The commercial potential of Aphanizomenon flos-aquae, a nitrogen-fixing edible cyanobacterium

Abstract

Aphanizomenon flos-aquae (AFA) is a Gram-negative nitrogen-fixing freshwater filamentous cyanobacterium with a rich nutritional profile that is approved for human consumption by both the Food and Drug Administration and the European Food and Safety Authorities. It has a high protein content (60-70%) and contains numerous vitamins, minerals and trace elements together with several high-value compounds with nutraceutical properties, such as C-phycocyanin and β-phenylethylamine. 500-1000 t of AFA dry biomass are currently wild harvested from natural blooms that occur seasonally in Klamath Lake, Oregon, USA, and distributed as a nutritional supplement worldwide. The requirements and unreliability of wild harvesting, owing to the dependence of AFA growth on environmental conditions and potential contamination by microcystin toxins, threaten the availability of biomass supply and restricts commercial expansion. In this review we propose AFA cultivation in open ponds or closed photobioreactors to obtain a reliable production of unialgal biomass to resolve the supply issue and enhance AFA as a feedstock for specific high-value by-products. Furthermore, we explore potential strategies for maximising overall yield and seasonal robustness by using a synergistic combination of cultivation and wild harvesting. Controlled AFA cultivation would also facilitate the use of genetic manipulation to generate bespoke strains with improved commercial applications, such as increasing the cyanobacterium’s nitrogen-fixation rate to enhance its value as a biofertiliser. Ultimately, realising the untapped biotechnological potential of AFA requires a better understanding of its fundamental biology, robust methodologies for laboratory and large-scale cultivation, and the development of AFA-specific genetic engineering techniques.

Introduction

Cyanobacteria, also known as blue-green algae, are some of the oldest and most abundant photosynthetic microorganisms on our planet and contributed to the so-called “oxygenic revolution” 2.4 billion years ago (Zahra et al. 2020). With the increasing demand for sustainable and environmentally friendly production of high-value compounds, cyanobacteria have emerged as attractive candidates for biotechnological applications (Zahra et al. 2020; Sproles et al. 2021). Cyanobacteria are capable of producing multiple secondary metabolites and can be considered as sunlight-driven microcellular factories (Al-Haj et al. 2016). Their relatively simple growth requirements allow inexpensive maintenance and cultivation, and modern genome analysis and genetic engineering tools make possible the creation of bespoke strains for the production of high-value products (Sreenikethanam et al. 2022). Nevertheless, the biotechnological exploitation of cyanobacteria is currently limited to a few model species such as Synechococcus sp. PCC 7002 (CP000951) or Anabaena sp. PCC 7120 (BA000019) (Zahra et al. 2020). The majority of cyanobacterial species have either not been studied, or are unable to adapt to laboratory growth conditions and be genetically manipulated (Lea-Smith et al. 2021). Cyanobacteria are also well-known for their high nutritional value and protein content, which have made them important players in the food and supplement markets. Among these, we find several species of filamentous cyanobacteria, such as Arthrospira spp., which are commercially marketed as Spirulina, and colonial species such as Nostoc, with a notable presence in Asian markets (Li and Guo, 2018, Zhu et al., 2023, Gentscheva et al. 2023). Currently, Spirulina dominates the cyanobacterial food market, together with the green algae from the genus Chlorella (Andrade et al. 2018). However, there exist other cyanobacteria that have an undiscovered and untapped biotechnological potential. One such example is Aphanizomenon flos-aquae (AFA).

Aphanizomenon flos-aquae Ralfs ex Bornet & Flahault is a filamentous nitrogen-fixing obligate phototroph that was first discovered towards the end of the 19th century (Cirés and Ballot 2016). It belongs to the Nostocacae family and is the type species for the genus Aphanizomenon (Cirés and Ballot 2016). Its most distinguishing feature is its ability to form fascicles: bundles of filaments that can reach lengths of 2 cm and can be seen with the naked eye (Carmichael et al. 2000). AFA is widely distributed in eutrophic freshwater ecosystems at high latitudes where it has developed traits that are unique to the Nostocacae (Cirés and Ballot 2016). These include nitrogen fixation, tolerance to varying temperatures and pH levels, buoyancy-dependent adjustments in the water column and the ability to regulate the expression of light-absorbing pigments (Komárek and Komárková, 2006; Komárek 2013). Furthermore, the vegetative cells of AFA can specialise into akinetes: dormant cells that form under unfavourable conditions, and which have enabled AFA’s survival across the millennia (Wildman et al. 1975).

AFA has attracted commercial interest owing to its high nutritional value. It boasts a high protein content (60-70%). It is a rich source of minerals and vitamins and is characterised by a high concentration of omega-3 polyunsaturated fatty acids (PUFAs) (Sandgruber et al. 2021). Furthermore, AFA produces a number of bioactive compounds, the most impactful being C-phycocyanin (C-PC), a pigment with potent anti-inflammatory and antioxidant effects, and β-phenylethylamine (PEA), an organic and endogenous neurostimulator (Benedetti et al. 2006; Nuzzo et al. 2018; Billy et al 2023). AFA is approved for human consumption by both the Food and Drug Administration (FDA) and the European Food and Safety Authorities (EFSA). Currently, its biomass is marketed worldwide as a nutritional supplement (Schaeffer et al. 1999; Carmichael et al. 2000).

The AFA biomass used for commercial products is exclusively harvested from the wild; specifically from Klamath Lake in Oregon, USA (Fig. 1). Here, AFA forms seasonal blooms with dense mats of the cyanobacterium accumulating at the lake water surface. These blooms are harvested en masse (see: Online Resource 1), processed, and further commercialised (Carmichael et al. 2000). AFA is the second most consumed cyanobacterial nutritional supplement following Spirulina (Grewe and Pulz 2012). However, due to the limitations of wild harvesting, AFA’s market has not been able to expand within the last 20 years. AFA blooms are seasonal and dependent on specific natural conditions, making AFA harvesting viable only during the warmer months (Carmichael et al. 2000). Furthermore, Klamath Lake also harbours Microcystis aeruginosa, a cyanobacterium known to produce a class of hepatotoxins, termed microcystins (Carmichael et al. 2000; Sivonen 2009). As a result, companies harvesting from Klamath Lake must ensure that any recovered AFA biomass has a microcystin concentration that is below the threshold considered safe for human consumption, which further reduces the number of harvestable days per annum (WHO, 1999, Saker et al. 2005, EPA, 2016, Lyon-Colbert et al. 2018). Overall, AFA’s commercial expansion has been restricted by these harvesting requirements of quantity and quality. However, these restrictions could be overcome through cultivation under carefully controlled conditions using optimised media in photobioreactors (PBRs).

Fig. 1

Illustration of Aphanizomenon flos-aquae (AFA) harvesting in Klamath Lake, Oregon, and the derived products obtained. A) aerial photograph of Upper Klamath Lake (UKL) taken days before the explosion of an AFA bloom, B) photograph of AFA harvesting during a summer bloom taken from Online Resource 1, and C) AFA-based supplements sold by Nutrigea s.r.l.

Advances in the controlled, large-scale cultivation of microalgae and cyanobacteria are enabling the industrialisation of previously non-commercialisable microalgae, including strains of Chlorella, Arthrospira, and Nannochloropsis (Belay 2013; Safi et al. 2014; Al-Hoqani et al. 2016). Notably, the cultivation of Spirulina and Chlorella has resulted in a rapid increase in their global production, leading them to dominate the microalgal- and cyanobacterial-based industry (Araújo et al. 2021; Show 2022). Additionally, recent improvements in PBR design have facilitated successful large-scale cultivation of diverse microalgae and cyanobacteria with unique properties, allowing for enhanced utilisation of their high-value products (Deprá et al. 2019; Legrand et al. 2021). For instance, the mass cultivation of the green alga Haematococcus pluvialis in tubular PBRs for the extraction of the valuable astaxanthin pigment has been highly successful (Li et al., 2020). Considering these factors, the cultivation of AFA holds significant commercial potential, particularly due to its established market presence (Carmichael et al. 2000). Controlled growth of AFA in PBRs has the potential to not only address the issue of limited biomass supply but also opens avenues for expanding AFA-based products into new markets, such as animal feed, biofertilisers and pharmaceuticals (Kumar et al. 2019).

