Production

Producing Cyanobacteria

Anatoxin was first found in Dolichospermum (Anabaena) flosaquae strains originating from Canada (Carmichael et al., 1975; Devlin et ah, 1977) and later in Finland in Anabaena mendotae (Rapala et ah, 1993), and D. circinale and Anabaena sp. in Finland and Japan (Sivonen et ah, 1989; Park et ah, 1993). Since then, many papers have been published reporting its production by several cyanobacteria species in many geographic areas by a variety of cyanobacteria taxa belonging to Nostocales - that is, Cbrysosporum (Apbanizomenon) ovalisporum, Cuspidotbrix, Rapbidiopsis (Cylindrospermopsis), Cylindrospermum, Dolichospermum (Anabaena) circinale, D. flosaquae and D. lemmermannii - and to Oscil- latoriales, that is, Blennothrix, Kamptonema, Microcoleus, Oscillatoria, Planktotbrix, Phormidium and Tycbonema (for species names and taxonomic changes, see Chapter 3). Tables 2.6 and 2.7 give examples of ATX contents in strains and concentrations in environmental samples, respectively. For further details, see reviews by Funari and Testai (2008), Pearson et ah (2016), Testai et ah (2016) and Cires and Ballot (2016).

The production of ATX is species- and strain-specific. It is of interest that the American and European isolates of D. circinale investigated so far produce only ATX, while the Australian isolates exclusively produce saxitoxins, even if the two strains are reported to form a phylogenetically coherent group (Beltran & Neilan, 2000).

Homoanatoxin-a was first isolated from a Kamptonema (Oscillatoria) formosum strain in Ireland (Skulberg et ah, 1992). Subsequently, it was found to be produced by Rapbidiopsis mediterranea in Japan and Oscillatoria in Norway, isolated from Microcoleus (Phormidium) autumnalis in New Zealand and from species of Dolichospermum/Anabaena in Ireland (see Testai et ah, 2016).

Toxin Profiles

Anatoxin has been found to be produced alone by Microcoleus (Phormidium) cf. autumnalis (James et ah, 1997) as well as coproduced with HTX in Rapbidiopsis mediterranea (Watanabe et ah, 2003), Oscillatoria (Araoz et ah, 2005), and with microcystins in Arthrospira fusiformis (Ballot et ah, 2005), Microcystis sp. (Park et ah, 1993) and Dolichospermum/Anabaena spp. (Fristachi & Sinclair, 2008). M. autumnalis can contain high contents of F1TX (together with ATX), showing large differences in toxin contents from week to week, and in some cases also in the same day (Wood et ah, 2012). Non-axenic M. autumnalis strain CAWBG557 produces ATX, HTX and their dihydrogen derivatives dihydroanatoxin-a (dhATX) and dihydrohomoanatoxin-a (dhHTX; Heath et ah, 2014). Dihydro-anatoxin-a

Table 2.6 Neurotoxin contents reported from laboratory cultures of cyanobacteria

Table 2.6 (Continued) Neurotoxin contents reported from laboratory cultures of cyanobacteria

Numbers following taxa indicate the number of tested strains if more than a single strain was analysed. The taxonomic classification is listed according to the current nomenclature with earlier synonyms given in parentheses (for an overview on recent changes in taxonomy, see Chapter 3).

n.q.: not quantified, only qualitative detection reported.

¹ The taxon given here may deviate from that given in the publication. For changes in taxonomy, see Chapter 3.

^b If not specified otherwise.

^c Several strains of Aph. flosaquae have been reclassified as Aphanizomenon sp. or Aph. gracile, respectively.

^d « Estimated from figure in publication.

Table 2.1 Neurotoxin contents of biomass and concentrations in water reported from environmental samples

Table 2.7 (Continued) Neurotoxin contents of biomass and concentrations in water reported from environmental samples

Contents are given in |jg toxin per gram dry weight (dw) or wet weight (ww). For individual studies, maximum values are given. Samples were collected in different types of waterbodies (L: lakes, R: rivers, P: pond, Res: reservoirs) in countries as indicated. For saxitoxins, contents are reported as saxitoxin equivalents (STXeq) in some reports or as individual variants (see text). The taxonomic classification is listed according to the current nomenclature with earlier synonyms given in parentheses (for an overview on recent changes in taxonomy, see Chapter 3).

