Infection Dynamics of the Lungworm Rhabdias Pseudosphaerocephala in Its Natural Host, the Cane Toad (Bufo Marinus), and in Novel Hosts (Native Australian Frogs)

Journal of Wildlife Diseases, 46(4), 2010, pp. 1152–1164 # Wildlife Disease Association 2010

INFECTION DYNAMICS OF THE LUNGWORM RHABDIAS PSEUDOSPHAEROCEPHALA IN ITS NATURAL HOST, THE CANE TOAD (BUFO MARINUS), AND IN NOVEL HOSTS (NATIVE AUSTRALIAN FROGS)

Ligia Pizzatto,1,3 Catherine M. Shilton,2 and Richard Shine1 1 School of Biological Sciences A08, University of Sydney, NSW 2006, Australia 2 Berrimah Veterinary Laboratories, Northern Territory Government, Berrimah, NT 0801, Australia 3 Corresponding author (email: [email protected]) Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021

ABSTRACT: Host-parasite systems have often evolved over time, such that infection dynamics may become greatly modified from the time of initial contact of the host with the parasite. Biological invasions may be useful to clarify processes in the initial contact of hosts with parasites, and allow us to compare parasite uptake between the ancestral (coevolved) host and novel (noncoevolved) hosts. Cane toads (Bufo marinus) are spreading rapidly through tropical Australia, carrying with them a nematode lungworm (Rhabdias pseudosphaerocephala) congeneric with those found in Australian frogs. We investigated the dynamics of infections of the toad parasite by conducting histologic examinations of cane toads and three native Australian frogs (Litoria dahlii, Litoria nasuta, and Opisthodon ornatus) at 2, 6, and 10 days after experimental exposure to the toad lungworm. More worms were found in toads than in frogs, especially at longer periods postexposure. In toads, the infective larvae entered the skin and muscles within 2 days postexposure, passed into the coelom in 6 days, and reached the lungs at 10 days. In frogs, larvae were found in many organs rather than migrating to consistent target tissues; a few larvae reached the lungs of L. dahlii. Migratory larvae caused increasing inflammation (primarily granulomatous admixed with granulocytes then lymphocytes) through time, especially in frogs. Evolution has resulted in an enhanced ability of the lungworm to locate the target organ (the lungs) of the toad, and an increase in rates of parasite survival within this host. Key words: Anura, host switch, immunologic response, inflammation, migratory pathway, nematode, parasite.

INTRODUCTION environment along with their hosts. For example, the microsporidian parasite Fi- Invasive species can exert an impact on brillanosema crangonycis was introduced native taxa via many pathways, including to the UK with its host, the amphipod predation, competition, and habitat mod- Crangonyx pseudogracilis (Galbreath et ification (Mack et al., 2000; Sakai et al., al., 2004). Such parasites may be useful for 2001). One of the most important but controlling their host (alien) populations, poorly understood processes may be the but if the parasites are not host specific, introduction of novel pathogens into they can also pose a threat to the native highly susceptible populations (Prenter et fauna by switching to other hosts (Prenter al., 2004), a process responsible for the et al., 2004; Taraschewski, 2006). For immense impact of European colonization example, a pathogenic protozoan (the on the indigenous human population of rosette agent) was introduced with an many Pacific (Willcox, 1980) and South AsiancyprinidfishtoEuropeandis American (Willcox, 1980; Hopkins, 1983; threatening an endangered native fish Rothschild, 2005) countries. Although (Gozlan et al., 2005). The cattle and sheep accidents of the introduction process often liver fluke is native to the Palearctic region result in invasive species arriving in their (see Taraschewski, 2006) but has been new range without many of their native- introduced to North, Central, and South range parasites (Mitchell and Power, 2003; America, Africa, Australia, and New Zea- Torchin et al., 2003), there are many cases land, where it has switched hosts to native of parasites being introduced to a novel marsupials (Spratt and Presidente, 1981),

