Myxosporean Parasites in Australian Frogs and Tadpoles

Home , Agile frog, Allobates marchesianus, Booroolong Frog, Cane toads in Australia, Clinus cottoides, Dendropsophus brevifrons

Ashlie Hartigan

B Sc. (Zoology), M App Sci (Wildlife Health and Pop. Mgmt)

Thesis by published works

This thesis is submitted in full satisfaction of the requirements for the degree of Doctor of Philosophy at the University of Sydney.

Faculty of Veterinary Science University of Sydney

March 2012

Statement of authentication

This thesis is submitted to the University of Sydney in fulfilment of the

requirement for the degree of Doctor of Philosophy.

The work presented in this thesis is, to the best of my knowledge and belief,

original except as acknowledged in the text. I hereby declare that I have not

submitted this material, either in full or in part, for a degree at this or any other

institution.

Signed:

Date:

ii Abstract

The investigation of new threats to amphibian conservation is a priority of researchers and wildlife managers. Emerging infectious diseases are one of the most threatening processes to wildlife around the world including amphibians. Australian frogs have suffered large scale declines and extinctions from pathogens such as chytrid fungus (Batrochochytrium dendrobatidis). The once overly abundant Green and golden bell frog (Litoria aurea) has declined over 90% of its range with disease listed as a key threat. A routine pathogen screen of tadpoles from a captive breeding population of Green and golden bell frogs found an unknown parasitic infection in the brains, bile ducts and gallbladders of tadpoles (later confirmed as

Myxosporea). It was this preliminary identification that was the impetus for my thesis.

Myxosporean parasites found in Australian frog gallbladders were thought to be Cystodiscus immersus from Central and South America. The parasite was assumed to have been introduced to

Australia with the Cane toad (Rhinella marina, syn. Bufo marinus) in 1935. Cystodiscus immersus was supposed to have translocated with the Cane toad from native Brazil and was infecting a broad range of Australian frog species with no apparent host impact. The aim of this thesis was to challenge all aspects of this assumption, to establish the true myxosporean species present in Australian frogs, if it was an exotic infection, if it was an emerging pathogen and what threat did it pose to hosts.

The parasite found in Australian frogs and tadpoles was examined using histological and ultrastructural morphology as well as molecular identification. This revealed there was in fact two myxosporean species in multiple species of frogs and tadpoles. Sequencing of partial small

iii subunit and large subunit as well as the internal transcribed spacers (ITS1 and ITS2) of ribosomal

DNA regions confirmed these were distinct species (differing 9%, 7%, 34% and 37% at each region respectively). They were in fact separate species infecting multiple Australian host species, samples of C. immersus from Brazil were compared to establish if either of these was the exotic

C. immersus. We found instead that these were completely unrelated despite sharing similar spore morphology. In addition the Brazil material revealed the cryptic diversity of myxosporea in frog gallbladders around the world and the ambiguous identity of Cystodiscus immersus.

To determine if the parasites were emerging, the presence of these parasites in 130 years was established using voucher specimens. Four native species and the exotic Cane toad were examined for gallbladder myxospores from 1879 to 2009, the first positive was in 1966 in Green tree frog outside of Sydney. The prevalence of myxosporea increased in both native and exotic species over the last 40 years including the threatened Green and golden bell frog. The emergence of these parasites is a cause for concern, especially when found in endangered species.

Close scrutiny of these parasites prompted the formal description of Cystodiscus axonis and

Cystodiscus australis, resurrecting the original genus name proposed by Lutz in 1889. The description revealed diagnostic features previously unknown for frog gallbladder myxosporea species. Scanning electron microscopy revealed filiform polar appendages on the myxospores of

C. axonis, absent in C. australis. In addition, histological examination showed unique developmental stages in the nervous tissue of tadpoles and frog hosts infected with C. axonis, these stages are only found in hosts infected with this parasite. The use of development stages as a species diagnostic character has not been previously reported for any amphibian myxosporea as

iv yet. The genetic and morphological differences between C. axonis and C. australis prompted the development of a species specific multiplex PCR using ITS1, 5.8S and ITS2 ribosomal DNA.

Multifaceted diagnostic tools (morphology and species specific PCR) demonstrated significant disease associated with these pathogens in native and exotic frog species. Seven host species were morphologically examined for Cystodiscus infection showing endangered species to have significant disease associated with infection. Cystodiscus axonis was linked to neurological signs, haemorrhage, necrosis and gliosis in Southern bell frogs (Litoria raniformis), Green and golden bell frogs (Litoria aurea), Booroolong frogs (Litoria booroolongensis), and the recently re- discovered Yellow spotted bell frogs (Litoria castanea). Cystodiscus spp. infection in the bile ducts of tadpoles was associated with biliary hyperplasia, loss of hepatocytes and inflammation.

These lesions were statistically associated with the presence of infection in the Green and golden bell frog.

This research disproved an assumption about myxosporean parasites that had held for 25 years.

The disease found with Cystodiscus parasite infection in Australian frogs and tadpoles, as well as the previously unknown biodiversity in this cryptic species complex, highlight a number of areas in need of further research. This work has provided insight into the importance of multifaceted approaches to species identification and amphibian pathogen surveillance which uncovered a threat to endangered Australian amphibians, now listed as one of the key disease threats to frogs by the Department of Environment and Health (Australia) in 2011.

v Acknowledgements

This thesis has taken roughly three years to complete, during that time I had the opportunity to meet and collaborate with a number of fantastic people. My greatest gratitude is to my two supervisors Dr. Jan Šlapeta and Assoc. Prof. David Phalen. The two of you provided me with amazing support, wisdom and patience during the steepest learning curve of my life. Jan, you have been a constant source of inspiration and motivation as my mentor and I count myself lucky for the opportunity to have learnt from you. David, thank you for your endless patience, awful jokes and letting me soak up as much of your wisdom as I could during this time. I’ve appreciated the influence you both have had on my development as a researcher and the skills I’ve learned from you along the way. I’d like to thank Dr. Karrie Rose at the Australian Registry of Wildlife Health, she has been a champion for frog myxosporea since the beginning and her enthusiasm, collaboration and friendship have supported me immensely.

My PhD is the culmination of substantial international collaboration with the Institute of Parasitology, Biology Centre, Academy of Sciences of the Czech Republic. I was extremely lucky to work with Dr. Ivan Fiala during this time and I’d like to thank him for his expertise, helpful suggestions and support during my visits to České Budějovice. I am grateful to have had the opportunity to discuss myxozoa and gain valuable training from Dr. Iva Dyková whilst at the Institute.

None of this research would be possible without the support and unending patience of Mr. Lance Jurd. He has been a fountain of information and source of help to me during this time and has been absolutely essential to this project.

A number of people within the Faculty of Veterinary Science were vital to my sanity in my doctorate. I’d like to thank Mrs Denise McDonell and Dr. Graeme Browne for giving me invaluable advice and learning opportunities. Thanks to Dr. Jessica King and Ms Christie Foster for being great listeners and giving good advice and Ms Elaine Chew for her seemingly endless patience and knowledge on the histopathology.

vi I’m especially in debt to my friends and family who had to learn to love parasites and frogs by proxy. I am extremely grateful to Dr. Stephen Hoggard (Macquarie University) for his advice, encouragement and amazing friendship over the years. Thank you to Kristie Flannery who has been a fire under me, a comforting ear and an inspiration to me since I was five. I’d like to thank my sister Amber, Steph Snoyman, Rochanne Barratt, and Sarah Atkinson for understanding when I missed life whilst in the lab, for letting me vent and especially for giving me a good laugh when I needed it most. Thank you to my friend Ms Lauren Walsh (Rouge), who propped me up when I was down, fed me when I was hungry and listened when I needed it. A special mention to my Mum Pauline, I wouldn’t have done any of this without her support, motivation and inexhaustible faith in me.

Lastly, thanks to the frogs and tadpoles, without which I would be a parasitologist looking for a host.

vii

Publications

Hartigan. A, Peacock. L, Rosenwax. A, Phalen. D.N and Šlapeta. J, (2012), Emerging myxosporean parasites of Australian frogs take a ride with fresh fruit transport, Parasites & Vectors, 5: 208, 10.1186/1756-3305-5-208

Hartigan. A, Dhand. N, Rose. K, Šlapeta. J and Phalen. D.N, (2012) Comparative pathology and ecological implications of two myxosporean parasites in native Australian frogs and the invasive Cane toad, PLoS One, 7(10): e43780. 10.1371/journal.pone.0043780.

Hartigan. A, Sangster. C, Rose. K, Phalen. D.N and Šlapeta. J, (2012), Myxozoan parasite in brain of critically endangered frog [letter]. Emerging Infectious Diseases, 18 (4), 10.3201/eid1804.111606

Hartigan. A, Fiala. I, Dyková. I, Rose. K, Phalen. D and Šlapeta. J, (2012), New species of Myxosporea from frogs and resurrection of the genus Cystodiscus Lutz, 1889 for species with myxospores in gallbladders of amphibians, Parasitology, 139 , pp 478-496 10.1017/S0031182011002149

Hartigan. A, Fiala. I, Dyková. I, Jirků. M, Okimoto. B, Rose. K, Phalen. D.N and Šlapeta. J (2011), A suspected parasite spill-back of two novel Myxidium spp. (myxosporea) causing disease in Australian endemic frogs found in the invasive Cane toad . PLoS One 6 (4): e18871, 10.1371/journal.pone.0018871

King. J.S, McAllan. B, Spielman. D.S, Lindsay. S.A, Hůrková-Hofmannová. L, Hartigan. A, Al-Qassab. S.E, Ellis. J.T and Šlapeta. J, (2011), Extensive production of Neospora caninum tissue cysts in a carnivorous marsupial succumbing to experimental neosporosis, Veterinary Research, 42: (75), 10.1186/1297-9716-42-75

Hartigan. A, Phalen. D.N and Šlapeta, J. 2010: Museum material reveals a frog parasite emergence after the invasion of the Cane toad in Australia. Parasites & Vectors, 3 (1), 10.1186/1756-3305-3-50

Zhu. B.Y, Hartigan. A, Reppas. G, Higgins. D.P, Canfield. P.J, and Šlapeta. J, (2009), Looks can deceive: Molecular identity of an intraerythrocytic apicomplexan parasite in Australian gliders, Veterinary Parasitology, 159 (2): 105-111.10.1016/j.vetpar.2008.10.031

Awards and Grants

Australian Postgraduate Award (2009-2012)

ARC/NHMRC Network for Parasitology Travel Award (2009)

Eric Horatio Maclean Scholarship (2010)

JD Smyth Postgraduate Award (Australian Society for Parasitologists) (2011)

Best Student Presentation Award, Australian Society for Parasitologists conference (2011)

viii International conference presentations

Hartigan. A, Fiala. I, Dyková. I, O’Keefe. P, Jirků. M, Okimoto. B, Rose. K, Phalen. D and Šlapeta. J, Myxidium infections of Australian endangered frogs, ICOPA International Parasitology Conference, Melbourne, August 2010 (poster)

Hartigan. A, Fiala. I, Dyková. I, Rose. K, Phalen. D and Šlapeta. J, MYXOZOAN PARASITES AS AGENTS OF FROG DISEASE: The case of the green and golden bell frog, Emerging Amphibian Disease Conference, James Cook University Townsville, June 2010

Hartigan. A, Fiala. I, Dyková. I, Rose. K, Phalen. D and Šlapeta. J, Myxidium sp. threatening Litoria aurea recovery programs and Australian amphibian conservation, European Association of Fish Pathologists 2009 conference, Prague, September 2009

Invited presentations

Hartigan. A, Fiala. I, Dyková. I, Rose. K, Phalen. D and Šlapeta. J, MYXOZOAN PARASITES AS AGENTS OF FROG DISEASE: The case of the green and golden bell frog, Australian Registry of Wildlife Health Rounds, (online webinar), September 2010

Hartigan. A, Rose K, Phalen D and Šlapeta J, MYXOZOAN PARASITES AS AGENTS OF FROG DISEASE: The case of the green and golden bell frog, Australian Veterinary Association conference, (Invited presentation), Brisbane, May 2010

Other conference presentations

Hartigan. A, Fiala. I, Dyková. I, Rose. K, Phalen. D and Šlapeta. J, Emerging Myxosporean Parasites of Native Australian Frogs and the Invasive Cane Toad, Australian Society for Parasitologists Annual Conference, July 2011

Hartigan. A, Fiala. I, Dyková. I, Rose. K, Phalen. D and Šlapeta. J, Are Myxozoan Parasites threatening Australian amphibians? , Wildlife Health and Conservation Conference, Camden, August 2009

Hartigan. A, Fiala. I, Dyková. I, Rose. K, Phalen. D and Šlapeta. J, Ultrastructural Identification of Myxosporidian Parasite in Australian Frogs and Tadpoles, Australian Society of Parasitologists 2009 conference, July 2009

ix TABLE OF CONTENTS

Statement of authentication ______ii Abstract ______iii Acknowledgements ______vi Chapter 1: Introduction ______1 1.1 Wildlife disease and amphibians ______2 1.2 Myxosporea background ______4 1.3 Myxosporean life cycle ______5 1.4 Inter-host development, tissue tropism and host range ______6 1.5 Species infecting amphibians: range and tissue tropism ______6 1.6 Australian frogs and myxozoa ______7 1.7 Aims of this thesis ______9 1.8 References ______9 Chapter 2: Museum material reveals a frog parasite emergence after the invasion of the invasion of the cane toad in Australia ______14 2.1 Abstract ______15 2.2 Background ______15 2.3 Results ______16 2.4 Discussion______17 2.5 Conclusion ______18 2.6 Methods ______18 2.7 Author Contributions ______19 2.8 Acknowledgements______19 2.9 References ______19 Supplementary material: Summary of the Australian Museum frog specimens and presence of Myxidium cf. immersum ______21 Chapter 3: A suspected parasite spill-back of two novel Myxidium spp. (Myxosporea) causing disease in Australian endemic frogs found in the invasive Cane toad ______22 3.1 Abstract ______23 3.2 Introduction ______23 3.3 Results and Discussion ______24 3.4 Conclusion ______30 3.5 Material and Methods ______30 3.6 Acknowledgements______33 3.7 Author Contributions ______33 3.8 References ______33 Supporting Information: Table S1 Summary of catalogued cases at the Australian Registry of Wildlife Health (Taronga Conservation Society Australia, Mosman, NSW, Australia) from 1997–2009. ______35 Supporting Information: Table S2 Summary of sequence data collected from Myxosporea in frogs ______36 Supporting Information: Table S3 Summary of morphology of myxospores (Myxidium spp.) recovered from gall bladders in frogs ______38 Supporting Information: Table S4 Summary of SSU rDNA sequences used to reconstruct myxosporean phylogeny 40 Supporting Information: Figure S1 Phylogenetic trees inferred using SSU rDNA of Myxosporea. ______42 Chapter 4: New species of Myxosporea from frogs and resurrection of the genus Cystodiscus Lutz, 1889 for species with myxospores in gallbladders of amphibians ______43 4.1 Summary ______44 4.2 Introduction ______44 4.3 Materials and Methods ______45 4.4 Results ______47 4.5 Discussion ______55 4.6 Synopsis of Cystodiscus Lutz, 1889 emend ______58 4.7 Acknowledgements______61 4.8 References ______61 Supporting Table 1: Myxosporean species DNA used for specificity test of Cystodiscus spp. with species specific multiplex PCR ______63 Supporting Figure 1: Alignment of 18S, ITS1, 5.8S, ITS2 and 28S regions from Cystodiscus axonis and Cystodiscus australis annotated with diagnostic primers used in multiplex species specific PCR ______64 Chapter 5: Comparative pathology and ecological implications of two myxosporean parasites in native Australian frogs and the invasive Cane toad ______65 5.1 Abstract ______66 5.2 Introduction ______66 5.3 Methods ______67 5.4 Results ______70 5.5 Discussion ______75 5.6 Conclusion ______78 5.7 Acknowledgements______78 5.8 Author Contributions ______78 5.9 References ______78 Supporting Information: Figure S1. Sensitivity of species specific multiplex PCR for Cystodiscus axonis and Cystodiscus australis ______80 Supporting Information: Table S1. Summary of statistical analyses results for associating Cystodiscus axonis and Cystodiscus australis with lesions of disease. ______81 Chapter 6: Final discussion, future directions and conclusions ______83 6.1 A Dark Horse: Myxosporea in Australian amphibians ______84 6.2 The origins of gallbladder myxosporea in Australian frog species ______85 6.3 The more you look, the more you find: cryptic diversity in gallbladder myxosporea ______87 6.4 The impact of Cystodiscus species in Australian frogs and tadpoles ______88 6.5 Future Research Possibilities ______89 6.6 Conclusions ______93 6.7 References ______94

Appendices ______96 Appendix 1: Supporting letters from co-authors ______97 Appendix 2: Summary of Genbank accession numbers ______101 Appendix 3: List of frog host species identified with Cystodiscus axonis or Cystodiscus australis infection ______104

Chapter 1: Introduction

1 1.1 Wildlife disease and amphibians

The International Union for Conservation of Nature (IUCN) has reported that over 20 percent of all species assessed are currently threatened with extinction (IUCN, 2010). Infectious diseases pose a significant threat to wildlife species (Daszak et al., 2000) and investigation of infectious agents (pathogens) is essential before effects are irreversible. Amphibians are one of the most threatened wildlife groups, being at risk from many anthropogenic, environmental and pathogenic threats (Blaustein et al., 2011, Hof et al., 2011). A number of fungi, parasites, bacteria and viruses have been shown to infect a range of amphibians (Hoverman et al., 2011,

Berger et al., 1998, Daszak et al., 2003, Stuart et al., 2004). The effects of these pathogens vary in severity between hosts, populations and species (Retallick and Miera, 2007, Gray et al., 2009,

Briggs et al., 2005). In some cases infectious agents do not cause any disease or pathology until a change in the environment occurs which upsets the balance between host and pathogen (Mazzoni et al., 2003). The case of the trematode parasite Ribeiroia ondatrae illustrates the complexity of the host-parasite relationship. This infection causes limb malformations in 20 species of frogs, toads and salamanders in eleven U.S. States (Johnson et al., 2004). These limb malformations were first documented as isolated events in the 1940s and the incidence appears to have increased in recent years (Johnson et al., 2003). A correlation was subsequently found between increased numbers of Planorbella spp. snails in ponds that were subject to a high level of runoff from agricultural areas (elevated phosphorus contamination) and the frequency of frog limb malformations due to R. ondatrae in the USA (Johnson et al., 2002, Johnson et al., 2004). Heavy infections of R. ondatrae cause 97.5% mortality in Rana pipens (Schotthoefer et al., 2003).

Severe limb deformities increase the predation pressure on the amphibians, which enables the transmission of R. ondatrae to its final host (bird or mammal) (Johnson et al., 1999, Johnson et

2 al., 2001, Johnson et al., 2004). The flow-on effect of this environmental change has led to concern for populations of endangered frogs (Johnson et al., 2004).

The consequences of not recognizing a potential threat in a timely manner can be serious.

Chytridiomycosis (Batrachochytrium dendrobatidis) causes global declines in frog populations

(Berger et al., 1998, Stuart et al., 2004). The impact of this disease has been increased by a delay in response time caused by wildlife managers and policy makers debating conservation implications (Skerratt et al., 2007). It is now estimated that B. dendrobatidis has contributed to the extinctions of at least 200 amphibian species around the world (Skerratt et al., 2007). The fungus emerged in Australia in the late 1970s and has been linked to the declines of many species and the extinction of at least two species in Australia (Hines et al., 1999, Berger et al.,

1998, Murray et al., 2009). Detection is the key factor in understanding the epidemiology of disease such as chytridiomycosis and implementing quarantine control. Detection is also vital in the later development of diagnostic methods to screen populations (Hyatt et al., 2007). Whilst surveillance of infectious agents is useful in detecting pathogens, investigation into pathogens is required in order to assess any potential clinical impact (Young et al., 2012).

The investigation of wildlife disease is a high priority for conservation importance species, particularly when considering the additional effects of environmental change, habitat loss and related threats. Investigation into agents of disease provides the basis for management policies

(Murray et al., 2011) and the determination of the potential impact of the threat to conservation and wildlife health.

3 1.2 Myxosporea background

Myxosporeans are metazoan parasites that infect a wide range of hosts (Lom and Dyková, 1992,

Kent et al., 2001, Lom and Dyková, 2006). Development occurs in two hosts to produce myxospores in the vertebrate (fish, frog, reptile etc.) which infect invertebrates (oligochaetes, polychaetes) and develop actinospores. Taxonomists were originally confused about the evolutionary origins of myxosporeans, as, although myxospores are multicellular, their presporogonic plasmodia are single celled. The controversy concerns which Metazoan group

Myxosporeans are related to, Cnidaria (jellyfish, corals, anemones etc.) or Bilitaria (flatworms, chordates, acanthocephalan etc.) (Kent et al., 2001, Siddall et al., 1995, Ringuette et al., 2010,

Jiménez-Guri et al., 2007). Despite the lack of consensus on the taxonomic placement, the group itself is of interest mostly due to the vertebrate hosts they infect. Myxosporean parasites cause large scale economic losses in commercial fisheries as well as becoming a conservation concern for wild populations (Lom and Dyková, 2006, Lom and Dyková, 1992, Vincent, 1996, Nehring and Walker, 1996). Despite their importance there is little known about the development of the majority of identified myxosporean species. It is assumed that all myxosporean parasites complete their life cycle through alternation between invertebrate (definitive) and vertebrate hosts (intermediate) (Køie, 2000, Lom and Dyková, 1992). There are over 2,000 myxosporean parasite described, but the complete life cycle is known for only for 34 species (Feist and

Longshaw, 2006).

4 1.3 Myxosporean Life Cycle

Myxosporean parasites undergo asexual development within a vertebrate host such as a fish, bird, reptile or frog (Canning and Okamura, 2004). After replicating and producing transmissible myxospores, the host either passes the spores out of its body (digestive or reproductive system), or suffers damage to the infected tissue, possibly deteriorating enough to cause host death

(predation, starvation, etc.) which releases infective spores from deep tissue sites (Dzulinsky et al., 1994). For example, in the case of Myxobolus cerebralis the spores have no chance to exit the salmonid host unless the fish is eaten or dies. This is because development is restricted to nervous tissue and cartilage, which does not allow release of the myxospores without harming the host. Once the myxospores are free in the environment they are ingested by an invertebrate host and development can occur within the body wall (Bartholomew et al., 1997), coelomic cavity (Rangel et al., 2009) or digestive tract (El-Matbouli and Hoffmann, 1998). The actinospores released from definitive hosts are usually morphologically distinct from their myxospore types in size, overall shape and organization (Canning and Okamura, 2004). In the case of Myxobolus cerebralis, actinospores actively seek out hosts, and attach to hosts based on chemical cues (dermal mucous) and physical stimulation (Kallert et al., 2005). Actinospores have many varied morphotypes (Canning and Okamura, 2004) with more still being described

(Hallett et al., 2002, Holzer et al., 2004, Eszterbauer et al., 2006). There is some indication that there may be a large amount of morphological variability even within actinospores of the same species (Hallett et al., 2002). The majority of new species descriptions using myxospore or actinospore material use genetic identification as well as morphology. There are many myxosporean species with unknown alternate hosts or actinospore types. To date the only life

5 cycles that have been described are fish infecting species, despite myxosporeans being found in a number of other species including reptiles and amphibians (Eiras, 2005).

1.4 Inter-host development, tissue tropism and host range

Within vertebrate hosts, the process of replication and development of myxospores can occur in one or multiple host organs (e.g swimbladders, muscle, nervous tissue, kidney, intestine) (Feist and Longshaw, 2006, Lom and Dyková, 2006, Lom and Dyková, 1992, Canning and Okamura,

2004). Tissue tropism is suggested to be a factor in predicting the phylogenetic relationships of myxosporean species rather than spore morphology or host taxonomic groups (Burger et al.,

2007, Eszterbauer, 2004, Fiala and Bartošová, 2010). Aquatic vertebrates (fish, amphibians and reptiles) were thought to be the only intermediate hosts for Myxosporea, however, birds have recently been identified as a potential host group (Bartholomew et al., 2008). In addition, three shrew species have now been identified as hosts of Soricimyxum fegati (Dyková et al., 2007,

Dyková et al., 2011), which is interesting as they are entirely terrestrial. Further investigation may reveal more terrestrial hosts for myxosporean parasites, which could possibly expand the known diversity of invertebrate hosts.

1.5 Species infecting amphibians: range and tissue tropism

The myxosporean species first identified in an amphibian was Cystodiscus immersus from the gallbladder of a Bufo marinus toad in Brazil (Lutz, 1889). This species was renamed numerous times (Cordero, 1919, Lühe, 1899) and eventually was placed within the genus Myxidium -

Myxidium immersum (syn. Cystodiscus immersus), and identified in a number of South American

6 toads (Kudo and Sprague, 1940). Since then, numerous myxosporean species have been observed in amphibian species on every continent on which amphibians are found (Eiras, 2005).

Myxosporea have been identified in the skin, intestine, urogenital system and gallbladder of amphibians (Eiras, 2005, Lom and Dyková, 2006, Sitjà-Bobadilla, 2009). Despite the large number of identified species there has been little investigation about the impact of myxosporean parasitism on the host. Some infections have been associated with moderate pathological changes, for example testicular tissue lost to developing plasmodia (Mubarak and Abed, 2001), hepatitis (Hill et al., 1997) and renal tubule dilation and necrosis (Duncan et al., 2004).

Population impacts of myxosporean parasites have not yet been evaluated, though the potential exists for significant implications for conservation of threatened species (Sitjà-Bobadilla, 2009,

Murray et al., 2011).

1.6 Australian frogs and myxozoa

The first myxosporean parasite identified in Australian amphibians was Myxobolus hylae, found in tissue from the testes, oviducts and kidneys of the Green and golden bell frog (Litoria aurea)

(Johnston and Bancroft, 1918). This infection was associated with sickly and lethargic behaviour and cysts were found within the gonads. Another myxosporean species, Myxobolus fallax was isolated from the testes of the Eastern dwarf tree frog (Litoria fallax) in 2006. Spore shedding in this species was induced by forced spermiation and the infection was deemed to be non- pathogenic (Browne et al., 2006). In the 1980s an infectious agent isolated from Australian frog gallbladders was determined (based on spore morphology) to be Myxidium immersum (syn.

Cystodiscus immersus). Delvinquier (1986) sampled 924 frogs of 62 families, including the invasive Cane toad (Bufo marinus syn. Rhinella marina) from the north-east coast of Australia

7 between 1983 and 1985. A high prevalence of animals with myxospores resembling M. immersum in their gallbladders was found. It was suggested that the Cane toad introduced M. immersum to Australian frogs when it was introduced as a biological control in (Easteal) A

Myxidium species was identified (based on spore morphology) as the causative agent of hepatitis in a cohort of adult Green tree frogs from Queensland, Australia (Hill et al., 1997). The lesions associated with infection included significant inflammation and proliferation of the liver bile ducts that was attributed to the presence of plasmodia within the bile ducts of adult frogs.

Molecular identification of this pathogen was not performed. Although M. immersum was thought to be exotic and assumed to be introduced from South America by the Cane toad, there was no further investigation of this parasite and its impacts on Australian frogs.

In 2007 the Australian Registry of Wildlife Health (ARWH) reported an infection of the nervous tissue and bile ducts in Green and golden bell tadpoles with an unknown multi-celled organism.

The infection was found when animals were screened for pathogens prior to their use as founding animals in a captive breeding project. Similar infections had been noted in other species of frogs and tadpoles by ARWH pathologists however the infectious organisms were never identified. The Green and golden bell frog is an iconic Australian frog species. It is the subject of numerous government funded translocation projects and yet remains a highly threatened species. Infectious diseases were noted as an important threat to the species in the

2005 recovery plan drafted by the NSW Department of Environment and Heritage (DEH, 2005).

