Diatoms for Carbon Sequestration and Bio-Based Manufacturing

Carbon dioxide (CO2) is a major greenhouse gas responsible for climate change. Diatoms, a natural sink of atmospheric CO2, can be cultivated industrially in autotrophic and mixotrophic modes for the purpose of CO2 sequestration. In addition, the metabolic diversity exhibited by this group of photosynthetic organisms provides avenues to redirect the captured carbon into products of value. These include lipids, omega-3 fatty acids, pigments, antioxidants, exopolysaccharides, sulphated polysaccharides, and other valuable metabolites that can be produced in environmentally sustainable bio-manufacturing processes. To realize the potential of diatoms, expansion of our knowledge of carbon supply, CO2 uptake and fixation by these organisms, in conjunction with ways to enhance metabolic routing of the fixed carbon to products of value is required. In this review, current knowledge is explored, with an evaluation of the potential of diatoms for carbon capture and bio-based manufacturing.

recombinant antibodies) as a viable cell factory, whilst enabling strategies to reduce CO 2 in the atmosphere [23].
This review showcases the potential of diatoms for CO 2 sequestration, coupled with bio-based manufacturing, highlighting the challenges to be overcome for a commercially viable, sustainable manufacturing solution. A strong emphasis is given to the mechanisms responsible for carbon acquisition, transport, and processing in diatoms, as a target for improvement of carbon fixation.

Diatoms for Bio-Based Manufacturing
Diatoms are unicellular microalgae possessing a silicon-based cell wall, and belong to the class Bacillariophyceae. They are an ecologically successful taxonomic group of phytoplankton. They contribute heavily to the global primary productivity [17,24], and play fundamental roles in the global nutrient cycling of carbon, nitrogen, phosphorus, and silicon [25,26]. The silica exoskeleton provides diatoms with structural integrity and protection in the ocean environment. Silicification increases cell density, enabling the cells to sink; possibly a selective evolutionary trait to move the cell to more optimal growth environment deeper in the water column, and evolved as a selection pressure against parasitism [27] that can be useful in establishing cost-effective harvesting methods. Their ability to prosper in the natural environment indicates their suitability for large scale cultivations in less sterile environments, to enable viable industrial scale operations. Diatoms have considerable metabolic diversity attributable to their evolution that involved endosymbiosis of diverse lineages. As a result, they can be employed to produce diverse chemicals. Manipulation of CO 2 supply can also be used to improve the accumulation of both lipids and carbohydrates, as has been studied in T. pseudonana, P. tricornutum, Asterionella formosa and Navicula pelliculosa [28]. The presence of efficient uptake systems for CO 2 and bicarbonate (HCO 3 − ) have been identified in the diatoms T. weissflogii and P. tricornutum, at concentrations typically encountered in ocean surface waters. The ability to adjust uptake rates to a wide range of inorganic carbon supply has also been reported [29]. Nevertheless, there is paucity of information and evidence regarding CO 2 uptake and there are many unanswered questions. In addition to photo-autotrophy, mixotrophic cultivation regimes can help yield higher biomass concentrations and productivities. Diatoms can be cultivated in both indoor and outdoor settings, as suspension cultures (in open ponds, flat panel, tubular and airlift photobioreactors (PBRs)), as well as immobilized cultivation systems to avoid dewatering costs. P. tricornutum biomass productivity was found to be doubled in high-technology photobioreactors to 21 ± 2.3 g m −2 d −1 , compared to cultivation in open ponds, and resulted in a CO 2 fixation rate of 35.5 g −1 m −2 [30]. Overall, this gives flexibility in cultivation, as different cultivation methods can be used to enhance productivity. Novel culturing media, such as FDMed medium, have been used for high biomass, fucoxanthin and EPA production yields in freshwater diatoms, such as Sellaphora minima and Nitzschia palea in autotrophic batch cultures [31]. Media for cultivation of freshwater diatoms include: FDMed medium [31], WC [32] and modified COMBO (MCOMBO) (modified COMBO (MCOMBO)) medium of the UTEX Culture Collection of Algae). F/2 media [33], DAM (diatom artificial medium; [34]), ASW (artificial sea water; [35]), Walne media [36] can be used for the culturing of marine diatoms. Yeast extract supplementation of F/2 media has been reported to result in increased biomass concentration (3.48 fold), TAG content (2.13 fold) and fucoxanthin content (1.7 fold) in the stationary phase [37].
Optimization of operational conditions has been shown to be useful in increasing product yields. Several such studies have been reported with the model diatom, P. tricornutum, for example, light shift with tryptone addition to improve fucoxanthin production [38], UV mutagenesis to improve EPA productivity by 33% [39], adaptive laboratory evolution to improve neutral lipid and carotenoid accumulation [40]. Marginal improvement in total lipid contents, in association with reduced poly unsaturated fatty acids, have been observed in Cyclotella cryptica as a result of silicate deprivation [41]. In the case of P. tricornutum, a weakly silicified diatom, the required quantities of silicon can be obtained from silicon dissolved from glass vessels in alkaline culture media [42]. P. tricornutum grown in the presence and absence of silicon showed little difference in growth, except under low light and green light conditions [43].
In addition to their fundamental role in global nutrient cycles, diatoms represent a potential bioprocess platform, for synthesizing biofuels and other value-added products. Microalgae, in general, are of considerable interest, because many accumulate significant amounts of energy-rich compounds, such as triacylglycerol (TAG), or other lipids that can be used as biofuel precursors [44]. Diatoms have been showcased to produce both homologous and heterologous compounds, proteins and other products (Table 1). Table 1. Exemplar products from diatoms for bio-manufacturing, using autotrophy or mixotrophy.

Carbon Assimilation in Diatoms
Carbon can be found in many forms in the natural environment. In the oceans, the dynamics of chemical dissolution of CO 2 and its biological uptake creates an interplay between chemical and biological equilibria that requires further elucidations and understanding. For terrestrial photosynthetic organisms, atmospheric CO 2 is the main form of inorganic CO 2 assimilated, but in water, the dissolution of CO 2 results in carbonic acid, which dissociates into bicarbonate and carbonate. In the oceans, 90% of inorganic carbon is in the form of bicarbonate [80]. Prior to the industrial revolution, CO 2 concentrations in the atmosphere were~280 ppm [81], but today they have increased to~420 ppm in 2020 (https://www.co2.earth/), with an increasing proportion of CO 2 sequestered in the oceans and on land. At pre-industrial concentrations of atmospheric CO 2 , the seawater concentration of bicarbonate was 1757 µmol kg −1 , but elevated levels of bicarbonate are now being observed, contributing to ocean acidification and a higher solubility of carbonate [80].