Knowledge of the optimal growth requirements for AFA remains limited. There are only a handful of studies that have investigated laboratory cultivation of AFA, and, to date, its mass cultivation has never been achieved or even attempted (Gerloff et al. 1950, Gentile and Maloney 1969, O'Flaherty and Phinney, 2008, Debella 2007). Therefore, the greatest risk associated with the deployment of a commercial facility for mass cultivation is technological feasibility (Kumar et al. 2021). Furthermore, the capital and operating costs associated with cyanobacterial mass cultivation would not be able to compete with the cost-effectiveness of Klamath Lake harvesting unless the PBR-based products carried a significant premium (Zhu et al. 2018; Clippinger and Davis 2019). Therefore, a more attractive strategy involves a combination and integration of both AFA cultivation in PBRs and wild harvesting from Klamath Lake, leveraging the synergistic benefits of both approaches. By cultivating AFA in PBRs the overall annual yield of AFA can be increased, providing a more reliable baseline production, while the bulk AFA biomass from wild harvesting could help off-set the overall costs associated with PBR cultivation.

Another advantage of successful cultivation of AFA in the laboratory is the opportunity to investigate strain improvements by developing molecular tools for genetic manipulation (Santos-Merino et al. 2019). Cyanobacteria such as Synechococcus and Anabaena have been successfully genetically engineered with varying degrees of development of the molecular toolkits and technologies for strain improvement (Kumar et al. 2020; Sproles et al. 2021; Grama et al. 2022). The establishment of DNA transformation technology for AFA would allow the generation of optimised strains for a range of biotechnological purposes. For example, as AFA is one of only a few safe and commercially distributed cyanobacterial species that is a diazotroph – i.e. able to convert atmospheric nitrogen into bioavailable forms of nitrogen – the enhancement of AFA’s nitrogen fixation rates would greatly promote its performance as a biofertiliser (Chaurasia and Apte 2011; Grewe and Pulz 2012). Nevertheless, genetic modification of filamentous cyanobacteria presents numerous challenges. For instance, a high concentration of extracellular polymeric substances can impede DNA entry into cells, while a substantial number of endogenous restriction endonuclease systems within the cell may degrade incoming DNA. Some filamentous species, such as Anabaena sp. 90 (CP003284), have proven recalcitrant to genetic modification, even after extensive characterisation. Nevertheless, as no transformation attempts have been made with AFA thus far, the feasiblity of its genetic modification has yet to be determined.

The primary objective of this review is to highlight AFA’s untapped biotechnological potential. We examine the biology and nutritional benefits of AFA and address the existing knowledge gaps that may impede a broader understanding of its possible application in the field of biotechnology. For example, we examine whether AFA represents a natural source of vitamin B₁₂ (cyanocobalamin) or merely a source of pseudocobalamin, an analogue that lacks bioactivity in humans (Miyamoto et al. 2006; Helliwell et al. 2016). We also detail the potential DNA transformation methods employed for other cyanobacteria which may be applicable to AFA. Finally, we present a strategic plan for the establishment of an integrated AFA wild harvesting and cultivation workflow at Klamath Lake, aiming to facilitate further translational research on this promising microorganism.

Aphanizomenon flos-aquae Characterisation

Morphology, physiology and life cycle

Aphanizomenon flos-aquae is composed of long trichomes (= filaments) that aggregate into 1.5 – 2 cm long fascicles, which can be described as bundles of trichomes (Fig. 2) (Carmichael et al. 2000). It is believed that the formation of these fascicles in AFA serves as a defence mechanism against grazing. This hypothesis is based on the fact that Daphnia pulex, a common water flea, is unable to graze AFA filaments longer than 1.5 cm (Dawidowicz 1990). Each trichome in AFA consists of intercalary cylindrical vegetative cells, typically straight but occasionally bent. The ends of each trichome are composed of a short series of terminal cells, which are also called hyaline apical cells (Komarek 2006; Komárek 2013). Each vegetative cell is enwrapped in a peptidoglycan cell wall, while each trichome is subsequently enveloped in a polysaccharide cellulose-like sheath (Cirés and Ballot 2016). The dimensions of the different cell types are detailed in Table 1.

Fig. 2

Micrographs of Klamath Lake Aphanizomenon flos-aquae (AFA). A), B) and C) show Klamath AFA fascicles and filaments at 1x, 10x and 100x magnification, respectively.

Table 1 Description of the length and width of the different Aphanizomenon flos-aquae -type cells.

Importantly, AFA forms heterocysts and akinetes, both of which are specialised vegetative cells (Fig. 3). The former, which is characterised by the presence of an active nitrogenase enzyme, fixes dinitrogen (N₂) from the air into ammonia (NH₃) which is then used in the production of other nitrogenous compounds including nitrites and nitrates. These bioavailable compounds are used to build key cellular components such as amino acids and nucleotides (Haselkorn 1978). Akinetes, instead, can be thought of as “AFA spores”. These are dormant cells which form under unfavourable conditions such as during the winter season, and germinate to form new vegetative cells when favourable conditions return (Wildman et al. 1975). They develop from vegetative cells under stress, which can occur, for example, under conditions of low temperature, inadequate irradiance, and low phosphorus concentration. Akinetes are a crucial part of AFA’s life cycle, as seen in Fig. 4, and have been reported to survive for up to 66 years (Kaplan-Levy et al. 2010).

Fig. 3

AFA cell type morphology. B) micrograph of Aphanizomenon flos-aquae NIES 4052 trichomes at 40x magnification showing its vegetative cells, terminal cells, heterocyst (shown in A) at 100x magnification, and akinete (shown in C) at 100x magnification. Scale bar: 10 µm

Fig. 4

Illustration of Aphanizomenon flos-aquae (AFA) life cycle within freshwater lakes, with its bloom occurring in the summer and autumn seasons and its dormancy in the winter seasons. Following akinite germination, vegetative cells divide exponentially via binary fission allowing for the formation of AFA blooms. Adapted from (Yamamoto and Nakahara 2009a)

AFA discovery, ecology and distribution

Aphanizomenon flos-aquae was first isolated in 1883 from Lake Tetonka, Minnesota, USA (Cirés and Ballot 2016). Since then, 122 different AFA strains have been reported based on morphological and phylogenetic analyses, and have been isolated from sites around the world (Suppl. Table 1). AFA is found in North America, Asia, and all across Europe, from Portugal and Spain to the Baltic Sea and Scandinavia (Cirés and Ballot 2016). Currently, there are 40 lakes and 17 countries worldwide that are known to harbour AFA (Fig. 5). AFA is predominantly found in freshwater lakes at high latitudes (temperate zones) as it prefers temperatures between 23°C and 29°C (Tsujimura et al. 2001; Yamamoto and Nakahara 2009b). However, AFA is highly adaptable: a notable example is its ability to grow in the brackish waters of the Baltic Sea (Cirés and Ballot 2016). The cyanobacterium can endure temperatures up to approximately 30°C, and AFA trichomes are known to survive at 10°C (Yamamoto and Nakahara 2005; Reinl et al. 2023). One study from Lake Stechlin even describes “winter blooms” of AFA at 2°C (Üveges et al. 2012). Similarly, AFA has a high resistance to alkalinity, growing at a pH of 10 (Yamamoto and Nakahara 2005). Cyanobacterial blooms have actually been shown to directly increase lake pH levels as dissolved CO₂ is consumed during photosynthesis (Wood et al., 1996, Newton 2015, Zepernick et al. 2021). Like most cyanobacteria, AFA prefers low light intensities, around 40-60 µmol photons m^-2 s^-1 (Park et al. 2018). AFA's ability to control its position within the water column by the production of gas vesicles allows it to regulate the amount of irradiance it receives and, thus, to fine-tune and optimise photosynthesis (Yamamoto 2009).