1152 PIZZATTO ET AL.—NEMATODE INFECTION DYNAMICS IN NATURAL AND NOVEL HOSTS 1153 rodents (Fuentes et al., 1997), and even the lungs during larval migration? More humans (Esteban et al., 1999). Its patho- generally, how do infection dynamics of genicity in some new hosts can be as high the parasite in a coevolved (natural) host as in the original hosts (Viggers and Spratt, (the cane toad) differ from those in novel 1995). hosts (Australian frogs)? To answer these When the cane toad (Bufo marinus,or questions we looked at a system where Rhinella marina in an alternative nomen- toads have arrived within the last few clatural scheme; Pramuk, 2006) was intro- years, leaving little opportunity for coevo- duced to Australia in 1935, it brought with lution. it the rhabditoid lungworm Rhabdias Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021 pseudosphaerocephala (Dubey and Shine, MATERIALS AND METHODS 2008). These parasites are found in cane We used three species of abundant non- toads across most of the toads’ Australian arboreal frogs: Litoria dahlii, Litoria nasuta range (Barton, 1998), but have not been (family Hylidae), and Opisthodon ornatus reported in Australian frogs, which are (family Limnodynastidae), as well as the cane infected with other Rhabdias species toad, B. marinus (family Bufonidae). The (Dubey and Shine, 2008). Rhabdias pseu- anurans were collected as eggs in ponds near Fogg Dam (12.6uS, 131.3uE) about 60 km east dosphaerocephala have a direct life cycle, of Darwin, in the wet-dry tropics of the similar to other species in the genus. The Northern Territory, Australia, where the first larvae (L3) infect the host via cutaneous cane toads arrived in early 2005. Tadpoles penetration and migrate through the body were raised outdoors in plastic containers to the lungs, where they develop into (1103110364 cm; each species in a different container) containing 650 l of nonchlorinated hermaphroditic adults. The adult nema- water, fitted with screened lids. Tadpoles were todes lay eggs, which are released into the fed on frozen lettuce, fish flakes (TetraminH lungs and pass through the intestines. In Tropical Flakes, Blacksburg, Virginia, USA) and pellets for algae eaters (HBH Veggie the intestines, the embryonated eggs hatch TM into (L1) larvae (Baker, 1979), which are Wafers , Springville, Utah, USA). After completing metamorphosis, nine individuals usually first seen in the toad metamorph’s of each species were placed in a single feces after 30 days of infection (C. container (as above) that was placed over a Kelehear, unpubl. data [University of plastic tray containing iodine scrub (BetadineH Sydney]; Pizzatto, pers. obs.). After being Surgical Scrub; 7.5% weight/volume povi- defecated, L1 larvae develop into L3 done-iodine) to avoid contamination and kept indoors under ultraviolet light (natural light/ larvae or free-living dioecious adults that dark cycles), at 28.262.68 C. Each container produce L3 larvae (Baker, 1979). In had a substrate of preboiled sand and tree bark anurans, migrating rhabditoid larvae can (for refuge) and a pool of nonchlorinated induce granulomatous inflammation, and water. They were fed ad libitum on termites adult worms in the lung can cause and small cockroaches every third day throughout the experiment. In other work, inflammation and scarring (Williams, we have found that this maintenance environ- 1960; Green, 2001; Ladds, 2009). ment for the eggs, tadpoles, and metamorphs Infection with R. pseudosphaerocephala enables us to eliminate any accidental contact can reduce the fitness (survival, locomotor of our study animals with Rhabdias sp. performance, food intake, and growth (Pizzatto, unpubl. data). On 30 May 2009, cultured infective larvae of rates) of metamorphs of its natural host, R. pseudosphaerocephala (about 3,000) were the cane toad (Kelehear et al., 2009), but added to the container. These larvae originat- their interactions with native frogs remain ed from the eggs of three–four adult lung- unclear. Does the lack of records of R. worms from local cane toads, smeared in a pseudosphaerocephala in Australian frogs small amount of toad feces (for details on culture preparation see Kelehear et al., 2009). (Dubey and Shine, 2008) reflect a lack of Because metamorphs were obtained as eggs infection, or an inability of the lungworms and reared in uncontaminated conditions, any to survive in nontoad hosts, or to locate nematodes infecting the study animals were 1154 JOURNAL OF WILDLIFE DISEASES, VOL. 46, NO. 4, OCTOBER 2010 assumed to be R. pseudosphaerocephala (con- proximal humerus or femur distally), including firmed by DNA sequence analysis in compan- feet, were rarely visualized in section (total of six ion studies using the same methods of distal front leg sections, and we were unable to infection; L. Pizzatto, unpubl. data). The identify front or hind leg section in six frogs). duration of the experiment (10 days) was not Distal limb portions were generally lost in long enough to allow the larvae to develop into processing because they were very small adults in the lungs of metamorphs (minimum segments of tissue that were isolated after being is 30 days: Pizzatto, pers. obs.). Thus, the severed from the rest of the body during original infective dose was the only contact transverse sectioning for histology. Because of between hosts and parasites. Larvae usually this paucity of distal limb sections, and the lack survive for up to 7 days under the conditions in of larvae in any of these sections, data on distal the containers (Pizzatto, pers. obs.); therefore limbs were not included in the analysis. Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021 hosts were potentially exposed to infective For subadults free in the lumen of a lung, the larvae during the entire period. Because all total number of sections containing lungworms anurans were housed together, all individuals was used to quantify the number of individuals had the same exposure to parasites. in the lung. Our method of counting subadults At 2, 6, and 10 days postexposure (DPE) to may overestimate worm numbers if these the infective larvae, three individuals per worms are large enough to appear in multiple species were euthanized by immersion in 4 g/l sections, but provides an index to compare buffered tricaine methanesulfonate bath adult worm burdens among the four species. (MS222; Sigma-Aldrich Co., St. Louis, Mis- Inflammation associated with migrating souri, USA) and preserved in 10% neutral larvae was characterized by estimating pro- buffered formalin. For preparation of histologic portions of inflammatory cell types. Amphib- slides, metamorphs were decalcified in 10% ian species may exhibit subtle differences in formic acid/10% formalin solution for 48 hr. morphology of inflammatory (white blood) cell The head and body were cut in 1–2-mm serial types, particularly with respect to granulocytes transverse sections resulting in five–eight trans- (Wright, 2001), and histology does not reveal verse sections per toad. Tissues were sectioned sufficient morphology of granulocytes to char- into 5-mm sections and then stained with acterize these cells in detail for each of the hematoxylin and eosin. Slides were examined species in this study. We therefore limited our by one observer, blinded as to the specifics of terminology for granulocytes to neutrophils (to the slide (species and duration of exposure). indicate cells with segmented nuclei and a The number of worms (usually larvae, except moderate amount of pale eosinophilic non- for subadults in the lungs) and their location granulated cytoplasm), and eosinophils (to were recorded for each anuran. Multiple indicate cells approximately the same size as sections of larvae in close proximity encom- neutrophils but with round or slightly seg- passed by a single focus of inflammation were mented nucleus and eosinophilic cytoplasmic assumed to be multiple sections through one granules of any shape or size). Macrophages larva. Larval locations were initially described were identified in all species as cells that were by both body region (head, thorax, abdomen, larger than granulocytes, with abundant, pale, pelvis, limb) and tissue type or organ (subcuta- eosinophilic, slightly foamy cytoplasm and neous tissue, skeletal muscle or fascia, coelom or round or ovoid nucleus. Lymphocytes were coelomic membranes, eye or periocular tissue, identified in all species as relatively small cells lung). To simplify presentation of data, tissue having scant cytoplasm and small, round, dark types and organs were amalgamated into five nuclei with clumped chromatin. Severity of categories based on predilection of larvae for inflammation was scored as zero if the larva the tissues and the ability to encompass all was not associated with any inflammatory cells; larvae visualized. The categories were SM/SQ or mild, moderate, or severe if surrounded by body: skeletal muscle, fascia or subcutaneous one–two, three–four, or more than four layers tissue of the thoracic, abdominal, or pelvic of inflammatory cells, respectively. Larvae regions, including proximal humerus and shoul- were assessed as intact or degenerate based der and proximal femur and pelvis areas; head: on whether the larvae appeared whole and not any tissue in the head except the eye or short infiltrated by inflammatory cells (intact) or neck region (anterior to the start of the thoracic fragmented and infiltrated or partially degrad- coelom); eye: the eye or periocular tissue such ed by inflammatory cells (degenerate). as conjunctivae or retro bulbar muscles; coelom: larvae free in the coelomic space or embedded Data analyses in the coelomic lining, mesenteries, or serosae of viscera other than the lung; lung: subadults We used one-way analysis of variance within the lung. Distal portions of limbs (from (ANOVA) to compare the intensity of parasit- PIZZATTO ET AL.—NEMATODE INFECTION DYNAMICS IN NATURAL AND NOVEL HOSTS 1155

