Cyanobacteria as whole-cell factories: current status and future prospectives

Cyanobacteria bear great potential to function as cell factories for producing chemicals using sunlight as sole energy source via oxygenic photosynthesis 1, 2. These organisms use light to oxidize water and provide electrons to run an autotrophic metabolism based on CO2 fixation. Compared with terrestrial plants, the CO2 fixation efficiency of cyanobacteria is 10–50% higher 3, 4, giving them a high potential for carbon sequestration and the ecoefficient production of fuels and chemicals. Whereas eukaryotic microalgae bear a similar photosynthetic potential, cyanobacteria profit from a lower degree of structural complexity, which simplifies strain and metabolic engineering. Currently, cyanobacteria are engineered to produce diverse compounds ranging from short-chain hydrocarbons and low-molecular synthons to complex bioactive compounds of nutritional or pharmaceutical interest [2]. However, compared with heterotrophic production strains, achieved rates, yields, and titers are typically lower and often below the needs of a commercial deployment. In this regard, energy allocation, that is, light supply, constitutes a major constraint raising the question: does sunlight provide enough energy and is its photosynthetic exploitation efficient enough to sustainably drive production? Albeit sunlight is an abundant energy source, it is areawise rather ‘diluted’. In plants or microalgae, the maximum theoretical efficiency of the photosynthetic apparatus to convert light energy into chemical energy is 26%, but the overall photosynthetic CO2 fixation reaches a real maximum of only 1–3% [5]. Thus, the development of efficient photosynthesis-based production processes possibly requires an enhancement of photosynthetic efficiency in addition to the implementation of productive pathways and their efficient coupling to native metabolism. Thereby, genetic tools and knowledge on regulatory networks are required to control pathway operation as well as electron and carbon fluxes toward products. Besides strain design, reaction/reactor engineering is crucial to attain high active biomass concentrations and resilience to temperature and light intensity variations.

Against this background, recently developed strategies to improve productivities and yields in cyanobacteria are reviewed. For the production of chemicals, one can distinguish electron- and CO2-based product formation. The former only uses light-driven water oxidation to fuel redox-intensive bioconversions, whereas the latter also requires diversion of CO2-derived carbon (Figure 1). Both approaches demand specific strategies. Overall, we aim to answer the following questions: What chemicals can be produced in cyanobacteria and with what efficiency? What are currently the best production strains? What are the limitations of/in cyanobacterial cells? What are promising metabolic, regulatory, and biochemical engineering targets?

Fast growth is considered a central selection criterium for cyanobacterial production strains and heavily depends on the conditions applied in terms of light and CO2 availability and medium composition [6]. While widely used models such as Synechocystis sp. PCC 6803 (Synechocystis, 4.3 h doubling time) show moderate growth rates, faster growth has been reported for several Synechococcus elongatus strains, namely PCC 11801, 11802, and 11901 and UTEX 2973, the latter reaching a minimal doubling time of 1.5 h. In-depth studies on the molecular basis of fast autotrophic growth revealed that it correlates with increased production of ATP and NADPH, high CO2 uptake rates, and elevated transcription of genes for precursor biosynthesis and protein translation [6]. Model-guided pathway design and host strain engineering based on such findings combined with advanced synthetic biology tools and high-cell-density (HCD) cultivation/bioprocessing technologies will be central to improve performance parameters (Figure 1). Thus, an integrated, multidisciplinary approach is required.

