Tetrodotoxin (TTX) is a non-protein, low molecular weight neurotoxin. It is highly potent when ingested, binding to sodium channels and interfering with muscle and nerve function [1]. TTX was long believed to be present only in pufferfish of the family Tetraodontidae, however, since its detection in the eggs of the California newt (Taricha torosa) in 1964 [2], TTX has been detected in a wide range of phylogenetically unrelated terrestrial and aquatic taxa including; one dinoflagellete, red calcareous algae, arthropods, echinoderms, molluscs, worms, newts and frogs [3]. Despite many decades of research there is uncertainty surrounding the biological origin of TTX, and its ecological function in TTX-containing organisms.

Many researchers have postulated that bacteria are responsible for producing TTX and multiple TTX-producing bacteria have been isolated from marine organisms [4,5]. However, the TTX concentrations produced by these bacteria are too low to account for the levels found in toxic organisms [3]. This, in concert with a lack of controls, non-specific techniques and cross-reactivity problems with the various methods used to detect TTX have introduced doubt around TTX’s microbial origins [6,7,8]. Evidence supporting endogenous production of TTX has been derived predominately from terrestrial taxa [9,10,11,12]. In captivity, TTX levels in the terrestrial newt T. granulosa increase with a TTX-free diet, and when induced to release TTX from their skin by electrical stimulation, regeneration of TTX occurs within nine months [11]. Additionally, Lehman et al., [13] found no evidence of bacterial symbionts (mtDNA signatures) in TTX-laden organs of T. granulosa.

Regardless of its origin, there are likely physiological costs to the TTX-conatining organisms associated with harbouring, accumulating and maintaining TTX and/or its producers, and thus any benefits must outweigh these energetic expenditures. Research suggests multiple ecological roles for TTX, which differ among organisms. These include: a defensive agent, e.g., the eggs and skin of many TTX-containing organisms have high levels of the toxin [12], an offensive weapon e.g., venom [14], and within- and between-species communication about the location of mates and potential food sources [15,16]. Among well-studied TTX-containing species both toxic and non-toxic populations have been identified (e.g., puffer fish and newts [17,18]), confounding these hypotheses somewhat. In T. granulosa marked differences in TTX between populations have been explained through chemically mediated co-evolution with its prey, the common garter snake–Thamnophis sirtalis [19]. Despite TTX having been identified in at least six phyla of organisms [5] T. granulosa are the only animals that contain TTX for which a resistant predator is known [19].

In 2009, an investigation into a series of dog poisonings on beaches in Auckland (North Island, New Zealand), led to the identification of TTX in Pleurobranchaea maculata (grey side-gilled sea slug). Pleurobranchaea maculata are highly opportunistic carnivores that scavenge on a range of invertebrate organisms (sea anemones, marine worms, other molluscs [20]). They are hermaphrodites and are found in shallow subtidal areas around New Zealand, and have also been recorded in southeastern Australia, China, Sri Lanka and Japan [21]. There are several features of P. maculata that make them a very amenable marine organism for investigating the origin and ecological function of TTX. They are found in relatively confined, easily accessible, shallow sub-tidal areas, as opposed to many other TTX-containing marine organisms, and they can be reared in captivity under very controlled conditions [21].

The aims of this study were to investigate temporal, spatial, individual and generational variability in TTX concentrations from New Zealand populations of P. maculata. Samples were collected from ten populations around New Zealand (Figure 1) and their TTX levels assessed using liquid chromatography-mass spectrometry (LC-MS). Pleurobranchaea maculata were also collected monthly over six months at Narrow Neck Beach (Auckland, New Zealand), to study temporal variability in TTX concentrations. To ensure differences in TTX (see results) were not due to the presence of cryptic species between North and South islands, we extracted DNA and sequenced a region of the cytochrome oxidase subunit 1 (CO1). Generational variability was examined by rearing P. maculata from eggs to maturation in a TTX-free environment. These data may provide insights into the origins and ecological function of TTX in P. maculata.