toads and L. dahlii, with a trend toward more larvae through time (an average 10– 16 larvae per individual at 6 and 10 DPE). The two other species of native frogs, L. nasuta and O. ornatus, had lower intensities of infection, especially at 10 DPE (ANOVA—6 DPE: F3,1151.63, P50.258; 10 DPE: F3,1155.40, P50.016; Fig. 1).

Larval location Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021 The proportion of larvae inside the host’s body showed significant or near- significant differences among host species

FIGURE 1. Mean6SE number of lungworms per (F3,7852.67, P50.054), DPE (F1,7854.05, anuran (host) species at postexposure times, seen in P50.048), and tissue infected (F3,7852.5, histologic slides. P50.067), suggesting differences in migratory pathways. In toads at 2 and 6 ism among species, both by period postexpo- DPE, larvae were present primarily in the sure and with all periods combined; three- subcutaneous tissue and muscle of the factor ANOVA to compare the proportion of body, and tissues of the head and eye larvae per tissue, per species, per DPE; the proportion of inflammatory cases per host (Fig. 2A). By 6 and 10 DPE there were a species, per DPE, per inflammation severity; few larvae in the coelom, and at 10 DPE the proportion of cells involved in the subadults in the lung predominated, and inflammatory response in different hosts, were present in all toads. One species of DPE, and type of cells; and two-factor native frog, L. dahlii, followed a similar ANOVA to compare the proportion of intact worms in different host species and DPE. All pattern to toads: Larvae were primarily in proportions were arcsine-square-root trans- the muscle and subcutaneous tissue of the formed prior to analysis. Means are presented body and tissues of the head at 2 and 6 with standard error. DPI, and by 10 DPI some larvae were in the coelom and subadults were in the RESULTS lungs of all frogs. However, this species Intensity of parasitism differed from toads in that more larvae were found in the eye, and fewer reached Infective R. pseudosphaerocephala lar- the lung (Fig. 2B). For the other two vae successfully located and penetrated species, L. nasuta (Fig. 2C) and O. the bodies of cane toads and native frogs, ornatus (Fig. 2D), the pattern of larval although with different intensities, de- migration was less consistent, with larvae pending on the host species (mean num- found in similar proportions in the muscle ber of worms: toads, 7.262.1; L. dahlii, and subcutaneous tissue of the body, head, 8.061.5; L. nasuta, 2.460.83; O. ornatus, eye, and coelom, and subadults never 3.561.4; Fig. 1). At 2 DPE, all species had appearing in the lung. Larvae in O. low intensities of parasitism (ANOVA—2 ornatus were less likely to migrate through DPE: F3,1150.57, P50.650). Litoria na- the eye region. suta had the lowest infection intensity at In all anuran species, most larvae in this period, and levels in this species subcutaneous tissues were immediately remained low throughout the experiment, beneath the skin in the loose connective similar to O. ornatus (ANOVA—all DPE tissue between the skin and the underlying combined F3,3552.96, P50.045). At all skeletal muscle. Most larvae in the skeletal postexposure periods, migrating larvae muscle were within or immediately be- were found in similar numbers in cane neath the superficial fascia (epimysium) of 1156 JOURNAL OF WILDLIFE DISEASES, VOL. 46, NO. 4, OCTOBER 2010 Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021

