (WT), as well as with reduced levels of intracellular pigment levels of chlorophyll a and carotenoids. A higher lipid content was noted in all the engineered strains compared to WT cells, especially in OXAas, with maximal content and production rate of 34.5% w/DCW and 41.4 mg/L/day, respectively, during growth phase at day 4. The OXAccDACB/KOLipA strain, with an impediment of phospholipid hydrolysis to FFA, also showed a similarly high content of total lipid of about 32.5% w/DCW but a lower production rate of 31.5 mg/L/day due to a reduced cell growth. The knockout interruptions generated, upon a downstream flow from intermediate fatty acyl-ACP, an induced unsaturated lipid production as observed in OXAas/KOAar and OXAccDACB/KOLipA strains with 5.4% and 3.1% w/DCW, respectively.

Conclusions: Among the three metabolically engineered Synechocystis strains, the OXAas with enhanced free fatty acid recycling had the highest efficiency to increase lipid production.

Keywords: Acetyl-CoA carboxylase; Acyl-ACP reductase; Acyl–acyl carrier protein synthetase; Lipase A; Synechocystis sp. PCC 6803; Total lipid; Unsaturated lipid.

Fig. 1

The fatty acid biosynthesis and its neighboring pathways in Synechocystis sp. PCC 6803. Key enzyme genes include accABCD, multi-subunit acetyl-CoA carboxylase gene; aar, acyl-ACP reductase gene; aas, acyl-ACP synthetase; ado, aldehyde oxidase; fabD, malonyl coenzyme A-acyl carrier protein transacylase; lipA, lipolytic enzyme genes; plsX, plsY, plsC, putative phosphate acyl-transferases; phaA, polyhydroxyalkanoates specific beta-ketothiolase gene. The thick arrow is represented as the overexpression (OX) of that gene whereas the cross symbol is represented the knockout (KO) of that gene

Fig. 2

Outline maps representing gene locations in all Synechocystis engineered strains. OXAas strain (upper) was singly recombined with Aas gene locus whereas OXAas/KOAar strain (middle) was generated by interrupting aar gene with aas gene fragment insertion. Finally, OXAccDACB/KOLipA strain (bottom) was constructed by inserting a cassette fragment of accD, accA, accC and accB to disrupt lipA gene

Fig. 3

Confirmation of each gene location by PCR analysis using specific pairs of primers (Table 2) in each engineered strain including OXAas (A), OXAas/KOAar (B) and OXAccDACB/KOLipA (C) strains in this study. The location of aas gene fragment in OXAas was checked using a pair of pE_SF and pE_SR primers for pEERM core structure (a). Lane M: GeneRuler™ DNA ladder (Fermentas), lane 1: negative control using WT as template and lanes 2–10: clone numbers 1 to 9, For Cm_SF and Aas_SR (b) primer, lane M: GeneRuler™ DNA ladder (Fermentas), lane 1: negative control using WT as template and lanes 2–6: clone numbers 1 to 5. In (c), the pair of Aas_F6 and Cm_SR (c) primers was used, lane M: GeneRuler™ DNA ladder (Fermentas), lane 1: negative control using WT as template and lanes 2–6: clone numbers 1 to 5. The UUPSF_Aas and Cm_SR (d) primer, lane M: GeneRuler™ DNA ladder (Fermentas), lane 1: negative control using WT as template and lanes 2–6: clone numbers 1 to 5. Confirmation of gene location in OXAas/KOAar (B) using a pair of CAar_F and Aas_SR (a) primer, Lane M: GeneRuler™ DNA ladder (Fermentas), lane 1: negative control using WT as template and lanes 2–9; clone numbers 1 to 8 whereas UUPSF_Aar and Aas_SR (b) primers was used. Lane M: GeneRuler™ DNA ladder (Fermentas), lane 1: negative control using WT as template and lanes 2–11: clone numbers 1 to 10. The gene location in OXAccDACB/KOLipA (C) using pair of UUPSF_lipA and AccD_SR primer, Lane M: GeneRuler™ DNA ladder (Fermentas) and lanes 1–5; clone numbers 1 to 5

Fig. 4

Growth curve (a), oxygen evolution rate (b), chlorophyll a content (c) and carotenoid content (d) of wild type, OXAas, OXAccDACB/KOLipA and OXAas/KOAar Synechocystis strains grown in BG₁₁ medium. The error bars represent standard deviations of means (mean ± SD, n = 3). Means with the same letter are not significantly different (in b) with the significance level at P < 0.05

Fig. 5

Total lipid content (a) and unsaturated lipid content (b) in wild type, OXAas, OXAccDACB/KOLipA and OXAas/KOAar Synechocystis strains grown in BG₁₁ medium. The error bars represent standard deviations of means (mean ± SD, n = 3). Means with the same letter are not significantly different with the significance level at P < 0.05

Fig. 6

The transcript levels (a) of pha A, accA, aas, plsX, aar, lipA and 16S rRNA genes of WT, OXAas, OXAccDACB/KOLipA and OXAas/KOAar Synechocystis strains. The intensity ratios (b) of phaA/16S rRNA, accA/16S rRNA, aas/16S rRNA, plsX/16S rRNA, aar/16S rRNA and lipA/16S rRNA of all studied strains at log phase of cell growth analyzed by GelQuant.NET program

Fig. 7

The Nile red staining of neutral lipids in Synechocystis sp. PCC 6803 wild type (a) and OXAas (b) strain cells in BG₁₁ medium at day 4 of cultivation. The stained cells were visualized under light and fluorescent microscopes with a magnification of ×100. It is noted that the focus setting in panel B with Nile Red staining was directed to PHB granules whereas the focus setting in panel A with Nile Red staining was directed to the whole cells