COMMENTARY Chemical or nematocyst-based defence in the peregrina? – a reply to B.K. Penney

Arnaldo Marin Departamento de Ecologı´a e Hidrologı´a, Facultad de Biologı´a, Universidad de Murcia, 30100 Murcia, Spain

Aguado & Marin (2007) analysed the interaction between (Karuso, 1987; Rogers & Paul, 1991). Thus, metabolite Cratena peregrina (Gmelin, 1791) and predatory fish in labora- accumulations within opisthobranchs may represent a mechan- tory and field assays, using both live aeolids and artificial ism to store toxic chemicals with impunity for subsequent elim- models. Penney’s comment (2009) on the article reopens the ination, and not necessarily to defend the organism (Pennings old question of the role of nematocyst-based defence in nudi- & Paul, 1993). Downloaded from branchs. Penney suggests that the defensive mechanism of Penney suggests that because E. carneum from North C. peregrina is chemical because the nematocysts used in American possesses chemical extracts that deter feeding by assays could be unarmed. Penney argues that (1) Aguado & pinfish, extracts from employed with models of C. Marin (2007) obtained nematocysts by macerating Eudendrium peregrina could be unpalatable because they contain chemicals hydroids with a mortar and pestle, and this cannot be con- and not active nematocysts. Stachowicz & Lindquist (2000) sidered equivalent to kleptocnidae isolated from C. peregrina; indicated that crude extracts from Eudendrium deterred feeding, http://mollus.oxfordjournals.org/ (2) the method by which the authors attempted to incorporate ‘demonstrating’ that defensive chemistry plays a significant role nematocysts into the test food is unclear; and (3) regardless of in the unpalatability of these . To test their hypothesis how nematocysts were added to the artificial food models, it is that unpalatable secondary metabolites represent alternative possible that no functional nematocysts would remain by the antipredator defences for hydroids, they compared the palat- time fish encountered them. ability of intact polyps with those in which the nematocysts The main purpose of our article was to demonstrate that fish had been chemically stimulated to discharge. Hydroid polyps learn to avoid the warning coloration pattern of C. peregrina, defended solely by nematocysts should become palatable when due to associating bad taste or unpleasant experience with the their nematocysts are discharged before offering them to preda- colour pattern. Nematocysts are contained in the defensive tors, whereas the palatability of chemically defended hydroids should not change after treatment. However, neither the lipo- exudates and in the external part of the body of C. peregrina, at Western Washington University on March 3, 2014 which can only be regarded as circumstantial evidence that philic (DCM þ butanol soluble compounds) nor the water- these defensive cells or the molecules that they contain rep- soluble fractions of Eudendrium deterred feeding. These authors resent a deterrent for predators. Of course, the article does not suggested that there is an additive or synergistic effect among provide experimental evidences that the source of deterrence is compounds in the two fractions, or that the deterrent com- the nematocysts, but there is no proof, either, that chemical pounds decomposed during the partitioning of the crude defences play an active role. If we support the general idea extract. In my opinion these results do not demonstrated that that a positive result for deterrence alone cannot be taken as defensive chemistry plays a significant role in the unpalatabil- evidence for a nematocyst-based defence, the same argument ity because toxins contained in nematocysts could also act as a must apply to chemical-based defence. We must classify chemical deterrence. If there truly are active chemicals in nematocyst-based defences as a special form of chemical Eudendrium, the lipophilic and the water-soluble fractions must defences because nematocysts contain harmful molecules. From be clearly deterrent to fish, but this is not the case. Secondary this point of view, C. peregrina transfers chemicals from food to metabolites contained in the discharged nematocysts could be the cnidosacs at the tip of the ceras. responsible for fish rejection. There is no evidence that C. peregrina contains secondary The biosynthesis of defensive metabolites is costly for organ- metabolites other than the nematocyst toxins. Chemical analy- isms. It is not surprising that species showing de novo biosyn- sis supports the general hypothesis that the nematocysts are the thesis of defensive metabolites use these metabolites for other main weapons. Cimino et al. (1980) studied three species of the purposes. This is the case in the nudibranch fimbria, hydroid Eudendrium (E. rameum, E. racemosum, E. ramosum) in the which synthesizes prostaglandin derivates (Marı´n, Di Marzo & Bay of Naples (Italy), which are prey of C. peregrina. These Cimino, 1991). The structural variety of the lactones and the authors found the same pathway of polyhydroxylated steroids data on their distribution in the body of T. fimbria suggest a (Cholest-4-en-4, 16b, 18, 22R-tetrol-3-one 16,18-diacetate) in range of different biological functions: to contract smooth the hydroids and in the nudibranchs. Unfortunately these muscle fibres, defence allomones of the ceratal secretion, and in steroids do not play a defensive role in hydroids or nudi- basic physiological functions (e.g. ion regulation, renal function branchs. Some aeolids contain pigments and secondary meta- and reproductive biology). These arguments do not mean that bolites from their prey, which probably have no chemical other aeolidoidean nudibranchs should also contain defensive defence role either (e.g. Goodwin & Fox, 1955; McBeth, 1972: metabolites. The aeolid nudibranch Phyllodesmium guamensis the compounds are ubiquitous and conserved amongst numer- feeds on the soft corals Sinularia maxima and S. polydactyla. ous non-deterrent species; Cimino et al., 1980: the compounds Phyllodesmium guamensis sequesters a diet-derived diterpene are located within internalized structures and thus not quickly 11b-acetoxypukalide selectively within various body parts. encountered by predators). Sequestered compounds can occur Levels were highest in the , with moderate and low to at concentrations lower than threshold feeding deterrent levels non-existent levels in the and viscera, respectively. Laboratory tests showed that 11b-acetoxypukalide deters feeding by the puffer fish Canthigaster solandri at a given concen- Correspondence: A. Marin; e-mail: [email protected] tration (0.5% dry mass), while feeding field experiments with