FIGURE 2. Mean percentage of lungworms seen in histologic sections of body areas of (A) cane toads, and frogs Litoria dahlii (B), Litoria nasuta (C), and Opisthodon ornatus (D). SM/SQ BODY5larvae in the skeletal muscle or subcutaneous tissue, HEAD5larvae in any of the tissues of the head/neck except those associated with the eye, EYE5larvae in the eye or periocular tissues, COELOM5larvae within the coelom or coelomic membranes, LUNG5subadults in the lung. the muscle, rather than embedded deep free in the inner ear cavity (one toad at 6 within the muscle. In the anuran’s head, DPE), one larva in a cranial nerve and one larvae occurred along fibrous connective in the braincase exterior to the dura mater tissue planes between muscles, nerves, (in a L. dahlii at 10 DPE). All of these bones, and cartilaginous structures, rather larvae are included in the data summaries than within these tissues. Most larvae in in the ‘‘head’’ category. No larvae were the eye category were in periocular tissue, seen in the parenchyma of the brain, such as conjunctiva, extraocular muscles, spinal cord, or thoracic or abdominal or in the connective tissue between the viscera (heart, liver, spleen, kidneys). sclera and the skin, oral mucosa, and braincase that together form the orbital Severity and type of inflammatory response cavity. A few larvae penetrated the globe Inflammatory response depended on to be embedded in the sclera or beneath the host species (F3,6353.22, P50.028), the retina. No larvae were found free in DPE (F2,6256.20, P50.003) ,and severity the chambers of the eye or associated with of inflammation (F3,6353.24, P50.031). In the cornea. Other locations of larvae all species, the severity of inflammation included three embedded in the tongue associated with migrating larvae increased muscle (one toad at 2 DPE), one larva in a over the 10 days of the experiment, large peripheral nerve in the neck and one ranging from none or moderate inflam- PIZZATTO ET AL.—NEMATODE INFECTION DYNAMICS IN NATURAL AND NOVEL HOSTS 1157 Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021

FIGURE 3. Intensity of inflammation in response to migrating Rhabdias pseudosphaerocephala larvae. Inflammation severity ranged from absent to mild, moderate, or marked, the last three categories being based on presence of one–two, three–four, or more than four layers, respectively, of inflammatory cells surrounding a larva. mation at 2 DPE to moderate or marked embedded in aggregates of inflammatory inflammation at 6 and 10 DPE (Figs. 3, 4). cells (Fig. 4C). However, there were interspecific differ- With increasing DPE, migrating larvae ences. Toads at 2 and 6 DPE had more were less likely to appear intact cases of no or mild inflammation than did (F2,2855.46, P50.010). At 2 DPE, the the three native species (Fig. 3A vs. 4A). mean percentage of intact worms ranged In comparison, in L. dahlii (paradoxically, from 35% to 95% among species, whereas the native species most similar to toads, in this figure was less than 20% by 10 DPE. that larvae exhibited a consistent migra- Toads tended to have more intact larvae tion pattern and the worms were able to at 2 and 6 DPE than did the other reach the lung), there was a substantial species (Fig. 5), but differences among early inflammatory response. In this spe- species were not statistically significant cies, at 2 DPE most larvae were associated (F3,2851.82, P50.165). with moderate inflammation, increasing to The type of inflammatory infiltrate was marked inflammation at 6 DPE (Fig. 3B). similar among host species (F3,8350.207, At 2 DPE, the other two native species P50.891). At all stages, the inflammatory showed a level of inflammation interme- response was composed primarily of diate between the low level of toads and macrophages accompanied by moderate the relatively high levels of L. dahlii numbers of granulocytes (eosinophils and (Fig. 3C, D). Of all the species, L. nasuta neutrophils; DPE: F2,8350.48, P50.623; had the strongest inflammation associated cell type: F3,83569.6, P,0.0001). The only with larvae at 10 DPE, with larvae often notable interspecific difference was the 1158 JOURNAL OF WILDLIFE DISEASES, VOL. 46, NO. 4, OCTOBER 2010 Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021

FIGURE 5. Mean percentage of intact lungworms per anuran (host) species seen in histologic sections at different times postexposure. Worms were judged as intact versus degenerate on the basis of whether they appeared to be unfragmented and were not penetrated by inflammatory cells. Total represents the average of the percentage of intact worms considering all three postexposure times.

dominance of eosinophils as the granulocyte present in the inflammatory response in all native species, compared to toads in which neutrophils were the predominant granulocyte (data not shown). At 6 DPE and more notably at 10 DPE, low numbers of lymphocytes were evident within the inflammatory response (Fig. 6). Occasional multinucleated giant

FIGURE 4. (A) Intact nematode larvae (arrows) in the subcutaneous tissue and connective tissue spaces surrounding a large nerve in the caudal head and neck region, with no associated inflammation, in a cane toad 6 days postexposure. (B) Intact larva in the subcutaneous tissue of the head, adjacent to cartilage, with mild associated inflammation, low numbers of macrophages and eosinophils, in Opisthodon ornatus, 6 days postexposure. Note the rhabditiform esophagus with corpus (Co), isthmus (I), and bulb (Bu), typical of this group of nematodes. (C) Degenerate larva embedded in skeletal muscle, surrounded by marked inflammation consisting of macrophages with lesser numbers of eosinophils, in FIGURE 6. Mean percentage of inflammatory Litoria nasuta, 10 days postexposure. Refer to cells (see Methods for description of cell types) Methods for cell type descriptions. associated with the worms at different times postexposure. Because of the similarity in patterns, data for all species are combined for this figure. PIZZATTO ET AL.—NEMATODE INFECTION DYNAMICS IN NATURAL AND NOVEL HOSTS 1159

dahlii that had a subadult in the lung at 6 DPE.

DISCUSSION

Lung nematodes of the genus Rhabdias are widespread parasites of anurans around the world, with more than 40 known species. About 17 species are