Despite this, the pathogens of these species of parasite identified in the central nervous system and liver of these frogs had not been investigated.

8 1.7 Aims of this thesis

The presence of myxosporean parasites within populations of high conservation value prompted an investigation into the significance and impact of these parasites to Australian amphibians. The aims of this thesis were to:

(1) Resolve if this pathogen was linked to the introduction of the Cane toad in 1935 as

hypothesized, using museum material and wildlife health records as a retrospective

method of assessing if the pathogen has recently spread into frogs (Chapter 2, Parasites

& Vectors 2010).

(2) Identify the myxosporean species found in Australian frogs, and determine if it is an

endemic or exotic infection based on morphological and molecular characterization and

sampling of potential source locations (Chapters 3 and 4, PLoS One 2011 and

Parasitology 2012).

(3) Assess the potential impact this infection has on Australian frogs and tadpoles, based

on histological investigation of infected hosts to establish host species, host reactions and

the implications of these parasites in frog populations (Chapter 5, PLoS One 2012).

References

Bartholomew, J. L., Atkinson, S. D., Hallett, S. L., Lowenstine, L. J., Garner, M. M., Gardiner, C. H., Rideout, B. A., Keel, M. K. & Brown, J. D. 2008. Myxozoan parasitism in waterfowl. International Journal for Parasitology, 38, 1199-1207. Bartholomew, J. L., Whipple, M. J., Stevens, D. G. & Fryer, J. L. 1997. The life cycle of Ceratomyxa shasta, a myxosporean parasite of salmonids, requires a freshwater polychaete as an alternate host. Journal of Parasitology, 83, 859-868. Berger, L., Speare, R., Daszak, P., Green, D. E., Cunningham, A. A., Goggin, C. L., Slocombe, R., Ragan, M. A., Hyatt, A. D., McDonald, K. R., Hines, H. B., Lips, K. R., Marantelli,

9 G. & Parkes, H. 1998. Chytridiomycosis causes amphibian mortality associated with population declines in the rain forests of Australia and Central America. Proceedings of the National Academy of Sciences, 95, 9031-9036. Blaustein, A. R., Han, B. A., Relyea, R. A., Johnson, P. T. J., Buck, J. C., Gervasi, S. S. & Kats, L. B. 2011. The complexity of amphibian population declines: understanding the role of cofactors in driving amphibian losses. Annals of the New York Academy of Sciences, 1223, 108-119. Briggs, C. J., Vredenburg, V. T., Knapp, R. A. & Rachowicz, L. J. 2005. Investigating the population-level effects of chytridiomycosis: An emerging infectious disease of amphibians. Ecology, 86, 3149-3159. Browne, R. K., Li, H. & Vaughan, M. 2006. Sexually mediated shedding of Myxobolus fallax spores during spermiation of Litoria fallax (Anura). Diseases of Aquatic Organisms, 72, 71-75. Burger, M. A. A., Cribb, T. H. & Adlard, R. D. 2007. Patterns of relatedness in the Kudoidae with descriptions of Kudoa chaetodoni n. sp. and K. lethrini n. sp. (Myxosporea: Multivalvulida). Parasitology, 134, 669-681. Canning, E. U. & Okamura, B. 2004. Biodiversity and evolution of the Myxozoa. Advances in Parasitology, 56, 43-131. Cordero, E. H. 1919. Nota sobre 'Opalina antilliensis' Metcalf, ciliado parasito de los batracios del Uruguay. Physis Seccion A los Oceanos y sus Organismos, 531-535. Daszak, P., Cunningham, A. A. & Hyatt, A. D. 2000. Emerging infectious diseases of wildlife - Threats to biodiversity and human health. Science 287, 443-449. Daszak, P., Cunningham, A. A. & Hyatt, A. D. 2003. Infectious disease and amphibian population declines. Diversity and Distributions, 9, 141-150. DEH, D. o. E. a. C. 2005. Green and golden bell frog Litoria aurea (Lesson 1829): Draft recovery plan. Department of Environment and Conservation, Hurstville, NSW. Delvinquier, B. L. J. 1986. Myxidium immersum (Protozoa: Myxosporea) of the Cane toad Bufo marinus in Australian anura with a synopsis of the genus in amphibians. Australian Journal of Zoology, 34, 843-854. Duncan, A. E., Garner, M. M., Bartholomew, J. L., Reichard, T. A. & Nordhausen, R. W. 2004. Renal myxosporidiasis in Asian horned frogs (Megophrys nasuta). Journal of Zoo and Wildlife Medicine, 35, 381-386. Dyková, I., Tyml, T., Fiala, I. & Lom, J. 2007. New data on Soricimyxum fegati (Myxozoa) including analysis of its phylogenetic position inferred from the SSU rRNA gene sequence. Folia Parasitologica, 54, 272-276. Dyková, I., Tyml, T. & Kostka, M. 2011. Xenoma-like formations induced by Soricimyxum fegati (Myxosporea) in three species of shrews (Soricomorpha: Soricidae), including records of new hosts. Folia Parasitologica, 58, 249-56. Dzulinsky, K., Cone, D. K., Faulkner, G. T. & Cusack, R. 1994. Development of Myxobolus neurophilus (Guildford, 1963) (Myxosporea) in the brain of Yellow perch (Perca flavescens) in Vinegar Lake, Nova Scotia. Canadian Journal of Zoology, 72, 1180-1185. Easteal, S. 1981. The history of introductions of Bufo marinus (Amphibia: Anura); a natural experiment in evolution. Biological Journal of the Linnean Society, 16, 93-113. Eiras, J. C. 2005. An overview on the myxosporean parasites in amphibians and reptiles. Acta Parasitologica, 50, 267-275.

10 El-Matbouli, M. & Hoffmann, R. W. 1998. Light and electron microscopic studies on the chronological development of Myxobolus cerebralis to the actinosporean stage in Tubifex tubifex. International Journal for Parasitology, 28, 195-217. Eszterbauer, E. 2004. Genetic relationship among gill-infecting Myxobolus species (Myxosporea) of cyprinids: molecular evidence of importance of tissue-specificity. Diseases of aquatic organisms, 58, 35-40. Eszterbauer, E., Marton, S., Rácz, O., Letenyei, M. & Molnár, K. 2006. Morphological and genetic differences among actinosporean stages of fish-parasitic myxosporeans (Myxozoa): difficulties of species identification. Systematic Parasitology, 65, 97-114. Feist, S. W. & Longshaw, M. 2006. Phylum: Myxozoa. Fish Diseases and Disorders, Vol 1: Protozoan and Metazoan Infections, 2nd Edition. Wallingford Cabi Publishing. Fiala, I. & Bartošová, P. 2010. History of myxozoan character evolution on the basis of rDNA and EF-2 data. BMC Evolutionary Biology, 10, 228. Gray, M. J., Miller, D. L. & Hoverman, J. T. 2009. Ecology and pathology of amphibian ranaviruses. Diseases of Aquatic Organisms, 87, 243-266. Hallett, S. L., Atkinson, S. D. & El-Matbouli, M. 2002. Molecular characterisation of two aurantiactinomyxon (Myxozoa) phenotypes reveals one genotype. Journal of Fish Diseases, 25, 627-631. Hill, B. D., Green, P. E. & Lucke, H. A. 1997. Hepatitis in the green tree frog (Litoria caerulea) associated with infection by a species of Myxidium. Australian Veterinary Journal, 75, 910-911. Hines, H., Mahony, M. & McDonald, K. (eds.) 1999. An assessment of frog declines in wet subtropical Australia, Canberra: Environment Australia. Hof, C., Araujo, M. B., Jetz, W. & Rahbek, C. 2011. Additive threats from pathogens, climate and land-use change for global amphibian diversity. Nature, 480, 516. Holzer, A. S., Sommerville, C. & Wootten, R. 2004. Molecular relationships and phylogeny in a community of myxosporeans and actinosporeans based on their 18S rDNA sequences. International Journal for Parasitology, 34, 1099-1111. Hoverman, J., Gray, M., Haislip, N. & Miller, D. 2011. Phylogeny, life history, and ecology contribute to differences in amphibian susceptibility to ranaviruses. EcoHealth, 8, 301- 319. Hyatt, A. D., Boyle, D. G., Olsen, V., Boyle, D. B., Berger, L., Obendorf, D., Dalton, A., Kriger, K., Hero, M., Hines, H., Phillott, R., Campbell, R., Marantelli, G., Gleason, F. & Colling, A. 2007. Diagnostic assays and sampling protocols for the detection of Batrachochytrium dendrobatidis. Diseases of Aquatic Organisms, 73, 175-192. IUCN. 2010. Red List of threatened species [Online]. Available: http://www.iucnredlist.org [Accessed 08.01.12. Jiménez-Guri, E., Philippe, H., Okamura, B. & Holland, P. W. H. 2007. Buddenbrockia Is a Cnidarian worm. Science, 317, 116-118. Johnson, P. T., Lunde, K. B., Thurman, E. M., Ritchie, E. G., Simon, N. W., Sutherland, D. R., Kapfer, J. M., Frest, T. J., Bowerman, J. & Blaustein, A. R. 2002. Parasite (Ribeiroia ondatrae) infection linked to amphibian malformations in the western United States. Ecological Monographs, 72, 151-168. Johnson, P. T. J., Lunde, K. B., Haight, R. W., Bowerman, J. & Blaustein, A. R. 2001. Ribeiroia ondatrae (Trematoda: Digenea) infection induces severe limb malformations in western toads (Bufo boreas). Canadian Journal of Zoology, 79, 370-379.

11 Johnson, P. T. J., Lunde, K. B., Ritchie, E. G. & Launer, A. E. 1999. The effect of trematode infection on amphibian limb development and survivorship. Science, 284, 802-804. Johnson, P. T. J., Lunde, K. B., Zelmer, D. A. & Werner, J. K. 2003. Limb deformities as an emerging parasitic disease in amphibians: evidence from museum specimens and resurvey data Conservation Biology, 17, 1724-1737. Johnson, P. T. J., Sutherland, D. R., Kinsella, J. M. & Lunde, K. B. 2004. Review of the trematode genus Ribeiroia (Psilostomidae): Ecology, life History and pathogenesis with special emphasis on the amphibian malformation problem. Advances in Parasitology. Academic Press. Johnston, T. H. & Bancroft, M. J. 1918. A parasite, Myxobolus hylae sp. nov. of the reproductive organs of the golden swamp frog, Hyla aurea. Australian Zoologist, 1, 171. Kallert, D. M., El-Matbouli, M. & Haas, W. 2005. Polar filament discharge of Myxobolus cerebralis actinospores is triggered by combined non-specific mechanical and chemical cues. Parasitology, 131, 609-616. Kent, M. L., Andree, K. B., Bartholomew, J. L., El-Matbouli, M., Desser, S. S., Devlin, R. H., Feist, S. W., Hedrick, R. P., Hoffmann, R. W., Khattra, J., Hallett, S. L., Lester, R. J. G., Longshaw, M., Palenzeula, O., Siddall, M. E. & Xiao, C. 2001. Recent advances in our knowledge of the Myxozoa. Journal of Eukaryotic Microbiology, 48, 395-413. Køie, M. 2000. First record of an actinosporean (Myxozoa) in a marine polychaete annelid. Journal of Parasitology, 86, 871-872. Kudo, R. R. & Sprague, V. 1940. On Myxidium immersum Lutz and M. serotinum n. sp., two Myxosporidian parasites of Salientia of South and North America. Revista de Medicina Tropical Habana, 6, pp. 65-73. Lom, J. & Dyková, I. 1992. Protozoan parasites of fish, Elsevier Science Publishers. Lom, J. & Dyková, I. 2006. Myxozoan genera: definition and notes on taxonomy, life-cycle terminology and pathogenic species. Folia Parasitologica 53, 1-36. Lühe, M. 1899. Cystodiscus immersus Lutz. Verhandlungen der Deutschen Zoologischen Gesellschaft, 9, 291-293. Lutz, A. 1889. Ueber ein Myxosporidium aus der Gallenblase brasilianischer Batrachier. Centralblatt fiir Bakteriologie und Parasitenkunde, 5, 84-88. Mazzoni, R., Cunningham, A. A., Daszak, P., Apolo, A., Perdomo, E. & Speranza, G. 2003. Emerging pathogen of wild amphibians in frogs (Rana catesbeiana) farmed for international trade. Emerging Infectious Diseases, 9, 995-998. Mubarak, M. & Abed, G. H. 2001. Morphopathological changes of the testicular tissue of Egyptian toad, Bufo regularis (Amphibia, Bufonidae) infected with Myxobolus sp. (Myxozoa, Myxobolidae). Acta Parasitologica, 46, 113-118. Murray, K., Skeratt, L., Marantelli, G., Berger, L., Hunter, D., Mahony, M. & Hines, H. 2011. Guidelines for minimising disease risks associated with captive breeding, raising and restocking programs for Australian frogs. A report for the Australian Government Department of Sustainability, Environment, Water, Population and Communities. A report for the Australian Government Department of Sustainability, Environment, Water, Population and Communities. Murray, K. A., Skerratt, L. F., Speare, R. & McCallum, H. 2009. Impact and dynamics of disease in species threatened by the amphibian chytrid fungus, Batrachochytrium dendrobatidis. Conservation Biology, 23, 1242-1252.

12 Nehring, R. B. & Walker, P. G. 1996. Whirling disease in the wild: The new reality in the intermountain West. Fisheries (Bethesda), 21, 28-30. Rangel, L. F., Santos, M. J., Cech, G. & Székely, C. 2009. Morphology, molecular data, and development of Zschokkella mugilis (Myxosporea, Bivalvulida) in a polychaete alternate host, Nereis diversicolor. Journal of Parasitology, 95, 561-569. Retallick, R. W. & Miera, V. 2007. Strain differences in the amphibian chytrid Batrachochytrium dendrobatidis and non-permanent, sub-lethal effects of infection. Diseases of Aquatic Organisms, 75, 201-207. Ringuette, M. J., Koehler, A. & Desser, S. S. 2010. Shared antigenicity between the polar filaments of Myxosporeans and other Cnidaria. Journal of Parasitology, 97, 163-166. Schotthoefer, A. M., Koehler, A. V., Meteyer, C. U. & Cole, R. A. 2003. Influence of Ribeiroia ondatrae (Trematoda : Digenea) infection on limb development and survival of northern leopard frogs (Rana pipiens): effects of host stage and parasite-exposure level. Canadian Journal of Zoology-Revue Canadienne De Zoologie, 81, 1144-1153. Siddall, M. E., Martin, D. S., Bridge, D., Desser, S. S. & Cone, D. K. 1995. The demise of a phylum of protists: Phylogeny of myxozoa and other parasitic cnidaria. Journal of Parasitology, 81, 961-967. Sitjà-Bobadilla, A. 2009. Can Myxosporean parasites compromise fish and amphibian reproduction? Proceedings of the Royal Society B: Biological Sciences, 276, 2861-2870. Skerratt, L., Berger, L., Speare, R., Cashins, S., McDonald, K., Phillott, A., Hines, H. & Kenyon, N. 2007. Spread of chytridiomycosis has caused the rapid global decline and extinction of frogs. EcoHealth, 4, 125-134. Stuart, S. N., Chanson, J. S., Cox, N. A., Young, B. E., Rodrigues, A. S., Fischman, D. L. & Waller, R. W. 2004. Status and trends of amphibian declines and extinctions worldwide. Science, 306, 1783-6. Vincent, E. R. 1996. Whirling disease and wild trout: The Montana experience. Fisheries (Bethesda), 21, 32-33. Young, S., Skerratt, L. F., Mendez, D., Speare, R., Berger, L. & Steele, M. 2012. Using community surveillance data to differentiate between emerging and endemic amphibian diseases. Diseases of aquatic organisms, 98, 1-10.

Chapter 2: Museum material reveals a frog parasite emergence after the invasion of the invasion of the Cane toad in Australia

14 Hartigan et al. Parasites & Vectors 2010, 3:50 http://www.parasitesandvectors.com/content/3/1/50

RESEARCH Open Access MuseumResearch material reveals a frog parasite emergence after the invasion of the cane toad in Australia

Ashlie Hartigan, David N Phalen and Jan Šlapeta*

Abstract Background: A parasite morphologically indistinguishable from Myxidium immersum (Myxozoa: Myxosporea) found in gallbladders of the invasive cane toad (Bufo marinus) was identified in Australian frogs. Because no written record exists for such a parasite in Australian endemic frogs in 19th and early 20th century, it was assumed that the cane toad introduced this parasite. While we cannot go back in time ourselves, we investigated whether material at the museum of natural history could be used to retrieve parasites, and whether they were infected at the time of their collection (specifically prior to and after the cane toad translocation to Australia in 1935). Results: Using the herpetological collection at the Australian Museum we showed that no myxospores were found in any animals (n = 115) prior to the cane toad invasion (1879-1935). The green and golden bell frog (Litoria aurea), the Peron's tree frog (Litoria peronii), the green tree frog (Litoria caerulea) and the striped marsh frog (Limnodynastes peronii) were all negative for the presence of the parasite using microscopy of the gallbladder content and its histology. These results were sufficient to conclude that the population was free from this disease (at the expected minimum prevalence of 5%) at 99.7% confidence level using the 115 voucher specimens in the Australian Museum. Similarly, museum specimens (n = 29) of the green and golden bell frog from New Caledonia, where it was introduced in 19th century, did not show the presence of myxospores. The earliest specimen positive for myxospores in a gallbladder was a green tree frog from 1966. Myxospores were found in eight (7.1%, n = 112) frogs in the post cane toad introduction period. Conclusion: Australian wildlife is increasingly under threat, and amphibian decline is one of the most dramatic examples. The museum material proved essential to directly support the evidence of parasite emergence in Australian native frogs. This parasite can be considered one of the luckiest parasites, because it has found an empty niche in Australia. It now flourishes in > 20 endemic and exotic frog species, but its consequences are yet to be fully understood.

Background Museum collections of amphibians have been used Museum material is important in comparisons between effectively in the global investigation into amphibian historical and contemporary animal ranges in relation to decline and historical emergence of external frog malfor- environmental modifications [1]. Investigation of patho- mations [1,3]. Examination of museum material sup- gens in museum material has been scarce, despite their ported the 'out of Africa' origin of the amphibian chytrid offering a unique insight into parasite emergence (Batrachochytrium dendrobatidis) [4]. International trade throughout the museum's sampling period, often span- of amphibians, that began in the mid-1930's, spread the ning several centuries [2]. chytrid which decimated amphibian populations worldwide [4]. Exotic introductions of myxosporeans are well known to the aquaculture industry. The myxosporean, * Correspondence: [email protected] Myxobolus cerebralis (Plehn 1905), which causes whirling 1 Faculty of Veterinary Science, University of Sydney, New South Wales 2006, Australia disease in salmonid fish was introduced from Europe to Full list of author information is available at the end of the article 26 other countries [5] and has significantly impacted wild © 2010 Hartigan et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

15 Hartigan et al. Parasites & Vectors 2010, 3:50 Page 2 of 6 http://www.parasitesandvectors.com/content/3/1/50

and farmed fish populations in North America signifi- probability of observing zero positives in a sample of 115 cantly [6]. frogs from before 1935 is 0.003. These results are suffi- A myxosporean parasite Myxidium immersum (Lutz cient to conclude that the population was free from this 1889) (Myxozoa; Myxosporea) was suggested to have disease (at the expected minimum prevalence of 5%) at been introduced to a wide spectrum of Australian native 99.7% confidence level. Assuming only 90% test sensitiv- frogs during the 20th century with the exotic cane toad ity we have exceeded the 66 animals as the minimum (Bufo marinus) [7,8]. This assumption is based on the number (p = 0.05) to be taken from a population with a absence of any written record for myxosporean parasite negative result to consider the population as free of the in Australian endemic frogs in 19th and early 20th century disease (expecting minimum prevalence of 5%). Using [7]. Myxidium immersum was found in gallbladders of at 115 animals and test sensitivity of 90%, it is adequate to least 12 species of Litoria, 4 species of Limnodynastes, conclude, that the population is free from disease at the and one each of Mixophyes, Ranidella, and Uperoleia in 99.5% confidence level. eastern Australia (Queensland, New South Wales) [7]. No myxospores were found in a selection of museum Moreover, a Myxidium sp. (myxospores closely resem- specimens of the green and golden bell frog from New bling Myxidium immersum) infection was found in a Caledonia (n = 29). The green and golden bell frog was green tree frog (Litoria caerulea) with hepatitis [9], and introduced to New Caledonia in 19th century [14]. circumstantial evidence suggests that myxosporeans are important pathogens of frogs and despite being recogn- The earliest specimen positive for gallbladder myxospores ised as potential disease agents there has been little inves- is from 1966 tigation into their threat status [10,11]. Currently, no life In total, 112 voucher specimens and their catalogue cycle for any frog myxosporean parasite has been eluci- records were recovered for frogs after 1935 (Figure 1, dated [12]. It is speculated, that myxosporean parasites in Additional file 1); the green and golden bell frog (n = 34), frogs will require an invertebrate host, because the major- the green tree frog (n = 12), the striped marsh frog (n = ity of myxosporean parasites in fish alternate between 18) - representing Australian endemic species, and the vertebrate and invertebrate hosts [13]. introduced cane toad (n = 48). The earliest specimen pos- The historical emergence of Myxidium immersum in itive for myxospores spores was from 1966 (green tree Australian amphibians can be effectively investigated by frog, New South Wales, AM#99574). Myxospores spores directly showing its absence in frogs prior to the well doc- were found in eight (7.1%, 8/112) Australian frog vouch- umented cane toad introduction into Queensland, Aus- ers in the post cane toad introduction period (Figure 1, 2; tralia in 1935. To test the hypothesis that the emergence Additional file 1). The myxospore positive native frogs, of myxosporeans in Australian frogs occurred after the the green and golden bell frog (AM#148744, introduction of the cane toad, we dissected frog voucher AM#153967), the green tree frog (AM#99754, specimens from the Australian Museum to microscopi- AM#99575) and the striped marsh frog (AM#158418, cally examine the presence of myxosporean stages before AM#162439), were from New South Wales. Two cane and after the cane toad translocation. toads from New South Wales and Queensland were positive for myxospores (AM#158500, AM#59922), the earli- Results est is from 1967. No gallbladder myxospores found in frogs prior to 1935 Ultrastructurally, myxospores resembled Myxidium A catalogue of the Australian Museum, Sydney was que- spores for which the number of ridges on the surface of ried for five frog species; the green and golden bell frog Myxidium spores is a species diagnostic character. All (Litoria aurea), the green tree frog (Litoria caerulea), the examined myxospores from museum vouchers had 5-9 Peron's tree frog (Litoria peronii), the striped marsh frog ridges on the myxospore surface (Additional file 1). (Limnodynastes peronii), and the introduced cane toad Myxospore morphology was consistent under light (Bufo marinus, syn. Rhinella marina). All specimens microscopy (Figure 2A), histology (Figure 2B) and scan- prior to 1935, for the green and golden bell frog (n = 27), ning electron microscopy (Figure 2C). The myxospores the green tree frog (n = 48), the Peron's tree frog (n = 16) from individual host species were not identical in size, and the striped marsh frog (n = 24) representing Austra- likely due to the effect of the ethanol or formalin fixation lian endemic species were retrieved (Additional file 1). for an unknown period of time prior to being transferred No myxospores were found in the total of 115 voucher into 70% ethanol at the Australian Museum. Neverthe- specimens collected before 1935. The museum speci- less, the number of spore ridges, direction of the suture mens were from eastern Australia, mostly from New line and spore shape are compatible with Myxidium South Wales (Figure 1). immersum [7,15]. This species was originally described Assuming the test sensitivity and specificity to be per- from the cane toad specimens in Brazil, South America, fect and the expected minimum prevalence to be 5%, the and later it was re-evaluated and redescribed by Kudo

16 Hartigan et al. Parasites & Vectors 2010, 3:50 Page 3 of 6 http://www.parasitesandvectors.com/content/3/1/50

Figure 1 Frog species collection sites in eastern Australia. Maps of collection sites in eastern Australia with the distribution of frogs collected and catalogued at the Australian Museum were grouped according to their collection date. Locations of individuals with the presence of myxospores in gall bladders are indicated in red. Each host species is depicted with a different icon. When more than one specimen from a single location was examined, the positive/negative values are shown. The western border of the cane toad distribution is delineated by a broken line. and Sprague (1940) [15], who have provided the following with their hosts especially if successively translocated measurements ranges for myxospores found in toad gall- multiple times across a great geographical distance, as bladders: length and width, 11.8-14.2 × 7.5-10.0 μm, polar experienced in the case of the cane toad. The cane toad capsule length and width 3.5-4.5 × 3.3-4.2 μm, and was introduced around the world from its native range of recorded 7-9 transverse striations. Myxospores found in Central and South America through the Caribbean into Australian frogs including the cane toad by Delvinquier Queensland, Australia in 1935 [8]. The 101 translocated (1986) [7] had the following characteristics: length and cane toads were bred and their progeny released as a bio- width 12.3-13.3 × 7.3-7.8 μm and 5-10 transverse stria- logical control agent for a sugar cane industry suffering tions (polar capsule measurements were not reported). from beetle damage [8]. A single host lungworm parasite with a direct life-cycle (Rhabdias pseudosphaerocephala) Discussion is endemic to the cane toad in Central and South Ameri- All vertebrates harbour a large diversity of parasites with cas and it has been translocated with the cane toad into often intricate life histories. For parasites whose life cycle the Australia but has not infected Australian endemic comprises more than one specific host, the survival of frog species [17]. Translocation of a multi-host parasite such a parasite is dependent on sympatry and interaction with the cane toad has yet to be unambiguously demon- with its host [16]. Removing a host from its natural envi- strated. Host specificity is a fundamental property of all ronment where it has coevolved with its surroundings parasites, and parasite faunas of invasive species consist over a long historical trajectory may have major deleteri- of a mixture of authentic parasites translocated from its ous consequences for its parasites. However, parasites original range and those newly encountered parasites that that do not need an additional host and can be sustained had broad enough specificity to take the advantage of the easily in the environment will be more likely to travel invasive host [18].

