Toxicon 54 (2009) 197–207

Contents lists available at ScienceDirect

Toxicon

journal homepage: www.elsevier.com/locate/toxicon

Parotoid macroglands in ( jimi): Their structure and functioning in passive defence

Carlos Jared a,*, Marta M. Antoniazzi a, Amarildo E.C. Jorda˜o b, Jose´ Roberto M.C. Silva b, Hartmut Greven c, Miguel T. Rodrigues d a Laborato´rio de Biologia Celular, Instituto Butantan, Av. Vital Brasil 1500, 05503-900, Sa˜o Paulo, b Departamento de Histologia e Embriologia, Instituto de Cieˆncias Biome´dicas, Universidade de Sa˜o Paulo, Sa˜o Paulo, Brazil c Institut fu¨r Zoomorphologie und Zellbiologie der Heinrich-Heine-Universita¨t, Du¨sseldorf, Germany d Departamento de Zoologia, Instituto de Biocieˆncias, Universidade de Sa˜o Paulo, Sa˜o Paulo, Brazil article info abstract

Article history: When (Rhinella) are threatened they inflate their lungs and tilt the body towards the Received 7 January 2009 predator, exposing their parotoid macroglands. Venom discharge, however, needs Received in revised form 27 March 2009 a mechanical pressure onto the parotoids exerted by the bite of the predator. The structure Accepted 30 March 2009 of parotoids was described before and after manual compression onto Available online 15 April 2009 the macroglands mimicking a predator attack. Parotoids are formed by honeycomb-like collagenous alveoli. Each alveolus contains a syncytial gland enveloped by a myoepithe- Keywords: lium and is provided with a duct surrounded by differentiated glands. The epithelium Amphibia Bufonidae lining the duct is very thick and practically obstructs the ductal lumen, leaving only Rhinella jimi a narrow slit in the centre. After mechanical compression the venom is expelled as a thin Parotoid jet and the venom glands are entirely emptied. The force applied by a bite of a potential Macrogland predator may increase alveolar pressure, forcing the venom to be expelled as a thin jet Skin glands through the narrow ductal slit. We suggest that the mechanism for venom discharge Venom release within all bufonids is possibly similar to that described herein for Rhinella jimi and that Passive defence parotoids should be considered as cutaneous organs separate from the rest of the skin specially evolved for an efficient passive defence. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction 1940; Lutz, 1971; Toledo et al., 1992) and Urodela (Phisalix, 1922; Luther, 1971; Brodie and Smatresk, 1990; Toledo and skin glands are a synapomorphy of the Jared, 1995). Because parotoids consist of an aggregation of group and are usually present in great numbers on the numerous secretory units, they have been named mac- whole body of all species. Among amphibian skin glands roglands to differentiate them from the common mucous the most conspicuous are undoubtedly the parotoids. and granular glands of the remaining skin (see Toledo and Parotoids (for etymology and the recommendation to use Jared, 1995). parotoid instead of ‘‘paratoid’’ or ‘‘parotid’’ see Cannon and There seems to be no doubt that parotoids as well as the Palkuti, 1976; Tyler et al., 2001) are multiglandular struc- granular glands distributed over the body of a toad, from tures, which can be found as paired protuberances, post- which the parotoids are derived (Toledo et al., 1993; Toledo orbital in position, in a variety of Anura (Wilber and Carroll, and Jared, 1995), are involved in chemical defence including defence against microorganisms (Tempone et al., 2008). Behavioural studies in both Anura and Urodela, have * Corresponding author. Tel.:/fax: þ55 11 37267222x2234. demonstrated that many species show characteristic E-mail addresses: [email protected], [email protected] (C. Jared). defence behaviour when threatened. Urodela, for example,