reported for bufonids, of which 11 seem Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021 to be host specific to these toads (Kuzmin et al., 2007). In their native range (Central and South America), the cane toads are parasitized by three species of lungworms: R. pseudosphaerocephala, Rhabdias alabialis, and Rhabdias fuëlleborni (Kuzmin et al., 2007). Rhabdias fuëlleborni is a host generalist occurring in other toads (e.g., Bufo jimi: Anjos et al., 2008; Bufo schneideri: Gonzalez and Hamann, 2008; Cranopsis valliceps: Espıńola-Novelo and Guilleń-Hernańdez, 2008; Bufo ictericus: Luque et al., 2005) as well as frogs (e.g., Eleutherodactylus guentheri: Martins and de Fabio, 2005; Rana berlandieri: Leoń- FIGURE 7. (A) Degenerate subcutaneous larva surrounded by marked granulomatous inflammatory Re`gagnon et al., 2005; Taropa miliaris: response composed of primarily macrophages and Kloss, 1974). Rhabdias alabialis is cur- lesser numbers of lymphocytes in a toad at 10 days rently only known from cane toads (Kuz- postexposure. At three foci (arrows) the macrophages min et al., 2007) and R. pseudosphaeroce- have fused to form multinucleated giant cells. (B) Lung with mild interstitial pneumonia characterized phala has been recorded for B. marinus by infiltration with low numbers of eosinophils, and B. schneideri (see Dubey and Shine, macrophages, and lymphocytes into the septae. In 2008). Because of identification difficul- the pulmonary lumen, note the adult lungworm with ties and taxonomic instability, it is unclear anuran red blood cells within its intestine. See if native-range R. pseudosphaerocephala Methods for description of cell types. are restricted to toads belonging to the marinus group, or if they also occur in cells (formed by the fusion of multiple frogs (e.g., Rana berlandieri: Martıńez, macrophages) occurred, particularly at 10 1969 and Leoń-Re`gagnon et al., 2005; DPE, when the inflammatory response Atelopus bomolochus: Iannacone, 2003). was marked and the larvae degenerate In addition to the taxonomic confusion, (Fig. 7A). there have not been many studies con- In the lungs, the only notable finding ducted to investigate the effects of the was mild or patchy interstitial infiltration parasite on their hosts or to evaluate host with eosinophils or heterophils, macro- switching. phages, and lymphocytes. This change was Our experiments indicated that under present at 10 DPE in two toads and all L. laboratory conditions, R. pseudosphaero- dahlii (all of which had subadults in the cephala can infect Australian frogs; how- lungs: Fig. 7B), was noted in one L. ever, this phenomenon had not been nasuta at 2 DPE and one O. ornatus at observed in wild-caught frogs (Dubey 10 DPE, but was absent in the only L. and Shine, 2008; C. Kelehear and L. 1160 JOURNAL OF WILDLIFE DISEASES, VOL. 46, NO. 4, OCTOBER 2010