17 Hartigan et al. Parasites & Vectors 2010, 3:50 Page 4 of 6 http://www.parasitesandvectors.com/content/3/1/50

the introduction of the cane toad supports a recent emergence of the parasite. It is further supported by the absence of Myxosporea in the green and golden bell frog gallbladders from New Caledonia, where it is the only frog introduced in late 19th century. However, does this ultimately victimise the cane toad? First, we do not know the life cycle of the parasite, or if it requires an invertebrate host. The majority (possibly all) of myxosporean species, including the pathogenic Myxobolus cerebralis requires its aquatic oligochaete host being infected by myxospores produced in salmonid fish. The success behind the global distribution of Myxobolus cerebralis is in the cosmopolitan distribution of its dominant invertebrate host Tubifex tubifex and its transcontinental trade [6]. However, direct transmission, fish-to-fish, was documented for several Enteromyxum species [19]. We know nothing about the life cycle in frogs, but it is likely that Myxidium immersum uses an invertebrate host in its life- cycle, and therefore international trade may also be impli- cated in Myxidium immersum distribution across continents similar to the spread of the chytrid fungus [4]. Myxidium immersum is described as a species with an extremely broad host specificity [7,15]; however, what might seem like a single morphospecies may mask a cryptic species complex for which molecular methods may resolve this difficulty [20]. We are currently collecting material to genetically characterise extant isolates from Australian frogs (AH, DNP, JŠ unpublished data). Figure 2 Myxidium cf. immersum spores recovered from the Aus- tralian Museum. Light microscopy (A), histological section through Conclusion the gallbladder stained with Giemsa (B) and scanning electron micros- While the cane toad may appear as a prime suspect, the copy on black background. White arrows show the surface ridges and black arrows show two polar capsules per myxospore. Scale bar 10 μm deleterious effects of successive cane toad translocations (A, B), 2.5 μm (C); AM#158500 (A, C; Bufo marinus) and AM#162439 (B; seems unfavourable to a parasite with a two host life cycle Limnodynastes peronii). such as Myxidium immersum due to its complexity. The low prevalence (7%) in the museum material suggests that Almost all we know about myxosporean parasites only 7 animals would be Myxidium-positive when first comes from their investigation in fish. Out of more than translocated into Australia, because the founding cane 2200 species, about 2000 are fish Myxosporea, of which toad population was only 101 individuals [8]. If, indeed we know the life-cycle of only 33. The parasite alternates these were the founding animals for the dispersal of the between vertebrate and invertebrate hosts; asexual devel- parasite, than Myxidium immersum is one of the luckiest opment occurs in vertebrate hosts leading to the produc- parasites in the world because it is found flourishing in > tion of resistant myxospores, and sexually in the 20 endemic and exotic frog species and had even reached invertebrate host leading to the production of fragile a distribution beyond the current cane toad range. The aquatic actinospores [6,13]. The introduction of a myxo- origin of the parasite needs to be considered further with sporean parasite into Australian frogs from the cane toad regards to its status in the source population, an intro- has been assumed based on morphological evidence from duction with other frogs brought to Australia for labora- extant frogs and the absence of written records of such tory purposes and accidental introduction via infected gallbladder parasite in scientific records [7]. However, the invertebrate hosts. two-host life cycle of Myxosporea and the unknown specificity for either vertebrate or invertebrate hosts has Methods prompted us to re-examine this hypothesis using Voucher museum material museum material. Indeed, the absence of Myxosporea in A catalogue of the Australian Museum, Sydney was que- the gallbladders in frogs collected and catalogued prior to ried for the green and golden bell frog (Litoria aurea), the green tree frog (Litoria caerulea), the Peron's tree frog

18 Hartigan et al. Parasites & Vectors 2010, 3:50 Page 5 of 6 http://www.parasitesandvectors.com/content/3/1/50

(Litoria peronii), the striped marsh frog (Limnodynastes animals (p = 0.05) as the minimum number be taken from peronii), and the introduced cane toad (Bufo marinus, a population with negative result to consider the popula- syn. Rhinella marina). Vouchers were originally fixed in tion is free of the disease expecting minimum prevalence ethanol or formalin for an unknown period of time prior of 5%, 10% and 20%, respectively. to transfer into 70% ethanol. In total, 256 voucher speci- The minimum expected prevalence is the lowest level mens and their catalogue records (denoted by AM#) were of prevalence that we aimed to detect. Delvinquier (1986) recovered, including all specimens prior to 1935 (Addi- detected the parasite in 50% (6/12) specimens of the tional file 1). A selection of museum specimens of the striped marsh frog (Limnodynastes peronii), 46% (5/11) green and golden bell frog from New Caledonia was specimens of the green tree frog (Litoria caerulea), 42% included (n = 29). (8/19) specimens of the Peron's tree frog (Litoria peronii), and 13/34 (38%) specimens of the cane toad (Bufo mari- Extraction of frog gallbladders nus); the green an golden bell frog (Litoria aurea) was not A small incision (0.5-1 cm) on the belly of the frog surveyed [7]. Despite these numbers we restricted our- voucher was used to reach the gallbladder and preserve selves to a conservative 5% minimum prevalence. In other the integrity of the specimen. For sufficiently large speci- words, the minimum expected prevalence means that if mens the content of the gallbladder was aspirated with a parasite were to be present below 5% level, our survey disposable G22-needle. For smaller specimens the entire would not be able to detect it. This minimum expected gallbladder was removed. The individual gallbladder con- prevalence was considered sufficient, because (i) Delvin- tents and the gallbladder were stored in 70% ethanol until quier (1986) [7] has detected much higher prevalence in examination or processing. Australian frogs and (ii), we calculated 13% (7/60) prevalence using our survey of museum frogs from 1975 Examination of gallbladders for myxospores onwards. For a wet mount examination the entire gallbladder contents were examined using 20 × and 40 × objectives. Additional material Myxospores were measured/photographed using a 100 × objective BX51 microscope equipped with a DP70 cam- Additional file 1 Summary of the Australian Museum frog specimens era (Olympus Australia); a minimum of 10 myxospores and presence of Myxidium cf. immersum. from each individual was measured and morphological Competing interests details evaluated. Myxospores were prepared for scan- The authors declare that they have no competing interests. ning electron microscopy; fixed in 2% OsO4 in 0.2 M Authors' contributions sodium cacodylate buffer (pH 7.0), rinsed in distilled All authors contributed to this study. JŠ and AH designed the study. AH per- water, dehydrated with a graded acetone series, critical formed the experiments and collected data. AH and JŠ analysed the data. JŠ, point dried, coated with gold and examined using a Zeiss AH and DNP drafted the manuscript. All authors read and approved the final ULTRA plus FE. Individual gallbladders were selected for manuscript. paraffin embedding, sectioning, Giemsa staining and Acknowledgements examined for the presence of myxospores. This project was financially supported by the Australian Academy of Science, ARC/NH&MRC Network for Parasitology and Dr William Richards Award in Vet- Statistical analysis erinary Pathology (University of Sydney). We thank Iva Dyková, Ivan Fiala, Karrie Rose, Navneet Dhand, Michael Tyler and Glenn Shea for their insight, Ross Sad- The probability of observing no positive samples in the lier and Cecilie Beatson for access into the Australian Museum, technical sup- total number of samples collected was calculated using port from the Electron Microscopy Unit (Australian Key Centre for Microscopy FreeCalc v2, an epidemiological probability calculator and Microanalysis). assisting with the planning and analysis of surveys to Author Details demonstrate freedom from disease [21]; http://www.aus- Faculty of Veterinary Science, University of Sydney, New South Wales 2006, vet.com.au/content.php?page=software - distributed by Australia AusVet Animal Health Services. Received: 25 May 2010 Accepted: 10 June 2010 We assumed 100% specificity of our detection of myxo- Published: 10 June 2010 ParasitesThis© 2010 isarticle an Hartigan &Open Vectorsis available Access et 2010, al; licensee from:article 3:50 http://www.parasitesandvectors.com/content/3/1/50 distributed BioMed Central under Ltd. the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. spores, because it relied on microscopic of the whole gallbladder content coupled with histology and Giemsa References 1. Tingley MW, Beissinger SR: Detecting range shifts from historical species staining. We calculated that 57, 29 and 14 animals (p = occurrences: new perspectives on old data. Trends Ecol Evol 2009, 0.05) is the minimum number of animals to be taken from 24:625-633. a population with negative result to consider the popula- 2. Suárez AV, Tsutsui ND: The value of museum collections for research and society. BioScience 2004, 54:66-74. tion is free from disease using 5%, 10% and 20% minimum 3. Johnson PTJ, Lunde KB, Zelmer DA, Werner JK: Limb deformities as an expected prevalence, respectively. We also evaluated a emerging parasitic disease in amphibians: Evidence from museum conservative 90% test sensitivity yielding 66, 32 and 15 specimens and resurvey data. Conserv Biol 2003, 17:1724-1737.

19 Hartigan et al. Parasites & Vectors 2010, 3:50 Page 6 of 6 http://www.parasitesandvectors.com/content/3/1/50

4. Weldon C, du Preez LH, Hyatt AD, Muller R, Speare R: Origin of the amphibian chytrid fungus. Emerg Infect Dis 2004, 10:2100-2105. 5. Whipps CM, El-Matbouli M, Hedrick RP, Blazer V, Kent ML: Myxobolus cerebralis internal transcribed spacer 1 (ITS-1) sequences support recent spread of the parasite to North America and within Europe. Dis Aq Org 2004, 60:105-108. 6. Hedrick RP, Adkison MA, El-Matbouli M, MacConnell E: Whirling disease: re-emergence among wild trout. Immunol Rev 1998, 166:365-376. 7. Delvinquier BLJ: Myxidium immersum (Protozoa: Myxosporea) of the cane toad, Bufo marinus, in Australian Anura, with a synopsis of the genus in amphibians. Aust J Zoo 1986, 34:843-853. 8. Easteal S: The history of introductions of Bufo marinus (Amphibia: Anura) - a natural experiment in evolution. Biol J Linn Soc 1981, 16:93-113. 9. Hill BD, Green PE, Lucke HA: Hepatitis in the green tree frog (Litoria caerulea) associated with infection by a species of Myxidium. Aust Vet J 1997, 75:910-911. 10. Green DE, Gray MJ, Miller DL: Disease monitoring and biosecurity. In Amphibian Ecology and Conservation: A Handbook of Techniques Edited by: Dodd CK. Oxford: Oxford Press; 2009:481-505. 11. Sitjà-Bobadilla A: Can Myxosporean parasites compromise fish and amphibian reproduction? Proc R Soc B 2009, 276:2861-2870. 12. Eiras JC: An overview on the myxosporean parasites in amphibians and reptiles. Acta Parasitol 2005, 50:267-275. 13. Lom J, Dyková I: Myxozoan genera: definition and notes on taxonomy, life cycle terminology and pathogenic species. Folia Parasitol (Praha) 2006, 53:1-32. 14. Bauer AM: The Herpetofauna of New Caledonia Ithaca, NY: Society for the Study of Amphibians and Reptiles; 2000. 15. Kudo R, Sprague V: On Myxidium immersum (Lutz) and M. serotinum n. sp., two myxosporidian parasites of Salientia of South and North America. Rev Med Trop Parasitol Bacteriol Clin Lab 1940, 6:65-73. 16. Bush AO, Fernandez JC, Esch GW, Seed JR: Parasitism: The diversity and ecology of animal parasites Cambridge: Cambridge University Press; 2001. 17. Dubey S, Shine R: Origin of the parasite of an invading species, the Australian cane toad (Bufo marinus): are lungworms Australian or American? Mol Ecol 2008, 17:4418-4424. 18. Torchin ME, Laffrety KD, Dobson AP, McKenzie VJ, Kurtis AM: Introduced species and their missing parasites. Nature 2003, 421:628-630. 19. Diamant A: Fish-to-fish transmission of a marine myxosporean. Dis Aq Org 1997, 30:99-105. 20. Poulin R, Keeney DB: Host specificity under molecular and experimental scrutiny. Trends Parasitol 2007, 24:24-28. 21. Cameron AR, Baldock FC: A new probability formula for surveys to substantiate freedom from disease. Prev Vet Med 1998, 34:1-17.

doi: 10.1186/1756-3305-3-50 Cite this article as: Hartigan et al., Museum material reveals a frog parasite emergence after the invasion of the cane toad in Australia Parasites & Vectors 2010, 3:50

20 Additional file 1. Summary of the Australian Museum (AM) frog specimens and presence of Myxidium cf. immersum .

Green and golden bell frog (Litoria aurea ) (n=61) Green tree frog (Litoria caerulea ) (n=60) Striped marsh frog (Limnodynastes peronii ) (n=42) AM# Date Location Result AM# Date Location Result AM# Date Location Result 19424 1890 Richmond, NSW Negative 1899 1895 Sydney,NSW Negative 1879 1^ Waverley, NSW Negative 19425 1890 Richmond, NSW Negative 3538 1903 Mapoon, QLD Negative 1880 1^ Waverley, NSW Negative 4194 1908 Maroubra, NSW Negative 3833 1905 Sydney,NSW Negative 1881 1^ Waverley, NSW Negative 4195 1908 Maroubra, NSW Negative 4289 1908 Sydney,NSW Negative 1882 1^ Waverley, NSW Negative 4196 1908 Maroubra, NSW Negative 4290 1908 Sydney,NSW Negative 4333 1909 Killara, NSW Negative 4198 1908 Maroubra, NSW Negative 4292 1908 Darling Floods, NSW Negative 4336 1909 Killara, NSW Negative 4199 1908 Maroubra, NSW Negative 4293 1908 Darling Floods, NSW Negative 4338 1909 Killara, NSW Negative 4200 1908 Maroubra, NSW Negative 4294 1908 Darling Floods, NSW Negative 4360 1909 Killara, NSW Negative 4201 1908 Maroubra, NSW Negative 4520 1909 Murray Island, Torres Strait Negative 4382 1909 Killara, NSW Negative 4202 1908 Maroubra, NSW Negative 4522 1909 Murray Island, Torres Strait Negative 4383 1909 Killara, NSW Negative 4203 1908 Maroubra, NSW Negative 4524 1909 Murray Island, Torres Strait Negative 4384 1909 Killara, NSW Negative 4205 1908 Maroubra, NSW Negative 4614 1909 New Guinea Negative 4387 1909 Killara, NSW Negative 4208 1908 Maroubra, NSW Negative 5088 1909 Sydney,NSW Negative 4389 1909 Killara, NSW Negative 4251 1908 Nambucca, NSW Negative 5092 1911 Sydney,NSW Negative 4481 1909 Lindfield, NSW Negative 4665 1910 Woonona, NSW Negative 5177 1911 Sydney,NSW Negative 4482 1909 Lindfield, NSW Negative 4666 1910 Woonona, NSW Negative 5243 1911 Darling Floods, NSW Negative 4483 1909 Lindfield, NSW Negative 4667 1910 Woonona, NSW Negative 5245 1911 Darling Floods, NSW Negative 4484 1909 Lindfield, NSW Negative 4668 1910 Woonona, NSW Negative 5355 1911 Murray Island, Torres Strait Negative 4485 1909 Lindfield, NSW Negative 5389 1^ North Sydney, NSW Negative 5437 1911 Lismore, NSW Negative 4736 1910 Darling Flood Plains, NSW Negative 7446 1^ Pambula, NSW Negative 6115 1913 Unknown Negative 4739 1^ NSW Negative 7974 1^ Upper Colo, NSW Negative 6328 1913 Eidsvold, QLD Negative 5201 1^ Katoomba, NSW Negative 8454 1^ Unknown Negative 6329 1913 Eidsvold, QLD Negative 7496 1^ Richmond River, NSW Negative 8456 1^ Sydney, NSW Negative 6330 1913 Eidsvold, QLD Negative 9433 1^ Wentworthville, NSW Negative 8483 1^ Sydney, NSW Negative 6522 1913 Central Northern Queensland Negative 9946 1929 East Dorrigo, NSW Negative 18464 1^ Wollongong, NSW Negative 6523 1913 Central Northern Queensland Negative 12035 1937 Lindfield, NSW Negative 18768 1^ Shoalhaven Heads, NSW Negative 6524 1913 Central Northern Queensland Negative 15170 1954 Burrawang, NSW Negative 19428 1^ Hillgrove, NSW Negative 6648 1914 Herbert River, QLD Negative 15452 1959 Mt.Wilson, NSW Negative 19669 1958 Singleton, NSW Negative 6649 1914 Herbert River, QLD Negative 166523 1965 Rules Point, ACT Negative 19670 1958 Singleton, NSW Negative 6984 1917 Kimberleys, W.A Negative 156242 1999 Narooma, NSW Negative 19671 1958 Singleton, NSW Negative 7336 1921 Goangra, NSW Negative 158418 3^ Brennan's Creek, NSW Positive 19672 1958 Singleton, NSW Negative 7337 1921 Goangra, NSW Negative 158460 3^ Unknown Negative 19674 1958 Singleton, NSW Negative 7338 1921 Goangra, NSW Negative 158462 3^ Unknown Negative 19675 1958 Singleton, NSW Negative 7339 1921 Goangra, NSW Negative 161800 1995 Olney State Forest, NSW Negative 19676 1958 Singleton, NSW Negative 7464 1922 Dunoon, NSW Negative 161801 1995 Sweetmans Creek, NSW Negative 19677 1958 Singleton, NSW Negative 7594 1^ Unknown Negative 162436 2002 Taree, NSW Negative 19678 1958 Singleton, NSW Negative 9680 1928 Murray Island, Torres Strait Negative 162439 2002 Taree, NSW Negative 19679 1958 Singleton, NSW Negative 9721 1929 Groote Eylandt, NT Negative 162439 2002 Taree, NSW Positive 19680 1958 Singleton, NSW Negative 9722 1929 Groote Eylandt, NT Negative 162440 2002 Taree, NSW Negative 19681 1958 Singleton, NSW Negative 11021 1933 Groote Eylandt, NT Negative 162444 2002 Taree, NSW Negative 144869 1980 Royal National Park, NSW Negative 11053 1933 Moree, NSW Negative 162447 2002 Taree, NSW Negative 146888 1994 Homebush, NSW Negative 11070 1933 Wyong, NSW Negative 170167 3^ Unknown Positive 148744 1996 Homebush, NSW Positive 11154 1934 Tyalgum, NSW Negative 170178 3^ Unknown Negative 148745 1996 Homebush, NSW Negative 11161 1934 Lindeman Island, Whitsundays Negative 148746 1996 Homebush, NSW Negative 11593 1935 Pillaga, NSW Negative Cane toad ( Bufo marinus, syn. Rhinella marina ) (n=48) 149255 3^ Unknown Negative 11594 1935 Pillaga, NSW Negative AM# Date Location Result 150422 3^ Breadalbane, NSW Negative 11595 1935 Pillaga, NSW Negative 14306 2^ Queensland Negative 150424 3^ Breadalbane, NSW Negative 11706 1935 Lindeman Island, Whitsundays Negative 16373 1960 South Johnstone, QLD Negative 150425 3^ Port Stephens, NSW Negative 11781 1^ Nyngan, NSW Negative 16374 1960 South Johnstone, QLD Negative 150429 3^ Somersby, NSW Negative 12400 2^ Yirrkala, NT Negative 16376 1960 South Johnstone, QLD Negative 150994 1997 Rosebery,NSW Negative 12401 2^ Yirrkala, NT Negative 16377 1960 South Johnstone, QLD Negative 151330 1997 Rosebery,NSW Negative 12402 2^ Yirrkala, NT Negative 16378 1960 South Johnstone, QLD Negative 151994 1997 Rosebery,NSW Negative 12403 2^ Yirrkala, NT Negative 16384 1960 Sandy Creek, QLD Negative 153967 1997 Homebush, NSW Positive 13125 2^ Tenterfield, NSW Negative 17110 2^ Lappa Junction, QLD Negative 155970 1997 Homebush, NSW Negative 13126 2^ Tenterfield, NSW Negative 17111 2^ Lappa Junction, QLD Negative 159000 1993 Rosebery,NSW Negative 13645 1948 Cape Arnhem, NT Negative 27614 1964 Longford Creek, QLD Negative 166405 1974 Smiths Lake, NSW Negative 14802 2^ Bowen, QLD Negative 27615 1964 Longford Creek, QLD Negative 166407 1974 Smiths Lake, NSW Negative 15320 2^ Maclean, NSW Negative 27616 1964 Longford Creek, QLD Negative 166408 1974 Smiths Lake, NSW Negative 15708 1959 Goonoo State Forest, NSW Negative 27617 1964 Longford Creek, QLD Negative 170186 3^ Crescent Head, NSW Negative 99574 1966 Wallacia, NSW Positive 27618 1964 Gladstone, QLD Negative 171115 3^ captive reared Negative 99575 1966 Wallacia, NSW Positive 27619 1964 Gladstone, QLD Negative 175438 3^ Unknown Negative 27622 1964 Gladstone, QLD Negative Peron's tree frog (Litoria peronii) (n=16) 27623 1964 Gladstone, QLD Negative Green and golden bell frog (Litoria aurea ) NEW CALEDONIA (n=29) AM# Date Location Result 27624 1964 Gladstone, QLD Negative AM# Date Location Result 3938 1^ Unknown Negative 27625 1964 Gladstone, QLD Negative 165857 2003 Mt. Aoupinie, New Caledonia Negative 4295 1908 Darling River Floods, NSW Negative 27628 1964 Gladstone, QLD Negative 165858 2003 Mt. Aoupinie, New Caledonia Negative 4663 1910 Woonona, NSW Negative 56441 2^ Hilda Creek, QLD Negative 165859 2003 Mt. Aoupinie, New Caledonia Negative 5443 1^ Buddah Lake, NSW Negative 59918 1953 Munduberra, QLD Negative 165860 2003 Mt. Aoupinie, New Caledonia Negative 7528 1^ ClarenceRiver, NSW Negative 59919 1963 Gladstone, QLD Negative 165861 2003 Mt. Aoupinie, New Caledonia Negative 7529 1^ ClarenceRiver, NSW Negative 59920 1963 Gladstone, QLD Negative 165862 2003 Mt. Aoupinie, New Caledonia Negative 7964 1^ NSW Negative 59921 1963 Gladstone, QLD Negative 165863 2003 Mt. Aoupinie, New Caledonia Negative 8092 1^ Buladelah, NSW Negative 59922 1967 Mary River, QLD Positive 166009 2003 Plaine Des Lacs, New Caledonia Negative 8423 1^ Mount Horeb, NSW Negative 62794 2^ Palmwoods,QLD Negative 166026 2003 Plaine Des Lacs, New Caledonia Negative 9396 1^ Unknown Negative 63103 1977 Charters Tower, QLD Negative 166109 2003 Plaine Des Lacs, New Caledonia Negative 10690 1932 Yanco, NSW Negative 63104 1977 Charters Tower, QLD Negative 166110 2003 Plaine Des Lacs, New Caledonia Negative 10691 1932 Yanco, NSW Negative 107585 1982/3 Weipa, QLD Negative 166115 2003 Plaine Des Lacs, New Caledonia Negative 10693 1932 Yanco, NSW Negative 107830 1983 Weipa, QLD Negative 166139 2003 Plaine Des Lacs, New Caledonia Negative 10695 1932 NSW Negative 107831 1983 Weipa, QLD Negative 166141 2003 Plaine Des Lacs, New Caledonia Negative 10696 1932 Yanco, NSW Negative 107832 1983 Weipa, QLD Negative 166174 2003 Plaine Des Lacs, New Caledonia Negative 10697 1932 Yanco, NSW Negative 107843 1983 Weipa, QLD Negative 166175 2003 Plaine Des Lacs, New Caledonia Negative 107844 1983 Weipa, QLD Negative 166177 2003 Plaine Des Lacs, New Caledonia Negative Notes: 107845 1983 Weipa, QLD Negative 166178 2003 Plaine Des Lacs, New Caledonia Negative 107853 1983 Weipa, QLD Negative 166179 2003 Plaine Des Lacs, New Caledonia Negative Note that some vouchers are only dated based on a submission date, due to absence of 107854 1983 Weipa, QLD Negative 166180 2003 Plaine Des Lacs, New Caledonia Negative collection date in the Australian Museum records: 1^ collected from before 1935, 2^ collected 107855 1983 Weipa, QLD Negative 167109 2005 Bopope, New Caledonia Negative from 1936 to 1975, and 3^ collected from 1975 to present. 107857 1983 Weipa, QLD Negative 167110 2005 Bopope, New Caledonia Negative 107866 1983 Weipa, QLD Negative 167111 2005 Bopope, New Caledonia Negative AM# - catalogue number in the Australian Museum (Sydney, NSW, Australia) 107874 1971 Marlborough, QLD Negative 167112 2005 Bopope, New Caledonia Negative NSW - New South Wales, ACT - Australian Capital Territory, QLD - Queensland 107875 1971 Marlborough, QLD Negative 167113 2005 Bopope, New Caledonia Negative 107876 1971 Marlborough, QLD Negative 167114 2005 Bopope, New Caledonia Negative 107877 1971 Marlborough, QLD Negative 167115 2005 Bopope, New Caledonia Negative 140594 1992 Alstonville, NSW Negative 167117 2005 Bopope, New Caledonia Negative 147183 1995 Terry Hills, NSW Negative 167118 2005 Bopope, New Caledonia Negative 158500 1998 Burringar, NSW Positive

Chapter 3: A suspected parasite spill-back of two novel Myxidium spp. (Myxosporea) causing disease in Australian endemic frogs found in the invasive Cane toad

22 A Suspected Parasite Spill-Back of Two Novel Myxidium spp. (Myxosporea) Causing Disease in Australian Endemic Frogs Found in the Invasive Cane Toad

Ashlie Hartigan1, Ivan Fiala2, Iva Dykova´ 2, Miloslav Jirku˚ 2, Ben Okimoto3, Karrie Rose4, David N. Phalen1, Jan Sˇ lapeta1* 1 Faculty of Veterinary Science, The University of Sydney, Sydney, New South Wales, Australia, 2 Institute of Parasitology, Biology Centre, Academy of Sciences of the Czech Republic, Cˇ eske´ Budeˇjovice, Czech Republic, 3 Honolulu Zoo, Honolulu, Oahu, Hawaii, United States of America, 4 Australian Registry of Wildlife Health, Taronga Conservation Society Australia, Mosman, New South Wales, Australia

Abstract Infectious diseases are contributing to the decline of endangered amphibians. We identified myxosporean parasites, Myxidium spp. (Myxosporea: Myxozoa), in the brain and liver of declining native frogs, the Green and Golden Bell frog (Litoria aurea) and the Southern Bell frog (Litoria raniformis). We unequivocally identified two Myxidium spp. (both generalist) affecting Australian native frogs and the invasive Cane toad (Bufo marinus, syn. Rhinella marina) and demonstrated their association with disease. Our study tested the identity of Myxidium spp. within native frogs and the invasive Cane toad (brought to Australia in 1935, via Hawaii) to resolve the question whether the Cane toad introduced them to Australia. We showed that the Australian brain and liver Myxidium spp. differed 9%, 7%, 34% and 37% at the small subunit rDNA, large subunit rDNA, internal transcribed spacers 1 and 2, but were distinct from Myxidium cf. immersum from Cane toads in Brazil. Plotting minimum within-group distance against maximum intra-group distance confirmed their independent evolutionary trajectory. Transmission electron microscopy revealed that the brain stages localize inside axons. Myxospores were morphologically indistinguishable, therefore genetic characterisation was necessary to recognise these cryptic species. It is unlikely that the Cane toad brought the myxosporean parasites to Australia, because the parasites were not found in 261 Hawaiian Cane toads. Instead, these data support the enemy-release hypothesis predicting that not all parasites are translocated with their hosts and suggest that the Cane toad may have played an important spill-back role in their emergence and facilitated their dissemination. This work emphasizes the importance of accurate species identification of pathogens relevant to wildlife management and disease control. In our case it is paving the road for the spill-back role of the Cane toad and the parasite emergence.