The effect of increasing CO 2 concentration supply to diatoms leads to increased growth and biomass production, under growth optimal conditions. Carbon capture in P. tricornutum happens predominantly in the form of bicarbonates with bicarbonate transporters [82], and as mentioned above, CO 2 fixation rates of 35 g m −2 d −1 have been reported [30]. When cultivating P. tricornutum in air sparged cultures, a CO 2 consumption rate of 1 g g −1 DW, at pH 7.2 and 0.8 g g −1 DW, at pH 9, both resulting in 0.06-0.08 g CO 2 uptake per day (removal of 50-65% of CO 2 from the air), has been reported [83]. It has been identified that the optimal CO 2 concentration for biomass accumulation is in a narrow range, between 1% and 1.25% CO 2 in air (v/v), at a gas supply rate of 0.66 vvm and light intensity of 1000 µmol m −2 s −1 (16 h light period), 90% of CO 2 supplied leaving the medium unused [84]. When P. tricornutum was provided with bicarbonate as an inorganic carbon source, between 73-99.9% of the bicarbonate was consumed or remained dissolved in the medium, resulting in a CO 2 consumption rate of 0.31 g d −1 (2.3 g CO 2 g −1 biomass), albeit at the cost of reduced growth and biomass production [83]. Cultivations of P. tricornutum (PHAEO2) in modified F/2 seawater (enriched four-fold with nitrogen and phosphorus) with 15% CO 2 have been shown to increase biomass productivity to 0.15 g L −1 d −1 , whilst consuming 0.28 g L −1 d −1 of CO 2 in a batch operation [85]. A comparative assessment of CO 2 concentration mechanisms (CCMs) in a handful of freshwater and marine diatoms (P. tricornutum, As. formosa, N. pellicosa, T. pseudonana, T. weissflogii) revealed that, for all the species, at 20,000 ppm, the affinity for DIC was lower than at 400 ppm CO 2 (atmospheric concentrations), and the reliance on CO 2 was higher, and that species-specific differences were greater than environmental differences, in determining the effectiveness of the CCMs [86]. Negative effects of CO 2 on growth have also been recorded. For example, Attheya longicornis growth was hampered by high levels of CO 2 supply [87]. Another factor affecting marine species is temperature. Rising temperatures may also have a negative effect on the CO 2 uptake rate by diatoms. In Navicula distans, rising temperature and pCO 2 resulted in a reduction of diatom cell size, which inevitably relates to the ecological and physiological functions of diatoms, such as nutrient diffusion, intake and requirements, and even the metabolic rate [88]. There are also some cases recorded where no reaction to increased CO 2 levels could be observed, as seen with Chaetoceros brevis cultures supplemented with pCO 2 (750 ppmv (2 × ambient) and 190 ppmv (0.5 × ambient) CO 2 ), where little or no significant effect was observed on the diatom growth, pigment content and composition, photosynthesis, photoprotection and RuBisCO activity [89]. CO 2 uptake in the aquatic photosynthetic organisms, such as diatoms, cyanobacteria and other microalgae, take place with the involvement of the CCM. Carbon metabolism pathways in diatoms, like in plants and other algae, require the transportation of CO 2 across intracellular compartments like the peroxisomes, chloroplasts, mitochondria, endoplasmic reticulum and the cytosol, with concentration at the site where RuBisCO is located for CO 2 fixation ( Figure 1). This arrangement gives flexibility to the cell to adjust the carbon flux, enhance the concentration of CO 2 in a stepwise manner from low concentrations on the outside to levels required for RuBisCO activity and hence fix CO 2 [90,91]. There is limited information available on carbon metabolism in several diatoms, as of now. In order to obtain the appropriate design of the carbon flow in diatom cells under different conditions, information regarding the localization and functionality of the component diatom enzymes is a necessity. The cellular machinery involved in diatom photosynthesis includes the chloroplasts, carbonic anhydrases (CAs), RuBisCO, Calvin Benson Bassham (CBB) cycle proteins, transporters, phosphoglycerate kinase (PGK), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), CP12, fructose-1,6-biphosphatase (FBPase), sedoheptulose-1,7-biphosphatase (SBPase), phosphoribulokinase (PRK), basic leucine zipper (bZip) bZIP transcription factors family, and others [24,82,86,92,93]. Diurnal rhythms also affect the TCA Cycle and that influences the amount of CO 2 that is absorbed. Moreover, the bZIP14 protein family members are involved in CO 2 sensing and blue light signaling [94]. Our current knowledge of diatom CCMs is discussed in the section below, an understanding of which will help in devising strategies to maximize uptake of CO 2 by diatoms. Figure 1. Carbon dioxide (CO 2 ) enrichment in cell organelles. In diatoms, there are different membranes for CO 2 to cross, and it has to be enriched from low to concentrated levels near RuBisCO to enable CO 2 fixation. The periplasmic space faces the extracellular environment. Adjacent to the periplasmic space is the cytoplasm. Further inward is the chloroplastic endoplasmic reticulum (CER)/periplastidal compartment (PPC). The stroma is the layer beyond the CER/PPC. The innermost layer is the pyrenoid where in embedded is the pyrenoid penetrating thylakoid. This usually happens in general, with CO 2 concentration mechanisms (CCM) in microalgae.

The Diatom CCM and the Chloroplast Pump Model
The physical constraints on photosynthesis in the marine environment, especially low (dissolved) CO 2 in seawater, is partially mitigated by the CCM. Photosynthetic CO 2 fixation was found to be near saturation at external concentrations of 200 µM (as bicarbonate and CO 2 ) [95]. The CCM of P. tricornutum is reported to be moderately efficient, with around one-third of the carbon transported into the chloroplast being fixed by RuBisCO, and the remainder leaking out as CO 2 , attributable to the limited permeability of diatom membranes to CO 2 . The major driver of the CCM is believed to be the chloroplast pump, which actively transports bicarbonate into the chloroplast, where the bicarbonate flux into the chloroplast exceeds the net CO 2 and bicarbonate flux across the plasmalemma. Additional bicarbonate transporters are required in the membranes surrounding the chloroplast [82]. A large inorganic carbon pool is accumulated in the chloroplast, hence elevating CO 2 concentrations around RuBisCO and inorganic carbon is depleted from around the cytoplasm resulting in a diffusive influx of CO 2 from the extracellular environment into the cytoplasm [96]. Some of the proteins responsible for bicarbonate transport are embedded in the chloroplast membrane, but the full characterization of these proteins is ongoing. It is hypothesized that a different solute career 4 (SLC4) family of transporters in sequence move bicarbonate from the external environment to the chloroplast stroma [82].
Cyanobacteria and green algae, such as Chlamydomonas reinhardtii, developed CCMs ( Figure 2), to reduce the impact of the oxygenase activity of RuBisCO. The most characterized eukaryotic CCM is that found in C. reinhardtii. Inorganic carbon (as bicarbonate) is pumped into the chloroplast by active transport, where it is converted to CO 2 by carbonic anhydrase (CA), CAH3 (localized in the thylkaoid lumen). The stromal soluble protein complex (CrLCIB-LCIC) has the ability to re-capture and to prevent the leakage of CO 2 generated by CAH3. Moreover, CrLCIB-LCIC is not fixed by RuBisCO [97]. C. reinhardtii actively transports both CO 2 and HCO 3 − across the plasmalemma, but CO 2 appears to be the preferred form. Proton leakage through the thinner diffusive boundary layer is important in smaller organisms that have a smaller size [98]. The basic features of a cyanobacterial CCM include transport of inorganic carbon and the presence of carboxysomes that help in minimizing CO 2 leakage ( Figure 2). The induction of the CCM takes place at low CO 2 levels. DIC transporters are involved in maintaining the supply of CO 2 to RuBisCO and CAs are utilized for DIC accumulation. Diatoms use a CCM to overcome the difficulties of CO 2 limitation in alkaline and high-salinity seawater, by using SLC4 family transporters to take up HCO 3 − actively from the surrounding seawater, leading to the intracellular accumulation of DIC [99,100]. Multiple CAs maintain pH within each of the organelles, by maintaining a fine ratio of CO 2 and bicarbonate (Figures 2-4).   There is interconversion involved between CO 2 and HCO 3 − in the outermost and innermost layers also [82].