0041-0101/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.toxicon.2009.03.029 198 C. Jared et al. / Toxicon 54 (2009) 197–207 position their body, tilting it toward the source of contact confused with Rhinella schneideri (Stevaux, 2002). It was (for review see Brodie, 1983) and passively release unpal- originally named as Bufo jimi (Stevaux, 2002) and renamed atable or toxic cutaneous glandular contents onto the body as Rhinella jimi (Chaparro et al., 2007). Six female adult surface, rather than actively spraying them at predators specimens were collected at Fazenda Sa˜o Miguel, district of (e.g., Brodie, 1983; Brodie and Smatresk, 1990). Anura, Angicos, State of Rio Grande do Norte (5 390 4300S, 36 360 especially toads, inflate their lungs and also assume special 1800W), Brazil and brought to the vivarium of the Cell postures to present the parotoids to the source of danger, Biology Laboratory of Instituto Butantan, Sa˜o Paulo. but seem to spray their venom only in response to physical To mimic the bite of a possible predator (e.g., some pressure (e.g., Hinsche, 1928; Toledo and Jared, 1995). snakes, mammals and birds of prey), a large and constant Secretions of the granular cutaneous glands (and paro- pressure was applied to the right parotoids of the six live toids) may be very toxic to (Lutz, 1966; Toledo and individuals by squeezing them laterally between the thumb Jared, 1989a, 1995; Barthalmus, 1989; Clark, 1997). In and the forefinger. This process led to the expulsion of bufonids they contain steroids such as bufogenines and venom (see Fig. 1(5)). After that, all the animals were bufotoxins that, when in contact with the buccal mucosa sacrificed with an overdose of Thionenbutal and of many vertebrates, especially of snakes, have car- compressed and non-compressed parotoids were removed dioacceloratory properties increasing the strength of the for histological study. heart beat and decreasing heart rate (Habermehl, 1981). The venom can also exert a marked effect as a local 2.1. Anatomy anaesthetic (Habermehl, 1981). In the case of a bite, after venom ingestion, the potential predator can show intense For this study the parotoids were removed from existing salivation and excitation, paralysis, trembling and convul- specimens from the herpetological collection of the Insti- sions, often leading to death (Vital Brazil and Vellard, 1925, tuto Butantan, fixed in 10% formalin and kept in 70% 1926; Garrett and Boyer, 1993; Pineau and Romanoff, 1995; ethanol. The parotoids were divided on a frontal plane into Sakate and Lucas de Oliveira, 2000; Sonne et al., 2008). two pieces. The alveolar content was carefully removed Anuran granular or venom glands are syncytial in nature using a toothpick to expose the internal macrogland (Fox, 1986; Delfino, 1991; Terreni et al., 2003). They are framework. Pore disposition was examined by direct irregularly distributed over the whole body surface, but are observation of the parotoid surface or by light microscopy found also in aggregated forms in certain parts of the body using whole mount preparation of the parotoid skin surface such as the parotoid and the paracnemic macroglands of flattened on a glass slide. bufonids (Hostetler and Cannon, 1974; Toledo and Villa, 1987; Toledo et al., 1992) or the lumbar macroglands of 2.2. Histology some leptodactylids (Toledo and Jared, 1989b, 1995; Toledo et al., 1996; Lenzi-Mattos et al., 2005). Studies on the The parotoids removed from the six collected individ- structure and function of such aggregations are limited. uals were divided transversely to the gland larger axis into Specifically in toads, there are some articles that describe four pieces and fixed in 4% formaldehyde, 1% glutaralde- their secretory syncytia, the secretion granules and the hyde, buffered in 0.1 M phosphate buffer, pH 7.2 (McDowell myoepithelial layer (Hostetler and Cannon, 1974; Cannon and Trump, 1976) for 4 days. The pieces were transversely and Hostetler, 1976; Toledo et al., 1992; Barthalmus, 1994; oriented and embedded in paraffin, sectioned 4 mm thick, Almeida et al., 2007). In addition, a special vascularization stained with Haematoxylin–Eosin and Mallory trichrome net irrigating the toad parotoids has been described in for general observations, and submitted to Bromophenol detail (Hutchinson and Savitzky, 2004). On the other hand, Blue, and PAS combined with Alcian Blue pH 2.5 (Bancroft the biochemistry and biological effects of toad skin, and Steven, 1990), for detection of proteins in general and including the parotoids, are relatively well studied (Pas- mucosubstances, respectively. Picrosirius staining followed quarelli et al., 1987; Rossi et al., 1997; Maciel et al., 2003; by polarized microscopy was used in a few sections for the Shimada et al., 2006; Tempone et al., 2008). analysis of collagen fibres (Junqueira et al., 1979). Micro- In the present article we describe the parotoids of the graphs were taken with an Olympus BX60 light microscope toad Rhinella jimi focusing on the structure of their tissues connected to Olympus PM-C35DX photo equipment or before and after mechanical pressure. Our results show that with an Olympus BX51 light microscope equipped with the ducts of the individual granular glands are lined by a digital camera and with the software Image-Pro Express thick differentiated epithelia constituting a plug that (Media Cybernetics). hinders the passage of secretion. Further we suggest that the special arrangement of tissues around the glands and 2.3. Transmission electron microscopy inside the ducts, together with pressure applied on the macroglands, e.g., by the experimenter or a predator, lead Small cubes (8 mm3) of parotoid superficial tissue con- to venom discharge in the form of strong jets, character- taining single parotoid pores were fixed in a mixture of izing a peculiar type of passive defence in toads. 