Citation: Hartigan A, Fiala I, Dykova´ I, Jirku˚ M, Okimoto B, et al. (2011) A Suspected Parasite Spill-Back of Two Novel Myxidium spp. (Myxosporea) Causing Disease in Australian Endemic Frogs Found in the Invasive Cane Toad. PLoS ONE 6(4): e18871. doi:10.1371/journal.pone.0018871 Editor: Anastasia P. Litvintseva, Duke University Medical Center, United States of America Received October 31, 2010; Accepted March 22, 2011; Published April 25, 2011 Copyright: ß 2011 Hartigan et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This study was financially supported by the Australian Academy of Science, Dr William Richards Award in Veterinary Pathology, ARC/NH&MRC Network for Parasitology, and in part by the Grant Agency of the Academy of Sciences of the Czech Republic (KJB600960701) and the Grant Agency of the Czech Republic (204/09/P519). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction from Europe as the result of the translocations of infected fish and Infectious diseases entering naı¨ve populations are key threaten- the cosmopolitan distribution of its aquatic invertebrate host – an ing processes contributing to the precipitous global decline of oligochaete worm Tubifex tubifex [10,12]. biodiversity [1]. Currently, more than three quarters of critically Pathological changes associated with Myxidium sp. (Myxosporea) endangered species of amphibians are threatened by infectious were described in livers of green tree frogs (Litoria caerulea) and it disease [2]. The chytrid fungus, Batrachochytrium dendrobatidis [3,4], was suggested that investigations of declining frog populations and ranaviruses [5] have been documented as significant drivers of should consider Myxidium spp. as potential pathogens [13]. These frog population declines. Circumstantial evidence suggests that parasites were assumed to be Myxidium immersum of the invasive myxosporean parasites are also causing significant disease in Cane toad (Bufo marinus, syn. Rhinella marina) and thought to have amphibians and may be playing a role in their decline [6,7,8,9]. If spilled over into a wide range of Australian frogs [14]. Recently, myxosporean parasites of amphibians behave in a similar manner we confirmed the emergence of Myxidium parasites in native frogs to those of fish, then they would have a potential for dissemination in Australia, by documenting their earliest occurrence in a outside of their original range and could be pathogenic in naı¨ve specimen from 1966, therefore 31 years after the introduction of populations [7,10]. The infamous Myxobolus cerebralis (Myxosporea) the Cane toad in Australia [15]. These findings suggested that causes a whirling disease in salmonid fish and has been proven to genetic characterisation is required to confirm the identity of these have a devastating impact on wild and farmed fish populations in parasites in Australian frogs and in Cane toads along their North America [11]. Myxobolus cerebralis was able to spread globally translocation route [16].

PLoS ONE | www.plosone.org 1 April 2011 | Volume 6 | Issue 4 | e1887123 Novel Parasites in Australian Frogs

The aim of our study was to elucidate the identity of biliary ducts or other stages either in the liver parenchyma or myxosporean parasites in brains and livers of Australian native blood smears from the Green and Golden Bell frog specimens. frogs, the Green and Golden Bell frog (Litoria aurea) and the Given the apparent ability of Myxidium species to cause disease Southern Bell frog (Litoria raniformis) of south east Australia [17], in two species of endangered frog the Australian Registry of and the invasive Cane toad using histological, ultrastructural and Wildlife Health (ARWH) amphibian data base was surveyed for genetic characterisation. Because of the apparent identity of these evidence of infection in other species of frogs from New South parasites in both native frogs and invasive Cane toads we Wales. Twenty six cases, in nine native frog species, were identified investigated the possibility that these parasites were translocated during the period of 1997–2009 that had myxosporean develop- into Australian frog populations with the Cane toad. We surveyed mental stages either in the brain and nerve ganglia or the bile Cane toads in Hawaii and compared Australian myxospores with ducts (Table S1). Liver lesions associated with the parasite those from the Cane toads in South America. Our data supports a development were consistent with findings in green tree frogs, recent emergence of myxosporea in Australian native frogs, Litoria caerulea [13]. Lethargy, emaciation or behavioural abnor- however, we found no evidence to suggest that it was introduced to malities were reported in 26.9% (7/26) of these cases. Given that Australia with the Cane toad. Using these data we discuss the many of these frogs were examined as part of a pre-translocation mechanisms by which these parasites have dispersed among health screen, these findings raised a concern that these parasites Australian native and endangered frogs. may be impacting captive breeding colonies and animals released from captive breeding colonies may be serving to disseminate these parasites. Results and Discussion Myxosporean parasites in Australian endangered frogs Myxosporean development confined within myelinated A subset of Green and Golden Bell frog tadpoles (n = 38) from a axons of the central nervous system of frogs semi-captive population in greater Sydney, New South Wales were Transmission electron microscopy (TEM) of the brain stages submitted for necropsy as part of a routine health screening revealed extrasporogonic developmental stages exclusively within protocol required before animals from this facility could be myelinated axons (black arrows, Figure 2AB). Myelinated axons translocated. At the time it was reported that some tadpoles were containing these stages were consistently several times larger than exhibiting behavioural changes. Following the preliminary nec- the largest normal myelinated fibre (compare the parasitised axon ropsy findings, 10 adult Green and Golden Bell frogs were also with normal axons - white arrows, Figure 2A). These stages submitted for necropsy from this population. Wild caught, adult possessed characteristic cell in cell development consistent with Southern Bell frogs (n = 8) from southern New South Wales that myxosporean development [23]. They consisted of round primary had been held in a captive breeding program were also submitted cells ranging from 5–15 mm in diameter and distinct intracellular cleavages defining the secondary cell development within (blue vs. for diagnostic necropsy as part of an investigation into their weight red area, inset of Figure 2A). Extrasporogonic stages were seen to loss, lethargy, and failure to breed. Both frog species have expand to the absolute periphery of the axon membrane in some experienced dramatic population declines across all of their range cases appearing to cause axonal swelling. Axonal lesions may since the 1980s [18,19]. Chytridiomycosis has been hypothesised interfere with neurological function and could explain the to be a driver in the decline, but cannot account for declines across behaviour changes observed in the Green and Golden Bell frogs. the entire range of both species as chytrid-free populations have Myxosporean parasites develop in a variety of tissues, however, also been lost [19,20,21]. Captive breeding programs for these intra-axonal development is uncommonly reported and has only frogs have been established and may play a critical role in saving been described in a small number of species of fish [24,25]. Intra- them from extinction [17,22]. axonal stages resembling those described here were seen in 12 out Multinucleated, variably-shaped, 5–30 mm in diameter, organ- of 22 spinal cords of the South American toad (Bufo arenarum, syn. isms consistent with myxosporean plasmodia were found within Rhinella arenarum) from Uruguay using TEM, although at the time, bile ducts of tadpoles of the Green and Golden Bell frog (14/38) the authors were uncertain of the identity of the organisms they (Figure 1ABC). Accompanying the organisms were varying were describing [26]. Similarly, parasite stages in the brains of the degrees of hepatic lesions characterised by biliary duct hyperplasia, Cane toad in South America observed under light microscopy and moderate to severe loss of peribiliary hepatocytes, hepatitis with originally interpreted to be Toxoplasma gondii, resemble the brain lymphoplasmacytic infiltration, and fibrosis (Figure 1A). Round forms described here [27]. Despite the close morphological and oval myxosporean stages, presumed to be early myxosporean similarity between the organisms found in the brains of the plasmodia, were also identified throughout the central nervous Australian frogs and the South American toads, it is not possible to system (CNS; brain and spinal cord) and root ganglia of tadpoles determine their relationship because other tissues were not and adults of the Green and Golden Bell frog as well as adults of examined (i.e., liver and gallbladder) and they have not been the Southern Bell frog (Figure 1DE). The size of the CNS stages genetically characterised. ranged from 5–25 mm in their widest dimension (Figure 1E). The TEMs revealed plasmodia in the early stage of develop- Each stage in CNS consisted of numerous secondary cells enclosed ment in the bile ducts of the Green and Golden Bell frog tadpoles. in a primary cell wall. These brain stages were accompanied by a These stages also had cell in cell organisation, with no attachment multifocal nonsuppurative meningoencephalitis found in 7 of 8 to the biliary epithelium seen (Figure 2CD). The coelozoic adults of the Southern Bell frog (Figure 1D). Adults of both plasmodia in bile ducts ranged from 4–25 mm in diameter, had Green and Golden Bell frogs (n = 1) and Southern Bell frogs (n = 7) numerous mitochondria, pinocytic channels and dispersed lipid were found to have plasmodia containing myxospores in the gall inclusions in the cytoplasm (Figure 2D). Compared to the bladder. The myxospores morphologically resembled those of the extrasporogonic stages in the brain, the primary cells in the bile genus Myxidium. One Green and Golden Bell tadpole was found to ducts had fewer daughter cells. It is suspected that the liver stages have immature plasmodia in its gall bladder which did not contain were in an earlier stage of development than those found in the mature myxospores. The plasmodia were oval to round, ranged brain, because advanced primary cells of myxosporea typically from 0.5–2 mm diameter. We did not identify any myxospores in contain large numbers of daughter cells [28].

PLoS ONE | www.plosone.org 2 April 2011 | Volume 6 | Issue 4 | e1887124 Novel Parasites in Australian Frogs

Figure 1. Myxosporea in the Green and Golden Bell frog (Litoria aurea) and the Southern Bell frog (Litoria raniformis). Affected tadpoles of Green and Golden Bell frog suffered from chronic hepatitic lesions characterised by biliary hyperplasia, hepatitis with lymphoplasmacytic infiltration and fibrosis (A), parasite development was observed inside the bile ducts (arrows). Liver tissue of tadpoles from a semicaptive population of Green and Golden Bell frogs contained early plasmodia found within bile ducts (B, C). Round myxosporean stages (arrow), were identified in the brain and spinal cord and meninges of tadpoles of the Southern Bell frog (D, E). Brain stages were accompanied by a multifocal nonsuppurative meningoencephalitis (D). Each consisted of several secondary cells enclosed in a primary cell wall (E). Scale bars: A - 50 mm; B, C, D, E - 10 mm. A, D, E - H&E, B, C - Giemsa. doi:10.1371/journal.pone.0018871.g001

Australian native and exotic frogs are parasitised by the To determine if both genotypes were present in other species of same Myxidium spp. genotypes frogs, we collected Striped Marsh frogs (Limnodynastes peronii) and Sequence analysis confirmed the presence of myxosporea in Peron’s Tree frogs (Lit. peronii) from the property where the both brain and liver tissues – the liver and the brain genotypes of population of the tadpoles of the Green and Golden Bell frog were Myxidium spp. (Figure 3, Table S2). The SSU rDNA sequence raised. The Striped Marsh frog adults had a high prevalence of amplified from the liver and brain of the Green and Golden Bell myxospores and plasmodia (19/22) within their gall bladder and frog differed by 9% (855 pairwise alignment length). The two were found to have the brain stage (2/39). Plasmodia with sequences obtained from the Southern Bell frog brain and myxospores were found in bile ducts of 2/38 tadpoles of the myxospores were identical to each other and matched (100%) Striped Marsh frog and amplified sequences matched the liver the sequence from the brain of the Green and Golden Bell Frog. genotype. The Peron’s Tree frog adults (1/3) were found to possess The rapidly evolving ITS rDNA region spanning across ITS1 large numbers of brain extrasporogonic plasmodia as well as rDNA and ITS2 rDNA sequences differed by 34% (375 pairwise myxospores in its gall bladder. Sequencing confirmed identity with alignment length) and 37% (314 pairwise alignment length), the brain genotype of the Green and Golden Bell frog respectively. Similarly, more conservative LSU rDNA revealed (Figure 3AB). 7% (2,195 pairwise alignment length) difference between the two Both genotypes were sequenced from isolated myxospores in sequences. A query of the public repositories of SSU rDNA gall bladders of adult Cane toads from Lismore and Byron Bay in sequence data using ‘blastn’ returned myxosporean sequences as northern New South Wales (Figure 3AB). In Australia, the Cane the most closely related sequences including Myxidium melleni toad is an introduced species that is still expanding its range with a (DQ003031) from the gall bladder of the Western Chorus frog population confined to the north east corner of the New South (Pseudacris triseriata triseriata) in North America [29]. Wales [30,31]. Lismore is the southern most front of the invading

PLoS ONE | www.plosone.org 3 April 2011 | Volume 6 | Issue 4 | e1887125 Novel Parasites in Australian Frogs

Figure 2. Transmission electron microscopy (TEM) of the developmental stages in the Green and Golden Bell frog tadpoles (Litoroa aurea). Brain multinucleated (A) and single nucleated (B) extrasporogonic developmental stages found exclusively within axons (a), enclosed by myelin sheets (black arrows). Developmental stages possessed characteristic cell in cell development consistent with myxosporean extrasporogonic development. The histozoic primary (blue) cells, inside myelinated axon (green), had distinct intracellular cleavages defining the secondary (red) cells (an enlarged area in the white quadrangle; inset A). The parasitised myelinated axon is several times larger in size than the normal myelinated fiber (white arrows). Liver bile duct plasmodia (C) with cell in cell development, primary cell (p) and secondary cell (s). Based on TEM these stages appeared to be unattached to the biliary epithelium (c). The coelozoic plasmodia in bile ducts had numerous mitochondria (m) in the cytoplasm with pinocytic channels (gray arrows) and dispersed lipid inclusions in the cytoplasm (D). Scale bars: A - 1 mm; B - 1 mm; C, D - 2 mm. doi:10.1371/journal.pone.0018871.g002 toads [32]. Initially we collected gall bladders and found 10% (6/ gall bladder of the Southern Bell frog with brain myxosporean 60) and 20% (2/10) Cane toads from Byron Bay positive for development were identified as the brain genotype. Subsequently, myxospores in February and August 2009, respectively. In the myxospores in the gall bladder of Cane toads were February 2010, toad sampling revealed 42% (37/82) of toad gall morphologically indistinguishable from those above and sequenc- bladders positive for myxospores and presence of extrasporogonic ing revealed presence of both liver and brain genotypes plasmodia in 10% (3/30) of brains (sampled in and in between (Figure 3AB). Together with the morphological details of other Byron Bay and Lismore). Brain plasmodia seen in the toads Myxidium spp. of frogs (Table S3), our data suggest that the resembled those seen in the Green and Golden Bell frog, the myxospores of divergent amphibian Myxidium species have Southern Bell frog and Peron’s Tree frog. maintained similar structural characteristics and that this shared Using spore morphology, the myxosporea found in Australian morphology represents an optimal phenotype that is under strong frogs belonged to the genus Myxidium Bu¨tschli, 1882 (Myxosporea, stabilising selection pressure [33]. Myxozoa). All their spores share a characteristic fusiform shape with shell valves either smooth or ridged, suture line cross- sectioning the spore, and two polar capsules that lie one at each Australian Myxidium spp. are not found in Cane toads end of the spore (Figure 4) [23]. Myxidium-myxospores recovered from Brazil and Hawaii from the Green and Golden Bell frog, the Southern Bell frog, the The Cane toad was introduced to a large number of countries Striped Marsh frogs, the Peron’s Tree frog and the introduced and states across the world from South America [34]. In Brazil, Cane toad were of similar size and with overlapping morpholog- the Cane toad is the host of Myxidium immersum [35]. The parasite ical details using light microscopy (Table S3). Genotyping was speculated to have been introduced with the Cane toad and to revealed that the Myxidium-myxospores in the gall bladder of the have spread widely in native Australian frogs [14,36]. Genetic Striped Marsh frog, but with no detectable brain lesions, were characterisation of M. cf. immersum myxospores isolated from gall identified as the liver genotype and Myxidium-myxospores in the bladders of the Cane toad from Brazil, South America two distinct

PLoS ONE | www.plosone.org 4 April 2011 | Volume 6 | Issue 4 | e1887126 Novel Parasites in Australian Frogs

Figure 3. Relationship of genotypes of Myxidium spp. in Australian frogs. (A) Phylogenetic tree based on SSU rDNA sequences including the North American M. melleni and rooted using a shrew parasite Soricimyxum fegati. Parasite hosts and their status in Australia are indicated for the brain and liver genotype. (B) Phylogenetic trees based on SSU rDNA, LSU rDNA, ITS rDNA and ITS2 rDNA with genotype names and mean distances are indicated on the right. (C) The intraspecific and interspecific distance of the brain genotype (blue N), liver (red n), Brazil-1 (green n), Brazil-2 (brown n) and pooled liver+Brazil-1+Brazil-2 genotypes (black outlined n) were placed on the graph to evaluate whether they represent candidate species. The graphs are divided into four quadrants that represent different categories of ‘‘species’’ [37]: top left - species concordant with current taxonomy; top right - probable composite species, i.e. candidates for taxonomic split; bottom left - species that have undergone recent divergence, hybridization, or synonymy; bottom right - probable specimen misidentification. Notice that if the liver+Brazil-1+Brazil-2 genotypes (black outlined n) are treated as a single species they are resolved in the top right quadrangle suggestive of cryptic species; however when split into the three individual genotypes they became resolved in the top left quadrangle supporting their species status. Trees and distances were inferred using the Minimum Evolution: Maximum Composite Likelihood method in MEGA4 with bootstrap test (1,000 replicates, .50% are shown). doi:10.1371/journal.pone.0018871.g003 groups of sequences (Brazil-1 and Brazil-2 genotypes, Figure 3 were not found in any of the toads (n = 261) collected on Oahu in AB). However, neither of these M. cf. immersum genotypes matched 2010 (Figure 5). The probability of observing zero positives in a the Myxidium spp. brain or liver genotype from Australian frogs sample of 261 frogs is 261026, using the expected minimum (Figure 3, Table S2). The Cane toad arrived in Australia from prevalence of 5%. Based on these findings we concluded that the the Hawaiian island of Oahu in 1935 [34]. However, myxosporea Hawaiian population of Cane toads is free of these myxosporeans

PLoS ONE | www.plosone.org 5 April 2011 | Volume 6 | Issue 4 | e1887127 Novel Parasites in Australian Frogs

Figure 4. Myxospores of Myxidium species - liver genotype. Scanning electron microscopy of myxospores recovered from the gall bladder of the Striped Marsh frog (Limnodynastes peronii). Myxospores are ellipsoidal, shell valves have ridges and suture line cross-sectioning the spore. Scale bar, 5 mm. doi:10.1371/journal.pone.0018871.g004 and that they were not the source of the myxosporeans that we identified. However, pooling together the liver, Brazil-1 and have identified in Australian frogs and in Cane toads in Australia. Brazil-2 sequences suggests that this group consisted of multiple All four DNA markers have resolved identified Myxidium spp. ‘candidate’ species (black outlined n, Figure 3C). We have genotypes as ‘candidate’ species using barcoding analysis used to arbitrarily applied 2–3% threshold, used in COI of mtDNA unmask cryptic species complexes (Figure 3C) [37,38]. The barcoding studies [37,39], for myxosporea more conservative brain, liver, Brazil-1 and Brazil-2 sequence group appeared in the regions (SSU rDNA, LSU rDNA), and suggest using even higher top left quadrangle representing species that are correctly thresholds (5–10%) for more variable regions (ITS1 rDNA, ITS2

Figure 5. Collection sites of Cane toads (Bufo marinus, syn. Rhinella marina) surveyed for the presence of Myxidium species. Names of the localities are shown together with the number of animals surveyed in February 2010. The Cane toad was introduced (149 individuals) to the Hawaiian island of Oahu from Puerto Rico in April 1932 and introduced to the Manoa Arboretum at the upper end of Manoa Valley and a taro patch adjoining the HSPA Waipio substation and immediately spread throughout the island [34]. In 1935, juvenile and adult toads were collected and in total 101 live individuals arrived in Australia [34]. Two localities recorded as source populations for translocation to Australia are indicated by w. doi:10.1371/journal.pone.0018871.g005

PLoS ONE | www.plosone.org 6 April 2011 | Volume 6 | Issue 4 | e1887128 Novel Parasites in Australian Frogs rDNA). This diversity observed between these sequence groups genotyping and morphological analysis of myxospores revealed exceeds intra-species variation within myxosporean species great diversity of the genus Ceratomyxa in fish on the Great Barrier [12,40]. The within Myxidium spp. group diversity was below Reef, with 10 new species varying from each other by 1.3% to 1%, except for the liver sequences at ITS1 rDNA (1.54%) and 28.0% at SSU rDNA [42]. ITS2 rDNA (1.39%), and for Brazil-2 sequences at ITS1 rDNA Phylogenetic reconstructions demonstrated monophyly of the (1.26%) (Figure 3B). Within a naturally dispersed cosmopolitan liver, brain, Brazil-1, Brazil-2 genotypes and M. melleni (Figure 6, parasite such as Kudoa thyrsites genetic analysis revealed 0.3% to Figure S1). The multiple sequence alignment included a selection 7.5% variation of ITS1 rDNA in an individual fish sample [40]. of key myxosporean species with attention to select taxa from Furthermore, major separations of the marine K. thyrsites and the different host groups and those with different tissue tropisms to freshwater Tetracapsuloides bryosalmonae were correlated with geo- map the biology of the frog myxosporean species groups (see graphic regions and exhibited regional differentiation [40,41]. The Material and methods; Table S4). The myxosporean sequences

Figure 6. Phylogenetic tree of Myxozoa based on SSU rDNA gene sequence. The Bayesian tree was reconstructed using MrBayes 3.1.2 with a GTR+G+I nucleotide model from the stringent Alignment 3, for details about alignment and phylogenetic reconstruction see Materials and Methods and Figure S1. For details on biological traits and GenBank accession numbers see Table S4. Bayesian posterior probabilities are shown at the nodes. The myxosporean sequences fall into marine and fresh clades; the exceptions are Sphaeromyxa spp. and Zschokkella icterica sequences from marine fish that cluster in the fresh water clade [43]. Myxosporean species with documented actinospore development are underlined. doi:10.1371/journal.pone.0018871.g006

PLoS ONE | www.plosone.org 7 April 2011 | Volume 6 | Issue 4 | e1887129 Novel Parasites in Australian Frogs were separated into mostly marine and freshwater species clades likely present in a low prevalence or have been in a geographically [43,44,45]. Moreover sequences tended to group according to host localised area, which may not have been sampled in past frog groups within the freshwater clade rather than their tissue tropisms surveys [15]. The requirement in the parasite spill-back theory is (Figure 6). Tree reconstruction using distance method and that (i) the invasive host, in our case the Cane toad, acquires and LogDet distance model that takes into account the nonstationary shares a native parasite and (ii) subsequently can successfully nucleotide frequencies to eliminate artificial grouping of sequences disseminate it, due to the high parasite prevalence in the invasive with similar nucleotide frequencies irrespective of true evolution- host. Alternatively, if the invasive host does not sufficiently ary relationship revealed trees congruent with the maximum multiply the parasite, the exotic host may serve as a sink for likelihood and Bayesian trees (Figure S1). The stringent parasites and reduces the overall burden of parasites in the native alignment (Alignment 3) revealed a robust phylogenetic position species due to a dilution effect [57]. Our situation satisfies both (ML 99% bootstrap, BP 100% posterior probability and NJ 100% conditions of the spill-back theory [55]; (i) the native Australian bootstrap) of the brain genotype as a sister group to the frogs share the parasites with the invasive Cane toad, and (ii) there monophyletic clade of liver, Brazil-1 and Brazil-2 genotypes, is a high prevalence of Myxidium-myxospores in the gallbladders of together with M. melleni. Interestingly, these frog Myxosporea Cane toads in Australia 38% (13/34) from 1980’s [14] and up to formed a sister group to the only sequenced myxosporean, 42% in our data from 2010. Therefore, we hypothesize that the Soricimyxum fegati (EU232760), from a mammalian host, the invasive Cane toad was a susceptible and amplifying host for Common shrew, Sorex araneus. Despite its low support values Myxidium spp. under Australian conditions. This hypothesis, (ML 52% bootstrap, BP 66% posterior probability and NJ 80% however, cannot account for all the data, as both the brain and bootstrap) the topology was reconstructed using all methods used liver genotype have been detected outside the current Cane toad’s (Figure S1). range, i.e. greater Sydney and southern New South Wales. Possible explanations for this discrepancy include a second spill- Conclusion over event where these parasites infected a naı¨ve frog species Accurate species identification is critical for informed wildlife allowing the parasite to spread beyond the range of the Cane toad. management, as well as having major implications for diagnosis, An investigation into the quantitative capacity of each susceptible prevention and control of disease agents in animals and humans frog species to produce myxospores will be required before it is [46,47]. Distinguishing exotic and invasive pathogens from native possible to start answering this question. Similarly, investigation pathogens requires pre-existing knowledge of their historical leading to the elucidation of the Myxidium life cycle in frogs will be distribution – a term ‘cryptogenic’ species has been coined to essential to fully understand the extent of the distribution and risk those species that we can not unambiguously place in either that these Myxidium species pose to the already diminishing frog category [48]. This distinction is particularly challenging for diversity. This study highlights the insight that can be acquired Myxidium spp. in Australian frogs where no historical baseline data from studying host-parasite systems disturbed by an invasive host, exists and where an invasive host may have carried parasites to but it also illustrates the limitations associated with absence of base new places, because the enemy-release hypothesis predicts that not line data on parasite distributions. all parasites are translocated with their hosts [49,50]. Moreover, Myxidium spp. have the extra challenge of finding the invertebrate Materials and Methods host necessary for successful establishment in a new environment unless the invertebrate host has been introduced concurrently or is Samples cosmopolitan [10]. Ethics statement. Tadpoles and frogs of the Green and To answer the question from where the cryptogenic myxospor- Golden Bell frog (Litoria aurea), Peron’s tree frog (Litoria peronii), ean parasites in Australian frogs emerged, we have evaluated Striped Marsh frog (Limnodynastes peronii) and the Cane toad (Bufo whether they are either (i) native or (ii) exotic - they have invaded marinus, syn. Rhinella marina) from New South Wales (NSW), Australia. Traditional concepts in invasive biology revolve around Australia were sampled over a period of two years (2008–2010) an invasive species bringing a pathogen along [16,47,51,52]. under the Department of Environment and Climate Change and Assuming this agent of disease is not strictly specific to its original Water NSW (Australia) scientific license (S12686 and S12969) and host it spills-over to the local naı¨ve fauna with potentially the University of Sydney Animal Ethics Committee approval devastating consequences [51,52,53,54]. An alternative to the (N00/9-2008/3/4855 and N00/9-2009/3/5134). All animals in spill-over concept is a recently revived theory in invasive ecology NSW were euthanised with AQUI-S (isoeugenol; AQUI-S Ltd., called parasite spill-back [55,56]. The native parasite spill-back New Zealand) according to standard protocol [58]. The Cane toad requires only an invasive host species that is susceptible to a native (B. marinus) in Hawaii, USA is a pest animal; individuals were parasite. Thus, the potentially serious consequences are not caused euthanised immediately in FINQUEL MS-222 (tricaine by the exotic parasite but the native parasite that is being amplified methanesulfonate; Argent Laboratories, WA) and isoflurane by by the susceptible invasive host species. The majority of parasites the Honolulu Zoo veterinary staff according to a standard humane in exotic species (67%) are native parasites implying that exotic euthanasia protocol at the Zoo facility (February 2010). The Cane species commonly have the potential to cause parasite spill-back to toad (B. marinus) material from Brazil was kindly donated by Prof. native hosts, with effects at both the host individual and population Ralph Lainson (Departamento de Parasitologia, Instituto Evandro scale [55]. Because, (i) our investigation of Hawaiian Cane toads Chagas, Bele´m, Para´, Brazil) to the laboratory in the Czech suggests, that they have cleared myxosporean parasites during Republic (MJ). All animal work was conducted according to their previous translocations, and (ii) the parasites in Brazilian relevant national and international guidelines. Cane toads are genetically distinct from the gall bladder Australian native frogs. Tadpoles and frogs of the Green myxosporea in Australian frogs, we suggest that the most plausible and Golden Bell frog, Peron’s Tree frog, Striped Marsh frog were explanation is the parasite spill-back. collected from a semi-captive Green and Golden Bell frog The question remains why there were no myxosporean parasites breeding site in greater Sydney, NSW, Australia. This detected prior to the introduction of the Cane toad in 1935 population is actively managed to facilitate increased breeding [14,15]. We offer the following explanation, the parasites were but individuals of any species can move back and forth between