Carbon Transport Systems
Diatoms can actively take up HCO 3 − and/or CO 2 and the uptake of DIC across the plasma membrane is the critical first step in using DIC for photosynthesis [101]. The most well studied CCMs in diatoms are in P. tricornutum and T. pseudonana ( Figure 4). Adequate flux of CO 2 to ensure an optimal CO 2 to RuBisCO ratio (amount of CO 2 as substrate for the RuBisCO molecule) is generated and is facilitated by two pyrenoidal β-carbonic anhydrases (PtCA1 and PtCA2) [96,102,103]. The amount of CO 2 is complemented by α-type CAs that are located in the vicinity and in the sub-cellular spaces of the four-layered chloroplast membranes, which prevent the leakage of CO 2 from the chloroplast in T. pseudonana and P. tricornutum [103,104]. In a diffusion-based CO 2 uptake system, HCO 3 − uptake is by plasma membrane based SLC4, and CO 2 is taken up from the external environment directly through the cell membranes, as they are permeable, and CO 2 cannot passively pass through a transporter. CO 2 can only be taken up by generating a CO 2 deficit inside the cell through a diffusive mechanism, which leads to the suction of CO 2 from the external environment [86]. The active transport of HCO 3 − out of the cytoplasm and into the chloroplast leads to the formation of a low HCO 3 − concentrated environment in the cytoplasm of diatoms. The action of a cytoplasmic CA leads to a reduction in the cytoplasmic CO 2 . The conversion of CO 2 to HCO 3 − occurs when the HCO 3 − concentration is below equilibrium with CO 2 . The CO 2 gradient leads to its passive diffusion into the cell across the plasma membrane, and the continued export of HCO 3 − from the cytoplasm maintains a constant cytoplasmic CO 2 deficit, resulting in sustained CO 2 uptake. In order to maintain the inward CO 2 suction, the activities of the transporters exporting HCO 3 − from the cytoplasm should be more than that of CO 2 and HCO 3 − influx into the cell [101]. The SLC4 family have been found to represent a major group of bicarbonate transporters which have an important role in pH regulation. It has been found that the N-terminal (Nt) domain (involved in functional regulation of transporters) accounts for 32-55% of the entire polypeptides of the SLC4 transporters. Interestingly, the SLC4 like transporters in plants lack the large Nt domain found in mammalian homologs, which has been found to be non-essential for the transport of anion exchangers of the SLC4 family [105]. Bicarbonate transport at the molecular level has predominantly been studied in the model diatoms P. tricornutum and T. pseudonana [82,106]. Ten putative bicarbonate transporters have been identified in P. tricornutum, which are similar to those found in mammalian protein families (SLC4 and SLC26) [80].
Three different transport systems can be identified in diatoms (plasma membrane, plastid and aquaporin). The first is the plasma membrane-based bicarbonate transport system (PtSLC4). Under CO 2 -limiting conditions, PtSLC4-2 in the plasmalemma is induced, and its expression increases with DIC uptake and photosynthetic activity. In the presence of a high concentration of sodium ions, PtSLC4-2 transports bicarbonate and it has a saturation limit of~100 mM sodium ions. PtSLC4-1 appears to be a sodium dependent bicarbonate transporter, and its function was inhibited by the addition of an anion-exchanger inhibitor. The bicarbonate uptake rate of PtSLC4-2 is highest at pH 8.2, equivalent to the pH of seawater [100]. PtSLC4-2 is involved in the direct uptake of bicarbonate, and is actively involved in acquisition of extracellular DIC under low CO 2 conditions, in P. tricornutum. SLC4 homologs have also been characterized in T. pseudonana [100]. Two putative bicarbonate transporters, PtSLC4-1 and PtSLC4-4, are found to be highly conserved in comparison to PtSLC4-2. PtSLC4-1 and PtSLC4-4 are also induced specifically under CO 2 -limiting conditions, and it has been suggested that these transporters contribute heavily to bicarbonate influx into the cell in seawater in a CO 2 -limiting environment [101,107]. Other SLC4 homologs are currently being studied in the chloroplast membranes of P. tricornutum, and more types of transporters are awaiting identification of location and functionality [101]. It has been suggested that different SLC4 transporters translocate bicarbonate from the environment to the chloroplast stroma [82]. A different group of SLC4 transporters in the chloroplast envelope are thought to transport bicarbonate to the chloroplast stroma [80].
In plastid membrane-based HCO 3 − transport systems, the DIC that is imported into the cytosol is not able to freely diffuse into the chloroplast for fixation. This is due to the chloroplast membrane, which is four-layered. At each of the four-layers, HCO 3 − transporters should be present to regulate the transport, aided by the CAs that should be present in close proximity to the chloroplast membranes and the transporters [96,101]. However, such HCO 3 − transporters are only proposed, and yet to be identified in diatoms. From genome annotations in P. tricornutum, it has been elucidated that there are three putative PtSLC26 genes, and seven PtSLC4 genes [100]. This displays the limited information regarding its location, structure, and the overall functionality in the scheme of carbon transport within the system. PtSLC4-6 and PtSLC4-7 genes have been found to be active and upregulated under high CO 2 and low CO 2 conditions [100], implying that PtSLC4-6 and PtSLC4-7 act as DIC transporters, and are involved in the regulation of the DIC flow from the cytosol to the plastid, and remains unaffected by the ambient CO 2 levels. Presently, the information regarding the mechanism of activity of the DIC transporters, such as PtSLC4-6 and PtSLC4-7, still remains highly limited [101]. CO 2 permeation through diatom membranes is very rapid [102], and this high permeability may be, in part, due to the presence of channels such as aquaporins (AQPs) [101]. In the aquatic environments, AQPs are ubiquitous water channels that have been known to facilitate the transport of many small molecules such as CO 2 and ammonia. AQPs are involved primarily in mechanisms that are responsible for maintaining the transmembrane fluxes of important small molecules, that are yet to be studied in marine photoautotrophic organisms. In silico analysis has revealed the presence of five AQP orthologs in P. tricornutum and two in T. pseudonana [108]. It is also believed that lipid bilayers being inherently permeable to CO 2 , also do present some resistance to diffusion that could be reduced by the presence of AQPs, as a result of which DIC can be easily imported into the cell and fixed, albeit with minimal energy expenditure.

Carbonic Anhydrase-Isoforms and Activation
The primary function of CAs is to interconvert CO 2 and bicarbonate, to enable transport across membranes and prevent loss of CO 2 . In addition, they are responsible for recovering CO 2 leaked from the chloroplast. This implies that a CCM is primarily regulated by CAs. CO 2 uptake is due to the internal CO 2 deficit generated by CA-catalyzed hydration of CO 2 to HCO 3 − in the cytoplasm [101]. CO 2 uptake by its synthesis from HCO 3 − at close proximity to the cell surface can also be done by external CAs, located in the periplasmic space [109]. There is a complex plethora of possible carbonic anhydrase isoforms from diatoms. They are found in diatoms, with varying capabilities in terms of carbon absorption, assimilation and utilization. All of the identified CAs are at different stages of confirmation, with respect to localization, functionalization and identity [82]. Ten putative CA genes have been identified in P. tricornutum [110], five α-CA, two β-CAs, two γ-CAs and one θ-CA [103,111].