Journal of Molluscan Studies (2009) 75: 201–202 # The Author 2009. Published by Oxford University Press on behalf of The Malacological Society of London, all rights reserved COMMENTARY nudibranch tissues exhibited species-specific reactions to the comparing past studies of isolated nematocyst discharge nudibranch tissues and their isolated metabolite, which ranged (Greenwood et al., 2002). from non-deterrence to emesis (Slattery et al., 1998). Penney also suggests that nematocysts obtained by macerating Even if active deterrent metabolites are present in C. peregrina, Eudendrium cannot be considered equivalent to kleptocnidae iso- it is very unlikely that chemical defence is the sole defensive lated from C. peregrina. However, Martin (2003) in an exhaustive mechanism, because retaining functional nematocysts must be study of C. peregrina found, when feeding on polyps of the hydro- energetically expensive. Several studies in opisthobranchs have zoan Eudendrium, that phagocytosis of nematocysts and food par- indicated that chemical and physical defences commonly ticles by digestive cells is rather non-selective. In the course of co-occur and can function either additively, or synergistically, over two years of this study there was no evidence of selective to reduce susceptibility to predators. In the nudibranch exclusion of specific nematocyst types by cnidophagous cells. verrucosa, several defensive strategies occur at the same time. This nudibranch is perfectly mimetized in its habitat and its mantle is protected by spicules and by the presence of two toxic REFERENCES molecules, verrucosin-a and -b (Cimino et al., 1988). Many AGUADO, F. & MARIN, A. 2007. Warning coloration associated algae and invertebrates display redundant defence mechanisms with nematocyst-based defences in aeolidiodean nudibranchs. against (Hay et al., 1994; Schupp & Paul, 1994). Journal of Molluscan Studies, 73: 23–28. Nudibranchs utilize both physical (spicules, tunic toughness)