PLoS ONE | www.plosone.org 8 April 2011 | Volume 6 | Issue 4 | e1887130 Novel Parasites in Australian Frogs the property and the external environment. Immediately after manipulations were made at the original dimensions and at euthanasia, all animals were dissected and internal organs 300 dpi. processed for transmission electron microscopy (liver and brain), histopathology and storage at 220uC. All paraffin blocks were Transmission and scanning electron microscopy submitted to Australian Registry of Wildlife Health (ARWH, Small portions of brain and liver (,1 mm3) were processed for Taronga Conservation Society Australia, Mosman, NSW, transmission electron microscopy (TEM) immediately after Australia). The captive Southern Bell frog adults (Litoria euthanasia; fixed in 2.5% glutaraldehyde in 0.1M sodium raniformis) originally from southern New South Wales, Australia (pH 7.0) cacodylate buffer and then postfixed with 2% osmium were submitted for health screening through the ARWH; tetraoxide (OsO4), dehydrated through graded ethanols and formalin-fixed histological section and tissues in 280uC were embedded in Spurr’s Resin. The blocks were sectioned, stained retrieved from this study. Gallbladders were removed from all and observed at either the Electron Microscopy Unit (University of animals and inspected for the presence of myxospores, if present, Sydney) on a Philips CM12 TEM or at the Institute of myxospores were washed in sterile phosphate buffered saline and Parasitology (Academy of Sciences of the Czech Republic) on a kept at 4uC. JEOL JEM 1010 TEM. Myxospores from Striped Marsh frog’s The Australian Registry of Wildlife Health records. We gallbladder were prepared for scanning electron microscopy queried the database for all frog entries with a diagnosis of (SEM); placed on poly-L-lysine coated coverslips, fixed in 1% unknown liver and brain ‘protozoan’ infections. Paraffin block and OsO4 in 0.2 M sodium cacodylate buffer (pH 7.0), rinsed in slides were retrieved and inspected for the presence of distilled water, dehydrated with a graded acetone series, critical myxosporean-like stages. point dried, and coated with gold. Myxospore examination was at Australian invasive Cane toad. Cane toads were sampled either the Electron Microscopy Unit (University of Sydney) using a from the southern end of their expanding range, Lismore and Philips XL-30 CP SEM or at the Institute of Parasitology Byron Bay, New South Wales, Australia. Animals were euthanised (Academy of Sciences of the Czech Republic) on a JEOL JEM and immediately processed for histopathology and their gall 7401F FE-SEM. Both FE-SEM allowed topographic imaging to bladders inspected for the presence of myxospores as described ,1 nm resolution by an efficient in-lens detector. Greyscale above. images were acquired as TIFF files. Images were imported as Hawaiian Cane toads. Cane toads from Oahu, Hawaii, US greyscale images into Creative Suite CS3 (Adobe Systems, San were captured (n = 261) by net and hand from areas close to the Jose, CA). Adjustment layers for levels and brightness/contrast original release and capture sites according to Easteal [34]. were used to find the optimal grey balance for investigated features Animals were euthanised in FINQUEL MS-222 and isoflurane by with the original image in the background. For SEM images the Honolulu Zoo staff according to a standard protocol and gall individual myxospores were traced and presented on a black bladders were examined immediately for the presence of background. All manipulations were made at the original myxospores by wet mount. Portions of liver/gall bladder dimensions and at 400 dpi. (n = 179) and brain (n = 47) were formalin-fixed for histological processing. The probability of observing no positive samples out of DNA extraction, amplification protocols and sequencing the total number of samples collected was calculated using Tissues for DNA extraction were collected from brain and liver FreeCalc v2; an epidemiological probability calculator and frozen at 220uC prior to DNA isolation. Total DNA was demonstrating freedom from disease [59]. extracted from 20–40 mg of tissue using the PureLink DNA Kit Brazilian Cane toads. A pooled 70% ethanol fixed sample (Invitrogen Australia, Victoria, Australia) or the DNeasy Blood from 10 individual gall bladders with myxospores of Myxidium cf. and Tissue Kit (Qiagen, Victoria, Australia) according to the immersum dissected from Cane toads collected in the vicinity of manufacturers’ instructions. DNA from myxospores was extracted Manaus, Brazil was shipped to and genetically characterised in the using the FastDNA Soil Kit Protocol with a Fast Prep-24 Czech Republic. Homogenisation System equipped with QuickPrep Adapter (MP Biomedicals, Australia); the speed setting used was 6.0 for 40 s, Microscopic and histopathological examination otherwise the manufacturer’s instructions were followed. Extracted The entire gall bladder contents were examined (wet mount) DNA from all samples was resuspended in a final volume of 200 ml using 206 and 406 objectives. Myxospores were measured/ (Purelink, DNAeasy) or 50 ml (FastDNA Soil). DNA was quantified photographed using a 1006 oil-objective BX60 microscope and purity assessed using Nanodrop 1000 spectrophotometer equipped with a DP70 camera (Olympus Australia); a minimum (Thermo Scientific, NSW, Australia). Total DNA aliquots were of 30 myxospores from each individual was measured and stored at 220uC prior to PCR analysis. morphological details evaluated according to Lom and colleagues The small subunit rDNA (SSU rDNA) was amplified using [28]. Animal tissue or entire individuals were processed for MyxoSpecF (59-TTC TGC CCT ATC AAC TWG TTG-39) and histology; fixed in buffered formalin (10%), processed using MyxoSpec R (59-GGT TTC NCD GRG GGM CCA AC-39) standard technique, embedded in paraffin, sectioned and stained primer pair [43]. The internal transcribed spacers (ITS) and 5.8S with haemotoxylin and eosin (H&E) or Giemsa (Veterinary rDNA was amplified using G-ITS-for (59-GGG ATC CGT TTC Histopathology Laboratory, University of Sydney). Individual gall CGT AGG TGA ACC TGC-39) annealing to the 39-end of SSU bladders were selected for paraffin embedding, sectioning, Giemsa rDNA and ITS-Zschok-rev (59-GAT TCT CAT AGT AAC TGC staining and examined for the presence of myxospores. All sections GAG TG-39) annealing to the 59-end of LSU rDNA [60]. The were observed with an Olympus BX60 microscope (Olympus large subunit rDNA (LSU rDNA) was amplified using NLF1050 Australia) equipped with a DP70 camera. All images were (59-AAT CGA ACC ATC TAG TAG CTG G-39), NLFMyxid acquired as RGB, TIFF files. Images were imported into Creative (59-AAT GCT AGG GTT CCR AGT GG-39), NLR3113 (59- Suite CS3 (Adobe Systems, San Jose, CA). Adjustment layers for GTC TAA ACC CAG CTC ACG TTC CCT-39) and NLR3284 levels and brightness/contrast were used to find the optimal colour (59-TTC TGA CTT AGA GGC GTT CAG-39) in a nested PCR balance for investigated features with the original image. All [61]. The expected PCR products of the partial SSU rDNA,

PLoS ONE | www.plosone.org 9 April 2011 | Volume 6 | Issue 4 | e1887131 Novel Parasites in Australian Frogs complete ITS rDNA and partial LSU rDNA was ,950 nt, MA: Sinauer Associates), and Bayesian using MrBayes3.1.2 [68]. ,900 nt and ,2,200 nt, respectively. Substitution models for DNA evolution were selected using AIC in PCR amplification mix contained 26EconoTaq PLUS GREEN ModelTest3.7/ModelTestJ [69]; GTR+G (for original Alignment 1) MasterMix (Lucigen, USA) or 26MasterMix (Fermentas, USA). and GTR+G+I (and moderately stringent Alignment 2 and the Alternatively, PCR was carried with 1 U of Taq-Purple polymer- stringent Alignment 3). The neighbour joining tree branch ase and dNTPs (Top-Bio, Czech Republic). Primers were added at robustness was evaluated using bootstrapping method with 1,000 0.25 mM concentration each. The PCR for SSU rDNA and ITS replicates [63]. The maximum likelihood phylogeny tree rDNA was run for 35 cycles (95uC for 0.5 min, 55uC for 1 min, (PhyML) was generated and its branch robustness assessed by 72uC for 1 min). The nested PCR for LSU rDNA was run for 30 the bootstrapping method with 500 replicates [67]. Maximum cycles in the first reaction (95uC for 0.5 min, 48uC for 1 min, 72uC likelihood tree (PAUP*) was reconstructed using heuristic search for 2 min); for the second PCR an 1 ml aliquot of the first PCR was with the NNI swap algorithm with parameters from ModelTest; used as template and run for 35 cycles (95uC for 0.5 min, 52uC for due to prohibitively long calculation times bootstraps were not 1 min, 72uC for 1 min). All PCR reactions were initiated at 95uC calculated. Bayesian support (posterior probability) for the nodes for 5 min and concluded at 72uC for 7 min. All reactions (25 ml) [68] were inferred using two independent runs and four Markov were run with negative controls (distilled water). chains and 5,000,000 Markov Chain Monte Carlo (MCMC) Aliquots of PCR reactions were resolved on a 1% agarose gel, steps, discarding the first 2,000,000 steps (40%) as a burn-in. stained with GelRed (Biotium) and visualised using UV-transillu- Sampling was performed every 250 generations. Mixing of the minator. PCR reactions with amplicons of expected mass were chains and convergence was properly checked after runs. either purified using the PCR Micro Kit (Invitrogen Australia, Victoria, Australia) and sent for direct sequencing or cloned using GenBank Accession numbers submitted with this the TA-TOPO for Sequencing (pCR-4) cloning kit (Invitrogen manuscript Australia, Victoria, Australia), pGEM-T Easy (Promega) or pDrive The following sequences were submitted to the GenBank: SSU (Qiagen) according to the manufacturer’s instructions. Plasmids rDNA HQ822149-HQ822173, LSU rDNA HQ822174- were purified with the MiniPrep Plasmid Kit (Qiagen, Australia). HQ822192 and ITS rDNA HQ822193-HQ822260 from Myx- PCR product and plasmids with target inserts were sequenced idium species of frogs and toads. using amplification primers, internal primers (LSU rDNA [61]) or primers (T7, SP6, M13R, M13F) within the cloning vector at Supporting Information Supamac (University of Sydney and Prince Albert Molecular Analysis Centre) or Macrogen Inc. (Seoul Korea). Chromato- Table S1 Summary of catalogued cases at the Australian graphs were visualised, inspected, assembled, aligned with related Registry of Wildlife Health (Taronga Conservation Society sequences and analysed using the CLC Main Workbench 5.5 Australia, Mosman, NSW, Australia) from 1997–2009. (CLC bio, Denmark). (XLS) Table S2 Summary of sequence data collected from Myxos- Phylogenetic analyses porea in frogs. To infer SSU rDNA phylogeny of the liver, brain and Brazil (XLS) genotypes of Myxidium spp. we selected taxa based on Fiala [43] and selected additional sequences from additional taxa to select Table S3 Summary of morphology of myxospores (Myxidium wide Myxosporea (Myxozoa) sampling including diverse groups, spp.) recovered from gall bladders in frogs. hosts and tissue tropism. We used blastn and taxonomy browser at (XLS) NCBI [62] to select representative sequences and progressively Table S4 Summary of SSU rDNA sequences used to reconstruct analysed them in MEGA4 [63]; Buddenbrockia plumatellae (Myxozoa, myxosporean phylogeny. Malacosporea) sequence was selected as an outgroup [64], for (XLS) each taxon formal biological traits have been collected from published sourced (Table S3). The final selection of sequences Figure S1 Phylogenetic trees inferred using SSU rDNA of was aligned using Clustal W1.83 [65] using default parameters, Myxosporea. The alignments (Alignment 1–3) are described in followed by manual selection of hypervariable regions and Material and Methods. Alignment 1: 2711 positions, 975 constant, realignment using more relaxed parameters (gap opening: 6, gap 437 variable parsimony-uninformative, 1,300 parsimony-informa- extension: 4) and minor adjustment done by eye. The original tive. Alignment 2: 1756 (64% of the original 2711 positions), 689 alignment (Alignment 1) with all 2,711 residues included (1,300 constant, 298 variable parsimony-uninformative, 769 parsimony- parsimony informative sites) was subject to gblocks 0.91b [66] informative. Alignment 3: 1206 (44% of the original 2711 positions), processing to systematically remove ambiguously aligned positions. 551 constant, 209 variable parsimony-uninformative, 446 parsimo- A moderately stringent process (Alignment 2) eliminated ny-informative. The trees were reconstructed using PhyML 3.0 variable sites and retaining 1,756 (64% of the original 2,711 (GTR+G+I) [67], PAUP*4b10 (D.L. Swofford, 2001, PAUP*, positions; 769 parsimony informative sites) and a stringent process Sunderland, MA: Sinauer Associates) with models selected using (gblocks: Alignment 3) retained 1,206 highly conservative residues AIC in ModelTest3.7 (Alignment 1: GTR+G, Alignment 2: (44% of the original 2,711 positions; 446 parsimony informative GTR+G+I, Alignment 3: GTR+G+I) [69], MrBayes3.1.2 sites). Alignments 2 and 3 were produced using parameters set to (GTR+G+I) [68]. The maximum likelihood phylogeny (PhyML) 20 and 30 for the minimum number of sequences for a conserved tree robustness was assessed by the bootstrapping method with 500 position as well as for the minimum number of sequences for a replicates [41]. Bayesian support (posterior probability) for assessed flanking position, respectively. For both alignments the minimum (MrBayes) using two independent runs and four Markov chains and length of a block was 10 and gapped positions were included. 40% steps discarded as a burn-in [68]. All trees were rooted using All alignments were subject to distance phylogenetic recon- Buddenbrockia plumatellae sequence. The sequence taxa are on the struction in MEGA4 [63], maximum likelihood using PhyML3 right of the tree (for details see Table S4). [67] and PAUP*4b10 (D.L. Swofford, 2001, PAUP*, Sunderland, (PDF)

PLoS ONE | www.plosone.org 10 April 2011 | Volume 6 | Issue 4 | e1887132 Novel Parasites in Australian Frogs

Acknowledgments Derek Spielman, Lance Jurd, David Newell, Reiner Mann and Ross Hyne for their insight and input to the project. We thank Toma´sˇ Scholz (Institute of Parasitology, Czech Republic) for his support and the clinic and zoo staff at Honolulu zoo for their advice and Author Contributions assistance. We acknowledge the Electron Microscopy Unit at The University of Sydney (Australian Key Centre for Microscopy and Conceived and designed the experiments: AH DNP JSˇ Performed the Microanalysis) and Microscopy unit at the Institute of Parasitology, Cˇ eske´ experiments: AH IF ID DNP JSˇ Analyzed the data: AH IF ID KR DNP JSˇ Budeˇjovice (Academy of Sciences of the Czech Republic). We thank Elaine Contributed reagents/materials/analysis tools: AH IF ID MJ BO KR DNP Chew for expert histopathology processing, Navneet Dhand, Glenn Shea, JSˇ Wrote the paper: AH DNP JSˇ.

References 1. Keesing F, Belden LK, Daszak P, Dobson A, Harvell CD, et al. (2010) Impacts 24. Lom J, Feist SW, DykovaÍ, Kepr T (1989) Brain myxoboliasis of bullhead, Cottus of biodiversity on the emergence and transmission of infectious diseases. Nature gobio L., due to Myxobolus jiroveci sp. nov.: light and electron microscope 468: 647–652. observations. Journal of Fish Diseases 12: 15–27. 2. Smith KF, Sax DF, Lafferty KD (2006) Evidence for the role of infectious disease 25. DykovaÍ, Lom J (2007) Histopathology of Protistan and Myxozoan Infections in in species extinction and endangerment. Conservation Biology 20: 1349–1357. Fishes. An Atlas. Praha: Academia. 220 p. 3. Blackburn DC, Evans BJ, Pessier AP, Vredenburg VT (2010) An enigmatic 26. Stensaas LJ, Stensaas SS, Sotelo JR (1967) An intra-axonal protozoon in the mortality event in the only population of the Critically Endangered spinal cord of the toad Bufo arenarum (Hensel). Journal of Protozoology 14: Cameroonian frog Xenopus longipes. African Journal of Herpetology 59: 585–595. 111–122. 27. Alencar AA (1957) Toxoplasmose espontanea e inaparente em amfibios dos 4. Briggs CJ, Knapp RA, Vredenburg VT (2010) Enzootic and epizootic dynamics generos Leptodactylus e Bufo. Jornal Brasileiro de Neurologia 9: 137–146. of the chytrid fungal pathogen of amphibians. Proceedings of the National 28. Lom J, McGeorge J, Feist SW, Morris D, Adams A (1997) Guidelines for the Academy of Sciences of the United States of America 107: 9695–9700. uniform characterisation of the actinosporean stages of parasites of the phylum 5. Teacher AGF, Cunningham AA, Garner TWJ (2010) Assessing the long-term Myxozoa. Diseases of Aquatic Organisms 30: 1–9. impact of Ranavirus infection in wild common frog populations. Animal 29. Jirku˚ M, Bolek MG, Whipps CM, Janovy J, Kent ML, et al. (2006) A new species Conservation 13: 514–522. of Myxidium (Myxosporea: Myxidiidae), from the western chorus frog, Pseudacris 6. Green DE, Gray MJ, Miller DL (2009) Disease monitoring and biosecurity. In: triseriata triseriata, and Blanchard’s cricket frog, Acris crepitans blanchardi (Hylidae), Dodd CK, ed. Amphibian Ecology and Conservation: A Handbook of from Eastern Nebraska: Morphology, phylogeny, and critical comments on Techniques. Oxford: Oxford Press. pp 481–505. amphibian Myxidium taxonomy. Journal of Parasitology 92: 611–619. 7. Sitja`-Bobadilla A (2009) Can Myxosporean parasites compromise fish and 30. Sutherst RW, Floyd RB, Maywald GF (1996) The potential geographical amphibian reproduction? Proceedings of the Royal Society B: Biological distribution of the cane toad, Bufo marinus L in Australia. Conservation Biology Sciences 276: 2861–2870. 10: 294–299. 8. Mutschmann F (2004) Pathological changes in African hyperoliid frogs due to a 31. Phillips BL, Brown GP, Webb JK, Shine R (2006) Invasion and the evolution of myxosporidian infection with a new species of Hoferellus (Myxozoa). Diseases of speed in toads. Nature 439: 803–803. Aquatic Organisms 60: 215–222. 32. Urban MC, Phillips BL, Skelly DK, Shine R (2007) The cane toad’s (Chaunus 9. Duncan AE, Garner MM, Bartholomew JL, Reichard TA, Nordhausen RW [Bufo] marinus) increasing ability to invade Australia is revealed by a dynamically (2004) Renal myxosporidiasis in Asian horned frogs (Megophrys nasuta). Journal of updated range model. Proceedings of the Royal Society B: Biological Sciences Zoo and Wildlife Medicine 35: 381–386. 274: 1413–1419. 10. Hedrick RP, Adkison MA, El-Matbouli M, MacConnell E (1998) Whirling 33. Bickford D, Lohman DJ, Sodhi NS, Ng PKL, Meier R, et al. (2007) Cryptic disease: re-emergence among wild trout. Immunological Reviews 166: 365–376. species as a window on diversity and conservation. Trends in Ecology and 11. Steinbach-Elwell LC, Eagle-Dtromberg K, Ryce EKN, Bartholomew JL (2009) Evolution 22: 148–155. Whirling Disease in the United States: A Summary of Progress in Research and 34. Easteal S (1981) The history of introductions of Bufo marinus (Amphibia: Anura) - Management. BozemanMO: Whirling Disease Initiative, Montana Water a natural experiment in evolution. Biological Journal of the Linnean Society 16: Center, Montana State University. 93–113. 12. Whipps CM, El-Matbouli M, Hedrick RP, Blazer V, Kent ML (2004) Myxobolus 35. Kudo R, Sprague V (1940) On Myxidium immersum (Lutz) and M. serotinum n. sp., cerebralis internal transcribed spacer 1 (ITS-1) sequences support recent spread of two myxosporidian parasites of Salientia of South and North America. Revista the parasite to North America and within Europe. Diseases of Aquatic de Medicina Tropical y Parasitologıá, Bacteriologıá, Clinica y Laboratoria 6: Organisms 60: 105–108. 65–73. 13. Hill BD, Green PE, Lucke HA (1997) Hepatitis in the green tree frog (Litoria 36. O’Donoghue PJ, Adlard RD (2000) Catalogue of protozoan parasites recorded caerulea) associated with infection by a species of Myxidium. Australian Veterinary in Australia. Memoirs of the Queensland Museum 45: 1–163. Journal 75: 910–911. 37. Hebert PDN, Stoeckle MY, Zemlak TS, Francis CM (2004) Identification of 14. Delvinquier BLJ (1986) Myxidium immersum (Protozoa: Myxosporea) of the cane birds through DNA barcodes. Plos Biology 2: 1657–1663. toad, Bufo marinus, in Australian Anura, with a synopsis of the genus in 38. Waugh J (2007) DNA barcoding in animal species: progress, potential and amphibians. Australian Journal of Zoology 34: 843–853. pitfalls. Bioessays 29: 188–197. 15. Hartigan A, Phalen DN, Sˇlapeta J (2010) Museum material reveals a frog 39. Hemmerter S, Sˇlapeta J, van den Hurk A, Cooper R, Whelan P, et al. (2007) A parasite emergence after the invasion of the cane toad in Australia. Parasites & curious coincidence: mosquito biodiversity and the limits of the Japanese Vectors 3: 50. encephalitis virus in Australasia. BMC Evolutionary Biology 7: 100. 16. Dubey S, Shine R (2008) Origin of the parasites of an invading species, the 40. Whipps CM, Kent ML (2006) Phylogeography of the cosmopolitan marine Australian cane toad (Bufo marinus): are the lungworms Australian or American? parasite Kudoa thyrsites (Myxozoa : Myxosporea). Journal of Eukaryotic Molecular Ecology 17: 4418–4424. Microbiology 53: 364–373. 17. Mahony M (1996) The decline of the Green and Golden Bell Frog Litoria aurea 41. Henderson M, Okamura B (2004) The phylogeography of salmonid proliferative viewed in the context of declines and disappearances of other Australian frogs. kidney disease in Europe and North America. Proceedings of the Royal Society Australian Zoologist 30: 237–246. of London Series B: Biological Sciences 271: 1729–1736. 18. Wassens S (2008) Review of the past distribution and decline of the southern bell 42. Gunter NL, Whipps CM, Adlard RD (2009) Ceratomyxa (Myxozoa: Bivalvulida): frog Litoria raniformis in New South Wales. Australian Zoologist 34: 446–452. Robust taxon or genus of convenience? International Journal for Parasitology 19. Mahony M (1999) Review of the declines and disappearances within the bell frog 39: 1395–1405. species group (Litoria aurea species group) in Australia. In: Campbell A, ed. 43. Fiala I (2006) The phylogeny of Myxosporea (Myxozoa) based on small subunit Declines and disappearances of Australian frogs: Environment Australia, ribosomal RNA gene analysis. International Journal for Parasitology 36: Canberra. pp 81–93. 1521–1534. 20. White AW, Pyke GH (2008) Green and Golden Bell frogs in New South Wales: 44. Kent ML, Andree KB, Bartholomew JL, El-Matbouli M, Desser SS, et al. (2001) current status and future prospects. Australian Zoologist 34: 319–333. Recent advances in our knowledge of the Myxozoa. Journal of Eukaryotic 21. Berger L, Speare R, Daszak P, Green DE, Cunningham AA, et al. (1998) Microbiology 48: 395–413. Chytridiomycosis causes amphibian mortality associated with population 45. Holzer AS, Sommerville C, Wootten R (2004) Molecular relationships and declines in the rain forests of Australia and Central America. Proceedings of phylogeny in a community of myxosporeans and actinosporeans based on their the National Academy of Sciences of the United States of America 95: 18S rDNA sequences. International Journal for Parasitology 34: 1099–1111. 9031–9036. 46. Hufbauer RA, Torchin ME (2008) Integrating ecological and evolutionary 22. McFadden M, Duffy S, Harlow P, Hobcroft D, Webb C, et al. (2008) A review theory of biological invasion. In: Nentwig W, ed. Biological Invasions. of the green and golden bell frog Litoria aurea breeding program at Taronga Heidelberg: Springer. pp 79–96. Zoo. Australia Zoologist 34: 291–296. 47. Tompkins DM, Poulin R (2006) Parasites and biological invasion. In: Allen RB, 23. Lom J, DykovaÍ (2006) Myxozoan genera: definition and notes on taxonomy, Lee WG, eds. Biological Invasions in New Zealand. Heidelberg: Springer. pp life-cycle terminology and pathogenic species. Folia Parasitologica 53: 1–36. 67–84.

PLoS ONE | www.plosone.org 11 April 2011 | Volume 6 | Issue 4 | e1887133 Novel Parasites in Australian Frogs

48. Carlton JT (1996) Biological invasions and cryptogenic species. Ecology 77: 61. Bartosˇova´P, Fiala I, HypsˇaV (2009) Concatenated SSU and LSU rDNA data 1653–1655. confirm the main evolutionary trends within myxosporeans (Myxozoa: 49. Torchin ME, Lafferty KD, Dobson AP, McKenzie VJ, Kuris AM (2003) Myxosporea) and provide an effective tool for their molecular phylogenetics. Introduced species and their missing parasites. Nature 421: 628–630. Molecular Phylogenetics and Evolution 53: 81–93. 50. Marr SR, Mautz WJ, Hara AH (2008) Parasite loss and introduced species: a 62. Wheeler DL, Barrett T, Benson DA, Bryant SH, Canese K, et al. (2008) comparison of the parasites of the Puerto Rican tree frog, (Eleutherodactylus coqui), Database resources of the national center for biotechnology information. Nucleic in its native and introduced ranges. Biological Invasions 10: 1289–1298. Acids Research 36: D13–D21. 51. Dunn AM (2009) Parasites and biological invasions. Advances in Parasitology 63. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular Evolutionary 68: 161–184. Genetics Analysis (MEGA) software version 4.0. Molecular Biology and 52. Prenter J, MacNeil C, Dick JTA, Dunn AM (2004) Roles of parasites in animal Evolution 24: 1596–1599. invasions. Trends in Ecology and Evolution 19: 385–390. 64. Canning EU, Tops S, Curry A, Wood TS, Okamura B (2002) Ecology, 53. Daszak P, Cunningham AA, Hyatt AD (2000) Emerging infectious diseases of development and pathogenicity of Buddenbrockia plumatellae Schroder, 1910 wildlife - Threats to biodiversity and human health. Science 287: 443–449. (Myxozoa, Malacosporea) (syn. Tetracapsula bryozoides) and establishment of 54. Power AG, Mitchell CE (2004) Pathogen spillover in disease epidemics. Tetracapsuloides n. gen. for Tetracapsula bryosalmonae. Journal of Eukaryotic American Naturalist 164: S79–S89. Microbiology 49: 280–295. 55. Kelly DW, Paterson RA, Townsend CR, Poulin R, Tompkins DM (2009) 65. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG (1997) The Parasite spillback: A neglected concept in invasion ecology? Ecology 90: CLUSTAL_X windows interface: flexible strategies for multiple sequence 2047–2056. alignment aided by quality analysis tools. Nucleic Acids Research 25: 56. Hershberger P, van der Leeuw B, Gregg J, Grady C, Lujan K, et al. (2010) Amplification and transport of an endemic fish disease by an introduced species. 4876–4882. Biological Invasions 12: 3665–3675. 66. Talavera G, Castresana J (2007) Improvement of phylogenies after removing 57. Telfer S, Bown KJ, Sekules R, Begon I, Hayden T, et al. (2005) Disruption of a divergent and ambiguously aligned blocks from protein sequence alignments. host-parasite system following the introduction of an exotic host species. Systematic Biology 56: 564–577. Parasitology 130: 661–668. 67. Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to estimate 58. Rose K (2005) Wildlife Health Investigation Manual: The Australian Registry of large phylogenies by maximum likelihood. Systematic Biology 52: 696–704. Wildlife Health. 68. Ronquist F, Huelsenbeck JP (2003) MrBayes 3: Bayesian phylogenetic inference 59. Cameron AR, Baldock FC (1998) A new probability formula for surveys to under mixed models. Bioinformatics 19: 1572–1574. substantiate freedom from disease. Preventive Veterinary Medicine 34: 1–17. 69. Posada D, Crandall KA (1998) MODELTEST: testing the model of DNA 60. Coleman AW, Vacquier VD (2002) Exploring the phylogenetic utility of ITS substitution. Bioinformatics 14: 817–818. sequences for animals: A test case for abalone (Haliotis). Journal of Molecular Evolution 54: 246–257.