As an example for CO2-based formation of high-volume products, the combined production of isobutanol and 3-methyl-1-butanol has been tackled. NADPH-dependent and NADH-evolving production of these two alcohols in Synechococcus sp. PCC 7002 was improved by introducing NADH-dependent nitrate assimilation as an additional NADH sink [7]. Another example is ethylene, whose production from CO2 profited from a deeper understanding of associated metabolic limitations [8]. However, titers achieved for these two products remained rather low (∼300 mg L−1). Further optimization may be achieved by a truly integrated approach. For lactate production, CRISPRi (CRISPR interference)-mediated cycling between growth and production phases has been presented as an interesting strategy to increase product titer (to 1 g L−1) as well as process stability [9]. Sucrose represents another well-investigated potential high-volume product. Its synthesis and intracellular accumulation could be improved in several species by manipulating gene expression for two key enzymes 6, 10. Efficient sucrose export was achieved by introducing a heterologous sucrose permease [10]. Highest product titer (8 g L−1) and productivity (1.9 g L−1 day−1) were reached with UTEX 2973 under salt stress conditions, with a cell dry weight (CDW)-based sucrose yield of 3.1 g gCDW−1 [10]. A life cycle assessment (LCA) to investigate the environmental impact and limitations of cyanobacteria-based butanol production revealed a high cumulative energy demand in all tested scenarios, indicating that significant metabolic engineering toward a carbon partitioning of> 90% as well as improved light utilization are necessary to displace fossil fuels or even first- and second-generation bioethanol [11].

For higher-value products, cyanobacterial process commercialization is closer or even already has happened, with pigment production as a prominent example 12, 13. Phycobiliprotein production by Anabaena variabilis CCC421 was enhanced via medium engineering to a product titer of 408.5 mg L−1, demonstrating the potential to efficiently produce a specific protein in cyanobacteria [13]. Further, many studies have tackled lipid and fatty acid production in cyanobacteria with a special focus on the dietary omega-3, polyunsaturated acids with yields up to 100 mg gCDW−1 [14]. To better exploit cyanobacterial cell factories for high-volume product formation, combined production of ethylene together with carotenoids was targeted [15]. For this purpose, Synechococcus sp. PCC 7002 was engineered and applied in a 100 L air-lift photobioreactor (PBR), enabling ethylene and biomass productivities of 2.5 mL L−1 h−1 and 0.3 gCDW L−1 d−1, respectively, and a carotenoid titer of 64.4 mg L−1.

Bioconversions exclusively driven by the light reaction independently from CO2 fixation (Figure 1) represent an interesting novel approach to make efficient use of natural photosynthesis 16, 17, 18. Biocatalytic redox reactions depend on efficient and sustainable redox cofactor supply, which can be established via photosynthetic water oxidation — primarily in the form of reduced ferredoxin (Fd) and NADPH — making use of abundant, cheap, land-sparing, and thus highly sustainable resources. Fd-dependent cytochrome P450 monooxygenases (CYP450s) have been applied, with specific activities up to 40 U gCDW−1 achieved on bioreactor scale [19]. Further, Bayer–Villiger monooxygenases (BVMOs) and reductases have been applied with specific activities up to 150 U gCDW−1 18, 20, 21•. Oxygenase catalysis thereby also profits from O2 derived from water oxidation, a substrate typically limiting the performance of heterotrophic microbes 17, 19. Another main product targeted via light reaction coupling is H2, with efficient electron transfer and the O2 sensitivity of H2- producing enzymes as major challenges [22]. Recent advances include hydrogenase–photosystem I (PSI) fusion and the introduction of an O2-tolerant hydrogenase into Synechocystis 23, 24, 25. The combination of these approaches and/or the generation of a low O2 environment via technical measures for in situ gas removal constitute promising strategies to achieve efficient photosynthetic H2 production.

Electron drainage via biotransformation reactions may cause an ATP/NADPH imbalance, (possibly) affecting both the biocatalytic reaction and host metabolism [26]. Changes in electron demand have, for example, been shown to involve changes in PSI photochemistry [27]. Further, light stress constitutes a major limitation in nature. The photosynthetic apparatus is evolutionarily optimized to handle overexcitation rather than to maximize light usage efficiency [28]. The introduction of heterologous pathways relying on CO2 fixation and/or reducing power indicated that they can act as a photosynthetic sink and partially substitute photoprotective mechanisms, while potentially conserving otherwise lost energy within useful products 26••, 29. In the following, novel promising approaches to improve photosynthesis-based product formation are reviewed in more depth focusing on photosynthetic efficiency, regulatory network engineering, and biochemical process engineering.