Notably, α-CAs are located at the four-layered chloroplast membrane, β-CAs in the pyrenoid, γ-CAs in the mitochondria and the θ-CA in the thylakoid lumen. CA activity has only been verified with two β-CAs and one θ-CA [101]. There are no free stromal CAs, no external and no cytosolic CAs identified in P. tricornutum [111]. CAs are metallo-enzymes, and zinc is essential for their activity [80]. However, several coastal diatoms have cadmium containing CAs, and this is considered an evolutionary adaptation to low zinc in marine habitats, but the cadmium at the catalytic site of the ζ-CA can be exchanged for zinc [112,113]. CA enzymes are ubiquitous in nature, and provide an example of convergent evolution. They appear to have a diverse role in many biological processes, including CO 2 fixation, pH homeostasis, and the transport of CO 2 /bicarbonate. Seven distinct classes of CAs have been identified, i.e., α, β, γ, δ, ζ, η, and θ-Cas, out of which α, β, γ CAs are found in higher land plants, but δ and ζ CAs are restricted to marine diatoms [80]. The θ-CA has been found to be widely distributed in algae and cyanobacteria, and it has been reported to be essential for photosynthesis in P. tricornutum [111].
Two pyrenoidal β-carbonic anhydrases (CAs) have already been isolated and characterized in P. tricornutum [102,103,114], but the actual reason for the existence of different CAs is not known, but the diversity could be due to different locational requirements. New subsets of CAs in T. weissflogii and T. pseudonana, have been identified. They are δ-CAs, ζ-CAs and θ-CAs [111,113,[115][116][117]. It is predicted that P. tricornutum lacks periplasmic CAs, however surprisingly possesses an efficient CCM, which may suggest that periplasmic CAs are not necessary for the operation of a CCM in microalgae [82]. Furthermore, diatom-specific adaptations in chloroplast metabolism highlight beneficial traits. These include the completion of tocopherol synthesis via a chloroplast-targeted tocopherol cyclase, a complete chloroplast ornithine cycle, and the chloroplast-targeted proteins involved in iron acquisition and CO 2 concentration not shared between diatoms and their closest relatives in the Stramenopiles family [118].
Ambient CO 2 is required to trigger the transcription of ptca1, and light affects the extent of acclimation [119]. An appropriate combination of CO 2 /cAMP-responsive elements i.e., CCRE1/2 or CCRE2/3 at proper distances from the minimal promoter are required as a potential target of the Zip protein PtbZIP11 for an effective CO 2 response of the ptca1 gene [120]. The detailed analysis of the promoter region of ptca2 appears to indicate that both CCRE2s are cis-elements governing the CO 2 /light response of ptca2 promoter [120]. The transcriptional activation of the two ptca promoters in CO 2 limitation was evident under illumination with a photosynthetically active light wavelength. An artificial electron acceptor from the reduction side of PSI efficiently inhibited ptca promoter activation, while neither inhibition of the linear electron transport from PSII to PSI, nor inhibition of ATP synthesis, showed an effect on the promoter activity, strongly suggesting a specific involvement of the redox level of the stromal side of the PSI in the CO 2 /light cross talk [121].

Pyrenoid Matrix
The pyrenoid is a protein body containing RuBisCO found in most algal chloroplasts (in the stroma) and RuBisCO is often coated with a starch sheath. Pyrenoids are associated with operation of CCM. The main function of pyrenoids is to act as centers of CO 2 fixation, generating and maintaining a CO 2 rich environment around RuBisCO. Bicarbonate supply could be used to elevate CO 2 reacting with CA in the pyrenoid [92]. In diatoms, pyrenoids are generally present, but in some species, their presence can be variable, even within a single genus.
There is a strong correlation between the presence of a pyrenoid and an active CCM in algae [92,100]. However, the presence of a pyrenoid does not always confirm the operation of a CCM. Of all the microalgae examined in any detail, it is clear that those with the highest affinity for CO 2 and clear CCM characteristics have a pyrenoid, and most probably a single chloroplast per cell. This is true of both the green and non-green algae. In the pyrenoid, bicarbonate is converted into CO 2 , resulting in a localized elevation in CO 2 concentration, favoring carboxylation by RuBisCO, over oxygenation. The three-dimensional structures of the chloroplast-pyrenoid in C. reinhardtii have been revealed, using in situ cryo-electron tomography [122]. It has been found that some of the thylakoid membrane can penetrate into the pyrenoid, called the pyrenoid-tubules and may have a role in the carbon capture process. The diatom, P. tricornutum, has a pyrenoid-based CCM, and contains a cluster of the genes homologous to C. reinhardtii (LCIB), PtLCIBI-4 [97].
As carbon availability is often the limiting factor for microalgal growth, most microalgal chloroplasts contain a pyrenoid with a high concentration of RuBisCO, for an effective CCM [123]. In addition, CO 2 responsive CAs occur in the pyrenoid of P. tricornutum [103]. For CO 2 fixation by RuBisCO, the transported HCO 3 − has to be converted to CO 2 . CO 2 can also be produced in P.
tricornutum, by importing HCO 3 − into the pyrenoid-penetrating thylakoid lumen. There, the activity of θ-CA and the low pH converts HCO 3 − to CO 2 for RuBisCO. β-CAs can also convert bicarbonate into CO 2 , thereby increasing the CO 2 concentration around RuBisCO [101]. The amount of CO 2 supplied to RuBisCO is fixed, but the rest of the CO 2 leaks out of the chloroplast [96]. The leaked CO 2 is recovered by CA-catalyzed conversion to HCO 3 − [82]. P. tricornutum lacks cytosolic and stromal CAs, but has numerous chloroplast envelope CAs. Such a design implies that the main recovery points of carbon are in the four-layered chloroplast envelope [103]. The essential pyrenoid component 1 (EPYC1) is a low complexity repeat protein which links RuBisCO to form the pyrenoid. EPYC1 is of comparable abundance to RuBisCO, and colocalizes with RuBisCO throughout the pyrenoid. EPYC1 is essential for normal pyrenoid size, number, morphology, RuBisCO content and efficient carbon fixation at low CO 2 . The most abundant proteins in the low CO 2 pyrenoid fraction are RuBisCO large (rbcL) and small (rbcS) subunits, and RuBisCO activase (RCA1) [124]. It has been found that pyrenoid-based CCM emerges as the most effective in achieving the greatest elevation of CO 2 [92]. CCMs increase the cellular CO 2 concentration around RuBisCO, resulting in a higher carbon fixation rate in a CO 2 limited environment. To avoid carboxylation/decarboxylation, and to ensure compartments in which CO 2 can be concentrated for carbon fixation by RuBisCO, diatoms may utilize their pyrenoid and four-layered plastids [125].