PLoS ONE | www.plosone.org 12 April 2011 | Volume 6 | Issue 4 | e1887134 Table S1. Summary of catalogued cases at the Australian Registry of Wildlife Health (Taronga Conservation Society Australia, Mosman, NSW, Australia) from 1997-2009. Myxosporea Species IUCN Status AWRC Year Location Age History Brain Liver S0087 1997 Taronga Zoo captive breeding Tadpole Euthanised due to ascites Yes Yes Long reef (ex Zoo captive S0117 1998 Tadpole Found dead Yes No Green and Golden Bell Frog breeding population) Vulnerable (Litoria aurea) S0131 1998 Adult Euthanised due to listlessness and anorexia. Yes No Sydney Olympic Park Euthanised due to deteriorated body S0133 1998 Adult Yes No condition with poor limb motor skills. Euthanised due to deterioration of body 6562.1 2008 Adult Yes No condition. Southern Bell Frog (Litoria DECCW captive breeding ex Euthanised due to deterioration of body Endangered 6919.1 2009 Adult Yes No raniformis) Coleambally condition. 6919.2 2009 Adult Euthanasia due to non-breeding. Yes No 5607.4 2007 Maragle Creek, NSW Adult Euthanised for disease screening. Yes No 5690.1 2007 Adult Found dead. Yes No Paralysis in hind limbs, no righting reflex, 5694.1 2007 Adult Yes No poor response to stimuli. 5702.1 2007 Adult Found dead. Yes No Booroolong Frog (Litoria Critically Endangered 5780.1 2007 Adult Found dead. Yes No booroolongensis) 5997.1 2007 Taronga Zoo captive breeding Adult Found dead. Yes No Signs of progressive lethargy and poor 6175.1 2008 Adult Yes No motor skills. 6191.1 2008 Adult Found dead. Yes No 6276.1 2008 Adult Found dead. Yes No 6504.1 2008 Adult Paresis of the hind limbs. Yes No Euthanised prior to species cohort 6096.1 2007 Adult Yes Yes Peron's Tree Frog (Litoria translocation. Least Concern Scheyville National Park, NSW peronii) Euthanised for envinronment health 6096.2 2007 Adult Yes Yes assessment for frog traslocation. Green Tree Frog (Litoria Wild caught for Taronga Zoo Sighns of oedema, poor posture, slow hind Least Concern S0086 1997 Adult Yes No caerulea) captive breeding limb reflexes. Disease screening project prior to species 6095.4 2007 Adult Yes Yes Eastern Dwarf Tree Frog translocation Least Concern Scheyville National Park, NSW (Litoria fallax) Disease screening project prior to species 6095.6 2007 Adult No Yes translocation Broad Palmed Rocketfrog Disease screening project prior to species Least Concern 6095.7 2007 Scheyville National Park, NSW Adult No Yes (Litoria latopalmata) translocation Euthanised prior to species cohort 6096.3 2007 Adult Yes No Striped Marsh Frog translocation. Least Concern Scheyville National Park, NSW (Limnodynastes peronii) Euthanised prior to species cohort 6096.4 2007 Adult Yes No translocation. Spotted Marsh Frog Least Concern 3669.1 2003 Sydney Olympic Park Tadpole Signs of progresive lethargy. Yes No (Limnodynastes tasmaniensis) 35 Table S2. Summary of sequence data collected from Myxosporea in frogs. A. All marker sequences and alignment data produced in this work ALL SSU rDNA ITS1 rDNA 5.8S rDNA ITS2 rDNA LSU rDNA Sequence length 875-884 321-377 176 279-301 2187-2195 Alignment length 887 386 176 317 2199 Number of variable sites 118 145 21 129 302 Number of parsimony informative sites 95 117 9 111 201 Number of singletons 23 28 12 18 101 Average nt genetic distance (S.E.) in % 5.63 (1.53) 20.10 (9.25) 2.63 (0.85) 19.48 (6.90) 4.77 (0.73)

B. Comparison of brain and pooled liver + Brazil-1 + Brazil-2 clade sequence data produced in this work clade liver + Brazil-1 + Brazil-2 SSU rDNA ITS1 rDNA 5.8S rDNA ITS2 rDNA LSU rDNA Number of variable sites 78 84 7 77 189 Number of parsimony informative sites 57 67 5 66 109 Number of singletons 21 17 2 11 80 Average nt genetic distance (S.E.) in % 3.68 (0.70) 7.17 (1.97) 0.85 (0.38) 7.63 (1.46) 2.96 (0.29) clade brain SSU rDNA ITS1 rDNA 5.8S rDNA ITS2 rDNA LSU rDNA Number of variable sites 3 23 11 18 30 Number of parsimony informative sites 0 3 1 1 1 Number of singletons 3 20 10 17 29 Average nt genetic distance (S.E.) in % 0.07 (0.04) 0.52 (0.10) 0.51 (0.17) 0.49 (0.12) 0.40 (0.08)

C. Comparison of individual liver, Brazil-1 and Brazil -2 clades sequence data produced in this work clade liver SSU rDNA ITS1 rDNA 5.8S rDNA ITS2 rDNA LSU rDNA Number of variable sites 4 26 3 19 45 Number of parsimony informative sites 2 13 0 12 6 Number of singletons 2 13 3 7 39 Average nt genetic distance (S.E.) in % 0.14 (0.07) 1.54 (0.36) 0.11 (0.07) 1.39 (0.36) 0.66 (0.10) clade Brazil-1 SSU rDNA ITS1 rDNA 5.8S rDNA ITS2 rDNA LSU rDNA Number of variable sites 8 2 0 3 12 Number of parsimony informative sites 0 0 0 0 0 Number of singletons 8 2 0 3 0 Average nt genetic distance (S.E.) in % 0.61 (0.21) 0.31 (0.19) 0.00 (0.00) 0.54 (0.28) 0.79 (0.21) clade Brazil-2 SSU rDNA ITS1 rDNA 5.8S rDNA ITS2 rDNA LSU rDNA Number of variable sites 13 10 0 6 24 Number of parsimony informative sites 0 3 0 0 0 Number of singletons 13 7 0 6 24 Average nt genetic distance (S.E.) in % 0.60 (0.18) 1.26 (0.39) 0.00 (0.00) 0.73 (0.27) 0.74 (0.16)

D. Between groups mean distances (bottom-left) and mean between group net-distances (top-right) SSU rDNA LIVER Brazil-1 Brazil-2 BRAIN LIVER - 4.2% 4.5% 9.3% Brazil-1 5.5% - 5.1% 8.2% Brazil-2 4.6% 5.1% - 7.6% BRAIN 8.3% 9.7% 8.0% -

LSU rDNA LIVER Brazil-1 Brazil-2 BRAIN LIVER - 3.3% 3.9% 6.8% Brazil-1 4.0% - 3.7% 6.9% Brazil-2 4.6% 4.4% - 6.6% BRAIN 7.3% 7.5% 7.2% -

ITS1 rDNA LIVER Brazil-1 Brazil-2 BRAIN LIVER - 13.5% 14.8% 34.1% Brazil-1 14.4% - 15.0% 37.5% Brazil-2 16.2% 15.7% - 37.2% BRAIN 35.2% 37.9% 38.1% -

ITS2 rDNA LIVER Brazil-1 Brazil-2 BRAIN LIVER - 13.9% 16.9% 32.7% Brazil-1 14.9% - 20.1% 37.3% Brazil-2 17.9% 20.7% - 34.0% BRAIN 33.7% 37.8% 35.0% -

Table S3. Summary of morphology of myxospores (Myxidium spp.) recovered from gall bladders in frogs. A - Unpreserved Myxospores (Myxidium spp.) recoved from frog gall bladders in Australia (this study) Green and Golden Bell Frog Peron's Tree Frog Striped Marsh Frog* Southern Bell Frog Cane Toad Litoria aurea Litoria peronii Limnodynastes peronii Litoria raniformis Bufo marinus (Sydney, NSW, Australia) (Sydney, NSW, Australia) (Sydney, NSW, Australia) (southern NSW, Australia) (Byron Bay, NSW, Australia) Spore shape Ellipsoidal Ellipsoidal Ellipsoidal Ellipsoidal Ellipsoidal Spore length 14.7 (13.0-16.5) 14.8 (13.0-16.0) 16.0 (15.0-18.0) 14.1 (13.0-15.5) 14.5 (12.0-17.0) Spore width 8.9 (8.0-10.0) 9.2 (8.0-10.0) 8.7 (8.0-10.0) 8.5 (8.0-10.5) 9.1 (8-10.5) PC length 4.1 (3.5-5.0) 3.9 (4.0-6.0) 5.3 (5.0-6.0) 3.8 (3.0-5.0) 3.9 (3.0-5.0) PC width 3.9 (3.5-4.5) 4.6 (4.0-5.0) 4.8 (4.0-5.5) 3.7 (3.0-5.0) 4.0 (3.0-5.0) PF coils n.a 4-5 5-6 4-5 4-6 SR extent To poles To poles To poles To poles To poles SR shape S-curve S-curve S-curve S-curve S-curve No. of LR 2 2 2 2 2 No. of TR 5-9 5-10 5-11 5-12 5-13 TR extent n.a n.a Whole valve n.a n.a TR branched n.a n.a No n.a n.a No. spores measured 30 30 30 30 30 Note: PC= polar capsule; SR= sutural ridge; LR=longitudinal ridge; TR=transverse ridge; n.a.=not measured; all measurement are in micrometers. * Myxidium sp. ex Striped Marsh Frog myxospores fixed in 10% buffered formalin (n=10) measurements (in micrometers): 13.4 (13.0-17.0) x 8.1 (7.0-9.0), PC 4.0 (3.0-5.0) x 4.4 (3.0-5.0) * Myxidium sp. ex Striped Marsh Frog myxospores fixed in 10% buffered formalin (n=10) measurements (in micrometers): 14.7 (13.0-16.0) x 7.8 (7.0-9.0), PC 4.4 (4.0-5.0) x 4.5 (4.0-5.0)

B - Myxospores characteristics of myxosporean species described from amphibians Myxidium melleni Myxidium serotinum Myxidium typhonius Myxidium immersum Myxidium cf. immersum (Jirku et al., 2006) (Kudo and Sprague, 1940) (Grey, 1993) (Kudo and Sprague, 1940) (Delvinquier, 1986) North America America Peru, South America America Australia Spore shape Ellipsoidal Ellipsoidal Ellipsoidal Spindle Ellipsoidal Spore length 12.3 (12.0-13.5) 15.6 (13.0-18.0) 10.9 (9.8-12.2) 11.8-14.2 12.3-13.3 Spore width 7.6 (7.0-9.0) 9.8 (9.0-11.0) 7.2 (5.7-8.8) 7.5-10.9 7.3-7.8 PC length 5.2 (4.8-5.5) 5.0 3.8 (2.5-5.2) 3.5-4.5 n.a PC width 4.2 (3.8-4.5) 6.0 3.6 (2.5-5.2) 4.5 n.a PF coils 7-8 3-5 4-5 n.a n.a SR extent To poles Not to poles To poles To poles To poles SR shape S-curve S-curve anteriorly Straight Straight n.a No. of LR 2 4-8 2 2 2 No. of TR 2-5 ca.15 9-11 7-9 5-10 TR extent Whole valve Whole valve Whole valve 2/3 valve Whole valve TR branched No Yes No No n.a Number of spores measured 30 n.a 325 n.a n.a Materal (fixed/unfixed) fresh fresh fixed formalin fixed in ethanol unfspecified Reference Jirku et al., 2006 Kudo and Sprague, 1940 Grey, 1993 Kudo and Sprague, 1940 Delvinquier, 1986 Note: PC= polar capsule; SR= sutural ridge; LR=longitudinal ridge; TR=transverse ridge; n.a.=not measured; all measurement are in micrometers.

C - Museum material recovered myxospore measurements and details* Green and Golden Bell Frog Striped Marsh Frog Green Tree Frog Cane Toad Litoria aurea Limnodynastes peronii Litoria caeruela Bufo marinus (Homebush, NSW, Australia) (Taree, NSW, Australia) (Wallacia, NSW, Australia) (Burringar, NSW, Australia) Spore shape Ellipsoidal Ellipsoidal Ellipsoidal Ellipsoidal Spore length 15.0 (14.0-16.0) 16.1 (15.5-16.5) 16.0 (15.5-17.0) 13.4 (10.5-15.0) Spore width 8.4 (8.0-9.0) 7.9 (7.5-8.5) 8.9 (8.0-10.5) 6.5 (6.5-7.0) PC length 4.3 (4.0-4.5) 4.2 (3.5-4.5) 3.6 (3.0-4.5) 4.0 (3.5-4.0) PC width 4.5 (4.5-4.5) 4.9 (4.5-5.5) 4.1 (3.0-5.0) 3.9 (3.5-4.0) PF coils n.a n.a n.a n.a SR extent To poles To poles To poles To poles SR shape S-curve S-curve S-curve S-curve No. of LR 2 2 2 2 No. of TR 6-8. 5-7. 5-8. 5-9. TR extent n.a n.a n.a. n.a. TR branched n.a n.a n.a n.a Number of spores measured 10 10 10 10 Note: PC= polar capsule; SR= sutural ridge; LR=longitudinal ridge; TR=transverse ridge; n.a.=not measured; results from Hartigan et al. (2010); all measurement are in micrometers. * - material originally fixed in formalin later transferred into 70% ethanol

References

Clark JG, Shoemake. Jp (1973). Eurycea bislineata (Green), 2-lined salamander, a new host of Myxidium serotinum Kudo and Sprague, 1940 (Myxosporida, Myxidiidae). Journal of Protozoology 20: 365-366. Gray ME (1993) A new species of Myxidium (Myxozoa: Myxidiidae) from the gall bladders of anuran amphibians from Peru. Transactions of the Kansas Academy of Science 96: 152-157. Hartigan A, Phalen DN, Slapeta J (2010) Museum material reveals a frog parasite emergence after the invasion of the cane toad in Australia. Parasites & Vectors 3: 50.

Jirku M, Bolek MG, Whipps CM, Janovy J, Kent ML, Modry D (2006). A new species of Myxidium (Myxosporea: Myxidiidae), from the Western Chorus frog, Pseudacris triseriata triseriata, and Blanchards's Cricket frog, Acris crepitans blanchardi (Hylidae), from Eastern Nebraska: morphology, phylogeny, and critical comments on amphibian Myxidium taxonomy. Journal of Parasitology 92: 611-619. Kudo R, Sprague V (1940). On Myxidium immersum (Lutz) and M. serotinum n. sp., two myxosporidian parasites of Salientia of South and North America. Revista de Medicina Tropical y Parasitología, Bacteriología, Clinica y Laboratoria 6: 65-73.

Table S4. Summary of SSU rDNA sequences used to reconstruct myxosporean phylogeny.

Type host Tissue tropism Host Myxozoan taxon Acc. No. Common L/C Disease Geographic location Ecology Scientific name group Presporogony Sporogony/Spores name Terestrial / aquatic Sphaerospora ranae EF211975 Rana dalmatina Agile frog n.a. Amphibian No n.a. Kidney Europe (Czech Republic) development Myxidium sp. clade Green and Brain and Spinal Terestrial / aquatic BRAIN - Litoria aurea golden bell frog n.a. Amphibian Yes cord Gallbladder Australia (New South Wales) development Myxidium sp. clade Green and Terestrial / aquatic LIVER - Litoria aurea golden bell frog n.a. Amphibian Yes Bile Ducts Gallbladder Australia (New South Wales) development Pseudacris Western chorus Terestrial / aquatic Myxidium melleni DQ003031 triseriata triseriata frog n.a. Amphibian n.a. n.a. Gallbladder North America (US) development Myxidium sp. clade Terestrial / aquatic Brazil 1 - Bufo marinus Cane toad n.a. Amphibian n.a. n.a. Gallbladder South America (Brazil) development Myxidium sp. clade Terestrial / aquatic Brazil 2 - Bufo marinus Cane toad n.a. Amphibian n.a. n.a. Gallbladder South America (Brazil) development Sphaerospora Europe (Germany, Czech truttae AJ581915 Salmo salar Atlantic salmon Yes Fish n.a. Kidney Kidney Republic, UK) Aquatic development Chloromyxum Europe (Russia, Czech truttae AJ581916 Salmo salar Atlantic salmon Yes Fish Yes n.a. Gallbladder Republic and Spain) Aquatic development Myxobolus arcticus AF085176 Oncorhynchus nerka Sockeye salmon Yes Fish No n.a. Brain Asia (Japan) Aquatic development Oncorhynchus Myxobolus kisutchi AB469988 kisutch Soho salmon n.a. Fish No n.a. Spinal cord North America (US) Aquatic development Gills, urinary bladder and Myxidium giardi AJ582213 Anguilla anguilla European eel Yes Fish Yes kidney Gills and kidney Worldwide distribution Aquatic development Myxobolus Oncorhynchus Brain and spinal cerebralis U96492 mykiss Rainbow trout Yes Fish Yes cord Brain and spinal cord Worldwide distribution Aquatic development Fresh water / aquatic Myxidium sp. U13829 Cottus bairdii Mottled sculpin n.a. Fish n.a. n.a. n.a. North America (US) development Chloromyxum Hypophthalmichthys Fresh water / aquatic cyprini AY604198 molitrix Silver carp n.a. Fish Yes Liver Gallbladder Europe and North America development Myxidium Cyprinus carpio Fresh water / aquatic cuneiforme DQ377709 haematopterus Common carp n.a. Fish n.a. n.a. Gallbladder Asia (Japan, China) development Oncorhynchus Fresh water / aquatic Ceratomyxa shasta AF001579 mykiss Rainbow trout Yes Fish Yes Intestine Intestine North America (US) development Chloromyxum Pomoxis Fresh water / aquatic trijugum AY954689 nigromaculatus Black crappie n.a. Fish Yes n.a. Gallbladder North America (US) development Pumpkin seed Pancreas and bile Pancreas and Fresh water / aquatic Myxobolus osburni AF378338 Lepomis gibbosus fish n.a. Fish Yes ducts gallbladder North America (US, Canada) development Oncorhynchus North America (US) and Fresh water / aquatic Myxidium truttae AF201374 kisutch Soho salmon Yes Fish Yes n.a Gallbladder Europe (UK) development Fresh water / aquatic Henneguya doori U37549 Perca flavescens Yellow perch n.a. Fish No Gills Gills North America (US, Canada) development Sphaeromyxa Helicolenus Blue mouth North Atlantic and Barents Marine / aquatic hellandi DQ377701 dactylopterus rockfish n.a. Fish n.a. n.a Gallbladder Sea development Marine / aquatic Sphaeromyxa longa DQ377691 Trisopterus minutus Poor cod fish n.a. Fish n.a. n.a. Gallbladder North Sea development Chloromyxum Common Fresh water / aquatic auratum AY971521 Carassius auratus goldfish Yes Fish n.a. n.a. Gallbladder Worldwide distribution development Marine / aquatic Myxobolus sandrae EU346379 Sander lucioperca Pike perch n.a. Fish Yes n.a. Brain and Spinal cord Europe (Hungary, UK) development Chloromyxum Marbled electric Europe, Mediteranean and Marine / aquatic leydigi AY604199 Torpedo marmorata ray n.a. Fish n.a. n.a. Gallbladder NW Atlantic development Myxidium Marine / aquatic lieberkuehni X76639 Essox spp. Pike n.a. Fish Yes n.a. Gallbladder Holarctic development Paralichthys Japanese Marine / aquatic Kudoa yasunagai AY302741 olivaceus flounder n.a. Fish n.a. n.a. Brain Asia (Japan) development Melanogrammus North Atlantic, Barents and Marine / aquatic Myxidium gadi DQ377707 aeglefinus Haddock n.a. Fish Yes Gall bladder Gallbladder White Sea development Sphaeromyxa Marine / aquatic zaharoni AY538662 Pterois miles Lion fish n.a. Fish No n.a. Gallbladder Red Sea development Marine / aquatic Zschokkella icterica DQ333434 Siganus spp Rabbit fish n.a. Fish Yes Bile Ducts Gallbladder Red Sea development Marine / aquatic Kudoa thyrsites AY542481 Sardinops ocellatus Pilchards n.a. Fish Yes Muscle Muscle Worldwide distribution development Myxidium Coryphaenoides Roundnosed Gallbladder and Marine / aquatic coryphaenoides DQ377697 rupestris grenadier n.a. Fish n.a. bile ducts Gallbladder Worldwide distribution development Myxidium Common Marine / aquatic incurvatum DQ377708 Callionymus lyra dragonet n.a. Fish n.a. n.a. Gallbladder Worldwide distribution development Europe (Poland, Czech Soricimyxum fegati EU232760 Sorex araneus Common shrew n.a. Mammal Yes Bile ducts Gallbladder Republic) Terestrial Crowned river Bile ducts and Bile ducts, kidney and Asia (India, Pakistan) and Myxidium hardella AY688957 Hardella thurjii turtle n.a. Reptile Yes kidney gallbladder North America (US) Fresh water Trachemys scripta Bile ducts and Bile ducts, kidney and North America (US) and Myxidium scripta DQ851568 elegans Red eared slider n.a. Reptile Yes kidney gallbladder North Sea Fresh water Myxidium Brown roofed North America and South chelonarum DQ377694 Kachuga smithii turtle n.a. Reptile Yes n.a Gallbladder East Asia Fresh water Buddenbrockia Europe, Asia (Japan) and plumatellae* AY074915 - - - Bryozoans - - - North America Fresh water * - outgroup

L/C - life cycle, actinosporeans stage identified and confirmed n.a. - not available / known note: disease, development and location is assined according to the cited publication

SUPPLEMENTARY FIGURE (TREE SSU rDNA - S1)

PhyML 3.0 (GTR+G+I)

Alignment 3 BRAIN Alignment 1 BRAIN Alignment 2 BRAIN 99 100 Brazil-1 Brazil-1 LIVER 97 53 Brazil-2 Myxidium melleni LIVER 100 52 81 Myxidium melleni Brazil-2 Brazil-2 LIVER 61 Brazil-1 Myxidium melleni 89 97 Soricimyxum fegati Myxidium coryphaenoideum Myxidium coryphaenoideum 100 Sphaeromyxa hellandi Sphaeromyxa hellandi Chloromyxum trijugum 62 62 100 Myxidium chelonarum 100 Sphaeromyxa longa 99 Sphaeromyxa longa 75 Sphaeromyxa zaharoni Sphaeromyxa zaharoni Myxidium hardella 100 100 69 Soricimyxum fegati Soricimyxum fegati 54 Myxidium scripta Myxidium sp. 100 Myxidium sp. 100 Myxidium scripta 62 78 Myxidium truttae Myxidium chelonarum Myxidium truttae 69 Myxidium cuneiforme 55 Zschokkella icterica 100 Myxidium hardella 99 0.2 Myxidium cuneiforme 0.2 Myxidium sp. 0.2 72 Zschokkella icterica 58 96 Myxidium coryphaenoideum 100 Myxidium hardella 89 Myxidium cuneiforme Myxidium chelonarum Zschokkella icterica 57 Sphaeromyxa hellandi 100 Myxidium scripta Myxidium truttae 100 Sphaeromyxa longa Chloromyxum trijugum 100 Sphaeromyxa zaharoni 82 Chloromyxum trijugum 94 100 97 Chloromyxum truttae Myxidium lieberkuehni 100 Chloromyxum truttae Chloromyxum cyprini 100 Myxidium giardi Chloromyxum cyprini 100 Chloromyxum auratum 100 Myxobolus cerebralis 100 Chloromyxum auratum 94 75 Myxidium lieberkuehni Myxidium lieberkuehni 89 Myxobolus arcticus 100 Myxobolus sandrae 100 Myxobolus sandrae Myxobolus kisutchi 89 100 86 100 99 74 93 Myxobolus sandrae 95 Myxobolus osburni 88 Myxobolus osburni 93 90 Henneguya doori 74 Henneguya doori Myxobolus osburni 79 83 100 Myxobolus cerebralis Myxobolus kisutchi 65 Henneguya doori 100 Myxobolus cerebralis 97 Myxobolus arcticus Chloromyxum truttae 66 69 100 Myxobolus arcticus Myxobolus kisutchi Chloromyxum cyprini 88 Myxidium giardi 100 Chloromyxum auratum Myxidium giardi Chloromyxum leydigi Chloromyxum leydigi Chloromyxum leydigi 100 Sphaerospora ranae 100 Sphaerospora truttae 100 Sphaerospora ranae Sphaerospora truttae Sphaerospora ranae Sphaerospora truttae 89 Myxidium incurvatum 88 Myxidium incurvatum 96 Myxidium incurvatum 68 Myxidium gadi 67 Myxidium gadi 64 Myxidium gadi 100 Kudoa thyrsites 100 100 Kudoa thyrsites 100 100 Kudoa thyrsites 100 Kudoa yasunagai Kudoa yasunagai Kudoa yasunagai Ceratomyxa shasta Ceratomyxa shasta Ceratomyxa shasta Buddenbrockia plumatellae Buddenbrockia plumatellae Buddenbrockia plumatellae

PAUP*4b10 / ModelTest3.7

BRAIN Alignment 1 (GTR+G) Alignment 2 (GTR+G+I) BRAIN Alignment 3 (GTR+G+I) BRAIN Brazil-1 Brazil-1 Brazil-2 Myxidium melleni Brazil-2 Brazil-2 Myxidium melleni Myxidium melleni LIVER Brazil-1 LIVER LIVER Myxidium coryphaenoideum Myxidium coryphaenoideum Sphaeromyxa hellandi Soricimyxum fegati Sphaeromyxa hellandi Sphaeromyxa hellandi Sphaeromyxa longa Sphaeromyxa longa Sphaeromyxa zaharoni Sphaeromyxa longa 0.2 0.2 Sphaeromyxa zaharoni Sphaeromyxa zaharoni Soricimyxum fegati Soricimyxum fegati 0.2 Myxidium chelonarum Chloromyxum trijugum Myxidium chelonarum Myxidium chelonarum Myxidium hardella Myxidium hardella Myxidium cuneiforme Myxidium hardella Myxidium cuneiforme Myxidium scripta Zschokkella icterica Zschokkella icterica Myxidium truttae Myxidium cuneiforme Myxidium truttae Zschokkella icterica Myxidium sp. Myxidium sp. Myxidium scripta Myxidium sp. Myxidium scripta Myxidium truttae Chloromyxum trijugum Chloromyxum trijugum Chloromyxum auratum Myxidium coryphaenoideum Chloromyxum auratum Chloromyxum auratum Chloromyxum cyprini Chloromyxum cyprini Chloromyxum truttae Chloromyxum cyprini Chloromyxum truttae Chloromyxum truttae Myxidium lieberkuehni Myxidium lieberkuehni Myxidium giardi Myxidium lieberkuehni Myxidium giardi Myxidium giardi Myxobolus kisutchi Myxobolus kisutchi Henneguya doori Myxobolus kisutchi Henneguya doori Henneguya doori Myxobolus osburni Myxobolus osburni Myxobolus sandrae Myxobolus osburni Myxobolus sandrae Myxobolus sandrae Myxobolus arcticus Myxobolus arcticus Myxobolus cerebralis Myxobolus arcticus Myxobolus cerebralis Myxobolus cerebralis Chloromyxum leydigi Chloromyxum leydigi Sphaerospora ranae Chloromyxum leydigi Sphaerospora ranae Sphaerospora ranae Sphaerospora truttae Sphaerospora truttae Ceratomyxa shasta Sphaerospora truttae Ceratomyxa shasta Ceratomyxa shasta Kudoa yasunagai Kudoa yasunagai Kudoa thyrsites Kudoa yasunagai Kudoa thyrsites Kudoa thyrsites Myxidium gadi Myxidium gadi Myxidium incurvatum Myxidium gadi Myxidium incurvatum Myxidium incurvatum Buddenbrockia plumatellae Buddenbrockia plumatellae Buddenbrockia plumatellae