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 M cacodylate buffer, pH 7.2, post-fixed in 1% osmium 2. Material and methods tetroxide, dehydrated and embedded in Spurr resin. Samples were oriented to obtain transverse ultrathin Rhinella jimi (Fig. 1) is a Brazilian toad restricted to the sections of the pores. Sections were cut in a Sorvall MT6000 semiarid Caatinga of northeastern Brazil and previously ultramicrotome, stained with uranyl acetate and lead C. Jared et al. / Toxicon 54 (2009) 197–207 199 citrate (Reynolds, 1963) and examined in a LEO 906E dermal strata, underlining the glands (Fig. 2(7)). The transmission electron microscope. unpigmented ventral skin shows a flat surface, a much thinner stratum compactum and a small number of glands, 3. Results mainly mucous glands (Fig. 2(8)). The histological analysis of the parotoids show that they 3.1. Anatomy of the parotoid glands and venom discharge are thickened portions of the dorsal skin in which, besides induced by external pressure the presence of mucous and regular granular glands, very large bottle-shaped syncytial granular glands are accumu- The parotoids are conspicuous glandular structures, lated in the stratum compactum. Each gland is connected to which are easily dissected and removed from the body the exterior through a narrow aperture (slit) in the centre when the dorsal skin around them is cut. Externally the of the ductal lining epithelia. In the parotoid the differen- parotoid skin seems similar to the rest of the dorsal skin, tiation of the two dermal strata (spongiosum and com- except for the presence of a large number of well defined pactum) is not as clear as in the rest of the dorsal skin. pores (Fig. 1). When the macroglands are horizontally cut However, similarly to the dorsal skin, the mucous and into two pieces and the pasty secretion is removed, regular granular glands are located just under the a honeycomb-like structure is revealed (Fig. 1(2,3)), epidermis (Fig. 2(9)). Moreover, epidermis and dermis composed of many alveoli. Transverse histological section are traversed by the large ducts connecting the surface to through the parotoid shows that each alveolus contains the large syncytial granular glands. These ducts are sur- a very large bottle-shaped gland, full of secretion and rounded by differentiated glands exclusively present in the closed on the top by a thick ductal epithelium apparently parotoid (Fig. 2(9–11)). Differently from the dorsal skin, the forming a plug (Fig. 1(4)). Serial sections, however, reveal calcified dermal layer in the parotoid is much thicker and that there is a very narrow ductal lumen forming a slit (Figs. spread among the connective tissue fibres, forming an 1(4) and 3(14)) through which the secretion can pass when irregular band about 150 mm thick below the epidermis the gland is squeezed. Since the parotoid forms a convex (but above the bottle-shaped granular glands), in which the structure, the bottle-shaped glands are larger and more mucous glands are immersed (Fig. 2(9)). elongated in the central region and gradually decrease in The bottle-shaped syncytial granular glands of the size toward its borders. The pores on the surface of the parotoid are filled with a network of secretion products parotoids measure from 0.5 to 1 mm, depending on the (Figs. 2(9) and 4(22,23)). This secretion is very reactive to individual, and appear as circular depressions, almost Alcian Blue and negative to Bromophenol Blue, differently entirely obstructed internally by solid tissue, except for the from the granular glands of the dorsal skin (compare Fig. small slit in the centre. Each pore is externally surrounded 2(10,11)). The syncytial nuclei are arranged in a single by smaller pores forming a special pore pattern on the peripheral row. Each gland is enveloped by a monolayer of parotoid surface (Figs. 1(6) and Fig. 3(16)). thin myoepithelial cells, with flat and elongated nuclei A constant manual pressure applied to the parotoids (Fig. 4(23)). forces the venom to be discharged only from the pores The large ducts of the bottle-shaped glands are sur- corresponding to the glands which were effectively rounded by a number of glands clearly different from squeezed between the fingers. The venom is expelled in the mucous and granular glands both in size and histochem- form of thin jets (Fig. 1(5)). It is sticky and has a colour istry (compare Fig. 2(9,10), and Fig. 2(12,13)). In histological varying from white to yellowish. transverse sections, these glands appear circularly arranged around the main duct (Fig. 3(15)), forming a rosette-like 3.2. Histology and histochemistry arrangement. In whole mount preparations the small pores of these glands can be seen surrounding the main pore of The dorsal skin has an irregular surface and is composed the large syncytial granular gland (Fig. 3(16)). Although of the epidermis containing around six cellular layers, and similar to the common mucous glands in terms of their the dermis with the stratum spongiosum, where mucous typical acinar structure and their positive reaction to PAS and granular glands are observed, and the stratum com- and Alcian Blue, these differentiated glands are much pactum, mainly constituted by thick collagen fibres larger, with diameters around 370 mm, whereas the (Fig. 2(7)). The mucous glands are arranged just below the common mucous glands have diameters around 100 mm epidermis and have a typical acinar form, with large lumina (Fig. 2(9,10)). Also, the acini of the differentiated glands are and thin ducts opening in the skin surface. The secretory formed by high prismatic cells, full of a homogeneous epithelium reacts differentially to PAS and Alcian Blue secretion very reactive to Bromophenol Blue (Fig. 2(10)) corresponding to the presence of two types of secretory and to PAS (Fig. 2(11,12)), and sparse cells, very positive to cells (Fig. 2(13)) but it does not react to Bromophenol Blue Alcian Blue, pH 2.5, located mainly in the apical secretory (Fig. 2(10)). The granular glands are larger than the mucous epithelium (Fig. 2(12)). Occasionally the lumen contains glands and consist of a syncytium with no lumen (Fig. secretion. 2(7,9)) and filled with secretion positive to Bromophenol Longitudinal serial sections reveal that bundles of Blue (Fig. 2(10)) and negative to PAS and Alcian Blue collagen fibres surround each duct, forming a distinct (Fig. 2(11)). Both types of gland are encased by myoepi- arrangement from the rest of the stratum compactum thelia and are connected to the skin surface through (Fig. 3(14)). At about 500 mm from the skin surface, the duct epithelial ducts. A continuous basophilic calcified dermal lining epithelium is thickened and practically obstructs the layer runs parallel to the epidermis, between the two ductal lumen, leaving only a narrow slit (measuring about