Fig. 5

Wild distribution of Aphanizomenon flos-aquae (AFA) across the globe. Each pin represents a specific lake from which AFA has been collected. All AFA strains have been confirmed morphologically and phylogenetically. All locations are listed in Suppl. Table 1. AFA strains FACHB 1200, 1208, 1209, 1249, 1259, 1260, 1265, 1287, 1290, 1040, 1168, 1169, 1170, 1171, Zayi, SAG 31.87 and 0HA35S1 are not listed as their exact locations have not been determined. Figure made via Google Maps. To view the interactive map and their associated Google Earth location use this URL: https://www.google.com/maps/d/viewer?mid=1gsHAkbBdW-BF0tiqCrcit6JdEA5ZLZk&ll

Aphanizomenon flos-aquae typically cohabits with other cyanobacteria, most commonly Microcystis species: mainly M. aeruginosa, Dolichospermum/Anabaena flos-aquae and Gloeotrichia echinulata (Aavad 1994; Eldridge et al. 2012) (Fig. 6). However, the underlying factors governing their ecological interactions within their relative ecosystems are poorly understood. Recent research has highlighted some key factors which may explain the seasonal dynamics between AFA and Microcystis. The latter has a preference for higher temperatures and light intensities, unlike AFA, favouring Microcystis blooms in the warmer months (Wu et al. 2010a; Wen et al. 2022). Relative concentrations in the lakes are also greatly affected by nutrient availability: Microcystis growth is favoured under low phosphorus conditions as it has a more effective phosphorus uptake and storage system; while AFA, being a nitrogen-fixer (unlike Microcystis), is greatly favoured under low nitrogen conditions (Wu et al. 2010a; Eldridge et al. 2012). Microcystis blooms generally occur following the decline of an AFA bloom, which most likely reflects a release of bioavailable nitrogen into the environment as AFA cells break down (Yamamoto and Nakahara 2009b). This hypothesis is supported by a study conducted in Lake Dianchi, China, where coexisting AFA and M. aeruginosa were observed. In 2010, the nitrogen levels in the lake increased significantly following industrial wastewater discharges, leading to a gradual decline in AFA blooms and a rapid take over by M. aeruginosa, ultimately resulting in the complete disappearance of AFA blooms in 2012 (Wu et al. 2016). Nitrogen may therefore serve as a limiting factor for the growth of Microcystis. However, it is unclear whether external inputs such as temperature cause AFA blooms to decline naturally and favour those of Microcystis, implying a facilitative relationship, or whether Microcystis strains may directly influence the decline of the AFA bloom, making their relationship competitive (Eldridge et al. 2012). Laboratory experiments point towards the latter hypothesis: Microcystis growth is directly correlated with the growth inhibition of AFA cells (Ma et al. 2015) (Wen et al. 2022). It is thought that Microcystis may release a compound which suppresses AFA proliferation. However, this has yet to be confirmed (Ma et al. 2015). The interplay between AFA and Anabaena/Dolichospermum strains is much less understood, as they have similar nutrient and growth condition preferences (Yamamoto 2009).

Fig. 6

Micrographs of cyanobacterial isolates from Upper Klamath Lake. a) Gloeotrichia echinulata, b) Anabaena flos-aquae, c) Microcystis aeruginosa. Scale bars: 100 µm

Taxonomy, phylogeny and potential toxicity

The genus Aphanizomenon was first established in 1886 (Bornet and Flahault 1886). However, the taxonomy of the genus has been the subject of several revisions, which, until the advent of molecular tools, were based purely on morphology (Komarek 2006). The most well-defined and accepted morphological criteria were established by Komarek in 1989 (Komárek 1989): it subdivided the genus into four subgroups depending on their fascicle formation, cell size, heterocyst frequency, trichome symmetry, and the shape of the terminal cells (Li et al. 2000; Komárek and Komárková, 2006; Komárek 2013). While these criteria did provide an accurate taxonomic framework for the majority of Aphanizomenon species, it did not for some. Species within different sub-groups were found to have overlapping characteristics, especially between A. flos-aquae and Aphanizomenon gracile strains, which have been often misidentified as one another due to their trichome similarity (Cirés and Ballot 2016). An example of this is observed with the taxonomic categorisation of Aphanizomenon sp. NH-5 (AY196086) over the years: it was initially identified as an AFA strain, later as an “atypical non-fasciculated strain”, and ultimately as A. gracile strain (Li et al. 2000). The sole morphological attribute that makes AFA obviously distinguishable, especially in comparison with A. gracile, is its ability to form fascicles.

With the advent of molecular phylogenetic analysis, the taxonomic arrangement of the Aphanizomenon genus underwent a complete revision. Investigation into conserved DNA sequences – e.g. genes for 16S rRNA, phycocyanin, ribulose-bisphosphate carboxylase and the sequence of rRNA internal transcribed spacers, found that several strains were only distantly related to each other (Wu et al. 2010b; Komárek 2013; Cirés and Ballot 2016). This led to the re-classification of several species, such as A. gracile or Aphanizomenon issatchenkoi, into different or novel genera (Rajaniemi et al. 2005). The genus Aphanizomenon itself was restricted to AFA and seven of its morphospecies: A. klebahnii, A. yezoense, A. paraflexuosum, A. flexuosum, A. slovenicum, A. platense and A. hungaricum (Cirés and Ballot 2016). However, AFA and A. klebahnii are the only strains within the genus that have had their genomes or highly conserved genomic regions sequenced and made available. As a result, morphology, with special attention to heterocyst and akinete size, is the sole basis for intra-genera identification (Cirés and Ballot 2016). As such, there is an urgent need for further phylogenetic studies that include all eight morphospecies to validate the legitimacy of this genus.

Despite the establishment of the genus Aphanizomenon, there remains considerable disagreement concerning the taxonomic classification of species that share genetic similarity with AFA but differ in morphology (Dreher et al. 2021a). Examples of this are observed with Anabaena and Dolichospermum species: these have been shown to cluster with AFA in phylogenetic analyses, indicating that they are not monophyletic (Cirés and Ballot 2016; Driscoll et al. 2018; Dreher et al. 2021b). The three genera, in fact, are collectively referred to as the “Anabaena, Dolichospermum and Aphanizomenon (ADA) clade”. Recent papers have proposed that the ADA clade become a single genus (Wacklin et al. 2009; Dreher et al. 2021a; Dreher et al. 2021b). However, the morphology of Anabaena and Dolichospermum, defined by oval/globular and elongate-barrel shaped vegetative cells and heterocysts, differs from the cylindrical structure of AFA (Prasanna et al. 2006). This further underscores the significant misalignment between morphological and phylogenetic taxonomical categorisations and highlights the necessity of employing a combined approach for a more comprehensive and accurate identification process.

The taxonomic uncertainties may have been further exacerbated by the misidentification of filamentous diazotrophic strains during the 20th century, as observed with A. gracile NH-5 (AY196086), and the submission of unverified DNA and protein sequences to official databases, without being associated with a peer-reviewed published paper (Li et al. 2000; Palinska and Surosz 2014). Deposition of incorrectly identified strains to culture collections, which may subsequently feature in phylogenetic trees and be erroneously used as a proxy for an genus of interest, has also potentially skewed our understanding of various cyanobacterial genera, including Aphanizomenon (Komárek and Komárková, 2006). For example, studies have highlighted the ability of certain Aphanizomenon species to produce cyanotoxins: specifically, paralytic shellfish poisons such as saxitoxins, anatoxin-α, and cylindrospermin (Mahmood and Carmichael 1986; Li et al. 2000; Pereira et al. 2000; Ferreira et al. 2001; Li et al. 2003; Liu et al. 2006; Preußel et al. 2006; Zhang et al. 2013). However, a recent review by Cires and Ballot (2016) demonstrated that the majority of the investigated strains had been misidentified as AFA rather than morphologically similar cyanobacteria such as A. gracile and A. issatchenkoi, which are known to be important cyanotoxin producers. The few remaining cyanotoxin-producing species that have been investigated have not been confirmed as AFA due to a lack of available phylogenetic data making their taxonomic assignment uncertain (Cirés and Ballot 2016). To date, AFA is yet to be confirmed as a cyanotoxin producer and Klamath Lake AFA is, in fact, approved as edible by both the FDA and EFSA (Schaeffer et al. 1999; Lähteenmäki-Uutela et al. 2021). Nevertheless, further research is required to confirm the safety of the genus Aphanizomenon, especially due to the morphological or phylogenetic similarity to other cyanotoxin-producing nitrogen-fixing filamentous cyanobacteria. One initial approach would be to investigate the presence of genes for biochemical pathways associated with cyanotoxin production within the available AFA genome sequences accessible through the NCBI database.