CIMINO, G., DE ROSA, S., DE STEFANO, S. & SODANO, G. Downloaded from and chemical (secondary metabolites, acidity) defences and 1980. Cholest-4-en-4, 16b, 18, 22R-tetrol-3-one 16, 18-diacetate a suffer relatively little predation by generalist predators. novel polyhydroxylated steroid from the hydroid Eudendrium sp. The method by which we incorporated nematocysts into the Tetrahedron Letters, 21: 3303–3304. test food was clearly described: ‘the artificial models were CIMINO, G., GAVAGNIN, M., SODANO, G., PULITI, R., bathed in hydroid sauce’ (Aguado & Marin, 2007). We used MATTIA, C.A. & MAZZARELLA, L. 1988. Verrucosin-a and -b, the word impregnation incorrectly because it suggests that the ichthyotoxic diterpenoic acid glycerides with a new carbon skeleton hydroid sauce entered and spread completely through the arti- from the dorid nudibranch Doris verrucosa. Tetrahedron 44: http://mollus.oxfordjournals.org/ ficial food. The Eudendrium sauce was applied as a coat to the 2301–2310. outside of models, because if nematocysts were mixed with the GOODWIN, T.W. & FOX, D.L. 1955. Some unusual carotenoids from artificial food before solidification the high temperature could two nudibranch slugs and a lamprid fish. Nature, 175: 1086–1087. have degraded them. This method was chosen because nema- GREENWOOD, P.G. & MARISCAL, R.N. 1984. The utilization of tocysts are not released into the cnidosac of the aeolid, but into cnidarian nematocysts by aeolids nudibranchs: nematocysts large epithelial cells called cnidophages. The release of nema- maintenance and release in Spurilla. Tissue and Cell 15: 719–730. tocysts from aeolids is caused by a contraction of the muscle GREENWOOD, P.G., BALBONI, I.M. & LOHMANN, C. 2002. A complex surrounding the cnidosac, and entire cnidophages are sea anemone’s environment affects discharge of its isolated nematocysts. Comparative Biochemistry and Physiology, 134A: 275–281. released (Greenwood & Mariscal, 1984). These cnidophages HAY, M.E., KAPPEL, Q.E. & FENICAL, W. 1994. Synergisms in are forced towards the ceras tip where they are pushed into a at Western Washington University on March 3, 2014 plant defense against herbivores: interactions of chemistry, cellular matrix and through the epithelium, before being calcification, and plant quality. Ecology, 75: 1714–1721. released from the ceras. The application of isolated nematocysts KARUSO, P. 1987. Chemical ecology of the nudibranchs. In: to artificial food might be considered inappropriate, because it Bioorganic marine chemistry I (P.J. Scheuer, ed.), pp. 31–60. does not reflect the defensive mechanism of aeolid nudi- Springer-Verlag, New York. branchs. The coating of artificial models with hydroid sauce is MARI´N, A., DI MARZO, V. & CIMINO, G. 1991. A histological more realistic, because it simulates the expulsion of the entire and chemical study of the cerata of the opisthobranch mollusc cnidophage in the aeolids, rather than a bioassay with isolated Tethys fimbria. Marine Biology, 11: 353–358. nematocysts. It is also suggested that any nematocysts that MARTIN, R. 2003. Management of nematocysts in the alimentary might have adhered to the models could be washed off when tract and in cnidosacs of the aeolid nudibranch gastropod Cratena the models came into contact with the water of the aquarium peregrina. Marine Biology, 143: 533–541. or field site and that it is possible that no functional nemato- SLATTERY, M., AVILA, C., STARMERB, J. & PAUL, V. 1998. A cysts would remain by the time fish encountered the models. sequestered soft coral diterpene in the aeolid nudibranch Penney’s criticism is in error, because our personal observation Phyllodesmium guamensis Avila, Ballesteros, Slattery, Starmer and indicates that a high percentage of nematocysts remains Paul. Journal of Experimental Marine Biology and Ecology, 226: 33–49. undischarged in the models. However, just a small percentage McBETH, J.W. 1972. Carotenoids from nudibranchs. Comparative of active nematocysts might be sufficient to deter fish in the Biochemistry and Physiology, 41B, 55–68. bioassays with artificial models. In addition, Martin (2003) in PENNEY, B.K. 2009. A comment on F. Aguado & A. Marin: ‘Warning a study of the management of nematocysts in C. peregrina found coloration associated with nematocyst-based defences in aeolidoidean that large proportions of these nematocysts were undischarged nudibranchs’. Journal of Molluscan Studies, 75: 199–200. in the alimentary tracts of the snails. This author even PENNINGS, S.C. & PAUL, V.J. 1993. Sequestration of dietary describes that the faeces from snails contained large numbers secondary metabolites by three species of sea hares: location, of undischarged nematocysts that appeared structurally intact. specificity, and dynamics. Marine Biology, 117: 535–546. Nematocysts were isolated from the sea anemone Calliactis ROGERS, S.D. & PAUL, V.J. 1991. Chemical defenses of three tricolor maintained under different feeding schedules or in Glossodoris nudibranchs and their dietary Hyrtios sponges. Marine different degrees of salinity, in an attempt to determine how Ecology Progress Series, 77: 221–232. these culture conditions influence the discharge of isolated STACHOWICZ, J.J. & LINDQUIST, N. 2000. Hydroid defenses nematocysts. The nematocysts isolated from individuals from against predators: the importance of secondary metabolites versus nematocysts. Oecologia, 124: 280–288. two separate populations of C. tricolor responded significantly differently to 5 mM EGTA and to de-ionized water, and these SCHUPP, P.J. & PAUL, V.J. 1994. Calcium carbonate and secondary metabolites in tropical seaweeds: variable effects on herbivorous responses also depended upon the isolation solution used. This fishes. Ecology, 75: 1172–1185. study provides evidence that the environment or feeding state of an anemone affects the discharge of isolated nematocysts. Inherent differences in ionic and osmotic characteristics among doi:10.1093/mollus/eyp007 nematocysts could explain some of the ambiguities when Advance Access Publication: 20 March 2009

202