Commentary

Avian : Toxic not of a feather

Paul J. Weldon

Conservation and Research Center, Smithsonian Institution, 1500 Remount Road, Front Royal, VA 22630

n 1992, Dumbacher et al. (1) substan- Itially altered prevailing views of avian physiology, biochemistry, and chemical ecology with their report of the potent homobatrachotoxinin in feathers and other tissues of several spe- cies of New Guinean birds of the . Their discovery was signif- icant not only for suggesting a protective mechanism rarely considered for birds (i.e., chemical defense) but for the nature of the compound they discovered, a struc- turally complex alkaloid that binds Naϩ channels and depolarizes electrogenic membranes. Alkaloids in tetrapods gen- erally had been thought to be confined to amphibians, whose skins have long been acknowledged as arsenals of these biolog- Fig. 1. Hornets (Vespa orientalis) attacking a freshly skinned carcass of a laughing dove (Steptopelia ically active compounds (2). Indeed, be- senegalensis)(Left) while ignoring that of a pied kingfisher (Ceryle rudis). This observation prompted H. B. fore its discovery in Pitohui, homobatra- Cott (4) to undertake an extensive investigation of avian chemical defense. [Reproduced with permission chotoxinin, a member of a family of from ref. 4 (Copyright 1947, The Zoological Society of London).] steroidal alkaloids called batrachotoxinins (BTXs), had been found only in skin se- cretions of Central and South American dichrous), the most toxic of the birds they An additional enigma described by poison-dart frogs (Dendrobatidae) of the examined, is aposematic and may be Dumbacher et al. (3) is the profound genus . In this issue of PNAS, mimicked by the variable pitohui (P. differences observed in the concentra- Dumbacher et al. (3) identify in birds the kirhocephalus) in parts of its range. tions of BTXs among and popu- three major alkaloids that occur in these This year, Dumbacher et al. (3) docu- lations. Indeed, more comparative data on frogs: homobatrachotoxinin, batracho- ment additional toxic species of New abundances among populations may toxinin, and their synthetic precursor, ba- Guinean birds and additional BTXs, some be needed to ascertain just what the spe- trachotoxinin-A. of which are newly described from nature. cies-typical are and at what levels Birds, in fact, had not escaped notice Newly documented among Pitohui are a they occur. Within the genus Pitohui, P. as bearers of defensive chemicals before series of BTXs, including four novel alka- dichrous and P. kirhocephalus contained Dumbacher et al.’s investigations. While loids. All but one of these compounds, the highest amounts of BTXs, although collecting and skinning birds in Egypt batrachotoxinin-A 3Ј-hydroxypentanoate, some individuals from some populations during 1941, the naturalist H. B. Cott (4) occur in the newly featured toxic , the were devoid of toxins. Trace or minor noted that hornets (Vespa orientalis) ar- blue-capped ifrita (Ifrita kowaldi), a mo- amounts of some BTXs were detected in riving on the scene differentially con- notypic species restricted to high montane the crested pitohui (P. cristatus) and the sumed the avian carcasses that he had rainforests (Ͼ1,500 m). (P. nigrescens); however, discarded (Fig. 1). Stimulated by this Ifrita, which as adults are approximately these compounds were not observed in the observation, Cott conducted an exten- 35 g, forages around tree trunks and few samples obtained from either the sive series of experiments on the palat- probes for insects in moss and under (P. ferrugineus) or the white- ability of bird flesh to hornets and do- branches. It overlaps little, if at all, geo- bellied pitohui (P. incertus). Individuals of mesticated cats. In addition, he examined graphically with the much larger (65–100 Ifrita ranged from harboring no detectable the literature and queried field biologists g) species of Pitohui, which are omnivo- to assess the acceptability of various toxins to containing major quantities of all birds for human consumption. The im- rous and occur throughout the vegeta- five of the BTXs that have been detected pression of widespread unpalatability tional column in lower forests down to sea from this species. One of two specimens of that emerged from these studies inspired level. The taxonomic position of Ifrita is the little -thrush (Colluricincla me- Cott to assert that conspicuous problematic, but systematists at least garhyncha, ) possessed might evolve as an aposematic feature agree that it belongs to a family other than trace amounts of only batrachotoxinin-A; signaling the distastefulness of birds to the Pachycephalidae, which contains Pito- several other pachycephalids contained predators. This hypothesis was rekindled hui. Thus, these genera, while sharing no detectable toxins. The widely varying by Dumbacher et al. (1), who suggested highly unique natural products, are not that the distinctive orange and black closely related or highly sympatric and coloration of the (P. occupy different niches. See companion article on page 12970.