Industrial role of AFA

Benefits

Klamath AFA, the only AFA strain to be characterised from a nutritional and nutraceutical perspective, is one of the richest foods on the planet and is considered a “superfood”. It has been reported to have average protein levels of 60-70% (Grewe and Pulz 2012) and contains a high concentration of essential amino acids (Sandgruber, 2021). In addition, Klamath AFA contains mycosporine-like amino acids (MAAs), specifically shinorine and porphyra, which are known to confer antioxidant properties and DNA damage protection (Righi et al. 2016). Furthermore, Klamath AFA is a source of all 14 vitamins (although, see the discussion on vitamin B₁₂ below), and 73 trace minerals (Sandgruber et al. 2021). More specifically, 1 g of dried Klamath AFA biomass contains Recommended Daily Allowance (RDA)-relevant amounts of vitamins A, K, B₁, B₃, B₅ and B₉, and the minerals, iron, molybdenum and iodine (Tables 2 and 3) (Kushak et al. 2001; McCommon et al. 2013). Klamath AFA is also an important producer of carotenoids with high levels of three xanthophylls: lutein, echinenone and canthaxanthin, and two carotenes: lycopene and β-carotene (Hertzberg and Jensen 1966; Krajewska et al. 2019). Moreover, Klamath AFA produces a number of polyphenols, specifically caffeic, vanillic and hydroxytyrosol acid (Righi et al. 2016). These are thought to contribute to the cyanobacteria’s nutraceutical effects by conferring anti-cancer, anti-inflammatory, antioxidant, cell growth promotion, and wound healing properties (Cory et al. 2018; Peng et al. 2023).

Table 2 Table summarising and comparing the macromolecules and vitamins of commercial Klamath AFA, Spirulina (Arthrospira) and Chlorella products. The lowest and highest values for each compound were chosen from the available literature, independent of the specific species and strains of Spirulina or Chlorella. Vitamin A only considers carotenes. All Klamath AFA information was obtained from the literature (McCommon et al. 2013; Sandgruber et al. 2021), and from unpublished data provided by Nutrigea s.r.l, a distributor of Klamath AFA (Suppl. Table 2). Spirulina references: Ali and Saleh, 2012, Soni, 2017, Andrade et al. 2018, Sangruber et al., 2021, Chlorella references: Safi et al. 2014, Andrade et al. 2018, Bito et al. 2020, Sandgruber et al. 2021

Table 3 Table summarising and comparing the mineral, essential metals & bioactive compounds of commercial Klamath AFA, Spirulina (Arthrospira) and Chlorella products. The lowest and highest values for each compound were chosen from the available literature, independent of the specific species and strains of Spirulina or Chlorella. All Klamath AFA information was obtained from the literature, and from unpublished data provided by Nutrigea s.r.l, a distributor of Klamath AFA (Suppl. Table 2)

Klamath AFA is also a rich source of polyunsaturated fatty acids (PUFAs). It produces 12-14 mg g^-1 of the anti-inflammatory omega-3 fatty acid, alpha-linoleic acid and this accounts for approximately 25% of its total lipid biomass (Sandgruber et al. 2021). The balance between omega-6 and omega-3 fatty acids plays a crucial role in regulating bodily inflammation (Simopoulos 2002). Omega-6 fatty acids tend to promote the production of pro-inflammatory mediators, while omega-3 fatty acids favour anti-inflammatory mediators (Simopoulos 2002, DiNicolantonio and O’Keefe, 2018). The ideal omega-6 to omega-3 ratio is believed to be between 1:1 and 4:1. The average omega-6/omega-3 ratio in the modern Western diet is between 10:1 and 20:1, largely due to the increased consumption of processed and fried foods within the past few decades (Simopoulos 2008, Di Nicolantonio and O’Keefe, 2018). This imbalance is known to cause chronic inflammation and obesity, which are important risk factors for many chronic diseases, such as heart disease, arthritis, and cancer. Regular consumption of Klamath AFA would help reduce bodily inflammation and the risk of developing chronic illnesses owing to its high concentration of omega-3 PUFAs (Sandgruber et al. 2021). In contrast, Arthrospira produces, for the most part, only omega-6s, whilst there is a wide and unknown variability between different Chlorella products, making it unreliable for the end-user. Even so, while Klamath AFA has a lower omega-6/3 ratio, some Chlorella strains actually have a higher amount of omega-3 PUFA due to their high lipid content (Table 2 and Fig. 7A) (Sandgruber et al. 2021).

Fig. 7

Bar charts illustrating the mean concentration of PUFAs (A) and bioactive compounds (B) in Klamath AFA, Spirulina (Arthrospira) and Chlorella supplements based on the minima and maxima reported amongst commercially sold products. The error bars detail the minima and maxima. For example, the maximum reported Omega-3 content in Chlorella products is 0.7 mg g^-1 of dry biomass, while the minima approach 0

Moreover, the pigments found in Klamath AFA, specifically chlorophyll and phycocyanin, possess well-known nutraceutical properties (Benedetti et al. 2006). Chlorophyll, characterised by a bitter taste, reduces inflammation, aids in weight loss, and acts as a prophylactic against certain cancers. Cyanobacterial phycocyanin (C-PC) is a powerful antioxidant, anti-inflammatory, anti-cholesterol and anti-cancer agent, with important promoting properties for the immune system and wound healing (Fig. 8A) (Jensen et al. 2000, Reddy et al. 2000, Romay et al. 2003, Madhyastha et al. 2008, Kuriakose and Kurup 2010, Xue et al., 2015, Kefayat et al. 2019, Zahiri et al. 2020, Blas-Valdivia et al. 2022, Liu et al. 2022). The inhibitory effect of C-PC on cycloxygenase-2 (COX-2), a pivotal mediator of inflammation, is analogous to the non-steroidal anti-inflammatory drug Celecoxib (Romay et al. 2003; Gong et al. 2012). Klamath AFA and Arthrospira both produce similar concentrations of C-PC, around 15% of the total biomass, while Chlorella as a green microalga does not produce any (Table 3). However, Chlorella does have higher levels of chlorophyll (Fig. 7B). Klamath AFA C-PC extracts, such as AphaMax®, also concentrate the related pigment phycoerythrocyanin (PEC) that is absent in Arthrospira. PEC is identical to phycocyanin in structure except in that it contains a phycoviobilin chromophore, giving it a deep violet colour (Bryant 1982; Galizzi et al. 2023). Further research is required to investigate the biological properties of PEC. Overall, the physiological effects of C-PC extracts from Klamath AFA and Arthrospira are noteworthy, and their use as a prophylactic agent against specific diseases – from inflammatory-based ones to cancers – is garnering increasing interest. Arthrospira C-PC extracts are in fact being considered as a potential drug candidate for the treatment of cancer (Braune et al. 2021).

Fig. 8

Mechanism of action of C-PC (A) and PEA (B). a) C-PC specifically inhibits the COX-2 enzyme, nitric oxide production, and cancer cell proliferation while also promoting the mobilisation of immune cells b) PEA’s nutraceutical effects are through its stimulation of the hTAAR receptor: hTAAR brain stimulation causes an increase in brain EPO, leading to angiogenesis and neurogenesis, catecholamine reuptake and VMAT2 receptor inhibition, leading to increased catecholamine concentrations. Red arrows indicate inhibition, while black ones indicate stimulation. Abbr.: COX-2: Cyclooxygenase 2; VMAT2: vesicular monoamine transporter 2. hTAAR: Human trace amine-associated receptor: SC: Stem Cell; NK: Natural Killer. Figure generated using Biorender

Another key compound that has gained notoriety in recent years is β-phenylethylamine (PEA), an endogenous phenolic compound primarily produced in humans during exercise. Amongst commercially available microalgal and cyanobacterial supplements, PEA is present only in Klamath AFA (Irsfeld et al. 2013; Nuzzo et al. 2018). PEA has important psychoactive properties which make it a powerful anti-anxiety and anti-depressant agent, natural painkiller, libido enhancer, hunger suppressant, and immune modulator (Barroso and Rodriguez 1996; Hart et al. 2007; Ohta et al. 2017). Klamath AFA PEA extracts, approved and sold alongside Klamath AFA biomass, contain approximately 1.4 mg g^-1 of PEA (McCarthy et al. 2022). PEA’s half-life when orally ingested is around 5-10 min as it is rapidly broken down by monoamine oxidases (MAO), specifically MAO-B enzymes (Irsfeld et al. 2013). However, Klamath AFA PEA extracts also contain potent MAO-B inhibitors in the form of MAAs (Scoglio et al. 2014). These have similar potency to the synthetic MAO-B inhibitor, Deprenyl, and can cross the blood-brain barrier, thereby greatly extending PEA’s overall half-life and allowing it to exert its mechanism of action over an extended period (Fig. 8B) (Magyar and Szende 2004; Scoglio et al. 2014). The Klamath AFA PEA extract (commercially marketed as Klamin®) has been shown to significantly improve the general well-being and self-esteem in individuals with post-menopause-induced depression (Scoglio et al. 2009; Genazzani et al. 2010). The extract has also been shown to improve the overall condition of children suffering from ADHD, both in terms of attention deficit and hyperactivity (Chen et al. 2007; Wang et al. 2012; Cremonte et al. 2017). In addition, the underlying mechanism of PEA’s neuronal and brain tissue regeneration/survival properties may also play a prophylactic role against the development of neurodegenerative diseases such as Alzheimer’s (Shytle et al. 2010; Borhani-Haghighi et al. 2012; Nuzzo et al. 2018; Galizzi et al. 2023). Klamath AFA PEA extracts hold substantial potential within the psychoactive space and the health benefits merit further in-depth research.