Significant variability in TTX concentrations occurred between P. maculata populations. No TTX was detected in the P. maculata population from Tasman Bay, Nelson and an individual from Kaikoura, while from Wellington to Whangarei populations showed a wide range of TTX levels. Historically there has been confusion as to whether there are multiple Pleurobranchaea species in New Zealand [20]. Additionally in this study morphological differences, e.g., mantel pigmentation, were observed between the Tasman Bay and other populations. To eliminate the possibility that the Tasman Bay population was a different species from the North Island population’s analysis of a segment of the CO1 gene was undertaken. This showed <1% sequence variability between the Tasman Bay and Narrow Neck populations providing evidence that these populations are the same species. Variability in TTX levels across populations has been documented elsewhere. For example, populations of the newt T. granulosa from north western USA were found to have highly variable levels of TTX, ranging from 0.007 (±0.004) mg g⁻¹ to 1.752 (±0.110) mg g⁻¹, while those collected further north around Vancouver Island, Canada, did not contain TTX [18]. In T. granulosa, variability in TTX levels is attributed to a coevolutionary interaction with the common garter snake T. sirtalis. In an apparent coevolutionary arms race, T. sirtalis is believed to have evolved TTX resistance in parallel with the toxicity level of T. granulosa [19]. Such a tight predator-prey relationship is an unlikely explanation for variability among P. maculata populations. In our studies we observed TTX-containing P. maculata in a diverse range of habitats ranging from sandy sediment to rocky reefs, where they would be exposed to predation from multiple marine organisms including fish and octopus. Geographic variability among marine TTX-containing organisms occurs in multiple species including: puffer fish [22,23], the horse-shoe crab Carcinoscorpius rotundicauda [24,25,26] and the starfish Astropecten polyacanthus [27]. Although definitive reasons for these variations are unknown the most likely explanation is that these organisms are exposed to TTX through the food chain [28] and that the abundance of this dietary source influences the presence and levels of TTX within an organism. Noguchi et al., [28] provided evidence to support this hypothesis by culturing over 5000 specimens of the puffer fish Takifugu rubripes in net-cages and land-based aquaria. These organisms were fed TTX-free diets for between 1–3 years and all livers and other tissues tested negative for TTX. A dietary source of TTX is a plausible explanation for the variation in TTX levels in P. maculata, however in-depth benthic surveys of the flora and fauna at sites with high P. maculata abundance have only detected trace levels of TTX in a few organisms (e.g., the Sand Dollar Arachnoides zelandiae and the crab Macrophthalmus hirtipes), all of which are unlikely to be an abundant food source for P. maculata [29]. It is possible that if the dietary source was microbial it may not have been detected in these surveys and attempts are currently underway to isolate TTX-producing bacteria from P. maculata. Recently latitudinal gradients in marine planktonic bacteria have been demonstrated [30], and it is possible that latitudinal and current-induced gradients may result in differences in bacterial diversity within P. maculata populations around New Zealand. While the distances between Kaikoura and Tasman Bay (non-toxic populations) and the Wellington site (toxic individual) are only 175–215 km respectively, there is limited opportunity for oceanic exchange between theses South and North Island sites, with the prevailing oceanic currents flowing east from Tasman Bay and north from Kaikoura [31,32].

In addition to between-population variability there were also significant difference in TTX concentration within toxic P. maculata populations, with the most pronounced difference (>60 fold) occurring among individuals collected from Illiomama Rock. Variability in TTX-concentrations among populations has been measured in many TTX-containing organisms. In the most extreme examples both non-toxic and highly toxic individuals are found in a single population e.g., the gastropods Rapana rapiformis and R. venosa venosa [33] and the horseshoe crab Carcinoscorpius rotundicauda [26]. In other species populations tend to be either toxic or non-toxic, but within the TTX-positive populations there is a marked range in toxin levels e.g., T. granulosa [34,35]. We considered the possibility that the concentration of TTX may be related to the size of individual P. maculata and this may partly explain the variability. If TTX is produced endogenously it would be logical to expect larger P. maculata capable of producing more TTX and thus higher concentrations would be associated with heavier individuals. However, no correlation between mass and TTX concentrations was observed (Figure 2). Similar patterns have been noted in other TTX-containing organisms. For example, there was no correlation between toxicity and body mass in specimens collected at the same place and time of the starfishes Astropecten polyacanthus and A. scoparius [27] or the ribbon worm Cephalothrix sp. [36]. Genetic variation among individual may also affect their ability to produce or sequester TTX and thus could possibly explain the lack of relationship between mass and TTX concentrations.