¹ The taxon given here may deviate from that given in the publication. For changes in taxonomy, see Chapter 3.

has been reported to be produced in amounts much higher than those of ATX by strains of M. autumnalis (Wood et al., 2017; Puddick et ah, 2021)

The few data available on ATX cell quota range from 90 fg/cell in Cuspidothrix issatscbenkoi (Selwood et ah, 2007) to 500 fg/cell in M. autumnalis (Heath et ah, 2014). Cell quota detected in Tychonema bour- rellyi were in a similarly wide range, 10-350 fg/cell (Shams et ah, 2015).

The highest contents within the wide variability of ATX contents reported from strains grown as laboratory cultures, in the order of a few mg/g dw, were found in strains of the genera Oscillatoria, Pbormidium, Apbanizomenon, Cuspidothrix and Dolichospermum. The maximum value (13 mg/g dw) was found in D. flosaquae and Oscillatoria sp., while much lower contents - generally by 1-2 orders of magnitude - of ATX are reported for cyanobacteria of other genera (Testai et ah, 2016).

Biosynthesis and Regulation

Cyanobacteria produce (+)ATX, but no specific studies have addressed the stereoselectivity of the biochemical reaction towards the positive enantiomer.

Anatoxin biosynthesis and regulation have been reviewed in Pearson et al. (2016). Mejean et al. (2009) reported the identification of the first gene cluster coding for the biosynthesis of ATXs (ana) within the sequenced genome of Oscillatoria sp. PCC 6506, producing mainly HTX. In the following years, five other ana clusters were identified within Dolichospermum/ Anabaena sp. 37, Oscillatoria sp. PCC 6407, Cylindrospermum stagnate sp. PCC 7417, Cuspidotbrix issatschenkoi RM-6, C. issatscbenkoi LBRI48 and C. issatschenkoi CHABD3 (Rantala-Ylinen et al., 2011; Shih et al., 2013; Mejean et al., 2014; Jiang et al., 2015).

Each cluster showed general similarities in the protein functions, with a high percentage of identity in nucleotide sequence (with the core genes anaB-G being conserved within all strains), but differences in the organisation of genes (Pearson et al., 2016), leading to different toxin profiles between the producing organisms.

The biosynthesis of the ATXs involves a polyketide synthase (PKS) family of multifunctional enzymes with a modular structural organisation as described in Mejean et al. (2014). A detailed biochemical description of the adenylation domain protein AnaC revealed the activation of proline as starter, and not glutamate as previously proposed (Dittmann et al., 2013). The biosynthetic pathway describes AnaB, AnaC and AnaD as acting in the first steps (which have been fully reproduced in vitro; Mejean et al., 2009; Mejean et al., 2010; Mann et al., 2011), and AnaE, AnaF, Ana J and AnaG catalysing the following steps, with the latter adding two carbons and methylating the substrate to produce HTX. The release of ATXs may be catalysed by the thioesterase AnaA, although this has not been experimentally verified (Pearson et al., 2016) or a spontaneous decarboxylation step may occur to yield the amine alkaloid ATX (Dittmann et al., 2013).

The molecular regulation of ATX has not been sufficiently studied so far. Under conditions where anaA, anaj, anaF and anaG transcripts were present in C. issatschenkoi CHABD3, no ATX was detected (Jiang et al., 2015). This result may indicate that the regulation of ATX occurs at the post-transcriptional level, but interpretation is limited by the lack of investigation of ATX dihydroderivatives production (Pearson et al., 2016).

An influence of light, temperature, phosphorous and nitrogen on cellular ATX content is reported, and it seems that the influence of environmental factors is strain-specific (Harland et al., 2013; Neilan et al., 2013; Boopathi & Ki, 2014; Heath et al., 2014). Overall, the influence of factors, such as light and temperature, reported for the ATX content in Dolichospermum/Anabaena and Aphanizomenon cultures varies around 2-4-fold, not exceeding a factor of 7 (Rapala &c Sivonen, 1998), and a similar range is reported for HTX in relation to phosphorus (Heath et ah, 2014). HTX production also seems to be linked to the culture growth phase in Raph. mediterranea strain LBRI 48 (Namikoshi et ah, 2004). However, the results of most studies were not strongly supported by statistical analyses; furthermore, determining the effect of nutrient limitation requires continuous culture systems or evaluating batch culture data in relation to growth rates, yet in few studies this was done.