Overall, Klamath AFA’s nutritional profile and nutraceutical properties, though comparable to Arthrospira and Chlorella, demonstrate more potent and comprehensive prophylactic and therapeutic effects, especially when considering PEA’s psychoactive properties (Sandgruber et al. 2021)(Fig. 8B). The nutritional value of Klamath AFA is only slightly diminished, as stated by Sandgruber et al., by its elevated levels of saturated fatty acids (SFAs), particularly C16:0, which are associated with heightened inflammation and cholesterol levels (Sandgruber et al. 2021). Finally, it is important to note that while many strains and species of Arthrospira and Chlorella have been characterised for nutritional content, Klamath AFA is the only Aphanizomenon species to be characterised (Sangruber et al., 2021). It would be worthwhile to characterise other AFA strains as they may possess different nutritional and nutraceutical properties and allow the expansion of AFA’s biotechnological opportunity and market.

Vitamin B₁₂

One market that has been receiving attention recently is vitamin B₁₂ supply for the vegetarian and vegan industries as there is now an important consumer demand for non-animal sources of vitamin B₁₂ supplements (Curtain and Grafenauer 2019). As green algae such as Chlorella are eukaryotesthey are not able to synthesise B₁₂. However, Chlorella is often grown in co-culture with B₁₂-producing bacteria and its cells are able to actively import the B₁₂ from the medium, making Chlorella and other edible green algae potential sources of B₁₂ (Bito et al. 2020). Nonetheless, a recent analysis by Bito et al. (2016) shows that only four out of thirteen commercially sold Chlorella tablets contained B₁₂, most likely due to the differences in their associated bacterial populations. Furthermore, the amount of B₁₂ in the Chlorella products is highly variable, with values ranging between 0.2 to 5 µg g^-1, complicating matters for potential customers (Bito et al. 2020). Cyanobacteria are also unable to synthesise the vitamin, but instead produce a B₁₂ analogue termed pseudocobalamin or pseudo-B₁₂ (Fig. 9) (Baroni et al. 2009; Helliwell et al. 2016). Arthrospira and Klamath AFA produce 1-2 µg g^-1 and 6-8 µg g^-1  of pseudo-B₁₂, respectively. Pseudo-B₁₂ is poorly absorbed by the body due to its lower binding affinity to the Intrinsic Factor protein which allows for B₁₂ absorption in the stomach. As a result, pseudo-B₁₂ should not be used for supplementation purposes (Wienhausen et al. 2022).

Fig. 9

Structure differences between B₁₂ (cyanocobalamin) and pseudo-B₁₂ (pseudocyanocobalamin). In general, the term cobalamin refers to any cobalt-containing corrin structure. Pseudo-B₁₂, produced by cyanobacteria, distinguishes itself from B₁₂ by featuring adenine in its lower axial ligand position (R2) instead of DMB.

Nevertheless, two spectroscopy studies have shown that Arthrospira and Klamath AFA also contain human-absorbable B₁₂, despite the lack of genes for its biosynthesis in these cyanobacteria. More specifically, Klamath AFA has been found to have 0.32-0.42 µg g^-1 of B₁₂ such that a few grams of dry Klamath AFA biomass per day would satisfy the current RDA for male adults (2.4 µg g^-1) (Miyamoto et al. 2006). Once again, the B₁₂ is thought to derive from bacterial communities present within Klamath AFA blooms with released B₁₂ present in the water being taken up by AFA cells (Watanabe et al. 1999; Miyamoto et al. 2006; Helliwell et al. 2016). For Klamath AFA, these bacteria are likely represented by a subset of the gammaproteobacteria (Driscoll 2016). An in vivo study by Baroni et al. (2009) carried out using vegan individuals provided evidence that Klamath AFA supplements are a reliable source of B₁₂. However, it was not established whether the B₁₂ is imported into the Klamath AFA cells, as with Chlorella, or whether the gammaproteobacteria are harvested from the wild alongside the Klamath AFA and are therefore a component of the end product. Recent research suggests that cyanobacteria possess a B₁₂ import system known as the btu import system (Pérez et al. 2016). The presence of this import system in AFA followed by experimental validation of B₁₂ translocation from the exogenous space into the AFA cell is required to confirm that the harvested Klamath Lake AFA biomass is the source of B₁₂, rather than the associated commensal bacteria. Confirmation of this would signify that fluctuations in the amount of bacterial contamination within the supplements would not affect the amount of bioavailable B₁₂, thereby making it a reliable source of the vitamin, so long as it is being produced by bacterial populations within Klamath Lake.

Klamath Lake AFA harvesting

Klamath AFA biomass has been harvested from Klamath Lake in Oregon, USA since the early 1980s (Fig. 1) (Carmichael et al. 2000). The cyanobacterium is harvested from the lake’s surface during the bloom seasons via rolling screens on self-powered barges. Each year 500-1000 t dry weight is harvested and distributed around the globe, with a particular focus on USA and European markets (Carmichael et al. 2000). The biomass is sold either as a whole-cell superfood supplement or as preparations of its bioactive compounds such as C-PC and PEA, which are sold as high-value nutraceuticals. Nevertheless, the global Klamath AFA market has remained fixed at around £100 millon and has not been able to expand since the early 2000s, unlike that of Spirulina and Chlorella (Carmichael et al. 2000; Pulz and Gross 2004; Grewe and Pulz 2012). The underlying reason is the lack of a continuous and reliable stream of Klamath AFA biomass supply owing to the dependence on wild harvesting. Firstly, the AFA blooms are dependent on nature and its seasons (Carmichael et al. 2000). These blooms predominantly occur, although sporadically, between the end of June and the beginning of July and, again, between the beginning of September and Mid-November (Eldridge et al. 2012). Secondly, harvesting companies must ensure that their AFA product is not contaminated with microcystins. The Federal State of Oregon set a stringent limit in the early 2000s of 1 µg g^-1 of microcystins in Klamath AFA-based nutritional supplements based on the World Health Organisation limit of 1 µg L^-1 of microcystin in water, thereby forcing harvesting companies to hold off any harvesting if M. aeruginosa is present and discard any Klamath AFA biomass that contains microcystins above the WHO limit (Schaeffer et al. 1999; Saker et al. 2005; Heussner et al. 2012; Vichi et al. 2012; Lyon-Colbert et al. 2018). Microcystis aeruginosa blooms do not coincide with those of Klamath AFA, as they occur between late July and the end of August, but they can certainly overlap or be present in low concentrations (Carmichael et al. 2000). These factors have limited the scope and range of AFA’s downstream applications. Klamath AFA is currently the only commercial cyanobacteria that is not grown in a controlled environment, and technologies for its cultivation in open ponds or PBRs would offer an important opportunity to resolve many of the issues associated with wild harvesting.

Future prospects

AFA cultivation parameters

Published studies on the cultivation of AFA in laboratory and photobioreactor (PBR) settings is limited and its optimal growth requirements have yet to be fully characterised. AFA prefers low light intensities (40-60 µmol photons m^-2 s^-1) but can tolerate up to a maximum of 115 µmol photons m^-2 s^-1, beyond which AFA experiences photoinhibition leading to reduced growth rates and eventually death (Gentile and Maloney 1969). Prolonged exposure to lower irradiances, above 60 µmol photons m^-2 s^-1, can also cause loss of buoyancy owing to the accumulation of polysaccharides within the cells, resulting in sedimentation and cell death (Konopka et al. 1987). To avoid these effects, most cultivation regimes advise a daily light-dark cycle between 12h:12h and 16h:8h (Debella 2007, Xue et al., 2015). AFA's growth is relatively unaffected by pH between 7.1 and 11, with its optimal at ~7.5 (Yamamoto and Nakahara 2005). AFA can also tolerate a range of temperatures but demonstrates faster growth rates when exposed to temperatures between 26°C and 28°C (Gentile and Maloney 1969; Debella 2007). The addition of CO₂ to the culture increases AFA growth and biomass yield, although the exact CO₂ levels required are not well-defined (Debella 2007). Various different cyanobacterial media have been recommended for AFA cultivation including BG11, CT, ASM, and JM media (Watanabe 1977, Jaworski et al. 1981, Castenholz 1988, Carmichael et al. 2000, Tsujimura et al. 2001, Debella 2007, O'Flaherty and Phinney, 2008).