MrBayes3.1.2 (GTR+G+I)

Alignment 1 Alignment 2 BRAIN Alignment 3 100 BRAIN Brazil-1 100 BRAIN Brazil-2 100 Brazil-1 Myxidium melleni Brazil-2 100 92 100 Myxidium melleni 75 Brazil-2 98 Myxidium melleni 66 LIVER 100 Brazil-1 Myxidium coryphaenoideum LIVER LIVER 99 Myxidium coryphaenoideum Soricimyxum fegati Sphaeromyxa hellandi 95 Sphaeromyxa longa Sphaeromyxa hellandi 0.2 56 Myxidium coryphaenoideum 100 0.2 100 Sphaeromyxa longa 0.2 100 Sphaeromyxa zaharoni 100 Sphaeromyxa hellandi 100 100 Sphaeromyxa zaharoni 100 51 Soricimyxum fegati 100 Sphaeromyxa longa Myxidium chelonarum Soricimyxum fegati 100 Sphaeromyxa zaharoni 100 63 Myxidium chelonarum 100 100 Myxidium hardella 100 Chloromyxum trijugum 99 Myxidium hardella 99 Myxidium chelonarum 97 Myxidium cuneiforme 100 Zschokkella icterica Myxidium cuneiforme 78 Myxidium hardella 100 Myxidium truttae 100 Zschokkella icterica Myxidium scripta Myxidium truttae 97 Myxidium sp. 98 Myxidium cuneiforme 100 98 Myxidium sp. 100 Zschokkella icterica Myxidium scripta 100 Myxidium scripta 100 Chloromyxum trijugum Myxidium sp. 100 Chloromyxum auratum Chloromyxum trijugum Myxidium truttae 100 100 100 100 Chloromyxum auratum 100 Chloromyxum cyprini 100 Chloromyxum auratum Chloromyxum cyprini 100 Chloromyxum cyprini Chloromyxum truttae Chloromyxum truttae 100 Henneguya doori Chloromyxum truttae 100 Myxobolus osburni 100 Henneguya doori 100 Henneguya doori 100 100 Myxobolus osburni 84 Myxobolus sandrae 100 98 100 Myxobolus osburni 100 100 Myxobolus sandrae 100 92 100 Myxobolus arcticus 99 Myxobolus sandrae 100 Myxobolus cerebralis 93 100 Myxobolus arcticus 81 100 Myxobolus arcticus Myxobolus kisutchi 100 Myxobolus cerebralis 100 Myxobolus cerebralis 94 100 Myxobolus kisutchi 91 Myxidium giardi 91 100 100 Myxobolus kisutchi Myxidium lieberkuehni Myxidium giardi Myxidium giardi Myxidium lieberkuehni Myxidium lieberkuehni Chloromyxum leydigi Chloromyxum leydigi 100 Sphaerospora ranae Chloromyxum leydigi Sphaerospora truttae 100 Sphaerospora ranae 100 Sphaerospora ranae Sphaerospora truttae Sphaerospora truttae 100 Ceratomyxa shasta Ceratomyxa shasta 100 Kudoa yasunagai 100 Ceratomyxa shasta 100 Kudoa yasunagai 100 100 Kudoa yasunagai 100 Kudoa thyrsites Kudoa thyrsites Myxidium gadi 99 97 Kudoa thyrsites 100 Myxidium gadi Myxidium gadi Myxidium incurvatum 100 Myxidium incurvatum Buddenbrockia plumatellae 100 Myxidium incurvatum Buddenbrockia plumatellae Buddenbrockia plumatellae

MEGA4.1 (NJ/LogDet)

Alignment 1 Alignment 2 Alignment 3 64 BRAIN 61 BRAIN BRAIN Myxidium melleni 56 Brazil-1 99 95 Brazil-1 100 Brazil-1 98 Brazil-2 Brazil-2 52 100 80 71 71 99 Brazil-2 74 Myxidium melleni Myxidium melleni LIVER LIVER 80 LIVER Soricimyxum fegati Soricimyxum fegati Soricimyxum fegati 67 100 Myxidium chelonarum Myxidium cuneiforme 67 Myxidium coryphaenoideum 0.02 67 Myxidium hardella 83 Zschokkella icterica Sphaeromyxa hellandi 57 0.02 99 100 Myxidium scripta Myxidium truttae 0.02 99 100 Sphaeromyxa longa Chloromyxum trijugum 83 Myxidium sp. Sphaeromyxa zaharoni 99 86 99 Myxidium cuneiforme 100 Myxidium chelonarum Chloromyxum trijugum Zschokkella icterica 100 Myxidium hardella 79 100 Myxidium chelonarum 62 Myxidium sp. Myxidium scripta 70 Myxidium hardella 93 Myxidium truttae 96 Chloromyxum trijugum 86 Myxidium scripta 77 Myxidium coryphaenoideum Myxidium coryphaenoideum 54 Myxidium cuneiforme Sphaeromyxa hellandi Sphaeromyxa hellandi 53 Zschokkella icterica 64 98 Sphaeromyxa longa 99 98 99 98 92 Sphaeromyxa longa Myxidium sp. Sphaeromyxa zaharoni Sphaeromyxa zaharoni 99 87 Myxidium truttae 97 Chloromyxum truttae 100 Chloromyxum auratum 100 Chloromyxum auratum 100 Chloromyxum auratum 97 Chloromyxum cyprini 100 Chloromyxum cyprini Chloromyxum cyprini Chloromyxum truttae Chloromyxum truttae Myxidium lieberkuehni Myxidium lieberkuehni 97 Myxidium lieberkuehni 62 Myxobolus osburni 85 97 100 Myxobolus arcticus Myxobolus kisutchi 77 94 99 94 Myxobolus sandrae Myxobolus cerebralis Myxidium giardi 95 Henneguya doori 95 Myxobolus kisutchi 100 Myxobolus arcticus 60 91 Myxobolus kisutchi Myxidium giardi Myxobolus cerebralis 56 Myxidium giardi 56 67 Henneguya doori 50 60 Henneguya doori 100 Myxobolus arcticus 67 Myxobolus sandrae 61 Myxobolus sandrae Myxobolus cerebralis Myxobolus osburni Myxobolus osburni Chloromyxum leydigi Chloromyxum leydigi Chloromyxum leydigi Ceratomyxa shasta 100 Sphaerospora ranae 100 Sphaerospora ranae 81 100 Sphaerospora ranae 53 Sphaerospora truttae Sphaerospora truttae Sphaerospora truttae Ceratomyxa shasta Ceratomyxa shasta 100 100 Kudoa yasunagai 99 100 Kudoa yasunagai Kudoa yasunagai Kudoa thyrsites Kudoa thyrsites 100 69 Kudoa thyrsites 56 99 Myxidium gadi 65 99 Myxidium gadi 99 Myxidium gadi Myxidium incurvatum Myxidium incurvatum Myxidium incurvatum Buddenbrockia plumatellae Buddenbrockia plumatellae Buddenbrockia plumatellae 42

Chapter 4: New species of Myxosporea from frogs and resurrection of the genus Cystodiscus Lutz, 1889 for species with myxospores in gallbladders of amphibians

43 478 New species of Myxosporea from frogs and resurrection of the genus Cystodiscus Lutz, 1889 for species with myxospores in gallbladders of amphibians

ASHLIE HARTIGAN1,IVANFIALA2,IVADYKOVÁ2, KARRIE ROSE3, DAVID N. PHALEN1 and JAN ŠLAPETA1* 1 Faculty of Veterinary Science, University of Sydney, New South Wales 2006, Australia 2 Institute of Parasitology, Biology Centre, Academy of Sciences of the Czech Republic, České Budějovice, Czech Republic 3 Australian Registry of Wildlife Health, Taronga Conservation Society Australia, Mosman, New South Wales, Australia

(Received 29 August 2011; revised 18 October and 26 October 2011; accepted 27 October 2011; first published online 19 January 2012)

SUMMARY

Two new myxosporean species in the gallbladders of frogs have recently spread across eastern Australia and cause disease. Cystodiscus axonis sp. n. and Cystodiscus australis sp. n. are species of Myxosporea (Myxozoa) identified from a range of Australian frogs and tadpoles including the introduced Cane toad (Rhinella marina). The new species are defined by their distinct genetic lineage, myxospore morphology and ultrastructure of the pre-sporogonic development. Spores of both species are produced in the gallbladder. Spores of C. axonis sp. n. possess distinct filiform polar appendages (FPA). The pre- sporogonic development of C. axonis sp. n. is within myelinated axons in the central nervous system of hosts, as well as bile ducts of tadpoles. Pre-sporogonic and sporogonic development of C. australis sp. n. is confined to tadpole bile ducts and myxospores of C. australis sp. n. are devoid of FPA. The genus Cystodiscus Lutz, 1889 introduced for Cystodiscus immersus Lutz, 1889 is emended to accompany myxosporean parasites affecting amphibians previously classified in the genus Myxidium sensu lato. A synopsis of described species within Cystodiscus is provided.

Key words: Cystodiscus axonis, Cystodiscus australis, new species, Myxidium, Myxosporea, amphibians, frogs, tadpoles, brain, liver, disease, phylogeny.

INTRODUCTION Australia in the second half of the 20th century. No spores were found in introduced or native frog species Accurate description and delimitation of parasite until 30 years after the introduction of the Cane toad species has implications for biodiversity assessment, (Rhinella marina) (Hartigan et al. 2010). Sequencing parasite-host evolutionary theory and the surveil- of small subunit (SSU) rDNA, internal transcribed lance of parasites with medical and veterinary spacer (ITS) and large subunit (LSU) rDNA of importance (de Leon and Nadler, 2010). Molecular M. cf. immersum (= C. cf. immersus) spores from characterization of species can reveal morphology South American and Australian spores confirmed masking multiple distinct cryptic species. Re- their distinct evolutionary trajectories (Hartigan et al. evaluation and formal naming of such cryptic species 2011). The 2 tentative new myxosporean species have reduces systematic confusion and improves under- been identified in a range of native frogs as well as the standing of diversity (Schlick-Steiner et al. 2007). introduced Cane toad; however, the full extent of frog We have discovered, using molecular phylo- species they can infect and the degree of disease they genetics, 2 tentatively new myxosporean species in cause in different species is yet unknown. It is deemed the gallbladders of frogs that have recently spread necessary to formally name these species to facilitate across eastern Australia and cause disease (Hartigan further research on myxosporean species affecting et al. 2011). Based on morphology, all myxospores amphibians. Such action will facilitate establishment from Australian frog gallbladders were assumed to of baseline data and approaches to delineate parasite represent an introduced species Myxidium immersum species in globally urgent conservation efforts of (= Cystodiscus immersus Lutz, 1889) originating from amphibian populations. South America (Delvinquier, 1986). Retrospective The aim of this study was to formally name the 2 analysis of museum specimens of frogs dating back to myxosporean species affecting Australian frogs and 1890 confirmed the emergence of this morphotype in to resurrect the genus Cystodiscus Lutz, 1889 for myxosporean species producing spores in the gall- * Corresponding author: McMaster Building B14, Faculty of Veterinary Science, University of Sydney, New South bladder of amphibians. We used morphological, Wales 2006, Australia. Tel: +61 2 9251 2025. Fax: ultrastructural and genetic techniques as a basis for +61 2 935 17348. E-mail: [email protected] our descriptions. We identified distinct differences in

Parasitology (2012), 139, 478–496. © Cambridge University Press 2012 doi:10.1017/S0031182011002149

44 Myxosporea in gallbladders of amphibians 479

Table 1. Summary of host species parasitized by Cystodiscus axonis sp. n. and Cystodiscus australis sp. n.

Multiplex PCR Myxospores Myxospores C. axonis/C. australis spp. n. Frog species Host life in gall in histological (family) Catalogue number stage bladder sections Brain Liver Myxospores Litoria raniformis ARWH 6985.1–7 Adult Yes Brain +/ −−/− +/− (Hylidae) Litoria aurea ARWH 6092.5a Tadpole No n.a. +/− +/+ n.a. (Hylidae) ARWH 6092.10a& Tadpole No Brain/Liver −/− +/+ n.a. ARWH 6092.14a Tadpole No Liver +/ − +/− n.a. ARWH 6092.36a* Adult No Brain +/−−/− n.a. ARWH 6092.37a Tadpole No Liver n.a. n.a. −/+◊ ARWH 6092.42m& Tadpole No Liver n.a. n.a. n.a. ARWH 6092.54a Adult Yes None −/−−/− n.a. Litoria peronii ARWH 6092.29a Adult Yes Brain +/− n.a. n.a. (Hylidae) ARWH 6092.63i Tadpole No Liver n.a. n.a. n.a. Limnodynastes peronii ARWH 6092.20f Adult Yes n.a. n.a. −/+ −/+ (Myobatrachidae) AWRH 6092.34b Adult Yes n.a. n.a. n.a. −/+ ARWH 6092.55b Adult Yes Brain +/− n.a. +/− ARWH 6092.57e Tadpole No Liver n.a. n.a. n.a. Rhinella marina ARWH 6092.45.7 Adult Yes Brain +/− +/− +/+ (Bufonidae) ARWH 6092.USC10& Adult Yes n.a. n.a. n.a. −/+

* TEM (brain); &TEM (liver); ◊plasmodia without myxospores; n.a., material not available; +, PCR positive; −, PCR negative. spore ultrastructure and present a species-specific the method of Lom and Arthur (1989) and terminol- multiplex PCR that unequivocally differentiates both ogy used as defined in Fig. 1. Sections and wet species within a single DNA sample. mounts were viewed using a X100 oil-objective on an Olympus BX60 microscope equipped with a DP70 camera (Olympus Australia). MATERIALS AND METHODS Collection of individuals Frogs and tadpoles collected for this study are listed Transmission electron microscopy in Table 1. Southern bell frog Litoria raniformis Materials for transmission electron microscopy adults (n=7) were retrieved from the Australian (TEM) were fixed overnight in cacodylate buffered Registry of Wildlife Health (ARWH). These animals 2·5% glutaraldehyde (pH 7·0), post-fixed with 2% were wild caught in Coleambally, NSW and kept in osmium tetroxide, dehydrated through graded etha- Sydney, NSW under laboratory conditions as part of nols (50–100%) and embedded in Spurr’s resin. a breeding study (Mann et al. 2010). All other frogs Ultrathin sections (70 nm) were stained with uranyl were wild-caught tadpoles and adult frogs collected acetate and lead citrate. Copper mesh grids (200 nm) under the University of Sydney animal ethics were viewed on a JEOL JEM 1010 transmission protocols (N00/9-2008/3/4855 and N00/9-2009/3/ electron microscope at the Institute of Parasitology, 5134) and Australian National Parks and Wildlife Biology Centre of the ASCR, Czech Republic. Service scientific licenses (S12686/12969). All host names follow the recommended nomenclature of Frost (2011). Scanning electron microscopy Myxospores for scanning electron microscopy Light microscopy (SEM, see Table 1) were fixed in cacodylate-buffered Samples submitted for histological examination were 2·5% glutaraldehyde (pH 7·0), allowed to settle on fixed in fresh 10% buffered formalin and processed poly-L-lysine (Sigma, Australia) coated coverslips into paraffin blocks (Veterinary Pathology Diagnostic and post-fixed with 2% osmium tetroxide in cacody- Services, University of Sydney). Blocks were sec- late buffer (pH 7·2). Specimens were dehydrated tioned (5 μm) and stained with haematoxylin and with a graded ethanol series (50–100%), critical-point eosin (H&E) or Giemsa. Wet mounts of spores in dried and coated with a 0·2 nm layer of gold (Au) for phosphate-buffered solution (PBS, pH 7·2) or 70% 2 min. Specimens were observed on a Phillips XL-30 ethanol were examined. Thirty spores were measured CP at the Australian Centre of Microscopy and (except in the case of ARWH 6092.55b) according to Microanalysis, Australia.

45 Ashlie Hartigan and others 480

A sutural ridge longitudinal depression (sutural ridge depression) longitudinal depression longitudinal ridge

transverse depressions

transverse ridges

suture of myxospore valves

sutural view valvular view

B 1 2 C 1 2 3

C. immersus C. lyndoyense C. serotinus C. haldari C. lesminteri C. typhonius C. melleni Fig. 1. Composite drawing of spores of Cystodiscus spp. (A) Major characteristics of the bivalved myxospore in sutural and valvular view. The suture is accompanied by a thickened ridge parallel to the suture and a longitudinal sutural ridge depression. (B) Depressions delineated by raised ridges. Valvular view shapes: ovoid (1) or ellipsoid (2). (C) Sutural view types indicating the course of the valve suture: straight (1), S-shaped (2), Z-shaped (3). (D) Characteristics of myxospores from valvular view.

Amplification of internal transcribed spacer rDNA directly sequenced with amplification primers at the Australian Genomic Research Facility (Sydney, DNA from frozen (−20 °C) frog brain and liver and Australia). Chromatographs were assembled and from myxospores was extracted using the PureLink aligned with related sequences using the CLC Main DNA Kit (Invitrogen Australia, Victoria, Australia) Workbench 5.5 (CLC bio, Denmark). The ITS according to the manufacturer’s instructions. rDNA region sequences were deposited into Extracted DNA from all samples was resuspended GenBank under the Accession numbers JN603234, in a final volume of 50–200 μl. DNA was quantified JN603235. using Nanodrop 1000 spectrophotometer (Thermo Scientific, NSW, Australia). Amplification of the ITS rDNA region using myxosporean specific Species-specific diagnostic PCR oligonucleotides S0108 (G-ITS-for) (5′-GGG ATC CGT TTC CGT AGG TGA ACC TGC-3′) and Species-specific primer sets were designed with S0109 (ITS-Zschok-rev) (5′-GAT TCT CAT AGT CLC Bio Workbench based on ITS region rDNA AAC TGC GAG TG-3′) was carried out as unit sequences described here and in Hartigan et al. previously described by Hartigan et al. (2011). PCR (2011), see Supplementary Fig. 1; online version products were purified with PCR Micro Kit only. Primer set 1 amplifying ITS region of C. axonis (Invitrogen Australia, Victoria, Australia) and S0116 (Br150 forward) (5′-GTA ATG TGT TTT

46 Myxosporea in gallbladders of amphibians 481

ACA GGC AAT TGA C-3′) and S0119 (Br760 RESULTS ′ reverse) (5 -TGC AGA GTA ACT AGA ATT GC- Scanning electron microscopy (SEM) and trans- ′ 3 ). Primer set 2 amplifying ITS region of C. australis mission electron microscopy (TEM) of myxosporean ′ S0110 (Liv200 forward) (5 -CTG AAC TTT CTA stages identified in Australian frogs have revealed yet ′ TGC CAG TTA C-3 ) and S0113 (Liv720a reverse) undescribed myxosporean species. Our current ultra- ′ ′ (5 -CAT TAT AAG CAA TAT CCG TAA GA-3 ). structural investigation and the molecular character- All primers were from Sigma-Aldrich (Australia). ization by Hartigan et al. (2011) form the basis for a fi Myxozoan DNA from sh (n=39, see Supplemen- formal description of 2 new parasite species belong- tary Table 1; online version only), bryozoa (Tetra- ing to the genus Cystodiscus Lutz, 1889. capsuloides bryosalmonae), amphibian (Sphaerospora ranae), reptile (Myxidium chelonarum) and mammal Myxozoa Grassé, 1881 (Soricymyxum fegati) were used to test the specificity Myxosporea Bütschli, 1881 of the diagnostic primers. Species were selected to Bivalvulida Shulman, 1959 represent a diverse selection from phylogenetic clades Variisporina Lom and Noble, 1984 in marine and freshwater Myxozoa (Bartošová et al. Myxidiidae Thélohan, 1982 2009). Cystodiscus Lutz, 1889 μ PCR reactions (25 l) were performed with 2x Cystodiscus axonis sp. n. EconoTaq PLUS GREEN MasterMix (Lucigen, USA) or 2x MyTaq RedMix (Bioline, Australia). Description: Ellipsoidal myxospores with 4–8 trans- Alternatively PCR was run using 1 U of Taq-Purple verse ridges (3–7 transverse depressions) on the polymerase and a 5x PCR buffer (Top Bio, Czech outside of 2 valves visible under light microscopy Republic) with 250 μM of each dNTP (Invitek, and SEM (Fig. 2A–E). Spore width and length 14·5 Germany). Primer sets were used individually in (13·0–15·0) ×8·7 (8·0–10·0) μm and polar capsules species-specific PCR or in combination in species- 4·5 (4·0–6·0)×4·5 (4·0–5·0) μm with 4–5 turns of specific multiplex PCR (each 10 pmol concentration). the polar filament (n=30, in PBS) (Table 2). Two Positive and negative controls were run with each spore valves joined by a raised ridge that varied reaction. The PCR cycling conditions were initiated from straight to curved (S-shaped in sutural view). with 95 °C for 5 min, then 35 cycles (95 °C for A sutural depression accompanied the raised ridge 0·5 min, 57 °C for 1 min, 72 °C for 1 min) and (sutural ridge) from pole to pole but was not concluded with 72 °C for 7 min. Results were encircling. The sutural depression 0·8–1·0 μm resolved on a 2% agarose gel stained with GelRed tapered at the polar ends of the spore. The sutural (Biotium) and visualized with a UV illuminator. ridge was thicker at the polar ends of the spore where Positive amplification for primer set 1 resulted in the polar filament pore was situated (Fig. 2B–E). The 595 bp amplicon and for primer set 2 resulted in depressions on the spore surface were centrally 497 bp; mixed infections were identified by the located and sometimes oriented diagonally rather presence of both amplicons. than parallel (Fig. 2A). The depressions did not always extend across the entire valve surface (Fig. 2A and C), some depressions only covered one third Phylogenetic analysis to one half of the spore surface. The ridges or The dataset for phylogenetic analysis consisted of depressions were never observed to be branching or SSU rDNA sequences of selected species of the longitudinal. Filiform polar appendages (FPA) were freshwater clade according to Fiala (2006) and attached to each end of the spore, these appeared to be included all Cystodiscus spp. and Myxidium spp. delicate and were subject to breakage during SEM Sequence of marine Chloromyxum leydigi was se- preparation (Fig. 2B). The FPA were visible using lected as an outgroup. The alignment of 27 sequences a bright-field microscope with a X100 objective, was constructed using MAFFT v6.626b (Katoh et al. all spores observed had 1 or 2 FPA on each spore 2005) with G-INS-i algorithm and default settings. rounded end (Fig. 2A–E). Using SEM, of the spores Ambiguous characters were removed using program with both polar ends seen, 64·3% (9/14) had 1 FPA Gblocks v0.91b (Castresana, 2000) with less stringent at one end and 2 at the other (Fig. 2C). Smaller parameters (14, 14, 8, 8 and with half). The final percentages had only 1 FPA seen at each end (21·4%, alignment consisted of 1853 positions (78% of the 3/14) or 2 FPA at each end (14·3%, 2/14) (Fig. 2B, E). original alignment). Maximum likelihood (ML) The majority of spores with only the end available for analysis was carried out by program RAxML observation had a single FPA (77·7%, 7/9) (Fig. 2E) (Stamatakis, 2006) with GTRGAMMA model and compared to having 2 FPAs at one end (22·2%, 2/9) 500 bootstrap replicates. Maximum parsimony (MP) (Fig. 2A). The FPA were single stranded with a was performed in program PAUP* 4.0b10 (Swofford diameter of 200–300 nm along their entire length et al. 2001) with heuristic search, random addition possessing a larger rounded terminus on SEM of taxa and Ts:Tv =1:2. Bootstrap support was (Fig. 2C, D). The longest observed FPA possessing calculated from 1000 replicates. the rounded terminus was 20 μm, the shortest seen

47 Ashlie Hartigan and others 482

Fig. 2. Myxospore morphology of Cystodiscus axonis sp. n. from Litoria raniformis (A–F, hapantotype, ARWH 6985.6) and Cystodiscus australis sp. n. from Limnodynastes peronii (G–K, hapantotype, ARWH 6092.20f). (A–C) Scanning electron micrograph (SEM) of C. axonis sp. n. myxospores with filiform polar appendages (FPA) (white arrows) emerging from sutural ridges (A) or spore valve (C, D), polar filaments are distinctly thicker (grey arrows) than appendages. (D) Magnification of B showing attachment of the FPA to the valve surface and rounded terminal end of FPA (black arrow). (E) SEM of appendage showing the fine structure of FPAs. (F) Light microscopy C. axonis sp. n. showing FPAs (arrows) and polar filament coils in polar capsules (arrowhead). (G, H) SEM of C. australis sp. n. myxospores. (I) Light microscopy of C. australis sp. n. with polar filament coils in polar capsules (arrowhead). (J, K) High magnification view of myxospore end showing rounded ends of sutural depression and polar filament eversion pore (J) and myxospore showing everted polar filament with twisted rope appearance (K). Scale bars: A–E, G, H=2 μm; F, I=10 μm; J, K=1 μm.

48 Myxosporea in gallbladders of amphibians 483 spp. n. sp. n. sp. n. species and sp. n. sp. n.

Fig. 3. Cystodiscus species-speciﬁc multiplex diagnostic C. australis sp. n. (hapantotype) sp. n. (hapantotype) C. australis C. axonis C. axonis C. axonis C. axonis C. australis C. australis Cystodiscus PCR based on ITS rDNA region. Presence of single approximately 600 bp (primers set 1) and 500 bp (primer set 2) product is indicative of C. axonis sp. n. and C. australis sp. n., respectively (see Materials and 6 5 5 6 6 Methods section). Presence of both products indicates lament – – – – –

Polar ﬁ coils mixed infection. (A) Negative control (water). (B) C. axonis sp. n. from brain tissue of type host Litoria raniformis (ARWH 6985.6). (C) C. axonis sp. n. from gallbladder myxospores of type host Litoria raniformis 5·0) n.a. 5·0) 4 5·0) 4 5·0) 4 4·5) n.a. n.a. 5·0) 5 5·0) 5 5·5) 5 – – – – – – – – (ARWH 6985.6). (D) C. axonis sp. n. from tadpole liver L. aurea (ARWH 6092.14a). (E) Brain tissue of Rhinella marina (adult, ARWH 6092.45.7) (F) Gallbladder myxospore sample showing mixed Cystodiscus infection (G–H). Mixed Cystodiscus spp. infection in tadpole liver 5·0)×4·5 (4·0 5·0)×4·0 (3·0 6·0)×4·6 (4·0 5·0)×3·7 (3·0 5·0)×3·9 (3·5 5·0)×4·7 (4·0 5·0)×4·8 (4·0 6·0)×4·8 (4·0 – – – – – – – – L. aurea (ARWH 6092.5a (G) and ARWH 6092.10a (H)). (I) C. australis sp. n. from immature gallbladder plasmodium in L. aurea tadpole (ARWH 6092.37a). 4·6 (4·0 3·9 (4·0 4·1 (3·5 4·3 (3·0 5·3 (5·0 Polar capsules (J) Rhinella marina (adult, ARWH 6092.USC10) with C. australis sp. n. infection. (K) Brain tissue of type host of C. australis sp. n. (Limnodynastes peronii, adult, ARWH 6092. 20f). (L) C. australis sp. n. myxospores from type host Limnodynastes peronii, adult, ARWH laments No n.a. 3·9 (3·0 n.a Yes 3·8 (3·0 No Yes 4·7 (4·0 No Caudal ﬁ No 6092. 20f. Size DNA ladder is on the left. Agarose gel (2%) stained with GelRed (Biotium).

was 1·5 μm. However, in many cases, measurement of the FPA was obscured by their conformation on 9 13 10 12 9 6 9 11

– SEM, i.e. curled, twisted or broken. The FPA were – – – – – – – Transverse ridges present on spores with (Fig. 2C) and without ejected

spp. myxospores in Australian frogs polar filaments (Fig. 2A, B, D–F). The polar filaments were easily distinguished from FPAs by greater length =30 6 =30 5 =30 5 =30 5 =30 5 =17 5 =30 5 =30 5 (>50 μm), width (0·5–0·7 μm) and the twisted rope n n n n n n n n appearance of the polar filament (Fig. 2C). Free spores

Cystodiscus were observed in the gallbladder, but no plasmodia 10·0), 10·5), 10·0), 10·5), 10·0), 10·0), 10·0), 10·0), – – – – – – – – containing spores were seen. Pre-sporogonic developmental stages seen in histology (H&E, Giemsa) appeared similar in CNS and meninges (Fig. 4C). The )

n species identity has been conﬁrmed using species- speciﬁc multiplex PCR (Fig. 3). The species was 16·0)×9·1 (8·0 17·0)×9·1 (8·0 16·0)×9·2 (8·0 15·5)×8·5 (8·0 16·5)×8·9 (8·0 16·0)×8·6 (8·0 16·0)×8·6 (8·0 18·0)×8·7 (8·0 – – – – – – – – previously reported as the “Myxidium sp. brain genotype” by Hartigan et al. (2011).