Pizzatto, unpubl. data). Although toad B. marinus with R. pseudosphaerocephala. lungworms located and penetrated the Within 24 hr the two infected toads had three novel host species tested, infective massive numbers of larvae in the mouth, larvae seemed to be ‘‘lost’’ inside the body esophagus, lung, deep in the heart, within in two of these frog species (L. nasuta and the liver, and within the vitreous of the O. ornatus), aberrantly migrating in the eye. However, Williams (1960) used a subcutaneous tissue, muscle, and connec- massive experimental dose, with hundreds tive tissue but rarely reaching the coelom of larvae in the infected toads at postmor- and never reaching the lungs. These two tem. frog species thus can be considered as One difference between Baker’s (1979) Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021 aberrant hosts, because the parasite failed study and our own is that in all of our to complete development inside them. In species, larvae had a tendency to enter and toads, the changing locations of migrating migrate in the head region, including larvae with DPE suggest a consistent surrounding the eyes, rather than the migratory pathway: They penetrate the pelvic or pectoral regions. This tendency skin and move through muscular tissues to is most obvious in toads at 6 DPE, where reach the coelom and then the lungs. more larvae were found in the head than Larvae in one native species (L. dahlii) in the rest of the toad, and in L. dahlii,in showed a pattern of migration similar to which at all stages postexposure, more that seen in cane toads. In all individuals larvae were present in the head than the of L. dahlii, adult worms eventually muscle or subcutaneous tissue of the body reached the lungs, albeit in lower numbers (Fig. 2A, B). The pectoral and pelvic than in toads (Fig. 2A). This frog therefore regions have extensive contact with the should be considered an accidental host, larvae-containing substrate, suggesting in which the worms are not commonly that larvae in our study moved over the found but can complete their develop- surface of the body until they reached the ment. periocular region before attempting to The migration route of R. pseudo- penetrate the skin. Direct observations sphaerocephala in cane toads is similar to confirm that the eyes are an important that described for Rhabdias americanus entry point for infective larvae in toad larvae in their natural host Bufo amer- metamorphs (Kelehear, 2007). The pref- icanus (Baker, 1979). Rhabdias ameri- erence of larval R. pseudosphaerocephala canus was found under the skin of for periocular entry might reflect moister American toads, particularly in the pelvic surfaces, or an easier point of entry and pectoral regions of the body, 1 day through the nonkeratinized conjunctive after exposure of the host. The larvae then epithelium of the eyes. moved through fascia around skeletal Another difference between Baker’s muscles of the limbs, pelvis, and pectoral (1979) results and our own was the region (days 3–4 postinfection), and thorax presence of only subadults or adults in and abdomen (days 5–6). At days 7–8 the coelom or lungs beyond 9 days post- postinfection, the worms were found in exposure, which may relate to differing the fascia close to the coelom and some methodology. In Baker’s study, toads were subadults were in the coelom and lungs exposed to larvae for only 4 hr, then (Baker, 1979). As in our study, Baker removed, whereas there was potential for found no evidence that larvae reached the continued exposure for the entire 10 days lungs via the blood circulatory system, a of the experiment in our study, which route suggested by other authors (Wil- likely resulted in larvae that were still liams, 1960; Ladds, 2009). migrating concurrent with adults in the Our results differ from those of Wil- lungs. Our data suggest that rates of larval liams (1960), who experimentally infected penetration and migration decreased after PIZZATTO ET AL.—NEMATODE INFECTION DYNAMICS IN NATURAL AND NOVEL HOSTS 1161 the first few days of exposure, because sponse towards a specific antigen (Cata˜o- inflammation around larvae increased Dias and Sinhorini, 1999; Wright, 2001; (Fig. 3) and the percentage of intact larvae Ackerman, 2007b; Snyder, 2007). The decreased (Fig. 5) with DPE. Overall, R. fusion of macrophages into multinucleated pseudosphaerocephala may present a giant cells, with marked inflammation much slower development than Rhabdias surrounding degenerate larvae, is a chron- ranae and R. americanus. Gravid females ic inflammatory response that may func- of the latter two species were found in the tion to break down, phagocytize, or wall lungs of their natural hosts at 9 days off large antigens or foreign material postinfection (and as early as 7 days in R. (Ackerman, 2007b). Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021 americanus: Baker, 1979), a period too One interesting interspecific difference brief for R. pseudosphaerocephala to in the inflammatory response was a mature (see Kelehear, 2007; Pizzatto, decreased intensity of the response in pers. obs.). the early DPE in toads (which had the In all host species in our study, the highest percentage of larvae with no or inflammatory response increased in sever- mild associated inflammation at 2 and 6 ity with DPE (Fig. 3), probably due to a DPE; Fig. 3A). The larvae appeared to longer duration for the inflammatory penetrate and migrate rapidly in toads, response to increase, plus adaptation in before the toad could mount an inflam- the immune response. The latter phenom- matory response, and also may be able to enon is suggested by the trend for initial, evade the toad’s inflammatory and innate acute, inflammatory responses to be lim- immune defenses more effectively than is ited to macrophages and granulocytes (in the case in a novel host (frogs). Nonethe- all species). Macrophages have diverse less, many larvae in the toads failed to roles in acute inflammation and innate reach the lungs before being engulfed by immunity (nonspecific early response to the host inflammatory or immune re- antigens) and chronic inflammations, sponse (Figs. 2A, 7). In contrast, Baker which often involve some degree of (1979) noted that at 3–4 days postinfec- adaptive immunity (specific response that tion, macrophages were attached to some develops with time following exposure to larvae that were not degenerate. Baker an antigen). Granulocytes are part of the (1979) also noted that after 9 days postin- innate immune system, coming in early to fection, all worms were present as adults respond nonspecifically to infectious and in the coelom or lung, with larvae no foreign agents by either phagocytosis of longer found migrating elsewhere. There- the antigen or degranulation and release fore, in Baker’s study (in which larvae of enzymes and other components into the were detected by dissection under a tissue in order to damage the foreign stereomicroscope, rather than histological- antigen, in the process contributing sig- ly), either all of the larvae eventually nificantly toward the acute inflammatory reached the lungs, or after 9 days, larvae response (Cannon et al., 1986; Wright, that had failed to reach the lungs were too 2001; Ackerman, 2007a; Snyder, 2007). degenerate, fragmented, or removed by an With chronicity, at 6 and 10 DPE, there inflammatory or immune response to be were still abundant macrophages, which recognizable. In the few toads experimen- release inflammatory cytokines to guide tally infected by Williams (1960), larvae the chronic inflammatory response and that did not reach the lungs became, in his initiate tissue repair, as well as assist words, ‘‘encysted’’ by the host reaction antigen processing for adaptive immunity. (presumably meaning isolated or walled By this time there were also lymphocytes, off rather than present in a true parasitic another part of the chronic inflammatory cyst). response and the adaptive immune re- Subadult worms in the lung induced 1162 JOURNAL OF WILDLIFE DISEASES, VOL. 46, NO. 4, OCTOBER 2010 mild, interstitial pneumonia with eosino- and the toads also acquired R. ranae from phils in all but one of the toads and L. the frogs (Baker, 1979). However, in both dahlii that had adults in the lung at 10 cases the intensity of parasitism in the new DPE. However, given that this change was hosts was lower than in the original hosts mild and also found in a few other under identical conditions (Baker, 1979). experimental animals with no worms in In Bermuda, Williams (1960) transmitted the lungs, the evidence is equivocal. There R. pseudosphaerocephala larvae obtained was no bronchopneumonia (inflammatory from B. marinus feces to a noneunatural exudate within the lumen of the lung) in host (the tree frog, Eleutherodactylus any of the individuals with adults in the johnstonei) but far fewer larvae penetrated Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021 lung, and the lungs were not obviously the tree frog, than the toad, when exposed scarred or hemorrhagic. However, this to a similar dose. These results also accord lack of pathology could be due to the early well with the hypothesis that coevolution- stage of infection, and inflammation could ary history shapes host-parasite interac- intensify with continued damage to the tions. lung via hematophagous feeding by the In conclusion, the toad lung nematode, parasite and irritation to the lung as gravid R. pseudosphaerocephala, can successfully females release eggs/larvae into the lu- enter the body of Australian native frogs, men. In Rhabdias bufonis infecting B. causing inflammatory response and even- marinus, the primary damage inflicted by tually reaching the lungs of some species. the parasite comes from worms penetrat- At present, there is no evidence that these ing the skin and migrating, rather than infections happen in natural conditions effects of adults in the lungs (Tinsley, (Dubey and Shine, 2008; C. Kelehear and 1995). L. Pizzatto, unpubl. data), and their Larval migratory pathways and inflam- physiologic and ecologic impacts are still matory responses to parasite exposure under investigation. differed among the frog species, and between frogs and toads. These differenc- ACKNOWLEDGMENTS es may be due to specific aspects of the We thank Crystal Kelehear and Greg Brown hosts’ physiology and ecology. For exam- for sharing their knowledge of host-parasite ple, the species we used differ in habitat biology and comments on the manuscript, use (L. nasuta is terrestrial, L. dahlii Christine Rioux for technical expertise in aquatic, O. ornatus is a burrower), and preparing histology slides, Michael Crossland for providing the clutches and caring for the this trait is likely to affect exposure rate. tadpoles, and Peter Hopkinson and the The lack of shared evolutionary history Coastal Plains Research Farm (DPI-NT) staff between the toad lungworm and Austra- for logistic support. The work was funded by lian frogs may be the primary reason why the Australian Research Council and by the lungworms in our study fared better in Western Australian Department of Environ- ment and Conservation. All procedures and toads than in frogs: Parasites tend to be animal use were approved by the University of more capable of exploiting hosts with a co- Sydney Animal Ethics Committee (Protocol evolutionary history (Poulin, 2007). Ex- L04/4/2008/2/4788), and the NT Parks and periments involving cross-infections of Wildlife Commission (Permit 36597). Rhabdias spp. in toads suggest that some species of the parasite are specific to their LITERATURE CITED hosts; species recovered from toads of the ACKERMAN, M. R. 2007a. Acute inflammation. In marinus group do not establish infections Pathologic basis of veterinary disease, M. D. in toads from the crucifer group, or vice McGavin and J. F. Zachary (eds.). Mosby Elsevier, St. Louis, Missouri, pp. 101–154. versa (Kloss, 1974). In other experiments, ACKERMAN, J. T. 2007b. Chronic inflammation and Rana sylvatica was successfully infected wound healing. In Pathologic basis of veterinary with R. americanus from American toads disease, M. D. McGavin and J. F. Zachary (eds.). PIZZATTO ET AL.—NEMATODE INFECTION DYNAMICS IN NATURAL AND NOVEL HOSTS 1163