These parameters allow for laboratory maintenance and cultivation of AFA. However, these parameters may not translate to the upscaling process and may require further optimisation (Acien et al. 2013). If one growth parameter was suboptimal in a PBR this would markedly affect overall productivity as all parameters are interdependent. For example, an inexact mass transfer of CO₂/O₂ could lead to a decreased growth rate and biomass density thereby altering the light exposure of AFA cells and potentially increasing photo-saturation/-oxidation (Acien et al. 2017). One sub-optimal parameter could generate a productivity-lowering positive feedback loop which may essentially impede AFA cultivation (Acien et al. 2017). One major concern, for example, is the mixing regime. As AFA is filamentous it is relatively fragile and susceptible to shear stress (Wang and Lan 2018). In Klamath Lake, AFA is generally mixed by small currents, methane springs and wind-induced waves. Reliance on continuous bubbling, mechanical stirring, or aeration mixing may compromise the culture (Wang and Lan 2018). Overall, the technological feasibility of AFA mass cultivation poses the greatest barrier to its realisation. Therefore, laboratory and pilot scale optimisation studies are required prior to AFA mass culture deployment. Low-cost proxies, such as vertical plastic hanging bags could be used for lab scale investigation of AFA cultivation feasibility and growth at commercial scale (Cui et al. 2022). Furthermore, while Klamath AFA is the obvious strain to cultivate, exploring the cultivation potential of other AFA strains is warranted. Such strains might possess different growth properties that lend themselves well to cultivation.

The selection of a suitable commercial cultivation system is a critical factor in the mass cultivation of microalgae and cyanobacteria. Currently, two main types of systems are employed: open and closed systems (Zittelli et al. 2013). The predominant open system is the raceway pond system (Acien et al. 2017). Drawbacks associated with this system include contamination, poor gas/liquid mass transfer, poor temperature control and light utilisation efficiency (Zittelli et al. 2013; Carney and Lane 2014). Contamination can be minimised by employing media conditions such as alkaline pH, thereby minimising fouling from other microorganisms, and worldwide Spirulina production occurs predominantly (<90%) in raceway ponds owing to its ability to tolerate high pH levels, similar to AFA (Acien et al. 2013). Raceway ponds would provide the best system to mimic Klamath Lake conditions and potentially obtain successful AFA mass cultivation.

Closed systems include tubular photobioreactors (tPBRs), flat panels and columns (Singh and Sharma 2012). These PBR systems provide accurate control of culture conditions and transfer of nutrients and gas between the inside and outside environment, thereby minimising any culture contamination and maximising production of high-value compounds (Acien et al. 2013; Acien et al. 2017). PBRs typically possess a biomass productivity rate ~5-10x higher than raceways. However, they are much more capital intensive: their deployment costs can be ~20x more expensive and operating costs as much as 15x that of a raceway pond (Lizzul 2016, Narala et al., 2016, Acien et al. 2017). PBR use is typically restricted to high-value products from microalgae or cyanobacteria, such as astaxanthin or phycocyanin (Ranjbar et al. 2008; Yuan et al. 2011). The most employed PBR system is a serpentine tubular PBR (stPBR) as it is easily cleanable, extremely reliable and automated. These stPBRs are composed of straight tubes interconnected by U-bends to create a flat loop that may either be arranged vertically or horizontally (Zittelli et al. 2013). Nevertheless, stPBRs can reach linear liquid velocities of 20-50 m s^-1, potentially causing unsuitable turbulent conditions for AFA (Singh and Sharma 2012). Another potential option for AFA cultivation are flat panel PBRs: cuboidal-shaped reactors with a width ranging between 0.07-5 cm which provides a minimal light path (Singh and Sharma 2012; Banerjee and Ramaswamy 2019). This latter characteristic allows high final biomasses as the light is able to reach cells even at the centre of a dense culture, namely the “aphotic zone”, which would otherwise be unable to receive light (Masojídek et al. 2015). Arthrospira grown in flat panels has reached biomass productivities of up to 12 g L^-1 day^-1, the highest reported of any PBR (Singh and Sharma 2012). The flat panel’s simple design, adaptable mixing rate and minimal light path could provide the optimal platform for maximising AFA production for high value bioactives and macromolecules.

Nutrigea s.r.l., a European distributor of Klamath AFA, have reported that 1 t of AFA currently generate approximately US$1 million in overall revenue. These sales figures are based on 1 t of dry AFA biomass being sold as a nutritional supplement for human consumption, together with C-PC extract sales that has been extracted from 60% of the biomass. Typical raceway pond yields of Arthrospira range from 10 to 15 t ha^-1 year^-1, while PBR yields average approximately 40 to 50 t ha^-1 year^-1 (Acien et al. 2013). Based on these results, the estimated maximum revenues obtainable from AFA cultivation range from US$10 –15 million for raceway ponds and US$32 – 40 million for PBRs. While these estimates highlight the commercial potential of growing AFA within PBRs, they may not be directly applicable to this cyanobacterium: unsuccessful cultivation, low productivities and/or low final biomass may not justify the high capital and operating costs of these cultivation platforms, especially when compared to the cost-effectiveness of AFA wild harvesting from Klamath Lake (Carmichael et al. 2000).

Towards Klamath Lake AFA cultivation

Klamath Lake AFA only requires sunlight and lake nutrients to bloom, both of which are provided by nature (Carmichael et al. 2000). According to Nutrigea s.r.l., the concentration of Klamath Lake AFA’s blooms range between 5-7% of total lake water content, or, in other words, 50-70 g L^-1. In comparison, typical final biomass Arthrospira concentrations obtained from open cultivation systems are 1-3 g L^-1 and, from closed cultivation systems, equal or above 10 g L^-1 (Costa et al. 2014). Klamath Lake AFA wild harvesting is more cost-efficient than any current PBR system. The only operating costs for wild harvesting are the energy to power the lake harvester and the workforce salaries (Carmichael et al. 2000). The density of the obtained biomass also lowers downstream processing costs compared to cultivated biomass, as the amount of dewatering required is less (Fasaei et al. 2018). In contrast, AFA cultivation would require supplementation with inorganic nutrients, provision of light emitting diodes (LEDs) to optimise light availability and regular PBR maintenance (Blanken et al. 2013). Nevertheless, cultivation would provide three main advantages over Klamath Lake harvesting: 1) the control and monitoring of AFA’s growth parameters; 2) a continuous yield of AFA, unaffected by seasonality; and 3) unialgal and microcystin-free biomass.

The optimisation of AFA growth parameters within PBRs such as flat panels, has the potential to increase the production of high-value compounds such as C-PC, carotenoids, and lipids compared to Klamath Lake AFA biomass (Sierra et al. 2008). Studies have shown that lower irradiances and wavelengths of 660 nm can increase C-PC production in cyanobacteria and could potentially raise AFA's C-PC concentrations above 15% of its whole cell dry biomass, a phenomenon already observed in Arthrospira cultivation (Takano et al. 1995; Rivera et al. 2021). Similar techniques could be applied to enhance the production of carotenoids, which prefer wavelengths between 500-600 nm (Maltsev et al. 2021). Temperature modifications can also be employed to favour the production of lipids, with lower temperatures known to increase the amount of PUFAs (Balakrishnan and Shanmugam 2021). Overall, controlled growth would allow a boost to final product pricing and enable better targeting of specific markets. For instance, AFA biomass with an elevated omega-3 content could be sold as an aquaculture product for fish oil applications (Remize et al. 2021). However, modification of cultivation parameters for the maximisation of secondary products could have knock-on effects on the growth rate and final biomass, thereby reducing any potential profit. Therefore, it will be important to focus on those products such as C-PC with already established markets that would yield the best net margins while ensuring its overall feasibility (Takano et al. 1995; Rivera et al. 2021). Most importantly, AFA cultivation would yield unialgal microcystin-free biomass all-year-round, providing a marketing and supply advantage compared to wild harvesting (Tan et al. 2020). In addition, the know-how underlying successful AFA mass cultivation would be a failsafe against any potential impacts of climate change on AFA blooms in Klamath Lake in the future. Currently, cultivated AFA products may not be able to compete with wild harvested AFA-based ones in terms of overall profitability, and it might be that a combination of controlled cultivation and wild harvesting would maximise the advantages of both methods (Table 4).