C. Jared et al. / Toxicon 54 (2009) 197–207 201

Fig. 2. (7) Rhinella jimi dorsal skin. The dermis shows many mucous (*) and granular (g) glands. The calcified dermal layer (ca) is present dividing the stratum spongiosum (ss) from the stratum compactum (sc). E, epidermis; v, blood vessels. Paraffin, HE. (8) Rhinella jimi ventral skin. When compared to the dorsal skin, it is much thinner and with a small number of glands, which are mainly mucous (*). The dermis is well vascularized, with many blood vessels (v) under the epidermis (E). sc, stratum compactum; ss, stratum spongiosum. Paraffin, HE. (9) Rhinella jimi parotoid macrogland. Apical portion of the macrogland where four glandular types are distinguish: common mucous gland (*), common granular gland (g), differentiated gland (dg) and parotoid granular gland (G). Note that the differentiated cells are much larger than the mucous cells. A thick calcified layer (ca) is seen in the upper dermis (D) underlining the epidermis (E). A small part of the ductal epithelium (d) is observed in the parotoid granular gland. Paraffin, HE. (10) Rhinella jimi parotoid macrogland. Equivalent section to Fig. 2(9) stained with Bromophenol Blue. Mucous glands (*) and parotoid granular gland (G) are negative, the common granular gland (g) is positive and the differentiated glands (dg) are highly positive to the method. E, epidermis; D, dermis, d, ductal epithelium. Paraffin. (11–13) Rhinella jimi parotoid macrogland. Comparison of the four gland types present in the macrogland by PAS-AB (pH 2.5). The parotoid granular gland (G) is positive to Alcian Blue, while the common granular glands (g) are negative. The cells of the mucous glands (*) are mainly positive to Alcian Blue (arrows, Fig. 2(13)) while in the differentiated glands (dg) most cells are positive to PAS, with a few cells positive to Alcian Blue (arrows, Fig. 2(12)). E, epidermis; D, dermis; d, duct; s, secretion. Paraffin.