RuBisCO and Its Activation: The Effect of Glycolate
CO 2 is of limited supply to marine phytoplanktons, because of the low partial pressure of CO 2 in the atmosphere, a faster CO 2 hydration rate compared with bicarbonate dehydration rate, and the high salinity and alkalinity of seawater [125]. RuBisCO in diatoms have a CO 2 half saturation constant of 23-68 µM [126]. The concentration of CO 2 in seawater is 10-15 µM at pH 8.2, and therefore diatoms are thought to operate a CO 2 -CCM, to improve the efficiency of carbon fixation [125]. Photosynthesis requires the carboxylation of RuBP by RuBisCO, but photorespiration occurs when RuBisCO oxygenates, RuBP forming the toxic by-product glycolate which needs to be removed by the cell. Glycolate accumulation, as a result of photorespiration, appears to influence RuBisCO activity more than the depletion of its substrates (CO 2 or RuBP) [127]. Photorespiration has been found to reduce the photosynthetic efficiency by 20-50% in C3 crops [128]. RuBisCO requires the concentration of CO 2 to be more than 25 µM for carbon fixation, but the rate of conversion of HCO 3 to CO 2 is generally found to be slow [92,126]. The catalytic step involving RuBisCO is the rate limiting step in the Calvin-Benson-Bassham (CBB) cycle. RuBisCO catalyzes the carboxylation of ribulose-1,5-bisphosphate (RuBP), to synthesize two molecules of 3-phosphoglycerate (PGA). RuBisCO has a very low affinity for CO 2 . Unlike other enzymes, the concentration of RuBisCO does not change the flux towards CO 2 fixation. The process is affected by the relative concentration of CO 2 and O 2 at the active site. The rate of CO 2 diffusion is low in aquatic systems, and the CO 2 concentration is often below the required threshold of RuBisCO. The ratio of CO 2 fixation rate and the photosynthetic electron transport rate regulate RuBisCO activity. For catalysis to occur, RuBisCO should be activated first. This happens with the help of RuBisCO activase (RCA). Otherwise, CO 2 binds at a lysine residue in RuBisCO for carbamylation. A change in the conformation of RuBisCO is brought about when RCA binds to the inactive RuBisCO and ATP hydrolysis occurs. This results in the synthesis of a highly active form of RuBisCO, based on cellular requirement. To maximize the performance of RuBisCO in photosynthetic CO 2 fixation, the kinetic properties of the enzyme have evolved over time. RuBisCO is a highly diverse biomolecule, four forms of which have been identified to date (I, II, III and IV), categorized based on the differences in the primary polypeptide sequence, along with the differences in the number of small and large subunits. RuBisCO form I is the most common form of RuBisCO found in nature. RuBisCO form I is further sub-divided into four subforms (A, B, C and D), as a result of their evolution. Forms IA and IB ("green-type") are found in higher plants, cyanobacteria, chlorophyceae and streptophytes [129,130], whilst forms IC and ID ("red-type") found in haptophytes, cryptophytes, rhodophytes and the heterokontophytes [129,131]. Which form of RuBisCO is better is a matter of further research, as both the RuBisCO types have their own preferences in terms of CO 2 needs in the environment. Currently, red algal RuBisCO is being studied in depth for crop improvement strategies, and is assumed to have better kinetics [132].
RuBisCO enzymes from microalgae have evolved a higher affinity for CO 2 when the algae have adopted a strategy for CO 2 fixation that does not utilize a CCM. This appears to be true for green and red algae form I RuBisCO enzymes. However, the red form I RuBisCO enzymes present in non-green algae appear to have reduced oxygenase potential at ambient concentrations of O 2 . This has resulted in a photosynthetic physiology with a reduced potential to be inhibited by O 2 and a reduced need to deal with photorespiration. Red form I RuBisCO enzymes appear to achieve superior kinetic characteristics when compared with the RuBisCO of C3 higher plants, which are derived from green algal ancestors [92]. The green-type RuBisCO activase functions as a canonical hexameric AAA + ATPase [133,134], and its higher plant homologs mostly occupy far larger polydisperse oligomeric states. RuBisCO catalyzes the fixation of atmospheric CO 2 in photosynthesis, but tends to form inactive complexes with its substrate ribulose 1,5-bisphosphate (RuBP). In plants, RuBisCO is reactivated by the AAA + ATPases (associated with various cellular activities) protein RuBisCO activase (Rca), but no such protein is known for the RuBisCO of red algae [135]. Understanding RuBisCO activation may facilitate efforts to improve CO 2 uptake and biomass production by photosynthetic organisms, by making more carbon available to the system and inducing utilization, being incorporated later into biomass.
The active site of the RuBisCO molecule is formed from the N-terminus of one of the monomers and of the C-terminus of another subunit, therefore having four dimers arranged together, in an optimum conformation with ideal structural stability [130,136,137]. It is also noteworthy that the conformational changes affect the specificity factor (CO 2 /O 2 ) of the RuBisCO enzyme, which is an important kinetic parameter. Premature binding of RuBP and other molecules such as other sugar phosphates inhibits RuBisCO activity; the biomolecule is also dogged by slow turnover rate, competition from oxygenase activity and low affinity towards CO 2 [129,138]. Coccolithophores and diatoms, as a virtue of their evolutionary process, have obtained the highly selective Rhodophyte form of 1D RuBisCO. The 1D RuBisCO form is better adapted to the oxygenic environment with a higher O 2 /CO 2 ratio. Furthermore, the 1D RuBisCO form is economical, and requires lower energetic or nutrient investment in a CCM to obtain high carboxylation rates under environmentally high O 2 /CO 2 ratios [139]. The CCM of diatoms are highly diverse, and are capable of concentrating very high levels of CO 2 in the pyrenoid. Diatom RuBisCO also displays very high variation in Michaelis constant for CO 2 , K C (23-68 µM), specificity for CO 2 over O 2 , S C/O (57-116 mol mol −1 ), and Michaelis constant for O 2 , K O (413-2032 µM), in comparison to plant and other algal RuBisCOs [126].

Evidence of C4 Metabolism
There are three main carbon fixation mechanisms employed by photosynthetic organisms: C3, C4 and Crassulacean acid metabolism (CAM), which have been extensively reviewed elsewhere [80,140].
The C4 type photosynthesis, as a carbon concentrating mechanism, has evolved more than 60 times, to address the inefficiencies of the ancestral C3 photosynthetic pathway. Inherent in the C4 pathway is a high rate of photosynthesis at low levels of stomatal conductance, especially at the low levels of CO 2 in the atmosphere [141].
Less than 4% of terrestrial plant species are believed to have a C4 pathway. The C4 pathway involves a CO 2 -CCM around RuBisCO, eliminating the oxygenase function of RuBisCO, and reducing the wastage of carbon assimilation to photorespiration [125]. The operation of a C4 pathway overcomes the tendency of RuBisCO to fix oxygen rather than CO 2 , and avoids the production of glycolate, thus minimizing photorespiration. C4 plants provide a CO 2 pump which results in an increased CO 2 /O 2 ratio at the site of RuBisCO, resulting in decreased oxygenase activity and RuBisCO operates at close to its V max , whereas RuBisCO in C3 plants only operates at around 25% of the V max [142]. RuBisCO in C4 plants is more efficient in terms of carboxylation than C3 plants; a higher light harvesting efficiency is observed as saturation under high light is avoided and higher nitrogen utilisation is observed, because less RuBisCO, and thus nitrogen, is required. The C4 CCM pathway in higher plants uses PEP to catalyze the first reaction in inorganic carbon fixation and requires compartmentalization called Kranz anatomy (bundle sheath and mesophyll areas). Comparatively, in single-celled plants, dimorphic chloroplasts contain a central compartment and peripheral chloroplasts. In C4-containing single-celled plants, the release of CO 2 in the direct vicinity of RuBisCO is critical for the activity of an efficient C4 photosynthesis [125]. Understanding microalgae with a C4 mechanism could enable further developments to improve photosynthesis and allow cultivation under more extreme conditions.