Table 4 SWOT (Strengths, Weaknesses, Opportunities and Threats) analysis of integration and combination of Klamath Lake AFA harvesting and mass cultivation

The productivity of Klamath Lake harvesting could be greatly increased by blending the wild harvested AFA biomass with that of microcystin-free cultivated AFA. This would significantly reduce the microcystin concentration within the wild harvested biomass, depending on the employed degree of dilution (Carmichael et al. 2000). In this scenario, Klamath Lake AFA could be harvested even with microcystins present, thereby increasing the amount of total harvestable days per annum and AFA supply, at minimal extra cost. Similarly, cultivated and wild-harvested AFA could be integrated according to the market need to ensure AFA biomass supply availability throughout the year. For example, if 10 t of dry AFA biomass were required to satisfy the estimated market demand and only 5 t of Klamath Lake AFA had been harvested, another 5 t of cultivated AFA could be blended in throughout the year , thereby also reducing the microcystin concentration by 50 %.

The successful integration of wild harvesting and cultivation would necessitate the deployment of AFA cultivation platforms close to the shore of Klamath Lake. This would save on capital investment for new equipment, as the machinery for downstream processing could be shared; save on overall transport costs, due to the proximity of cultivated and wild-harvested AFA biomass; and, most importantly, allow access to Klamath Lake’s natural resources. The lake water, which has the required nutrients for AFA growth, could be directly filtered and used as the employed PBR medium, thereby eliminating a large proportion of the cost for nutrient supply (Ganesh Saratale et al. 2022). It would be however necessary that correct filtering of all biological material is ensured, including bacteriophage that might target the AFA or commensal bacteria. Specific key nutrients such as phosphorus may need to be added to the lake water to improve the growth rate within PBRs (Yaakob et al. 2021). After biomass dewatering, the water could be returned to the lake, or, to a centralised treatment system, depending on the water quality, effluent, and discharge regulations. The changes in the lake's nutrient concentrations throughout the year suggest that the nutrients needed for AFA growth are, for the most part, present at the optimal concentrations during the bloom seasons (Wood et al., 1996, Eldridge et al. 2012, Walker et al. 2012). To address this, lake water collected during AFA blooms could be stored in large silos and subsequently combined with filtered lake water extracted at other times with specific key nutrients such as CO₂, phosphorus and nitrogen added to enhance growth (Debella 2007). The mineral and vitamins provided by Klamath Lake water would further ensure that the final biomass would have the same nutritional benefits as that harvested directly from the lake surface (Fig 10).

Fig. 10

Model for the integration of Klamath Lake AFA harvesting and cultivation. The biomass harvested from Klamath Lake and open ponds is combined and stored in the K.L. (Klamath Lake) AFA unit. Biomass derived from closed system photobioreactors (flat panels here), specifically enriched for particular products, is stored in the H.V. (High value) AFA unit. Importantly, the latter biomass could also be combined into the K.L. AFA storage unit to tackle any supply problems. The biomass processing facility then dewaters and dries all harvested biomass into a powder. The H.V. powder is subsequently transferred to a lab for extraction of high-value products such as C-PC, and the remaining AFA powder is further blended into the K.L. AFA powder. Finally, the whole-cell AFA biomass powder and the extracted high-value product go through product development and are distributed worldwide

As AFA is a diazotroph it may not necessarily require supplementation with nitrogen compounds such as ammonium or nitrate salts, unlike most other commercially grown microorganisms including heterotrophic species such as E. coli and yeast, and phototrophic species such as Chlorella and Arthrospira (De Nobel et al. 1997; Yaakob et al. 2021). Nitrogen is a primary element for cyanobacterial growth (together with carbon, phosphorus, iron, etc.) and the supply of bioavailable forms of nitrogen is one of the major costs in cultivation of microalgae and cyanobacteria (Peccia et al. 2013). For example, Zarrouk medium used for Arthrospira cultivation contains 2.5 g L^-1 nitrates (Madkour et al. 2012). However, the process of nitrogen fixation by diazotrophs is energy intensive and nitrate supplementation might result in faster AFA growth rates compared to cultivation in nitrate-free media (Touloupakis et al. 2022). Further research is required to verify this.

Another advantage of locating AFA cultivation close to Klamath Lake is the amount of sunlight it receives – circa 300 days annually (Nuzzo et al. 2018). This would facilitate the set-up of both open and closed cultivation systems. Open ponds could be completely sunlight-driven and employed during the summer months to generate AFA biomass (Slegers et al. 2013). As shown in Figure 10, open ponds could be used to maximise AFA supply and minimise the negative outcomes of a poor harvest from the lake. Closed systems could be reserved for the production of high-value compounds such as C-PC (Acien et al. 2017). The excess biomass could then be blended with the AFA from the lake or from the open ponds. Sunlight could also be captured using solar panels and the energy employed to power the majority of the associated closed-system operations, such as LED illumination and CO₂ bubbling (Nwoba et al. 2020). Importantly, the PBRs could also be repurposed to increase overall AFA biomass, especially during the winter months, in order to meet any AFA biomass yield targets. This would provide some harvesting and supply flexibility (Fig 10). Finally, Klamath Lake AFA may be used as the inoculum for cultivation, although adaptation of the cyanobacterium to a PBR environment might be necessary (Srivastava et al. 2013).

Overall, this model would increase the total harvestable AFA biomass from Klamath Lake, while ensuring a steady supply throughout the year. Additionally, closed-system cultivation would provide the opportunity to expand within high-value markets, as observed with the nutraceutical industry, and novel ones, including pharmaceuticals, sunscreens, and dyes or colourants (Rahman 2020; Stromsnes et al. 2021). C-PC extracts could be investigated for the treatment of cancer and/or inflammatory diseases, and PEA extracts for mood disorders and/or neurodegeneration (Nuzzo et al. 2018; De Caro et al. 2019; Braune et al. 2021; Galizzi et al. 2023). However, the drug development process is a lengthy and costly process which may threaten overall profitability (Mohs and Greig 2017). C-PC extracts could also be utilised as colourants within the food and drinks industry, or as fluorophores/markers in the biomedical industry (Kannaujiya et al. 2019). Another potential application is the use of AFA MAAs for the production of natural sunscreens as MAAs are low molecular weight compounds (<400 Da), which can dissipate light energy in the form of heat without generating ROS (Fig. 11) (Kumar et al. 2019). There are number of other potential compounds produced by AFA which may also hold important biotechnological potential, such as exopolysaccharide extracts for bioremediation treatments or lutein extracts for the prevention of eye diseases (De Philippis et al. 2011; Kumar et al. 2019).

Fig. 11

Aphanizomenon flos-aquae (AFA) biotechnological opportunities. Illustration of AFA’s potential market penetration following its successful mass cultivation. PC: Phycocyanin; PE: Phycoerythrin

Furthermore, an augmented supply of AFA would facilitate the entry of the cyanobacterium into the foods, animal feed and aquaculture markets. AFA’s nutritional properties could be delivered via its incorporation into everyday foods, such as pasta or baked goods. AFA PEA extracts have been shown to be heat-resistant and to survive the cooking process, a major challenge in terms of nutrient delivery (Caporgno and Mathys 2018; Nuzzo et al. 2019). The high protein, vitamin, and mineral content of AFA is likely to improve the growth, quality, and immune system of farmed fish and animals, as observed with Anabaena and Arthrospira species, although AFA-based animal studies are necessary to confirm this (Khatoon et al. 2010, Sudaporn et al. 2010, Fadl et al. 2020, Nagarajan et al. 2021, Altmann and Rosenau, 2022). Moreover, AFA's ability to fix nitrogen could make its whole cell biomass, or even its wet biomass, an excellent natural biofertiliser. AFA-based biofertilisers might provide an environmentally friendly alternative to chemical nitrogen fertilisers, which are known to cause water quality problems through fertiliser run-off (Righini et al., 2022, Alvarez et al., 2021, Pathak et al., 2018, Chittora et al. 2020).