Over seven months the population of P. maculata at Narrow Neck Beach (Auckland, New Zealand) showed a trend of decreasing levels of TTX from an initial peak in June-July to a low in December. In other TTX-containing organisms there are conflicting data regarding seasonal fluctuations. For example, although toxicity varied among samples there were no clear seasonal trends in studies of two gastropods; the lined-moon shell, Natica lineata [37] and Nassarius glans [38]. In contrast, ribbon worms (Cephalothrix sp.) collected from November 1998 to March 1999 showed mean toxicities of 2101–3278 MU g⁻¹. These decreased in April and March, 1999 (995 MU g⁻¹), before increasing sharply to (5843 MU g⁻¹) in June [36]. Likewise in their studies of the starfish Astropecten polyacanthus, Miyazawa et al. [39] found that all specimens collected in autumn and winter werenon-toxic. Recent studies on puffer fish have associated seasonal changes in TTX concentrations with different sexual maturity stages. Sabrah et al. [40] studied different sexual maturity stages in the puffer fish (Lagocephalus sceleratus) and found highest toxicity levels during the spawning stage of maturity. Most of the specimens in maturity stage II (developing stage) were non-toxic. A similar pattern was observed in female Takifugu poecilonotus where liver toxicity was highest in the “ordinary period” (just after spawning) and ovary toxicity was highest during the “maturation period” [41]. Although little is known about the ecology of P. maculata, they were most abundant and significant quantities of egg masses were observed during our surveys in June to August [29]. This was also the period where the highest TTX-levels were measured suggesting a link between maturation stage and TTX levels in P. maculata. Our previous research and ongoing studies have detected high levels of TTX in P. maculata eggs and larvae, and demonstrate TTX levels decrease in consecutive batches of eggs [42,43]. Thus, P. maculata may invest their offspring with TTX or inoculate them with TTX-producing bacteria as a defense mechanism.

To investigate whether TTX-producing bacteria are transmitted vertically between generations we attempted to rear P. maculata in captivity. The two sub-samples of planktonic larvae tested positive for TTX. Despite more than 20 pediveligers settling and more than 10 metamorphosing into juveniles, only one survived to full maturity and no TTX was detected. Unfortunately with such a limited sample size we cannot draw any definitive conclusions from this experiment. Similar experiments with captive-raised pufferfish [17] and riparian frogs (Atelopus variusdo) [44] have also found that second generation cohorts do not possess TTX. In both of these examples the researchers suggest that the lack of TTX in these captivity-reared organisms indicates that TTX may either have a dietary origin or may be dependent on environmental factors, such as microbial symbionts. Gall et al. [45] detected TTX in all developmental stages of T. granulosa, however, TTX concentrations declined steadily through metamorphosis. In contrast, TTX levels in the blue-ringed octopuses (Hapalochlaena spp.) continue to increase during development, suggesting that offspring or their symbionts begin producing TTX independently [46]. Further research tracking TTX levels through all development stages of P. maculata is required to understand it role and origin in each life stage.

In summary, our initial studies on TTX in P. maculata show that this organism conforms to the trends observed in other marine TTX-containing organisms. There is significant variability in TTX concentrations within and among populations. TTX concentrations were highest during the egg laying season (June–August) and this, in concert with the detection of high levels of TTX in eggs and early larval stages [42], suggests that TTX may be used for defensive purposes in P. maculata. Confounding this hypothesis is the existence of populations that do not contain TTX. It is possible that these populations contain alternative chemical defense mechanisms but this requires further investigation. Our present results suggest that the occurrence of TTX in P. maculata may involve symbiotic TTX-producing bacteria, though the possibility that TTX is accumulated through the food chain or produced endogenously cannot be fully excluded.