Of the few diatoms that have features of a biochemical CCM, T. weissflogii and P. tricornutum are the only ones that have been investigated in detail, to determine if they undergo C4 photosynthesis. P. tricornutum has a naturally slow rate of photorespiration. T. pseudonana and P. tricornutum both possess several carboxylases and decarboxylases, which could be used for a C4-like CCM. Both diatoms have a pyruvate phosphate dikinase (PPDK), which converts pyruvate into PEP, the initial three carbon molecule that accepts HCO 3 − in the C4 CCM of higher plants [125]. In low CO 2 concentrations, T.
pseudonana has been found to use a "closed loop biochemical model", an atypical C4-type CCM, where the back-reaction of the pyruvate carboxylation was presumed to be responsible for CO 2 release in the plastid [93]. During the transition from growth to lipid accumulation, pyruvate carboxylase in the mitochondrion is utilized as the primary inorganic carbon fixation stage in a C4 pathway, and malate undergoes decarboxylation by the malic enzyme in the peroxisome to concentrate CO 2 for diffusion into the chloroplast [143]. Whilst genome analysis has indicated that P. tricornutum possesses the necessary enzymes for operating a C4 pathway, recent evidence has suggested that a C4-like CCM is not present [125]. Neither radiolabeling with 14 C nor PPDK silencing experiments via RNAi with the aim of identifying primary products of inorganic carbon fixation were found to support the presence of a C4-like CCM in P. tricornutum, T. pseudonana, T. weissflogii, As. Formosa, N. pelliculosa [86]. None of the known or investigated decarboxylases nor a cytosolic PEP carboxykinase (PEPCK) has been found in the plastid (a prerequisite for a single cell C4-type CCM pathway). The CCM efficiency of P. tricornutum is not affected by a reduction in PEPCK activity, and therefore PEPCK does not appear to contribute to the CCM. It has been concluded that the C4-like CCM in P. tricornutum does not have an essential role in CO 2 fixation, and the enzymes investigated are more likely involved in similar functions (e.g., gluconeogenesis, amino acid synthesis or replenishing the TCA cycle), seen in C3 plants [125].
In photosynthesis, the efficient conversion of CO 2 into organic matter requires a tight control of the ATP/NADPH ratio which, in other photosynthetic organisms, relies principally on a range of plastid-localized ATP generating processes. Diatoms can regulate ATP/NADPH through extensive energetic exchanges between plastids and mitochondria. This interaction comprises the re-routing of reducing power generated in the plastid towards mitochondria and the import of mitochondrial ATP into the plastid, and is mandatory for optimized carbon fixation and growth [144].
In summary, as can be seen from the molecular mechanisms highlighted above, our knowledge of CO 2 fixation in diatoms is informed largely by investigations carried out in a handful of diatoms, and extrapolated from plants and other microalgal species. Given the diversity in diatoms and their metabolic capabilities, molecular level data from other diatoms under specific industrial CO 2 sequestering environments would be required to develop a broader picture of the diatoms CO 2 utilization capabilities. Nevertheless, the adaptive nature of these organisms suggests a broader capacity to CO 2 uptake and routing of fixed carbon to value-added products. A few innovative studies have come up recently for improving photosynthetic efficiency in diatoms. One of these strategies is the use of a high silicate medium along with blue light under high light conditions (255 µE m −2 s −1 ) for fucoxanthin production in P. tricornutum [145]. Another is the application of intracellular spectral recompositioning of light (ISR) on a genetically engineered P. tricornutum, with a green fluorescent protein (GFP) to enhance photosynthetic efficiency (by 50%) and biomass productivity, influencing fucoxanthin production in diatoms [146]. In addition, enhanced lipid production in genetically modified P. tricornutum stable strains has been achieved by the use of gene editing tools such as meganucleases and transcription activator-like effector nucleases (TALEN) to achieve targeted disruption of the UDP-glucose pyrophosphorylase gene, a step involved in carbohydrate accumulation that enabled routing carbon to lipid accumulation [147].

Opportunities and Challenges of CO 2 Sequestration by Diatoms-Direct Air Capture, Pure CO 2 or Flue Gases
CO 2 is an indispensable resource for autotrophic organisms such as diatoms. Effective utilization of diatoms for CO 2 sequestration in biomanufacturing requires in-depth consideration of issues, such as CO 2 supply, CO 2 source and the optimization of conditions for species-specific CO 2 uptake. CO 2 supply is an important aspect to be considered for growing autotrophic microorganisms. CO 2 can be supplied in either the solid form (as carbonate or bicarbonate salts) or in the gaseous form to autotrophic organisms. There are three potential gaseous sources; (a) flue gases or product streams from industry, (b) purified CO 2 available in cylinders and (c) direct air capture. Flue gas, a by-product of industrial production and power generation, can be a useful CO 2 resource. Flue gas from cement manufacturing has been tested on the diatom S. marinoi, and found to be non-toxic. In fact, high quality of microalgal biomass (lipids 20-30% DW, proteins 20-28% DW, carbohydrates 15-30%(DW)) and a higher biomass productivity has been demonstrated with flue gas addition, compared to aeration with atmospheric level of CO 2 [148]. Flue gas from industry (which might include SO x , NO x and other gases along with CO 2 ), typically contains CO 2 in the range of 6-15%, whilst product streams, such as from ethanol manufacturing and biogas, typically contain CO 2 in the range of 20-40%. CO 2 can also be supplied to the diatoms directly from air (which contains~0.04% CO 2 ), in the form of pure CO 2 , or CO 2 mixed with air or nitrogen in cylinders (available up to 100% CO 2 ). A challenge in providing CO 2 from air is in arriving at economically viable propositions for extracting the CO 2 from air. A commercial plant, where CO 2 is sucked from the air (before being resold), had opened up in Switzerland. It was founded by Climeworks (https://www.climeworks.com/); the direct air capture (DAC) plant is capable of removing 900 tons of CO 2 from ambient air annually. Unlike capturing emissions from industrial flue stacks, the technology by Carbon Engineering (https://carbonengineering.com/) captures CO 2 directly out of the air. From a pilot facility in Squamish, Canada they had fully demonstrated the Direct Air Capture (DAC) technology, and are now commercializing. The SOLETAIR project (https://soletair.fi/) is also involved in direct CO 2 capture from the air. For a successful diatoms-based bio-venture, the project should be environment friendly, sustainable, feasible on a large scale, and preferably work around emerging technologies such as DAC.