Towards genetic engineering of AFA

The establishment of a cultivation system for AFA would enable the application of genetic engineering strategies that could potentially be used to create bespoke strains with improved traits. Whilst the number of transformable microalgal and cyanobacterial species is growing, the development of more advanced genetic engineering technologies is still limited to a handful of model eukaryotes such as Chlamydomonas reinhardtii, Nannochloropsis sp., and the cyanobacterial models Synechocystis and Synechococcus (Doron et al. 2016; Poliner et al. 2018; Santos-Merino et al. 2019; Sebesta et al. 2019; Jackson et al. 2021; Baldia et al. 2023). A summary of studies reporting the successful transformation of filamentous cyanobacteria is given in Table 5. The most commonly adopted method of DNA delivery in these systems is electroporation, in which an electrical field is applied to cells to increase the permeability of the cell membrane to plasmid DNA. In systems such as AFA where transgenic DNA integration into the genome would be expected to occur in a targeted manner via homologous recombination, it is necessary to identify a ‘neutral site’ where genomic integration would not disrupt or affect the expression of essential genes (Chaurasia and Apte 2011). Another critical prerequisite for DNA transformation is the availability of a suitable selectable marker together with regulatory DNA elements to drive its expression. The identification of neutral sites and endogenous elements is highly dependent on the availability of a sequenced and well-annotated genome for the host organism, which for Klamath AFA has already been made available (NCBI entry: ASM159382v2) (Driscoll 2016).

Table 5 A summary of DNA deliverytr methods and related studies on the transformation and genetic engineering of filamentous cyanobacteria. Chlor: Chloramphenicol; GFP: green fluorescent protein; CAT: chloramphenicol acetyltransferase

Achieving successful transformation of filamentous cyanobacteria is technically challenging owing to the difficulties associated with quantifying cell numbers during the culturing and transformation process. Two possible methods to overcome this are the use of sonication to break the filaments into single cells to enable counting, or the use of optical density to estimate cell numbers (Toyomizu et al. 2001). Typically, individual colonies are sub-cultured following transformation to ensure the isolation of a clonal cell line, however, this is often not possible when transforming filamentous strains (Geada et al. 2018). It is also unclear how a filamentous structure affects the dynamics of transgenic DNA propagation and the ability to reach homoplasmy whereby all copies of the genome within a polyploid system contain the transgenic DNA. One study has reported the use of sonication to break filaments of Arthrospra into single cells prior to transformation in an attempt to generate discrete transformant colonies, although this was shown to be unsuccessful due to the gliding mobility of the strain (Geada et al. 2018). Arthrospira and Nostoc spp. have a high prevalence of restriction endonucleases within their cells, and these are likely to be present also in AFA. Such endonucleases are known to inhibit transformation by degrading transgenic DNA within the cell before it can integrate stably into the genome (Moser et al. 1993; Walker et al. 2005). However, the use of endonuclease inhibitors has been shown to significantly improve transformation efficiencies in systems with a high endonuclease activity (Jeamton et al. 2017).

One other potential approach involves natural transformation, a process whereby prokaryotes actively take up extracellular DNA into the cytoplasm and integrate the acquired DNA into their genome via homologous recombination (Jester et al. 2022). Naturally competent prokaryotes can also take up whole self-replicating plasmids and express them as their own: this has been successfully employed to transform the filamentous heterocystous cyanobacterium Chlorogloeopsis fritschii PCC 6912 (AJLN00000000) (Nies et al. 2022). Natural competency is an evolutionary mechanism that enhances species survival by providing genetic diversity. This process depends on the presence of competence proteins, particularly those associated with type IV pili (Wendt and Pakrasi 2019).

Natural uptake of exogenous DNA requires the competence proteins ComEA, ComEC, and ComF. Subsequent integration of the acquired DNA into the recipient genome requires proteins that bind to single-stranded DNA, such as the DNA processing protein DprA and the DNA recombination and repair protein RecA (Wendt and Pakrasi 2019; Jester et al. 2022). Cyanobacteria that possess genes (termed competence genes) for all these proteins, have been experimentally proven to be naturally transformable. A recent bioinformatic analysis of 345 cyanobacterial genomes revealed that approximately 70% of all genomes possess the genes required for natural competency, including AFA strain 2012/KM1/D3 (Wendt and Pakrasi 2019). Natural competency has also been proven experimentally in Nostoc muscorum and A. platensis (Jester et al. 2022; Nies et al. 2022). Therefore, it is plausible that Aphanizomenon is naturally competent. However, further investigation is needed to confirm this hypothesis.

Genetically engineered strains of AFA could be generated for the enhanced accumulation of high-value products such as C-PC. Alternatively, metabolic pathways could be introduced or modified to synthesise novel metabolites or increase flux, for example increasing the nitrogen fixation rate for biofertiliser purposes. The latter option, for example, has already been demonstrated in Anabaena sp. 7120 by targeting the HetR gene responsible for heterocyst differentiation (Chaurasia and Apte 2011). The engineered line showed a heterocyst frequency of 12-18% in cells of Anabaena filaments compared to the 6-8% in the wild-type strain, resulting in a 1.5 to 2-fold higher nitrogenase activity. This was obtained via insertion into the genome of an additional copy of the hetR gene driven by the high-expression light-inducible promoter from psbA1 (Chaurasia and Apte 2011). Importantly, Nostoc and Anabaena cultures are already known to fix a high amount of nitrogen, 20 to 30 kg N ha^-1 season^-1 (Chittora et al. 2020). It is therefore possible that similar AFA strains over-expressing hetR could potentially fix up to 60 kg N ha^-1 season^-1whilst being a natural, safe and sustainable option (Chaurasia and Apte 2011). Similarly, the introduction of another copy of the operon encoding C-PC, or genes for one of its precursors such as the haem oxygenase or uroporphyrinogen III gene could lead to increased cellular accumulation of C-PC, thereby increasing the biotechnological potential of such a strain (Manir AFAsha et al., 2018). Recent cyanobacterial research has also highlighted the potential of fine-tuning the synthesis of pseudocobalamin to produce bioactive vitamin B₁₂. More specifically, cyanobacteria lack the BluB enzyme required to make DMB from riboflavin and the CobT, CobC and CobS enzymes required for inclusion of DMB into the corrin ring (Helliwell et al. 2016). The expression of these four bacterial genes within cyanobacteria could offer a route to low-cost, light-driven production of B₁₂. This has recently been attempted in A. platensis and warrants an investigation as to whether this could be achieved in AFA as well (Morente 2021). Finally, AFA’s edible status offers the possibility of bespoke AFA strains being used for the oral delivery of high-value bioactive compounds and therapeutic proteins such as vaccines within the aquaculture and animal feed industry (Yan et al. 2016).

Conclusion

The commercial and biotechnological potential of the nitrogen-fixing filamentous cyanobacterium Aphanizomenon flos-aquae is largely untapped. Its comprehensive nutritional profile and high-value products such as C-PC and PEA, make it a promising candidate for markets beyond the nutritional supplement industry, including the animal feed, aquaculture and biofertiliser markets. Currently, AFA harvesting is limited to the natural occurrence of blooms in Klamath Lake, and this brings with it issues of seasonal fluctuations and the risk of contamination with microcystin toxins. The integration of artificial cultivation systems such as open ponds and closed photobioreactor systems with natural harvesting could help maximise and stabilise production. The blending of microcystin-free cultivated AFA and wild-harvested AFA biomass would increase the number of Klamath AFA harvestable days per annum and help resolve the current supply predicament. Furthermore, closed-system AFA cultivation would yield AFA biomass enriched for specific products, thereby facilitating the expansion into novel high-value industries, such as the dyes and colourants, sun cream, fish oil, and pharmaceutical sectors. Genetic manipulation of AFA could generate bespoke strains with increased commercial potential that could further expand its market. Nevertheless, AFA is a poorly studied cyanobacterium, and its controlled cultivation and genetic engineering potential remains largely unexplored. It therefore remains a possibility that AFA strains are unable to be domesticated, grown at scale or genetically manipulated. Therefore, it is crucial to undertake further fundamental research into growth parameters and genetic manipulation prior to attempting to cultivate this cyanobacterium at industrial scale.