Mosby Elsevier, St. Louis, Missouri, pp. 154– GREEN, D. E. 2001. Pathology of Amphibia. In 191. Amphibian medicine and captive husbandry, K. ANJOS, L. A., L. E. M. SILVA,W.O.ALMEIDA,J.G.M. M. Wright and B. R. Whitaker (eds.). Krieger, COSTA, AND A. VASCONCELLOS. 2008. Chaunus Malabar, Florida, pp. 402–485. jimi (NCN). Endoparasites. Herpetological Re- HOPKINS, D. R. 1983. Princes and peasants: Smallpox view 39: 337. in history. University of Chicago Press, Chicago, BAKER, M. R. 1979. The free-living and parasitic Illinois. pp. 384. development of Rhabdias spp. (Nematoda: IANNACONE, J. 2003. Helmintos para´sitos de Atelopus Rhabdiasidae) in amphibians. Canadian Journal bomolochus Peters 1973 (Anura: Bufonidae) de of Zoology 57: 161–178. Piura, Peru. Gayana 67: 9–15. BARTON, D. P. 1998. Dynamics of natural infections KELEHEAR, C. 2007. The effects of lung nematodes

of Rhabdias cf. hylae (Nematoda) in Bufo (Rhabdias cf. hylae) on metamorph cane toads Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021 marinus (Amphibia) in Australia. Parasitology (Chaunus marinus), and implications for biolog- 117: 505–513. ical control. BSc Honours Thesis, University of CANNON, M. S., H. W. SAMPSON, AND E. D. KAPES. Sydney, NSW, Australia. 1986. The blood leukocytes of Bufo marinus:A ———, J. K. WEBB, AND R. SHINE. 2009. Rhabdias light, phase-contrast, and histochemical study. pseudosphaerocephala infection in Bufo mari- Canadian Journal of Zoology 65: 1445–1453. nus: Lung nematodes reduce viability of meta- CATA˜ O-DIAS, J. L., AND I. L. SINHORINI. 1999. morph cane toads. Parasitology 136: 919–927. Influence of low environmental temperature on KLOSS, G. R. 1974. Rhabdias (Nematoda, Rhabditoi- inflammation in bullfrog (Rana catesbeiana): dea) from the marinus group of Bufo. A study of Qualitative and quantitatie evaluation. Brazilian sibling species. Arquivos de Zoologia 25: 61–120. Journal of Veterinary Research and Animal KUZMIN, Y., V. V. TKACH, AND D. R. BROOKS. 2007. Science 36, http://www.scielo.br/scielo.php?script5 Two new species of Rhabdias (Nematoda: sci_arttext&pid5S1413-95961999000200004&lng5 Rhabdiasidae) from the marine toad, Bufo en&nrm5iso. doi: 10.1590/S1413-95961999000200004. marinus (L.) (Lissamphibia: Anura: Bufonidae), DUBEY, S., AND R. SHINE. 2008. Origin of the parasites in Central America. Journal of Parasitology 93: of an invading species, the Australian cane toad 159–165. (Bufo marinus): Are the lungworms Australian or LADDS, P. 2009. Pathology of Australian native American? Molecular Ecology 17: 4418–4424. wildlife. CSIRO, Collingwood, Victoria, Austra- ESPIŃOLA-NOVELO, J., AND S. GUILLE´ N-HERNA´ NDEZ. lia. 648, pp. 2008. Helminth parasites in Chaunus marinus LEO´ N-RE` GAGNON, V., E. A. MARTıŃEZ-SOLAZAR,D. and Cranopis valliceps (Anura: Bufonidae) from LOZCANO-VILLAREAL, AND R. ROSAS-VALDEZ. 2005. Lagunas Yalahau, Yucatan, Mexico. Journal of Helminth parasites of four species of anurans Parasitology 94: 672–674. from Nuevo Leon, Mexico. Southwestern Natu- ESTEBAN, J. G., A. FLORES,R.ANGELES, AND S. MAS- ralist 50: 251–258. COMA. 1999. High endemicity of human fascio- LUQUE, J. L., A. N. MARTINS, AND L. E. R. TAVARES. liasis between Lake Titicaca and La Paz Valley, 2005. Community structure of metazoan para- Bolivia. Transactions of the Royal Society of sites of the yellow Cururu toad, Bufo ictericus Tropical Medicine and Hygiene 93: 151–156. (Anura, Bufonidae) from Rio de Janeiro, Brazil. FUENTES, M. V., J. R. COELHO, AND M. D. BARGUES. Acta Parasitologia 50: 215–220. 1997. Small mammals (Lagomorpha and Roden- MACK, R. N., D. SIMBERLOFF,W.M.LONSDALE,H. tia) and fascioliasis transmission in the northern EVANS,M.CLOUT, AND F. BAZZAZ. 2000. Biotic Bolivian Altiplano endemic zone. Research and invasions: Causes, epidemiology, global conse- Reviews in Parasitology 57: 115–121. quences and control. Issues in Ecology 5: 1–20. GALBREATH, J. G. M. S., J. E. SMITH,R.S.TERRY,J.J. MARTIŃEZ, V. J. M. 1969. Para´sitos de algunos BECNEL, AND A. M. DUNN. 2004. Invasion success anfibios colectados en diferentes a´reas de los of Fibrillanosema crangonycis, n. sp., n. g.: A municipios de Escobedo, Pesquerıá y Santiago, novel vertically transmitted microsporidian par- Nuevo Leoń, Me´xico. Tesis de Licenciatura. asite from the invasive amphipod host Crango- Universidad Nacional Autońoma de Nuevo nyx pseudogracilis. International Journal for León, Monterrey, Me´xico, 53 pp. Parasitology 34: 235–244. MARTINS, A. N., AND S. P. DE FABIO. 2005. Parasitism GONZALEZ, C. E., AND M. I. HAMANN. 2008. Nema- by nematodes in sympatric populations of tode parasites of two anuran species Rhinella Eleutherodactylus parvus (Girard, 1853) and schneideri (Bufonidae) and Scinax acuminatus Eleutherodactylus guentheri (Steindachner, (Hylidae) from Corrientes, Argentina. Revista de 1864) - (Anura: Leptodactylidae). Acta Biologica Biologıá Tropical 56: 2147–2161. Leopoldensia 27: 47–50. GOZLAN, R. E., S. ST-HILAIRE,S.W.FEIST,P.MARTIN, MITCHELL, C. E., AND A. G. POWER. 2003. Release of AND M. L. KENT. 2005. Disease threat to invasive plants from fungal and viral pathogens. European fish. Nature 435: 1046. Nature 421: 625–627. 1164 JOURNAL OF WILDLIFE DISEASES, VOL. 46, NO. 4, OCTOBER 2010