Diatoms are mostly cultivated either in submerged or immobilized reactors. There are many challenges in supplying CO 2 for cultivation and administered by the bubbling method. CO 2 bubbled into the medium needs to be dissolved and accessible to the diatoms. At high supply rates, and in saturated media, most of the CO 2 supplied is released back into the atmosphere [84], due to the low CO 2 solubility and low retention potential of CO 2 in the medium. Moreover, the larger the bubble size, the greater the buoyancy and faster the release of CO 2 bubbled out from the medium (and hence, lost from the system). The CO 2 utilization of photoautotrophic organisms is also slow and limited. Factors which affect CO 2 utilization efficiency and growth include CO 2 concentration, bubble size, aeration rate, mixing time, and the residence time of the bubble. Ensuring an extensive air/liquid interface is essential for ensuring good CO 2 mass transfer [149]. The CO 2 /O 2 balance is also a key factor in attaining a higher photosynthetic rate. The knowledge of CO 2 uptake in most of the diatom species is fragmentary. For CO 2 fixation, CO 2 can be injected as a gas into the culture, dissolved in a separate absorption column, or added as an alkaline solution in the form of bicarbonate [83]. Direct injection of CO 2 is known to lower the pH of the culture, and can adversely affect growth and biomass productivity at high CO 2 concentrations. There is a body of literature on high CO 2 tolerance for cultivation of diatoms with improved product yields. For example, the addition of 10% CO 2 (v/v) into the cultures of T. weisflogii and C. cryptica resulted in doubling the lipid content in comparison to air sparging, but induced only a modest increase of biomass. In the same set up, CO 2 also stimulated lipogenesis in both of the diatoms (T. weisflogii and C. cryptica). Moreover, TAG became the major lipid component, and accounted for more than 60% of total glycerolipids in C. cryptica. [150] Approaches such as microbubble generation [151] have enabled increased surface area for effective mass transfer in dissolving gaseous CO 2 into the culturing medium. There are also a variety of cultivation systems that focus on the method of CO 2 supply into the cultivation medium. The optimization of carbon use in pilot-scale outdoor tubular photo-bioreactors by application of effective control techniques, such as model-based predictive control (MPC), for reducing loss of CO 2 along with total supply of CO 2 volume, has been reported to potentially increase productivity by 15% and reduce the cost of producing biomass by >6% [152]. Different requirements may need to be considered for the supply of CO 2 into open cultivation systems, such as lakes, lagoons, ponds, constructed raceway ponds and closed systems, such as tubular photo-bioreactors, flat panel photo-bioreactor, fermenters, cascade raceways, raceways and Tic bag photo-bioreactors.
In its solid form, HCO 3 − can be a source of carbon. Bicarbonate is the main form of inorganic carbon utilized by P. tricornutum [153]. Sodium bicarbonate is readily available in large quantities due to coal fired power stations, using a CO 2 scrubbing system and generating bicarbonates. Bicarbonate is also easier to transport than gaseous CO 2 [123]. Sodium has been found to increase the affinity for inorganic carbon and facilitate the utilization of bicarbonate in P. tricornutum [154]. Sodium ions can enhance the rate of photosynthetic oxygen evolution, which could be due to the presence of a sodium dependent bicarbonate-transport system, and as the internal inorganic-carbon concentration is lower in relation to the external concentration, the effect of sodium is possibly at the plasmalemma [95]. P. tricornutum has been shown to have a similar growth rate and CO 2 uptake rate at extreme pH values compared with neutral pH [83]. Bicarbonate feeding was found to reduce the loss of CO 2 to the environment compared with gaseous CO 2 supply, however, the growth rate was reduced, along with the biomass yield [83]. Three different concentrations of NaHCO 3 (5, 15 and 25 mM) have been added at one of two different culturing phases, either at day 0 (during bioreactor inoculation) or at day 4 (~24 h before nitrate depletion). The cultures supplemented with 15 mM NaHCO 3 accumulated more carbohydrate than the control culture. The supplementation of 25 mM NaHCO 3 led to higher protein content for unknown reasons (Mus et al. 2013). Nitzschia plea is known to tolerate high concentration of NaHCO 3 (0.15 mol/L) and high pH (>10) [155].
A diatom that has the machinery for both HCO 3 − uptake and CO 2 capture will be indispensable for industrial applications, and enable the development of a sustainable biomanufacturing process.
There are a few diatoms that are known to have the capacity to utilize both CO 2 and HCO 3 − . N. palea kutzing is a very interesting diatom in this regard, and its cells were found to be capable of using HCO 3 − in addition to gaseous CO 2 , and the CO 2 enrichment decreased their affinity of HCO 3 − and CO 2 [156].
Another diatom, Chaetocerous muelleri, has the capacity to use bicarbonate to acquire inorganic carbon through one or multiple CCM. It also has the capacity to use HCO 3 − to acquire inorganic carbon, through one or multiple pathways [157]. Storage of DIC is also an area that requires detailed study in diatoms, an example being the preference of HCO 3 − by Cyclotella sp. and Nitzschia sp. [158].
The carbon capture potential of a diatom is directly related to its CCM design and the efficiency of its CCM enzymes, such as CAs and RuBisCO. The regulation of external CA activity and photosynthetic CO 2 affinity are dependent not only on CO 2 concentration, but also on light availability, as observed in S. costatum. The presence and activity of external CAs decide how well a CCM is designed. External CA activity has been detected in cells grown at 4 µmol L −1 CO 2 , but not at 31 and 12 µmol L −1 CO 2 , with its activity being about 2.5 times higher at high irradiance than at low irradiance. Further, the development of higher external CA activity and CO 2 affinity under higher light level could sufficiently support the photosynthetic demand for CO 2 , even at a low level of CO 2 [159]. Light has a pivotal role to play in CCM efficiency. At saturating light intensities, S. costatum cleve and P. tricornutum Bohlin maintain maximum photosynthetic rates under low CO 2 levels, but P. tricornutum is well adapted to rapid changes in irradiance and CO 2 availability. In P. triconutum and Nitzschia ovalis, acetate has been found to be the preferred carbon source for the formation of the sterols in the cytoplasm, via the mevalonate pathway. Also, CO 2 was regarded as the main source for phytol biosynthesis in the chloroplasts, via the mevalonate independent methyl erythritol 4-phosphate pathway. Both the diatoms, P. triconutum and N. ovalis, have been found to display the same compartmentation for isoprenoid biosynthesis, as previously found in higher plants, the red alga Porphyridium cruentum and the chrysophyte Ochromonas danica [160]. The extracellular carbohydrates of the diatom Cyclotella meneghiniana have been found to increase with elevated CO 2 and temperature [161]. Inlet pCO 2 have been found to enhance lipid production along with chitin formation in Cyclotella sp. in a photobioreactor setup [162].
Along with the biology of a diatom, the carbon capture potential is also dependent on the culture health and viability. Moreover, the efficiency of its CCMs depends on the physical, chemical and biological conditions in the culturing environment. The amount of carbon captured in the system is directly proportional to the density and growth phase of the diatom. For diatoms, the literature on CO 2 fixation is sparse. Buono et al. (2016) had found that there was no linear relationship between the CO 2 added to the culture and the CO 2 assimilated by the microalga [30]. A significant amount of CO 2 was found to be lost to the atmosphere when the gas was added to the culture. It has been found that closed systems, compared with open ponds, have a better ability to assimilate CO 2 (a 44.2% reduction in CO 2 fixation was observed in open ponds) [30,149], and had resulted in higher CO 2 fixation, biomass productivity, and a higher photosynthetic efficiency [30]. A high CO 2 fixation rate in P. tricornutum was observed in 1 L cultures when supplemented with 15% (v/v) CO 2 , resulting in a CO 2 fixation rate of 0.282 g L −1 d −1 , but the biomass productivity was low (0.15 g L −1 d −1 ) [85]. pH plays an interesting role, as evident when the culture pH under 15% CO 2 (pH 6.3) compared to a buffered system (pH 7) did not significantly affect the biomass productivity [85]. P. tricornutum takes up HCO 3 − predominantly, whilst T. pseudonana takes up CO 2 [86].