Fig. 1. (1) Female Rhinella jimi exhibiting the right parotoid macrogland localized at the postorbital region. The arrows point to the glandular pores. (2) A parotoid sectioned according to a frontal plane, from which the venom was withdrawn. Notice the alveolar, honeycomb-like internal structure. (3) Higher magnification of the alveoli showing the walls (arrows) and floors (asterisks). (4) Longitudinal section of two parotoid bottle-shaped glands (G). The arrows point to the pores obstructed by a thick epithelium. D, dermis. Paraffin, HE. (5) Venom jets squirting from the pores, after parotoid manual compression. (6) View of the pores on the parotoid surface. Note the small slit (large arrows) in the duct centre. Each pore is surrounded by smaller pores (small arrows). 202 C. Jared et al. / Toxicon 54 (2009) 197–207

Fig. 3. (14) Rhinella jimi parotoid macrogland. Histological longitudinal section of a duct obstructed by the ductal epithelium (d) and surrounded by the differentiated glands (asterisks) and wrapped by distinct bundles of collagen fibres (arrowheads). Note the thin canal (small arrow) running in the central portion of the duct. The larger arrow point to the external side of the pore. ca, calcified dermal layer; D, dermis; E, epidermis. Paraffin, HE. (15) Rhinella jimi parotoid macrogland. Histological transverse section of the duct surrounded by specialized mucous glands (asterisks), forming a rosette. E, epidermis. Paraffin, HE. (16) Rhinella jimi parotoid macrogland. Whole mount preparation of the parotoid epidermis in the region of the pore. Most of the duct volume is obstructed by the plug, leaving only a central pore (central arrow). Circling disposed arrows point to the differentiated glands around the central pore. (17) Rhinella jimi parotoid macrogland. Larger magnification of the duct epithelial cells, showing difference in shape between the peripheral and central cells. Paraffin, HE. (18) Rhinella jimi parotoid macrogland. The duct epithelial cells are positive to the PAS method. (19 and 20) Rhinella jimi parotoid macrogland. Transmission electron micrographs of the epidermal cells composing the thick duct lining, which are voluminous at the periphery (Fig. 3(19)), becoming flatter towards the lumen, with slender cytoplasm processes (asterisks) (Fig. 3(20)). All cells show electron transparent cytoplasm inclusions (i). Nu, nuclei.

40 mm) in the centre. Longitudinal sections through the vascularized stratum, closely adjacent to the gland epithelium reveal that the cells arranged towards the myoepithelia, and a dense stratum (Fig. 4(21,22)). The ductal slit are flatter than the basal cells (Fig. 3(17)). In dense stratum is continuous with the superficial stratum addition, the whole ductal epithelium is quite reactive to compactum, and envelopes each gland, structuring the PAS (Fig. 3(18)). The transmission electron microscopy of honeycomb-like framework of the parotoid (see Fig. 1(2,3)). these cells shows several cytoplasm inclusions containing The parotoid base is flat and constituted of the same dense material of medium electron density (Fig. 3(19,20)). The connective tissue. connections among the central flatter cells consist of very After parotoid compression, conspicuous structural loose interdigitations (Fig. 3(20)). changes are observed in the alveoli affected by squeezing: The connective tissue surrounding the large syncytial they are empty and almost completely free of secretion. The glands is divided into two strata, a loose and richly periphery, where the syncytial nuclei are located, collapses C. Jared et al. / Toxicon 54 (2009) 197–207 203

Fig. 4. (21) Rhinella jimi parotoid macrogland. Two granular parotoid glands, one non-compressed filled with secretory product (G) and another compressed, with the syncytium (sy) completely collapsed in the centre, appearing in orange. Note the difference in volume between the loose connective tissue (lt) in both glands. The dense connective tissue (dt) is unchanged. Blood vessels (arrows). Paraffin, Mallory trichrome staining. (22) Rhinella jimi parotoid macrogland. Connective tissue between two parotoid granular glands (G). The loose (lt) and dense (dt) connective tissues are clearly distinguish. Blood vessels (arrows) are abundant in the loose connective tissue. Paraffin, HE. (23) Rhinella jimi parotoid macrogland. Part of the syncytium (sy) of a non-compressed alveolus, sheathed by the myoepithelial layer (my) and full of granular secretory product (G) . The arrow points to a blood vessel. Paraffin, HE. (24) Rhinella jimi parotoid macrogland. A compressed parotoid gland, equivalent to the one observed in Fig. 4(21), with the syncytium (sy) completely collapsed in the center. The arrows point to blood vessels. lt, loose connective tissue, dt, dense connective tissue. Paraffin, HE. (25) Higher magnification of Fig. 4(24), where rests of the secretory product are seen inside the syncytium, which is enveloped by the myoepithelial layer (my). lt, loose connective tissue, dt, dense connective tissue. Paraffin, HE. 204 C. Jared et al. / Toxicon 54 (2009) 197–207