POULIN, R. 2007. Evolutionary ecology of parasites. tralian Journal of Experimental Biology and 2nd Edition. Princeton University Press, Prince- Medical Science 59: 713–721. ton, New Jersey, 342 pp. TARASCHEWSKI, H. 2006. Hosts and parasites as aliens. PRAMUK, J. B. 2006. Phylogeny of South American Journal of Helminthology 80: 99–128. Bufo (Anura: Bufonidae) inferred from com- TINSLEY, R. 1995. Parasitic disease in amphibians: bined evidence. Zoological Journal of the Control by the regulation of worm burdens. Linnean Society 146: 407–452. Parasitology 111: S153–S178. PRENTER, J., C. MACNEIL,J.T.A.DICK, AND A. M. TORCHIN,M.E.,K.D.LAFFERTY,A.P.DOBSON,V.J. DUNN. 2004. Roles of parasites in animal MCKENZIE, AND A. M. KURIS. 2003. Introduced invasions. Trends in Ecology and Evolution 19: species and their missing parasites. Nature 421: 628. 385–390. VIGGERS, K. L., AND D. M. SPRATT. 1995. The parasites

ROTHSCHILD, B. M. 2005. History of syphilis. Clinical recorded from Trichosurus species (Marsupialia: Downloaded from http://meridian.allenpress.com/jwd/article-pdf/46/4/1152/2239199/0090-3558-46_4_1152.pdf by guest on 01 October 2021 Infectious Diseases 40: 1454–1463. Phalangeridae). Wildlife Research 22: 311–332. SAKAI, A. K., F. W. ALLENDORF,J.S.HOLT,D.M. WILLCOX, R. R. 1980. Venereal diseases in the Pacific LODGE,J.MOLOFSKY,K.A.WITH,S.BAUGHMAN, Islands. Papua New Guinea. British Journal of R. J. CABIN,J.E.COHEN,N.C.ELLSTRAND,D.E. Venereal Diseases 56: 277–281. MCCAULEY, P. O’NEIL,I.M.PARKER,J.N. WILLIAMS, R. W. 1960. Observations on the life THOMPSON, AND S. G. WELLER. 2001. The history of Rhabdias sphaerocephala Goodey, population biology of invasive species. Annual 1924 from Bufo marinus L., in the Bermuda Review of Ecology and Systematics 32: 305–332. Islands. Journal of Helminthology 34: 93–98. SNYDER, P. W. 2007. Diseases of immunity. In WRIGHT, K. M. 2001. Amphibian haematology. In Pathologic basis of veterinary disease, M. D. Amphibian medicine and captive husbandry, K. McGavin and J. F. Zachary (eds.). Mosby M. Wright and B. R. Whitaker (eds.). Krieger, Elsevier, St. Louis, Missouri, pp. 193–251. Malabar, Florida, pp. 129–146. SPRATT, D. M., AND P. J. A. PRESIDENTE. 1981. Prevalence of Fasciola hepatica infection in Submitted for publication 3 March 2010. native mammals in southeastern Australia. Aus- Accepted 28 May 2010.