Unspecified Toxicity and Other Cyanobacterial Metabolites

Andrew Humpage and Martin Welker

Early studies on toxic cyanobacteria largely reported effects of extracts of cyanobacteria, isolated strains or bloom material collected in the field on test systems such as animals and plants. With the purification of individual compounds that cause toxic effects and the elucidation of their structure, whole organisms were often replaced as test systems with cell lines, tissues or enzyme/substrate systems. The mode of action of a number of toxins could be revealed by these tests and eventually lead to a good understanding of the human health risks associated with these toxins.

However, in a number of studies, toxic effects on whole animals or in vitro test systems were found that could not be explained by the activity of known and quantifiable cyanobacterial toxins. It is therefore likely that cyanobacteria produce metabolites toxic to humans - as well as animals and plants in general - other than the ones described in sections 2.1-2.6.

This section therefore addresses two aspects of cyanobacterial toxicity beyond the known toxins: compounds produced by cyanobacteria that have shown bioactivity in various test systems and toxic effects of cyanobacterial extracts that cannot be attributed to the well-known compounds. Both aspects are tightly linked and may lead to the identification of further cya- notoxins in future.

Bioactive Metabolites Produced by Cyanobacteria

The cyanotoxins described in sections 2.1-2.6 are only a tiny part of the total diversity of secondary metabolites produced by cyanobacteria. Many of these compounds show bioactivity in organismic or in vitro test systems, making cyanobacteria a potentially interesting source of pharmacologically active substances (Burja et al., 2001; Chlipala et ah, 2011; Welker et ah, 2012; Vijayakumar &c Menakha, 2015). It is beyond the scope of this book to review the diversity of cyanobacterial metabolites and their biosynthesis (as far as it is known) and the reader is referred to available reviews (Welker & von Dohren, 2006; Dittmann et ah, 2015; Huang & Zimba, 2019). In this context, it is worth to mention that heterologous expression of peptide or polyketide metabolites in cyanobacterial strains has become feasible (Videau et ah, 2019; Vijay et ah, 2019), potentially offering new opportunities for pharmacological research (Cassier-Chauvat et ah, 2017; Stensjo et ah, 2018).

Most known metabolites, including the known cyanotoxins, are synthesised by three biosynthetic pathways or hybrids thereof: nonribosomal peptide synthetases (NRPS), polyketide synthases (PKS) or ribosomal synthesis of peptides that are modified post-translationally (Ziemert et ah, 2008; Dittmann et ah, 2015). These pathways allow the synthesis of virtually hundreds of structural variants of a single basic structure by variations in amino acid composition, modifications such as methylation or dehydration, and others, as has been well documented for microcystins (Catherine et ah, 2017). Similar variability is known for several classes of nonribosomally synthesised peptides - for example, cyanopeptolins, aeruginosins and ana- baenopeptins (Rounge et ah, 2007; Ishida et ah, 2009) - and ribosomally synthesised peptides such as microviridins (Philmus et ah, 2008). The chemistry and biosynthesis of these peptides and that of the well-known cyanotoxins are very similar, suggesting that their role in contributing to the fitness of the producer organisms is also similar and the high toxicity of some molecules to humans (or mammals in general) is a mere coincidence. The latter point is also supported by the evidence that nonribosomal peptide synthesis by cyanobacteria and in particular, microcystin biosynthesis, is a very ancient trait, dating back to times long before mammals thrived on earth (Christiansen et ah, 2001; Rantala et ah, 2004). The comparison of gene clusters for biosynthetic pathways for peptide or polyketide synthesis, respectively, revealed a pattern of alternating regions with high conservation of variability between species (Cadel-Six et ah, 2008; Ishida et ah, 2009; Dittmann et ah, 2015). This may indicate that some of the metabolite variability arises from horizontal gene transfer and recombination events (Sogge et ah, 2013).

The production of particular metabolites is highly clone-specific, and clones within a population can be described as chemotypes. A high chemotype diversity has been reported for species of Microcystis, Planktothrix, Dolicbospermum (Anabaena) and Lyngbya, for example (Welker et ah, 2007; Rohrlack et ah, 2008; Leikoski et ah, 2010; Engene et ah, 2011; Harustiakova & Welker, 2017; Le Manach et ah, 2019; Tiam et ah, 2019). Since individual cyanobacterial clones can produce multiple variants of multiple classes of metabolites, a multiclonal bloom of cyanobacteria can contain hundreds of bioactive metabolites (Welker et ah, 2006; Rounge et ah, 2010; Agha &C Quesada, 2014). This diversity makes it difficult to relate an observed toxic effect that cannot be explained by the activity of known (and quantifiable) cyanotoxins to a particular compound in a specific sample. Hence, the key challenges for a comprehensive risk assessment of cyanopeptides are their structural diversity, the lack of analytical standards and complex requirements for their identification and quantification (Janssen, 2019).