Fig. 5. (26 and 27) Rhinella jimi parotoid macrogland. Connective tissue between two non-compressed parotoid granular glands (G) in non-polarized (Fig. 5(26)) and polarized (Fig. 5(27)) microscopy. The loose connective tissue (arrows) is compressed towards the dense connective tissue (dt). Paraffin, Picrosirius staining. (28 and 29) Rhinella jimi parotoid macrogland. Connective tissue between two compressed parotoid granular glands (G) in non-polarized (Fig. 5(28)) and polarized (Fig. 5(29)) microscopy. The loose connective tissue (lt) is expanded towards the dense connective tissue (dt). Paraffin, Picrosirius staining. together with the surrounding myoepithelium: both (probably consisting of a complex of species), lives in the (syncytium and myoepithelium) are now seen as a wrin- central Brazilian Cerrados (dry open woodlands). Both kled structure in the centre of the alveolus (compare full species, despite the presence of the paracnemic macro- and empty alveoli in Fig. 4(21)). The loose vascularized glands (characteristic of the Rhinella schneideri group) in connective tissue, adjacent to the myoepithelium, appears their hind limbs, comprise large individuals, up to 300 mm expanded occupying a larger space between the collapsed in adult females. One of the diagnostic features which gland and the dense connective tissue, which is unchanged enables the recognition of Rhinella jimi as a distinctive (Fig. 4(21,24,25)). The comparison of a gland full of secre- species in the Rhinella schneideri group is the presence of tion with a squeezed gland in Picrosirius stained sections macroglands also in the forelimbs (Stevaux, 2002). This viewed under polarized microscopy shows the difference characteristic has already been described by (Toledo and between non-expanded and expanded loose connective Jared, 1995, Fig. 2(13)) who referred to such glandular tissue surrounding each of the wrinkled syncytia (compare accumulations as radioulnar macroglands, in order to Fig. 5(26,27) with Fig. 5(28,29)). distinguish them from the paracnemic macroglands, Serial histological sections of the compressed glands did previously studied by Toledo and Villa (1987). Different not show any significant modification of the ducts after from other anurans inhabiting open and dry regions, secretion release except for a discrete widening of the Rhinella jimi and Rhinella schneideri remain completely ductal slit and the presence of secretion in the ductal exposed to the environment for long periods, without lumen. running the apparent risk of desiccation, a fact that is probably associated with the presence of a calcified dermal 4. Discussion layer (Elkan, 1976; Toledo and Jared, 1993b), covering the entire body, including the parotoids. In both species paro- Rhinella jimi has a wide distribution in Brazilian Caa- toid macroglands are very prominent, a fact commented on tinga. Rhinella schneideri, which is a close relative of R. jimi by Lutz (1925) in the original description of Rhinella C. Jared et al. / Toxicon 54 (2009) 197–207 205 schneideri. At that time he qualified those macroglands as The dense connective tissue surrounding each bottle- ‘‘enormous’’ although referring to them as ‘‘parotids’’. shaped gland forms a resistant framework responsible for Besides their large size, they occupy a prominent position the shape maintenance of the parotoid, which appears in relation to the body and, considering the combination of macroscopically unaltered even after venom release. On defensive behaviours usually found in these toads, it seems the other hand, the loose connective tissue seems to form evident that they depend on the parotoid macroglands in part of the secretion release system. Also, the loose tissue their defence. must have a significant role in syncytium and myoepithelial Rhinella jimi when threatened usually exhibits a stereo- maintenance, since it is highly vascularized and must be typed defensive behaviour, characterized by the inflating involved in gland nutrition and transport of precursor of the lungs and the assumption of a stiffened and volu- molecules for venom synthesis (Hostetler and Cannon, minous form. This behavioural posture is often followed by 1974; Cannon and Hostetler, 1976; Erspamer, 1994; the head-butting behaviour, in which the tilts the Hutchinson and Savitzky, 2004). Immediately after the body towards the threatening agent, exposing one of the explosive emptying, the syncytium is totally collapsed and parotoids (Toledo and Jared, 1995, referred to as Bufo par- the space between the dense connective tissue and the acnemis). Head-butting seems to be an important part of myoepithelium is filled by the obviously expanded loose the toad defensive strategy. The act of exposing the paro- connective tissue. Simultaneously with the syncytium toid certainly increases the chances of the animal having its secretion refilling, the loose connective tissue is gradually venom release system triggered when the predator bites it. pushed towards the dense connective tissue walls. Our observations strongly suggest that jets of venom are The ultrastructure of compressed and non-compressed only discharged from the parotoids after a considerable parotoids in (formerly Bufo ictericus), mechanical pressure. In fact, spontaneous discharge of jets suggests that the alveolus, when full of venom, is under from the parotoids has not been reported in Anura and has a constant internal pressure, since organelles are crowded rarely been in Urodela (Brodie, 1983; Brodie and Smatresk, in a small cytoplasm volume. After discharge, however, the 1990). In toads, however, pressure seems to be necessarily organelles are easily recognized and distributed in a larger exerted by an external force which in nature must corre- volume of cytoplasm (Toledo et al., 1992). Our histological spond to the attack or biting of a predator. Such external results strongly indicate that similar features must occur in pressure was mimicked herein by manual squeezing. the parotoid alveoli of Rhinella jimi. When the alveoli are Toads seem to have a graded sequence of defensive full, it is most probable that the normal pressure of the behaviours: first inflating the lung and lifting the body, venom on the loose connective tissue maintains a condi- which is followed by spraying the venom, if the predator tion of relative internal turgidity. The moment Rhinella jimi should attack and/or bite his victim. Even if the toad is feels threatened its defence mechanism of inflating its killed, the defensive strategy may be efficient at the level lungs may contribute to the increase of the internal pres- of species defence, since the predator may be able to sure of each alveolus. Since the collagen surrounding the associate the stereotyped defence behaviour with the alveolus and the duct appears very resistant, when an distasteful or harmful venom when grasping another toad. external force is applied to the parotoid, e.g., by a bite, the However, this system seems not to have been explored in individual alveolar pressure is increased to a threshold detail in toads and depends on the predator’s capabilities, level, forcing the venom to eject through the narrow slit by as many animals are known to have found ways to compressing the ductal epithelial lining cells. The ductal avoid poison. Experiments in captivity in which thick epithelium could be compared in this context with (Eupemphix nattereri, formerly Physalaemus nattereri) a small nozzle. Only the alveoli directly compressed cause were offered to the procyonid Nasua nasua (coati) showed the venom discharge, liberating simultaneous multi- that this mammal rejected the first but could ingest directed ‘‘shots’’ inside the predator’s mouth. This diffuse the second after vigorously rolling and rubbing it against mechanism of venom elimination in a spray form can the soil (Sazima and Caramaschi, 1986). We have many reinforce the envenomation through breathing mucosa. times observed in the field half-eaten dead toads where The parotoid alveoli which were not directly compressed the head and most of the skin were discarded by the remain practically intact and ready to be triggered in the predator. In urodelans, chickens learn avoidance of the event of a new attack. efts of the salamander Notophthalmus