To prevent losses in energy conversion, the photosynthetic light reactions need to be improved. Exemplary studies targeting antenna systems and other photosynthetic proteins as well as the content and ratio of photosynthetic pigments have been reviewed recently [30]. To enhance photosynthetic efficiency, two major strategies can be defined: (i) expanding the absorbed light spectrum, especially toward infrared light, and (ii) reducing the energy loss associated with photoinhibition or photodamage. Photosynthetic organisms developed multiple mechanisms to protect themselves from excessive light exposure at the expense of photosynthetic efficiency. The latter can be improved either via efficient rerouting of electrons to productive enzymes or by eliminating such protection mechanisms such as flavodiiron proteins consuming up 20% of the electrons originating from photosystem II 31, 32. Spontaneous electron rerouting to a CYP450 in Synechococcus sp. PCC 7002 has been reported to result in reduced expression of natural electron dissipation pathways [33]. Further, catalytic CYP450 performance could be fostered by deleting the respiratory cytochrome-c oxidase.

Direct coupling of desired redox reactions to the photosynthetic apparatus is another approach to increase productivity as recently demonstrated and mentioned above for hydrogenase-based H2 production [22]. Other examples are CYP450s anchored with a linker to the thylakoid membrane [34]. These approaches follow the hypothesis that reducing the distance to the photosynthetic apparatus gives attached enzymes an advantage over competing intrinsic pathways, such as the Calvin–Benson–Bassham (CBB) cycle or nitrate assimilation. As an alternative, the fusion of a CYP450 to a flavodoxin-like electron carrier, with a redox potential suitable for electron transfer to the CYP450 but not to other native electron sinks, has been shown to enable a nearly 25-fold improvement in in vivo electron transfer to the CYP450 on a per-protein basis [34]. Furthermore, modulation of the ferredoxin-NADP+ reductase (FNR) could enhance the electron flow into the desired pathway [35]. For the production of chemicals from CO2, CBB cycle operation must also be efficient. Several studies tried to enhance the CBB cycle and streamline carbon flow toward the desired product [36]. For this purpose, determination of bottlenecks via genome-scale modeling and fluxomics as well as detailed knowledge of product-forming enzymes are crucial.

Finally, cell shading and light scattering effects obviate light penetration at high cell densities, leading to low and/or highly fluctuating light conditions. These can be mitigated by reducing the light absorption per cell, for example, via the reduction of antenna systems, which simultaneously reduces light stress [37]. Strain engineering toward high light resistance also can improve productivity, as it has been achieved by adaptive laboratory evolution [38], adapting Synechocystis to cope with high light on account of light-harvesting capacity. Additionally, screening for and analyzing fast-growing cyanobacteria with high photosynthetic activity can disclose strategies to improve photosynthesis in workhorse strains such as Synechocystis [39].

Although we are only at the beginning of a full understanding of the regulation of cyanobacterial metabolism, recent research has opened the window toward a more comprehensive view on fundamental control principles. Obviously, small regulatory proteins of less than 100 amino acids play a crucial role [40]. For instance, the CBB cycle is mainly controlled by CP12 that interacts with and inhibits glyceraldehyde-3-phosphate dehydrogenase and phosphoribulokinase, when light availability is not sufficient 41, 42. Adequate assembly and hence activity of ribulose-1,5-bisphosphate carboxylase-oxygenase requires the small chaperone RbcX [43]. Efflux from the CBB cycle is then mainly controlled at the phosphoglycerate-mutase reaction via interaction with the recently discovered PirC protein [44]. Further examples of such regulatory protein–protein interactions have been revealed for the control of the TCA (tricarboxylic acid) cycle [45], nitrogen assimilation [46], the synthesis of key amino acids [47], ATP synthesis [48], and the photosynthetic apparatus itself [49].