12948–12949 ͉ PNAS ͉ November 21, 2000 ͉ vol. 97 ͉ no. 24 levels of BTXs observed among popula- Aside from the toxin origins, aspects of be produced in the epidermis (11), feath- tions of these birds prompted Dumbacher the distribution of BTXs in the different ers (12–14), and uropygial (preen) gland et al. (3) to consider that BTXs are derived tissues of birds and the mechanisms by (15), a prominent exocrine organ located from the environment, most likely from which they are disseminated warrant at the base of the tail of many birds. the diet. mention. In their initial study of Pitohui, Different lipids, or relative proportions of The emerging picture of avian toxicity, Dumbacher et al. (1) observed the high- them, occur among feather types as well while still early in its development, is est levels of homobatrachotoxinin in the (12). The separate extractions with meth- similar to that of another group of toxic skin and slightly lower levels in the feath- anol or ethanol of the epidermis, feathers, vertebrates, the dendrobatid frogs. Den- ers. Although high levels of other BTXs and uropygial gland of New Guinean birds drobatids exhibit great interspecific vari- have been confirmed in the epidermis, it provide preliminary information on the ation in the concentrations of some alka- is now suspected that the homobatracho- distribution of BTXs (1, 2). Histochemical loids present in their skin, as do natural toxinin detected there previously re- and radiolabeling techniques, however, populations within a species (5). Varia- sulted from contamination by down which are invaluable in studies of lipogen- tions in toxin levels among frog popula- feathers, the small lower shafts of which esis in vertebrate skin, are needed for tions have been attributed to different are difficult to remove completely from more definitive results on the genesis of diets because it has been established that the integument (3). The highest levels of avian skin toxins. laboratory-reared frogs sequester in- BTXs in Ifrita and Pitohui are observed Although having published only two gested alkaloids (6) and that compounds in the contour feathers of the breast, major papers on the chemistry of New identical to those observed in free-rang- belly, or legs. Guinean birds, Dumbacher and collabo- ing frogs occur in potential arthropod Feathers constitute birds’ first line of rators clearly have opened a window into prey (7). defense against consumers. Many pred- a vast realm of unexplored avian biology, The search for BTXs in organisms con- ators, in fact, such as raptors and carni- raising questions on many levels. The re- sumed by birds has yet to indicate an vores, pluck them from carcasses before cent demonstration that some ectopara- exogenous source. Stomach content stud- commencing to feed. Thus, it is not sur- ies reveal a variety of arthropods, mostly prising from a functional standpoint that sitic lice from New Guinea avoid the insects, and occasional fruits, but chemical feathers serve as a repository of defen- feathers of Pitohui or are killed by expo- analyses of these materials fail to reveal sive chemicals. Dumbacher et al. (3) sug- sure to them marks a beginning at eluci- the presence of toxins. One can only spec- gest that BTXs might be transferred from dating the survival value of avian toxins ulate on sources of avian BTXs if they are feathers onto eggs or nest materials, thus (16). Other consumers and the anticon- not synthesized de novo. The occurrence affording protection against nest-raiding sumer properties of naturally occurring of BTXs in muscle, viscera, and deep vertebrates. In addition, BTX-laden dan- BTXs need to be tested. A survey of birds regions of the skin argues against these der or feather pieces shed from the birds for BTXs and related compounds also is substances being topically applied, i.e., may impart these nonvolatile toxins to called for, particularly in light of the dis- through ‘‘anting,’’ a behavior common in other organisms—including humans, covery of alkaloids in another brightly where arthropods, fruits, or who may experience respiratory irrita- colored species, the (Ergaticus other materials are smeared directly onto tion or other untoward reactions to these ruber) from Mexico (17). Perhaps, as with COMMENTARY the plumage (8). Perhaps birds sequester birds. each of the avian species now established BTXs produced by microorganisms (9) in The relative involvement of various as toxic, the negative chemosensory re- a way analogous to that in which pufferfish avian integumentary structures in toxin sponses or other sensitivities of humans to (Fugu poecilonotus, Tetrodontidae) may sequestration (or synthesis) and release certain birds will point to taxa profitably obtain , another neurotoxin, needs further clarification. Studies of lip- investigated for biologically active natural from bacteria in their skin (10). ids reveal that different compounds may products (18).

1. Dumbacher, J. P., Beehler, B. M., Spande, T. F., Spande, T. F., Wisnieski, A. & Cover, J. F., Jr. 12. Bolliger, A. & Varga, A. (1961) Nature (London) Garraffo, H. M. & Daly, J. W. (1992) Science 258, (1994) Toxicon 32, 657–663. 190, 1125. 799–801. 7. Daly, J. W., Garraffo, H. M., Jain, P., Spande, 13. Jacob, J. & Grimmer, G. (1975) Z. Naturforsch. 30, 363–368. 2. Daly, J. W. (1998) J. Nat. Prod. 61, 162–172. T. F., Snelling, R. R., Jaramillo, C. &. Rand, A. S. 3. Dumbacher, J. P., Spande, T. F. & Daly, J. W. 14. Jacob, J. & Pomeroy, D. E. (1979) Comp. Bio- (2000) J. Chem. Ecol. 26, 73–85. 64, (2000) Proc. Natl. Acad. Sci. USA. 97, 12970– chem. Physiol. 301–303. 8. Simmons, K. E. L. (1966) J. Zool. London 149, 145–162. 15. Jacob, J. & Ziswiler, V. (1982) in Avian Biology, 12975. 9. Berkowitz, D. B. (1993) Science 259, 1238–1239. eds. Farner, D. S., King, J. R. & Parkes, K. C. 4. Cott, H. B. (1947) Proc. Zool. Soc. London 116, 10. Yotsu, M., Yamazaki, T., Meguro, Y., Endo, A., (Academic, New York), Vol. VI, pp. 199–324. 371–524. 16. Dumbacher, J. P. (1999) Auk 116, 957–963. Murata, M., Naoki, H. & Yasumoto, T. (1987) 5. Daly, J. W. & Spande, T. F. (1986) in Alkaloids: 17. Escalante, P., Emanuel, V. & Daly, J. W. (1994) Chemical and Biological Perspectives, ed. Pelletier, Toxicon 25, 225–228. J. Ornithol. 135, 410. S. W. (Wiley, New York), Vol. 4, pp. 1–274. 11. Wertz, P. W., Stover, P. M., Abraham, W. & 18. Weldon, P. J. & Rappole, J. H. (1997) J. Chem. 6. Daly, J. W., Secunda, S. I., Garraffo, H. M., Downing, D. T. (1986) J. Lipid Res. 27, 427–435. Ecol. 23, 2609–2633.

Weldon PNAS ͉ November 21, 2000 ͉ vol. 97 ͉ no. 24 ͉ 12949