viridescens (Bran- This putative mechanism does not contradict the fact don et al., 1979) and a hedgehog avoided Pleurodeles waltl that granular glands of the body skin and also the parotoid after biting a specimen several times on the tail (Nowak glands to some extent may discharge their secretion by and Brodie, 1978). contraction of the myoepithelium. Both mechanisms may Although the external pressure is probably the most complement each other at the moment the animal is important factor for parotoid venom release, lung inflation threatened. Contraction of the myoepithelium is triggered must be significant in the defence of Rhinella jimi since, by an adrenergic mechanism evoked by orthosympathetic besides making the animal appear larger, the lung pressure stimulation (Holmes and Balls, 1978; Delfino et al, 1982; against the body is possibly carried over to the parotoid Nosi et al., 2002). In addition, the myoepithelial cells in floor and transferred to the parotoid bottle-shaped glands. macroglands may have the additional function of homo- The state of turgidity thus produced and the probable aid of genizing the voluminous gland secretion by constant gentle the individual glandular myoepithelial layer make it such contractions. that, by external pressure, the secretion is expelled through The thick epithelium present in the parotoid ducts, the duct slit causing venom jets to squirt out. although previously shown elsewhere (Lobo, 2005; Lobo 206 C. Jared et al. / Toxicon 54 (2009) 197–207 et al., 2005; Jared et al., 2007) was herein described in morphology of these structures is quite similar in many detail, with suppositions of its role in venom release. In the toad species, we suggest that the mechanism for venom same way, although the large differentiated glands in the release within the genus is possibly similar to that we have parotoids and their special arrangement around each pore described for Rhinella jimi. In addition, supported by the of the bottle-shaped glands have already been shown structural complexity herein shown, the parotoid macrog- (Lobo, 2005; Lobo et al., 2005; Jared et al., 2007), they were lands can no longer simply be considered as mere gland herein described in more detail. The function of these aggregations but should be regarded as highly differenti- differentiated glands is entirely unknown as yet, but based ated structures, characteristic of bufonids, forming cuta- on their specific location, one should expect their secre- neous organs separate from the rest of the skin, which were tions to have a role in venom release or in making part of specially evolved for an efficient passive defence. the venom itself. In fact, Almeida et al. (2007) reported that in Rhinella icterica, the parotoid venom is composed of a mixture of products released by different gland types. Acknowledgements Almeida et al. (2007), however, did not make a distinction between the common mucous cells and the herein named This work was made possible due to the help and differentiated glands, which they called mixed glands due kindness of Mr Francisco de Assis Rodrigues, manager of to the protein and mucous nature of their secretory the Fazenda Sa˜o Miguel (District of Angicos, Rio Grande do product. We have shown the same type of histochemical Norte State). Thanks are due to CNPq (307247/2007-4 for results in R. jimi. Also, Almeida et al. (2007) did not notice CJ, and 302212/82-5 for MTR), FAPESP and Fundaça˜o the clear association of these mixed glands with the large Butantan. We would like to thank Prof. Reynaldo C. de parotoid ducts which have already been described by Lobo Toledo for important insights regarding the manuscript. (2005), Lobo et al. (2005) and Jared et al. (2007). We also thank Simone G.S. Jared and Maria Helena Ferreira Another morphological aspect of R. jimi skin deserving for their efficient technical assistance. Specimens were attention is the difference between the calcified dermal collected under 172/1999 and 193/2001 IBAMA permits. layer of the dorsal skin and of the parotoid macrogland skin. In the parotoid this layer is much thicker and is Conflict of interest located more superficially, just below the epidermis and above the bottle-shaped glands. Considering anuran The authors have no conflicts of interest. integument in general, this observation seems quite unexpected since the calcified layer is usually localized below the glands, even when they are large (personal References observations). This unusual more external position of the calcified dermal layer in the parotoid probably confers to Almeida, P.G., Felsemburgh, F.A., Azevedo, R.A., Brito-Gitirana, L., 2007. Morphological re-evaluation of the parotoid glands of Bufo ictericus the macrogland a more intense superficial mechanical (Amphibia, Anura, Bufonidae). Contrib. Zool 76, 145–152. resistance, and may help, at the moment of a bite, to Bancroft, J.B., Steven, A., 1990. Theory and Practice of Histological Tech- canalize the internal venom pressure towards the ducts, niques. Churchill Livingstone, Edinburgh. forcing the venom to be released through the ductal slits. Barthalmus, G.T., 1989. Neuroleptic modulation of oral dyskinesias induced in snakes by Xenopus skin mucus. Pharmacol. Biochem. Association of the calcified dermal layer distribution Behav 30, 957–959. pattern with the mechanism of gland discharge has already Barthalmus, G.T., 1994. Biological roles of amphibian skin secretions. In: been proposed for Physalaemus nattereri (presently Heathole, H., Barthalmus, G.T., Heatwole, A.Y. (Eds.), Amphibian biology – The integument. Surrey Beatty & Sons, Chipping Norton, pp. Eupemphix nattereri)(Lenzi-Mattos et al., 2005), whose 382–410. posterior dorsal skin is provided with a pair of black and Brandon, R.A., Labanick, G.M., Huheey, J.E., 1979. Learned avoidance of circular macroglands resembling two black eyes. These brown efts, Notophthalmus viridescens louisianensis (Amphibia, Uro- dela, Salamandridae), by chickens. J. Herpetol 13, 171–176. macroglands appeared to be the only region of the frog’s Brodie Jr., E.D., 1983. Salamander antipredator postures. Copeia 1977, integument where the calcified dermal layer is absent, 523–535. probably facilitating the venom to be expelled when the Brodie Jr., E.D., Smatresk, N.L., 1990. The antipredator arsenal of fire salamanders: spraying of secretions from highly pressurized dorsal attack of a predator is directed to the macroglandular skin glands. Herpetologica 46, 1–7. region (Lenzi-Mattos et al., 2005). Cannon, M.S., Hostetler, J.R., 1976. The anatomy of the parotoid gland Besides R. jimi, we have examined the parotoid of in Bufonidae with some histochemical findings. II. Bufo alvarius. J. Morphol 148, 137–160. many different species of Rhinella (R. schneideri, R. icterica, Cannon, M.S., Palkuti, G.A., 1976. The ‘parotoid’ gland of Bufonidae. Tox- R. margaritifera, R. crucifer and R. granulosa). In all the icon 14, 149–151. studied species the morphological and histochemical Chaparro, J.C., Pramuk, J.B., Gluesenkamp, A.G., 2007. A new species of patterns of the parotoid macroglands were very similar to arboreal Rhinella (Anura: Bufonidae) from cloud forest of south- eastern Peru. Herpetologica 63 (2), 203–212. what we have described for R. jimi, including the duct Clark, B.T., 1997. The natural history of amphibian skin secretions, their arrangement, the differentiated glands around the ducts normal functioning and potential medical applications. Biol. Rev. 72, and the location of the calcified dermal layer. This similarity 365–379. Delfino, G., 1991. Ultrastructural aspects of venom secretion in anuran is a strong indication that the structure of the parotoid cutaneous glands. In: Tu, A.T., Dekker, M. (Eds.), Reptile Venoms and macrogland herein described may constitute a basal char- Toxins. Handbook of Natural Toxins, vol. 5. Marcel Dekker Inc., New acter within the genus. York, pp. 777–802. Delfino, G., Amerini, S., Mugelli, A., 1982. In vitro studies on the ‘‘venom’’ Based on the fact that parotoid macroglands are emission from the skin of Bombina variegata pachypus (Bonaparte) common to the whole genus Rhinella, and that the general (Amphibia, Anura, Discoglossidae). Cell Biol. Int. Rep 6, 843–850. C. Jared et al. / Toxicon 54 (2009) 197–207 207