Beside synthetic biology tools to introduce pathway genes and trigger their expression [50], strategies to redirect carbon flux toward specific routes have come into focus, for example, the blockage of competing pathways or storage compound synthesis. This is usually achieved by enzyme/pathway deletion, whereby the permanent loss of certain functions often is a major disadvantage, for example, under (fluctuating or harsh) outdoor conditions, especially in the long term. Intrinsic regulators punctually addressing metabolic reactions pose interesting alternative engineering targets to fine-tune metabolic fluxes toward products. For example, deletion of a transcriptional repressor of the glucosylgycerol phosphate synthase gene in Synechocystis increased glucosylglyerol yields [51]. This also has been demonstrated for enzyme effectors such as CP12 or PirC. CP12 deletion, for example, resulted in higher production rates for 2,3-butanediol in S. elongatus PCC 7942 [52] or bisabolene and limonene in Synechocystis [53]. Similarly, PirC deletion in Synechocystis overexpressing genes for the plastic alternative polyhydroxybutyrate (PHB) resulted in increased production yields of up to 0.81 g gCDW−1 [54]. This impressively demonstrates the power of regulatory engineering as a rather new metabolic engineering concept for cyanobacteria. Further understanding of these control principles may allow a transfer to other pathways and the design of synthetic effectors to control fluxes through any enzyme of interest.

One of the critical bottlenecks for phototrophic bioprocesses is the lack of suitable PBRs allowing cost-effective HCD cultivation. The light-energy dependence resulted in diverse reactor concepts, including tubular, flat-panel, and vertical-column PBRs, to maximize light supply while satisfying constraints such as nutrient and CO2 supply, O2 removal, temperature and pH regulation, and scalability. Recent publications address available PBR concepts and modeling tools for optimizing reactor performance 55, 56. Illumination via the reactor surface and light sources distributed inside liquid cultures constitute two general light-supply strategies. Light sources inside liquid cultures can distribute light more uniformly and enable higher light utilization efficiencies [57], as investigated for the reduction of 2-methylmaleimide to (R)-2-methylsuccinimide by Synechocystis [20]. Thereby, suspended wireless light emitters enabled more than twofold improvement in specific activity compared with surface-illuminated systems. However, the specific activity dropped by 61% upon an increase in cell density from 0.48 to 2.4 gCDW L−1, restricting HCD cultivation. Additionally, technological complexity, scaling issues, and energy demands (for example for LEDs) can limit application.

Tubular PBRs featuring external illumination are recognized as mature cultivation systems for commercial applications [58]. Typically, tubes with 10–60 mm inner diameter offer a surface area to volume ratio (SA/V) of over 100 m2 m−3 but poor light penetration depth and limited gas transfer, that is, CO2 supply and removal of otherwise inhibitory O2 [59]. Recently, different geometric configurations, such as capillary reactors or tree-like structures, have been investigated to maximize the SA/V ratio. For example, capillary reactors with inner diameters of 3 mm provide a high SA/V ratio of 1333 m2 m−3, require low light penetration depth, and offer the option of slug flow mode operation with gas slugs enabling high gas mass transfer 60, 61•. Importantly, such systems enable cultivation and operation of immobilized cells, for example, making use of the native ability of microbes to immobilize as biofilms. In such a system, dual-species phototrophic biofilm cultivation resulted in a high cell density (above 30 g LCDW−1) of phototrophic microorganisms 61•, 62. Such capillary biofilm reactor operation enabled continuous cyclohexane oxidation with a volumetric productivity of 0.2 g L−1 h−1. Critical aspects for such systems include mass transfer and possible cell heterogeneities in the biofilm [63].

Often, outdoor cultivation is targeted for commercial application. Thus, challenges and limitations arising from the variation in sunlight intensity (including the day–night cycle), adequate CO2 availability, temperature variations, and contamination issues must be considered. Based on simulations, a V-shaped PBR has been proposed for low-latitude locations, giving 1.4-times higher biomass productivities than flat horizontal PBRs [64]. Such modeling approaches augur well to accelerate the design and scale-up of PBRs for efficient outdoor operation 55, 56, 64.

留言 (0)

沒有登入
gif