The species and strain of diatom to be used for large scale carbon capture needs to be able to grow under a day-night cycle, be suitable for large scale cultivation, and can be coupled directly with the CO 2 flue gas from a power plant [149]. Utilizing flue gases for microalgal cultivation would be beneficial as, in addition to high CO 2 , it also contains NO x and sulfur dioxide (SO 2 ), that can be used by the diatom as a nitrogen and sulphur source. However, there is a problem associated with it, as shown for P. tricornutum, where the addition of SO 2 at 50 ppm resulted in the growth being slightly inhibited, and a further increase to 400 ppm caused a cessation of growth [85]. When CO 2 from flue gas is used, a high pH is required to ensure the bicarbonate remains in solution and does not dissipate into the atmosphere [83]. An ideal cultivation system would enable appropriate carbon supply, carbon uptake and CO 2 fixation using diatoms. The construction of a suitable PBR appears to be essential for successful CO 2 supply, reduction in CO 2 wastage and carbon sequestration. Carbon sequestration by microalgae is itself dependent on characteristics of light (intensity, incident angle, photo period and wavelength) and carbon optimization, both required to improve photosynthesis for CO 2 uptake. Open ponds are often employed worldwide, due to their economic benefit, but are constrained by low biomass productivities, issues with evaporation, CO 2 diffusion to the atmosphere, a lack of temperature control, insufficient light transmission, and vulnerability to contamination. Comparatively, flat-plate and tubular PBRs have been shown to have a large surface area, good mixing, mass transfer and short internal light paths. However, tubular PBR systems can be constrained in their size and length, because of oxygen build up and CO 2 depletion [149]. Air-lift PBRs have been reported to have a higher CO 2 fixation rate, due to their better circulation and mass transfer through the use of risers and downcomers [163], but bubbles can cause a high attenuation of light and can create shear when they break. Several strategies have been employed, like degassers, air outlets through water traps etc. in PBRs to strip this excess oxygen from the culture medium with air or inert gases [164]. However, more efficient methods are required to alleviate the effects of O 2 inhibitory effects during microalgal growth. P. tricornutum has been successfully cultivated on a large scale (55 L) in a flat-plate customized photobioreactor, for the simultaneous synthesis of storage lipids, EPA, fucoxanthin and chrysolaminarin [56]. P. tricornutum has also been cultivated outdoors in 800 L bubble column PBR [165], and in 1250 L indoor open raceway pond [58].

Bio-Manufacturing with CO 2 Uptake
One of the diatoms which has been extensively evaluated for potential as a bio-based manufacturing chassis is P. tricornutum, a versatile diatom that has the capacity to produce a range of natural (fucoxanthin, EPA, DHA, oil, brassicosterol, and chrysolaminarin) and genetically engineered products (lupeol, betulin, arachidonic acid, antibodies, and polyhydroxybutyrate) [23]. It is a well characterized diatom, and can be routinely cultivated in the laboratory, and has been shown to perform well at scale (>1250 L) [58]. P. tricornutum is a saltwater diatom and offers potential as a sustainable cell chassis for multiple products of interest, and capable of performing well in constrained environments, including low light and high pH [83,166]. It has a relatively well annotated genome and demonstrated cases of downstream processing for the sequential extraction of multiple products of interest using a bio-refinery approach [155].
Another diatom, Chaetocerous gracilis, has been shown to accumulate TAG without nutrient deprivation, and has a great potential as a biofactory [167]. CO 2 addition is associated with increased protein content and lowered carbohydrates, but had no effect on lipid content in the marine diatom Chaetoceros cf. wighamii [168].
The marine diatoms, as a group, have potential for high CO 2 fixing capacity, being naturally evolved for this function. However, further elucidation of carbon fixation is required before strategies can be developed to maximize carbon uptake and route fixed carbon to products of value in industrial scale operations. Our knowledge of molecular pathways and strategies to improve carbon uptake are rudimentary at this stage, and will need to be developed for more diatoms than has been available so far. More information will enable ways to maximize CO 2 uptake and route it effectively, to increase the productivity of diverse products. Combined mixotrophic approaches might help further. Improvements in cultivation, PBR development, and strategies, including adaptive laboratory evolution, omics analyses and targeted genetic engineering, will be useful in taking the investigations forward towards sustainable implementations.
The steady states of specific intracellular levels of carbon metabolic intermediates affect the yield of bio-based manufacturing products, and can be increased by higher amounts of CO 2 concentrations. Acetyl CoA is the precursor for lipids, carotenoids, exopolysaccharides (EPS) and other valuable metabolites in most organisms, and shunting carbon through this metabolite would be a useful way to elevate product yields. Supplementation of 2.6% CO 2 has been shown to lead to increased Acetyl CoA in P. tricornutum, with a 41, 25% and 27% increase over the air-sparged controls in the lag, log and stationary phases, respectively [143]. Acetyl CoA (AcCoA) has also been shown to increase with exposure to increased CO 2 concentrations, and in turn lead to improved lipid accumulation [169]. However, long term exposure to elevated pCO 2 has been shown to have a detrimental effect on the diatom Cylindrotheca fusiformis [170]. Carbohydrate yields in different EPS fractions increased with elevated pCO 2 exposure. Although the proportions of monosaccharide sugars among total sugars did not change, higher abundances of uronic acid were observed under high pCO 2 conditions, suggesting the alteration of EPS composition [171]. An increase in CO 2 supplied from 400 ppm to 20,000 ppm has been shown to lead to a general increase in biomass productivity, by 11%, 28% and 21%, respectively, in T. pseudonana, P. tricornutum and N. pelliculosa (seawater strain), in the exponential phase, when other nutrients are expected to still be available in sufficient quantities [28]. Such positive results open up new avenues for CO 2 capture.
There are many advantages of culturing diatoms for large scale cultivation, especially those from the marine habitat. Sea water can be used as the culturing medium, thereby preventing the usage of scarce freshwater resources. Diatoms are robust organisms, competitive, and can be cultivated in less stringent and non-sterile conditions. They are encased with a silica layer that helps them to evade predators and grazers [172], except the less silicified oval stage of P. tricornutum [173,174]. The strong silica frustule may also help them to maintain structural integrity in the rough seas, and provide buoyancy for the cells to access nutrient and light enriched surfaces [175]. This can also create density, enabling implementable harvesting protocols. However, exposure to pests leading to cultivation crashes and subsequent economic losses still require addressing and the development of innovative solutions. With an ever-increasing human population, more food will need to be produced and more energy will be consumed as a result. In a transition to a bio-based economy with sustainable solutions, diatoms have the potential to be the futuristic and sustainable source for CO 2 sequestration and bio-based manufacturing.

Conclusions
Diatoms are responsible for 20% of the global CO 2 fixation, a major greenhouse gas responsible for climate change. In addition to sequestering CO 2 , diatoms can be utilized to produce a plethora of commercially viable products for food, feed, fuel, and nutraceuticals/pharmaceuticals applications. There is a diversity of products with market potential, ranging from high value low volume to low value high volume productions. In order to accommodate the diverse requirements of value addition and still be effective contributors to CO 2 mitigation, the development of strategies along the biorefinery concept will be necessary. Much of our knowledge on diatoms for use in carbon sequestration and biomanufacturing has been built on a handful of species, P. tricornutum being the dominant one. Carbon metabolism in more diatom species requires elucidation, especially with respect to exposure to industrial CO 2 supply compared to the natural environment, and the bottlenecks in the CCM and the overall carbon capture pathways understood. These remain as challenges to be overcome for diatoms to be developed as a microbial cell factory for a suite of products of commercial interest, whilst simultaneously making the most of their CO 2 utilizing capacity for CO 2 sequestration strategies. There is a huge potential in diatoms for carbon capture and utilization and in bio-based manufacturing that awaits research and development.