For a number of individual cyanobacterial metabolites or groups of metabolites, bioactivity data are available. Toxicity to zooplankton (Daphnia) has, for example, been observed for microviridin J (Rohrlack et al., 2004), but no data are available for other organisms or other structural variants. Other frequently occurring peptides, such as cyanopeptolins or anabaenopeptins, have been shown to inhibit proteases of herbivorous zooplankton (Agrawal et al., 2005; Rohrlack et al., 2005; Czarnecki et al., 2006; Schwarzenberger et al., 2010). This indicates that synthesis of these peptides by cyanobacteria may confer a grazing protection for cyanobacterial populations (Savic et al., 2020). However, other compounds isolated from cyanobacteria have been variously described as cytotoxic, immune suppressant or cardioactive, or been shown to inhibit key mammalian enzymes such as acetylcholine esterase, chymotrypsin and trypsin (Humpage, 2008; Nagarajan et al., 2013). Thus, “offtarget” effects also appear to be quite common.

Another hypothesis links the production of diverse (peptide) metabolites to the defence of cyanobacteria against bacteria, phages and parasitic fungi (Gerphagnon et al., 2015). In particular for the latter, evidence has been presented that particular peptides can protect strains of Planktothrix from being infected by Chytridomycota (Sonstebo & Rohrlack, 2011). The protection is apparently specific for the Planktothrix chemotype as well as for the infectious fungal strains (Rohrlack et al., 2013). This could explain the chemotype diversity and their wax and wane in populations of planktonic cyanobacteria with peptide diversity protecting populations from massive parasitic prevalence in a “Red Queen race” (Kyle et al., 2015). Protection from parasite infection may not be the only selective pressure triggering the high metabolic diversity of cyanobacteria, but surely is an interesting field, last but not the least, for the potential discovery of compounds of pharmacological interest, such as antifungal agents (Chlipala et al., 2011; Welker et al., 2012; Vijayakumar & Menakha, 2015).

Although the structure of hundreds of cyanobacterial metabolites is known, the number of compounds not yet known may be equally high or even higher. This could explain the toxic effects of cyanobacterial extracts that are discussed in the following section.

Toxicity of Cyanobacteria beyond Known Cyanotoxins

A number of researchers have reported toxic effects of cyanobacterial extracts that could not be explained by the compounds verifiably present in the extract. In addition, it has been noted that toxic effects of cyanobacteria that have been attributed to known cyanotoxins may actually have been caused by other toxic compounds (reviewed in Humpage (2008), with later examples included in Humpage (2008), Bernard et al. (2011), Froscio et al. (2011), and Humpage et al. (2012). Such unexplained effects include higher-rhan-expected acute toxicity in animal bioassays, effects on particular tissues or cell lines that are not observed using known cyanotoxins, and toxic effects which are not in agreement with established mechanisms attributed to known cyanotoxins (Falconer, 2007). For details of toxicity testing and possible pitfalls, see section 14.3.

To further complicate matters, many harmful effects described in human exposure events such as pneumonia and gastrointestinal symptoms, are not easily or solely explainable based on the described effects of cyanotoxins (Stewart et al., 2006). A cyanobacterial bloom provides an ideal habitat for concomitant growth of dependant bacteria, some of which may be pathogenic to humans (Chaturvedi et al., 2015).

From the observations on animals exposed to blooms in waterbodies or cyanobacterial culture material in laboratories, at the time of the publication of this book it appears likely that with the microcystins, cylindro- spermopsins, neuro- and dermatotoxins described in sections 2.1-2.6, the most potent and most frequently occurring cyanotoxins have been identified and their principle modes of action characterised. If these are absent or their concentrations are below their respective guideline values, major risks to human health from exposure to cyanobacteria therefore seem unlikely. Flowever, the evidence discussed above also implies that any cyanobacterial bloom may contain further, yet unknown substances or microorganisms that may be hazardous to exposed water users. This is a further reason to avoid exposure to high concentrations of cyanobacterial biomass, regardless of its content of known cyanotoxins.

 
Source
< Prev   CONTENTS   Source   Next >