Elkan, E., 1976. Ground substance: an anuran defense against desiccation. Nowak, R.T., Brodie, E.D., 1978. Rib penetration and associated antipred- In: Lofts, B. (Ed.), Physiology of the Amphibia, vol. 3. Academic Press, ator adaptations in the salamander Pleurodeles waltl (Salamandridae). London, pp. 101–111. Copeia 1978, 424–429. Erspamer, V., 1994. Bioactive secretions of the amphibian integument. In: Pasquarelli, P., Mendes, E.G., Sawaya, P., 1987. The action of parotoid Heathole, H., Barthalmus, G.T., Heatwole, A.Y. (Eds.), Amphibian venom on the heart of the toad (Bufo ictericus ictericus Spix 1824) and Biology – The Integument. Surrey Beatty & Sons, Chipping Norton, pp. its effects on the inhibition caused by vagal stimulation. Comp. Bio- 176–350. chem. Physiol 87, 393–399. Fox, H., 1986. Dermal glands. In: Bereiter Hahn, J., Matoltsy, A.G., Phisalix, M., 1922. Animaux Venimeux et Venins. Masson & Co., Paris. Richards, K.S. (Eds.), Biology of the Integument, vol. 2. Vertebrates. Pineau, X., Romanoff, C., 1995. Envenomation of domestic carnivorous. Springer, Berlin, pp. 116–135. Rec. Me´d. Ve´t 171, 182–192. Garrett, C.M., Boyer, D.M., 1993. Bufo marinus (Cane Toad). Predation. Reynolds, E.S., 1963. The use of lead citrate at high pH as an electron- Herpetol. Rev. 24, 148. opaque stain in electron microscopy. J. Cell. Biol. 17, 208–212. Habermehl, G., 1981. Venomous Animals and their Toxins. Springer Ver- Rossi, M.H., Blumenthal, E.E.A., Jared, C., 1997. Bufadienolides from the lag, Berlin. venom of Bufo paracnemis (Amphibia, Anura, Bufonidae). An. Assoc. Hinsche, G., 1928. Kampfreaktionen bei einheimischen. Anuren. Biol. Zbl Bras. Quı´m 46, 21–26. 48, 577–617. Sakate, M., Lucas de Oliveira, P.C., 2000. Toad envenoming in dogs: effects Holmes, C., Balls, M., 1978. Invitro studies on the control of myoepithelial and treatment. J. Venom. Anim. Toxins 6, 1–9. cell contraction in the granular glands of Xenopus laevis skin. Gen. Sazima, I., Caramaschi, U., 1986. Descriça˜odePhysalaemus deimaticus, Comp. Endocrinol 36, 255–263. sp. n., e observaço˜es sobre comportamento deima´tico em Phys- Hostetler, J.R., Cannon, M.S., 1974. The anatomy of the parotoid gland in alaemus nattereri (Steindachner) – Anura. Leptodactylidae. Rev. Biol. Bufonidae with some histochemical findings. I, Bufo marinus. J. Mor- 13, 91–101. phol 142, 225–240. Shimada, K., Miyashiro, Y., Nishio, T., 2006. Characterization of in vitro Hutchinson, D.A., Savitzky, A.H., 2004. Vasculature of the parotoid glands metabolites of toad venom using high-performance liquid chromato- of four species of toads (Bufonidae: Bufo). J. Morphol 260, 247–254. graphy and liquid chromatography–mass spectrometry. Biomed. Jared, C., Antoniazzi, M.M., Jorda˜o, A.E.C., Silva, J.R.M.C., Vasconcellos, T.P., Chromatogr 20, 1321–1327. Zanotti, A.P., Rodrigues, M.T., 2007. Morphology of the parotoid Sonne, L., Rozza, D.B., Wolffenbu¨ ttel, A.N., Meirelles, A.E.W.B., Pedroso, P.M.O., macroglands in the toad Chaunus jimi and the mechanism of venom Oliveira, E.C., Driemeier, D., 2008. Intoxicaça˜o por veneno de sapo em um expulsion. Annals of the Third Congress, Brazilian Society of Herpe- canino. Cienc. Rural 38, 1787–1789. tology, Bele´m, Brazil. Stevaux, M.N., 2002. A new species of Bufo Laurenti (Anura, Bufonidae) Junqueira, L.C.U., Bignolas, G., Brentani, R.R., 1979. Picrosirius staining plus from Northeastern Brazil. Rev. Bras. Zool 19, 235–242. polarization microscopy, a specific method for collagen detection in Tempone, A., Pimenta, D., Lebrun, I., Sartorelli, P., Taniwaki, N., Andrade Jr., H., tissue sections. Histochem. J. 11, 447–455. Antoniazzi, M.M., Jared, C., 2008. Antileishmanial and antitrypanosomal Lenzi-Mattos, R., Antoniazzi, M.M., Haddad, C.B., Tambourgi, D.V., activity of bufadienolides isolated from the toad Rhinella jimi parotoid Rodrigues, M.T., Jared, C., 2005. The inguinal macroglands of the frog macrogland secretion. Toxicon 52, 13–21. Physalaemus nattereri (Leptodactylidae): structure, toxic secretion Terreni, A., Nosi, D., Greven, H., Delfino, G., 2003. Development of serous and relationship with deimatic behaviour. J. Zool 266, 385–394. cutaneous glands in Scinax nasica (Anura, Hylidae): patterns of Lobo, S, 2005. Macroglaˆndulas paroto´ ides de sapos (Anura: Bufonidae): poisons biosynthesis and maturation with larval glands in specimens um estudo integrativo. Dissertaça˜o (Mestrado em Morfologia), of other families. Tissue Cell 35, 274–287. Departamento de Morfologia, Universidade Federal de Sa˜o Paulo, Sa˜o Toledo, R.C., Jared, C., 1989a. Consideraço˜es sobre o veneno de anfı´bios. Paulo. Cienc. Cult 41, 250–258. Lobo, S., Piffer, T.R.O., Antoniazzi, M.M., Carneiro, S.M., Jared, C., 2005. Toledo, R.C., Jared, C., 1989b. Estudo histolo´ gico das glaˆndulas lombares Morphophysiology of the paratoid macroglands in toads (Amphibia, de Pleurodema thaul (Amphibia, Anura, Leptodactylidae). Rev. Bras. Anura, Bufonidae). Mem. Inst. Butantan 61, 84. Biol. 49, 421–428. Luther, W., 1971. Distribution, biology and classification of salamanders. Toledo, R.C., Jared, C., 1993. The calcified dermal layer in anurans. Comp. In: Bu¨ cherl, W., Buckley, E.E. (Eds.), Venomous Animals and their Biochem. Physiol 104, 443–448. Venoms, vol. 2. Academic Press Inc., Boston, pp. 557–568. Toledo, R.C., Jared, C., 1995. Cutaneous granular glands and amphibian Lutz, A., 1925. Batraciens du Bre´sil. C.R. Se´ances Soc. Biol. Paris 93, venoms. Comp. Biochem. Physiol 111, 1–29. 211–214. Toledo, R.C., Villa, N., 1987. Estudo histolo´ gico das glaˆndulas tibiais de Bufo Lutz, B., 1966. Biological significance of cutaneous secretions in toads and paracnemis (Amphibia, Anura, Bufonidae). Rev. Bras. Biol. 47, 257–264. frogs. Mem. Inst. Butantan 33, 55–59. Toledo, R.C., Jared, C., Brunner, A., 1992. Morphology of the large granular Lutz, B., 1971. Venomous toads and frogs. In: Bu¨ cherl, W., Buckley, E.E. alveoli of toad (Bufo ictericus) parotoid glands before and after (Eds.), Venomous Animals and their Venoms, vol. 2. Academic Press, compression. Toxicon 30, 745–753. Boston, pp. 423–473. Toledo, R.C., Jared, C., Brunner, A., 1996. The lumbar glands of the frog Maciel, N.M., Schwartz, C.A., Pires, O.R., Sebben, A., Castro, M.S., Sousa, M. Pleurodema thaul (Amphibia, Anura, Leptodactylidae): an ultrastruc- V., Fontes, W., Schwartz, E.N.F., 2003. Composition of indolealkyl- tural study. Rev. Bras. Biol. 56, 451–457. amines of Bufo rubescens cutaneous secretions compared to six other Tyler, M.J., Burton, T., Bauer, A.M., 2001. Parotid or parotoid: on the Brazilian bufonids with phylogenetic implications. Comp. Biochem. nomenclature of an amphibian skin gland. Herpetol. Rev. 32, 79–81. Physiol 134, 641–649. Vital Brazil, O., Vellard, J., 1925. Contribuiça˜o ao estudo do veneno de McDowell, E.M., Trump, B.F.,1976. Histologic fixatives suitable for diagnostic batrachios do geˆnero Bufo. Braz. Med 2, 175–180. light and electron microscopy. Arch. Pathol. Lab. Med 100, 405–414. Vital Brazil, O., Vellard, J., 1926. Contribuiça˜o ao estudo do veneno de Nosi, D., Terreni, A., Alvarez, B.B., Delfino, G., 2002. Serous gland poly- batrachios. Mem. Inst. Butantan 3, 7–70. morphism in the skin of Phyllomedusa hypochondrialis azurea Wilber, C.G., Carroll, P.L., 1940. Studies on the histology of the glands in (Anura, Hylidae): response by different gland types to norepinephrine the skin of Anura. I. The parotoid gland of Bufo americanus Holbrook. stimulation. Zoomorphology 121, 139–148. Trans. Am. Microsc. Soc 59, 123–128.