Type host: The Southern bell frog, Litoria raniformis 15·1 (14·0 14·5 (12·0 14·8 (13·0 14·1 (13·0 14·7 (13·0 14·9 (14·0 15·1 (14·0 16·0 (15·0 Keferstein, 1867 (Hylidae, Anura), carcass not available for deposition. Type locality: Wild animal collected in Coleambally, New South Wales, Australia (−34° 47′ 49″ S,145° 59′ 39″ E) and kept captive in Sydney, New South Wales until May, 25th 2009. ARWH ARWH ARWH ARWH Species and genus type material (hapantotype): G.17758. Histological slides of the liver, CNS, gut and gonad (H&E stain), SEM stubs and spores in 70% ethanol deposited in the Australian Museum, 6092.USC10 6092.45.7 6092.29a ARWH 6985.6 6092.54a ARWH 6092.55b ARWH 6092.34b ARWH 6092.20f Bufo marinus Bufo marinus Litoria peronii Litoria raniformis Litoria aurea Limnodynastes peronii Limnodynastes peronii Table 2. Summary of morphological characteristics of Host species and catalogue number Myxopore size ( Limnodynastes peronii Sydney, New South Wales, Australia.

49 Ashlie Hartigan and others 484

B C

D E

Fig. 4. Histology of nervous tissue with pre-sporogonic development of Cystodiscus axonis sp. n. (A) Distribution of myxosporean plasmodial stages (arrows) within spinal cord of Litoria peronii (adult, ARWH 6092.29a). Note concentration of myxosporean development at the periphery. (B) Higher magnification of the spinal cord of L. peronii (adult, ARWH 6092.29a). (C) Pre-sporogonic plasmodium in the meninges of an adult Litoria raniformis (type host, ARWH 6985.6). (D) Dorsal root nerve of Litoria aurea (adult, ARWH 6092.36a) showing glial infiltration associated with the presence of Myxosporea. (E) Higher magnification of the pre-sporogonic plasmodia in the dorsal root nerve of L. aurea (adult, ARWH 6092.36a). Scale bars: A=500 μm, B–C=20μm, D=50 μm, E=20 μm.

Additional material: Formalin-fixed paraffin- Wildlife Health, Mosman, NSW, Australia (ARWH embedded blocks, unfixed tissues in −80 °C of the 6985.6); the type belongs to a cohort of adult type host L. raniformis at the Australian Registry of L. raniformis ARWH 6985.1-7.

50 Myxosporea in gallbladders of amphibians 485

Nucleotide signature sequenced data: HQ822168, Ultrastructure of intra-axonal plasmodia of HQ822169, HQ822171 – SSU rDNA; JN603234, Cystodiscus axonis sp. n.: Brain tissue of JN603235 – ITS1 rDNA, 5·8S rDNA, ITS2 rDNA L. aurea (adult ARWH 6092.36a), confirmed to be (GenBank Accession numbers); “Myxidium sp. brain infected with C. axonis sp. n. by PCR, showed genotype” sensu Hartigan et al. (2011). myxosporean stages in semi-thin sections. The stages Etymology: The species epithet is derived from the were characterized by a large plasmodium with anatomical term ‘axon’ to indicate the location of numerous daughter cells within, located exclusively the developmental stages of this species in the host. within myelinated axons (Fig. 5C). Brain tissue The name is derived as Latin singular genitive. samples from L. raniformis (ARWH 6985.4-5), L. aurea (ARWH 6092.10a), and Lim. peronii Other hosts and locality of Cystodiscus axonis (ARWH 6092.55b) were processed for TEM but no sp. n.: Table 1 lists frog species infected with myxosporean development was detected. Using C. axonis sp. n. Adults of L. raniformis from the TEM, multiple myxosporean development stages same cohort as the type host (ARWH 6985.6) were were observed throughout the sample of L. aurea infected with C. axonis sp. n. (ARWH 6985.1-5, 7). (ARWH 6092.36a). The myxosporean stages ob- Myxosporean stages were identified in the brain and served included a single cell enveloping one daughter CNS, occasionally associated with glial inflammation cell or more complex myxosporean cell formations, (Hartigan et al. 2011). Myxosporean developmental including secondary and tertiary cell development stages within myelinated axons were identified in (Fig. 5D and E). All stages observed were pre- the CNS and dorsal nerve roots (Fig. 4) of 3 native sporogonic. The secondary cells had only mitochon- frogs in Riverstone, outskirts of Sydney, NSW dria and glycogen granules and the tertiary cells and considered conspecific with stages detected appeared to have very little cytoplasm with 1 or 2 in L. raniformis (ARWH 6985.1-7). Numerous mitochondria (Fig. 5E). Microtubules were seen C. axonis sp. n. plasmodia were seen throughout throughout the large primary cell cytoplasm and the brain and CNS of Litoria peronii (adult, ARWH numerous mitochondria with tubular cristae 6092.29a) (Fig. 4A), the pre-sporogonic plasmodia (Fig. 5F). More than 1 plasmodial body was observed were found on the periphery of the CNS and were in an axon, often one was significantly smaller than associated with microgliosis (Fig. 4B). The immature the other (Fig. 5G). The individual plasmodia were stages in L. peronii ARWH 6092.29a were indis- never connected to each other or the axonal sheath. tinguishable from C. axonis sp. n. in the meninges of No pseudopodia were observed on the external L. raniformis (type host ARWH 6985.6) (Fig. 4C) surfaces of these stages. The primary cells had a and the dorsal root nerve of Litoria aurea (adult, high degree of multicellularity, the cytoplasm con- ARWH 6092.36a) (Fig. 4D and E). The CNS stages tained numerous vacuoles, glycogen granules and in L. peronii (adult, ARWH 6092.29a) and L. aurea Golgi apparatus (Fig. 5G). Despite the complexity of (adult, ARWH 6092.36a) were confirmed to be these developmental stages no sporogonic activity C. axonis sp. n. with species-specific PCR and was observed. nucleotide sequencing (Hartigan et al. 2011). A third frog from Riverstone NSW Limnodynastes Cystodiscus australis sp. n. peronii (adult, ARWH 6092.55b) was identified with C. axonis sp. n. developmental stages in the brain and Description: Ellipsoidal myxospores 5–11 transverse spores in the gallbladder (Table 2) (confirmed with ridges (4–12 transverse depressions) on the outside species-specificPCR)(Fig. 3). Pre-sporogonic plas- each of the 2 valves visible under light microscopy modia in the bile ducts of L. aurea (tadpoles ARWH LM and SEM (Fig. 2G–I). Spore width and length 6092.14a) (Fig. 4) were confirmed to be C. axonis. 16·0 (15·0–18·0) ×8·7 (8·0–10·0) μm and polar sp. n. with species-specific multiplex PCR (Fig. 3). capsules 5·3 (5·0–6·0)×4·8 (4·0–5·5) μm with 5–6 Infection with C. axonis sp. n. of this tadpole turns of the polar filament (n=30, in PBS) (Table 2). specimen was associated with lymphoplasmocytic The 2 spore valves were joined by a raised ridge that inflammation and hyperplasia of the biliary system. appeared straight or slightly curved (S-shaped in An adult R. marina (ARWH 6092.45.7) from sutural view) along the medial axis of the spore. The Lismore, NSW was identified with a mixed infection raised sutural ridge was accompanied by a depression of Cystodiscus species. The brain stages were unequi- that varied in thickness and depth from 0·5 to 1·5 μm vocally confirmed as C. axonis sp. n. only using at its widest point and tapered towards the poles species-specific multiplex PCR (Fig. 3). Mixed (Fig. 2H, J, K). Transverse depressions on the infection was confirmed for spores in the gallbladder surface of the spores varied in depth and length and using species-specific multiplex PCR (Fig. 3). appeared connected to the sutural ridge depression. Similarly, tadpole livers (ARWH 6092.5a and The majority of the depressions seen were parallel. ARWH 6092.10a) of L. aurea were identified by No FPAs were observed on any spores in either light species-specific multiplex PCR to contain a mixed microscopy or SEM. The polar filament eversion infection of Cystodiscus species (Fig. 3). pores were approximately 500 nm in diameter

51 Ashlie Hartigan and others 486

Fig. 5. Histological and ultrastructural morphology of Cystodiscus axonis sp. n. in L. aurea tadpole (ARWH 6092.14a) and adult (ARWH 6092. 36a). (A) Histological section of C. axonis sp. n. pre-sporogonic plasmodia (arrows) in bile ducts of L. aurea tadpole (ARWH 6092.14a). (B) Biliary plasmodia at higher magniﬁcation (ARWH 6092.14a). (C) Transmission electron micrograph of C. axonis sp. n. plasmodium in myelinated axon of L. aurea adult (ARWH 6092.36a) showing secondary and tertiary daughter cells within the primary cell. (D) Secondary cell (Sc) and nucleus (Sn) with internal tertiary cell (Tc) of C. axonis sp. n. (ARWH 6092.36a). (E) Microtubules (arrow) throughout the cytoplasm of the primary of C. axonis sp. n. plasmodium (ARWH 6092.36a). (F) A plasmodium budding oﬀ another plasmodium within a single axon, note absence of attachment to the axonal sheath (ARWH 6092.36a). (G) A primary cell seen with tubular cristae mitochondria (m), Golgi apparatus (arrow), and glycogen granules (Gg) in the cytoplasm of the cell. Scale bars: A=100 μm, B=20 μm, C=5 μm, D–E=1μm, F=2 μm, G=1μm.

52 Myxosporea in gallbladders of amphibians 487

A B C

D E F

G H I

J K

Fig. 6. Histological and ultrastructural morphology of Cystodiscus australis sp. n. (A–C) Biliary and gallbladder development of Myxosporea from Litoria aurea tadpole (ARWH 6092.37a). (A) Lymphoplasmocytic inflammation of the bile ducts and proliferation of the bile tree associated with myxosporean development (arrows), H&E. (B) Immature pre-sporogonic developmental stages (arrows) in the biliary system, H&E. (C) An immature pre-sporogonic plasmodium from the gallbladder on black background. (D–K) Transmission electron microscopy of sporogenesis and myxospores within plasmodia from Rhinella marina (adult, ARWH 6092.USC10). (D) Early stage of sporogenesis, differentiation of sporoplasm (Sp), capsulogenic cell (Cc) and valve cell (Vc), (ARWH 6092.USC10). (E) Myxospore with defined valvogenic cell (n=valve cell nuclei) and sporoplasm, note clearly defined sutural ridge and depression (arrow indicating depression) (ARWH 6092.USC10). (F) Internal structure of the myxospore sutural ridge, note the microtubules lining the sutural depression (ARWH 6092.USC10). (G) Early definition of myxospore polar capsule, spiral pattern of the

53 Ashlie Hartigan and others 488

(Fig. 2J). Polar filaments are 500 nm thick and had multiplex PCR (Fig. 3). The pre-sporogonic stages of an appearance of 3 or more single filaments twisted C. australis sp. n. were associated with lymphoplas- around each other (Fig. 2 K). Polysporic gallbladder mocytic inflammation and proliferation of the biliary plasmodia ranged from 0·5 to 1·5 mm in diameter, tree of the liver (Fig. 6A). The gallbladder contained <0·3 mm in thickness and were circular or oval in pre-sporogonic plasmodia confirmed by species- shape (Fig. 6C). The plasmodia were unattached to specific multiplex PCR as C. australis sp. n. The the gallbladder epithelium and could be seen without pre-sporogonic plasmodial stages were free floating a microscope when the gallbladder was held against in the bile, appeared highly vacuolated and ranged direct light. Spores were free floating or attached to in diameter from 20–150 μm(Fig. 6B). An adult the plasmodium. The species identity has been R. marina (ARWH 6092.USC10) from Lismore, confirmed using species-specific multiplex PCR NSW was identified with plasmodia and spores in the (Fig. 3). The species was previously reported as the gallbladder. The spores were morphologically ident- “Myxidium sp. liver genotype” sensu Hartigan et al. ified as C. australis sp. n. and confirmed with species- (2011). specific multiplex PCR (Fig. 3). Species-specific multiplex PCR was used to detect Type host: Striped Marsh Frog, Limnodynastes the presence of spores of both species (C. axonis peronii Dumeril and Bibron, 1841 (Myobatrachidae, sp. n. and C. australis sp. n.) in R. marina (ARWH Anura), carcass was deposited at the Australian 6092.45.7) and L. aurea tadpoles (ARWH 6092.5a Museum, Sydney, New South Wales, Australia and ARWH 6092.10a) (Fig. 3). (R.175753). Type locality: Wild animal caught in Riverstone, Ultrastructure of gall bladder C. australis New South Wales, Australia (−33°41′7″, 150°50′53″) sp. n. plasmodia with spores: Gall bladder plas- on 19 September 2008. modia of R. marina (adult, ARWH 6092.USC10) Species type material (hapantotype): G.17756. were confirmed as C. australis sp. n. by species- Formalin-fixed spores in 70% ethanol deposited in specific multiplex PCR (Fig. 3). The plasmodia were the Australian Museum, Sydney, New South Wales, fixed, embedded and sectioned to observe the Australia. ultrastructure of C. australis sp. n. The plasmodia Additional material: Myxospores in 70% ethanol and contained various stages of spore development SEM stubs of Lim. peronii (AWRH 6092.34b) (sporogony) including spores identified throughout deposited at the Australian Museum (G.17757) and the cytoplasm. Early stages of sporogenesis were formalin-fixed paraffin-embedded blocks, unfixed observed where valve, sporoblast and capsular cells tissues in −80 °C at the Australian Registry of had differentiated and started to envelope one another Wildlife Health, Mosman, NSW, Australia. (Fig. 6D). Valve cells with thick cytoplasm and nuclei Nucleotide signature sequenced data: HQ822152 – still present were seen covering sporoblast cells and SSU rDNA; HQ822193, HQ822194, HQ822195, capsulogenic primordium (Fig. 6D). Sporoblast cells HQ822196, HQ822203, HQ822208, HQ822211, developed a thick electron-dense layer, which defined HQ822212, HQ822215, HQ822216, HQ822221, the cellular boundary between sporoblast, valve cell, HQ822222, HQ822223 – ITS1 rDNA, 5·8S rDNA, and capsular cell (Fig. 6D). The valve cells enlarged ITS2 rDNA; HQ822178, HQ822179 – LSU rDNA to envelop the sporoplasm and began to develop (GenBank Accession numbers); “Myxidium sp. liver structural features (Fig. 6E). The sutural ridge and genotype” sensu Hartigan et al. (2011). depression of the spore could be seen before the Etymology: The species epithet is derived from a capsular primordium had begun to differentiate noun in the genitive case, i.e. australis, from the (Fig. 6E). The sutural ridge and adjacent depression geographical name Australia. was lined with microtubules (Fig. 6F). There was no indication of any FPAs or other external features to Other hosts and locality of C. australis sp. n.: the spore valve. The capsular primordium or polar Table 1 lists frogs infected with C. australis capsule development formed a circular structure sp. n. Myxosporean plasmodia in the bile ducts of a with thick rings in its cell wall (Fig. 6G). The angle L. aurea tadpole (ARWH 6092.37a) were confirmed of sectioning could make the capsular wall appear to be C. australis sp. n. only using species-specific spiralled, the external tube was still visible before

polar capsule wall is an artefact caused by sectioning angle. The external tube of capsular primordium is visible (*), this is prior to the formation of polar filaments. (H) Polar capsule wall thickened in a later stage of myxospore development, club-shaped polar filaments seen within the capsule. (I) Polar capsule development of the polar eversion pore (arrow), polar filaments appear S-shaped due to sectioning angle. (J) Cross-section of gall bladder plasmodium epithelium showing elongated projections on surface (arrow), and myxospore positioning adjacent to the primary cell wall. (K) Mature myxospore showing binucleate sporoplasm with the developed polar capsule, transverse ridges and depressions on the valve surface. Scale bars: A=20 μm, B–C=100 μm, D–E=1μm, F=500 nm, G–I=1μm.

54 Myxosporea in gallbladders of amphibians 489 development of the polar filaments (Fig. 6G). The complex cellular differentiation (Fig. 7C–F). The polar filaments appeared S shaped (Fig. 6I), the early stages of development were simple primary cells definition of the polar filaments coincided with a with a single secondary cell within (Fig. 7C). These thickening of the polar capsule wall (Fig. 6H–K) and early stages were small and free floating within a bile development of a discharge pore (Fig. 6I, arrow). duct. In contrast, the primary cells in advanced The surface of plasmodia was covered in villus-like developmental stages became more complex with an projections that may represent pseudopodia (Fig. 6J). increased number of secondary cells, the plasmodium A lattice of electron-dense material and microtubules expanded to the limits of the bile duct (Fig. 7D). connected the plasmodial walls, spores were found Sometimes there was very little intercellular space along the lattice and close to the primary cell wall between the plasmodial primary cell and the host’s (Fig. 6J). Ridges and depressions on the surface of the bile duct (Fig. 7D). Larger plasmodia were observed spore valve were easily recognizable (Fig. 6K). with higher levels of cellular organization, the primary cells in addition to secondary cells, had Remarks: Only 2 Myxidium-like species numerous mitochondria with tubular cristae, elec- (=Cystodiscus spp.) were detected in Australian tron-dense lipids and vacuoles (Fig. 7E). Secondary frogs (Hartigan et al. 2011). Both species have cells were relatively simpler in cellular organization. been identified in 3 major frog families (Hylidae, The cytoplasm of secondary cells mostly contained Myobatrachidae and Bufonidae). The spores of mitochondria (Fig. 7F) and vacuoles containing C. axonis sp. n. are distinguished from C. australis granules (Fig. 7F arrow). Despite the detail and sp. n. by the presence of FPAs and pre-sporogonic intracellular proliferation seen within myxosporean stages in the CNS of affected frogs. primary cells in the bile ducts, no tertiary cells or sporogenesis were observed. Ambiguous biliary stages of Cystodiscus in Phylogenetic position of the genus Cystodiscus tadpoles Sequences of all Cystodiscus species clustered to- Cystodiscus species development in the bile ducts gether in both ML and MP analyses with a maximum of tadpoles were deemed to be morphologically bootstrap support for the clade monophyly (Fig. 8). indistinguishable (tadpole livers of L. aurea, The sequence of C. axonis sp. n. is a well-supported ARWH 6092.5a and ARWH 6092.10a, see Table 1). sister branch to the rest of the Cystodiscus spp. Histology of liver showed myxosporean developmen- sequences. Relationships of C. australis sp. n., tal stages, multinucleated plasmodia, in the biliary C. melleni, and 2 sequences of C. cf. immersus had tree (Fig. 7A, B; ARWH 6092.10a). The plasmodia moderate or low bootstrap support but their positions appeared unattached to the biliary epithelium, oval were stable in both analyses. The Cystodiscus spp. to round and 5–50 μm in diameter. A single bile group clustered inside the freshwater gallbladder duct could contain multiple pre-sporogonic plasmo- clade according to Fiala (2006) with a close relation- dia. The myxosporean developmental stages were ship to the representative of the marine genus associated with inflammation and proliferation of Sphaeromyxa in ML (bootstrap <50%) and to the the bile ducts (Fig. 7A, B). Morphologically similar terrestrial Soricimyxum fegati (bootstrap <50%) in stages were observed in L. aurea tadpoles ARWH MP. The SSU rDNA alignment did not robustly 6092.14a (Fig. 5A and B) and ARWH 6092.37a resolve the phylogenetic position of the monophy- (Fig. 6A and B) confirmed with species-specific letic Cystodiscus spp. clade within the myxosporean multiplex PCR (Fig. 3) to be single species infections tree, as indicated by low bootstrap values (Fig. 8). of C. axonis sp. n. and C. australis sp. n respectively. No morphological features were noted to be species specific based on histology. Identical myxosporean stages were seen in tadpoles of L. aurea (ARWH DISCUSSION 6092.42 m), L. peronii (ARWH 6092.63i) and Lim. The description of these 2 new myxosporean peronii (ARWH 6092.57e), the liver tissues of these parasites was based on a large collection of native individuals were not available for PCR. and exotic frog material from eastern Australia. The Transmission electron microscopy was used to combination of histological, ultrastructural and investigate the ultrastructure of Cystodiscus spp. genetic characterization has unequivocally discrimi- developmental stages of L. aurea tadpole liver nated 2 Australian frog species from each other and (ARWH 6092.42 m) which was not available for from described myxosporean taxa from other con- PCR. Plasmodia were only observed within the bile tinents. Our investigation led to the re-examination duct. No pseudopodia or other attachment structures of frog gallbladder Myxosporea and subsequent were observed to connect the plasmodial cell to the resurrection of the genus Cystodiscus Lutz, 1889. biliary epithelium. The plasmodial development was The coupling of traditional myxospore morpho- asynchronous in ARWH 6092.42 m, stages observed logical description with molecular analysis has aided ranged from a single cell enveloping a cell to more the diagnostic methods to differentiate between these

55 Ashlie Hartigan and others 490

Fig. 7. Cystodiscus species developmental stages in the bile ducts of tadpoles. (A–B) Myxosporean development in bile ducts of Litoria aurea (tadpole, ARWH 6092.10a). (A) Histological section of biliary ducts showing multiple immature free-floating plasmodial forms of Cystodiscus spp. n. (arrows), H&E. (B) Higher magnification of multinucleated plasmodia in bile ducts, H&E. (C–F) Transmission electron microscopy of myxosporean development in bile ducts of L. aurea (tadpole, ARWH 6092.42 m). (C) Early plasmodial stage with primary cell (Pc) and nucleus (Pn) with a single internal secondary cell (Sc) with nucleus (Sn). (D) Bile duct filled with myxosporean plasmodium with numerous internal secondary cells, the plasmodium has expanded to the limits of the biliary duct (arrow). (E) Plasmodium in a bile duct showing cellular complexity of primary cell, numerous vacuoles (v), mitochondria (arrows) and secondary cells within the cytoplasm. (F) Secondary cell at higher magnification, internal organization included nucleus, mitochondria with tubular cristae (m) and vacuoles (arrow). Scale bars: A=50 μm, B=20 μm, C=1 μm, D=5 μm, E=2 μm, F=1 μm.

56 Myxosporea in gallbladders of amphibians 491

Cystodiscus australis HQ822149 59/69 C. australis 71/66 Cystodiscus cf. immersus Brazil-2 HQ822159 87/94 Cystodiscus cf. immersus Brazil-1 HQ822162 100/100 C. axonis Cystodiscus melleni DQ003031 Cystodiscus axonis HQ822169 Sphaeromyxa hellandi DQ377693 Soricimyxum fegati EU232760 Myxidium coryphaenoideum DQ377697 Myxidium anatidum EF602629 M. coryphaenoideum 100/100 100/69 Zschokkella nova DQ377688 Myxidium truttae AF201374 100/99 Zschokkella icterica DQ333434 99/88 M. anatidum Myxidium cuneiforme DQ377709 M. truttae 84/100 85/90 Zschokkella parasiluri DQ377689 M. cuneiforme 100/100 Myxidium hardella AY688957 Myxidium chelonarum DQ377694 53/56 Myxidium scripta DQ851568 M. hardella Chloromyxum cristatum AY604198 52/- Henneguya ictaluri AF195510 100/98 Myxobolus cerebralis MCU96492 Myxobolus lentisuturalis AY119688 99/72 Henneguya salminicola AF031411 Chloromyxum schurovi AJ581917 M. giardi 100/100 100/100 Myxidium giardi AJ582213 Myxidium lieberkuehni X76639 100/100 Sphaerospora oncorhynchi AF201373 M. lieberkuehni (type species) Chloromyxum leydigi AY604199 0.1

Fig. 8. Phylogenetic position of Cystodiscus spp. within the myxosporean freshwater clade. Maximum likelihood tree (−ln=14644.0851) with GTRGAMMA model of evolution. Numbers at the nodes represent bootstrap values (ML/ MP) for the nodes gaining more than 50% support. GenBank Accession numbers are next to the species names. Scale bar is given under the tree. Schematic drawings illustrate the diversity of spore morphology of Myxidium and Cystodiscus spp. species as suggested by Fiala and Bartošová (2010). these species is restricted entirely to CNS and In addition, the examination of in situ tissue associated cartilage and is dissimilar to the parasite development provided more information about development observed in frogs. The axonal stages the biology and development of these parasites. The are ultrastructurally very similar to parasite stages developmental stages of C. axonis sp. n. within the in the axons of the Bullhead Cottus gobio (Lom et al. axons of tadpoles are considered to be extraspor- 1989), a lungfish Polypterus endlicheri (Marquet ogonic, because no stages were observed to progress and Sobel, 1970) and a toad Bufo arenarum to sporogenesis. Extrasporogonic stages are suggested (Stensaas et al. 1967). The significance of the axonal to exist for proliferation of the parasite only development of Myxosporea remains unknown. (Lom, 1987; Lom and Dyková, 1992). For example, Therefore, it is likely that axonal development is extrasporogonic stages in the blood are thought to not restricted to C. axonis sp. n., and complete amplify and spread infection of Sphaerospora dykovae histopathology including CNS may reveal that axon (syn. S. renicola) in carp (Lom et al. 1985). Multiple tropism of myxosporeans is more common than attempts to identify blood cell stages were unsuccess- previously recognized. It remains unknown, if the ful in any of the animals infected with C. axonis axonal stages described by Stensaas et al. (1967)in sp. n. or C. australis sp. n. (A.H., unpublished data). South American toads represent C. axonis sp. n. or a Without blood cell stages, the presence of extraspor- closely related new species. ogonic plasmodia in myelinated axons raises at least Examination of host tissues and the developing two significant questions. It is yet unknown, if the stages of C. australis sp. n. and C. axonis sp. n. offered stages form a tissue pool that continually contributes important information about the significance of to the terminal bile duct development. It is yet to be infection to the host and revealed a significant discovered, if the swollen parasitized axons have species-specific diagnostic feature. These features decreased capacity for signal transmission and there- would otherwise be missed if only molecular data and fore an effect on motor function of frogs compared to spore measurements were used for species identifi- unaffected frogs. cation. Spore ultrastructure revealed the presence of Significant disease has been observed in FPAs on the spores of C. axonis sp. n. and serves Myxosporea affecting CNS in fish, such as as the distinguishing feature between otherwise Myxobolus cerebralis, Myxobolus neurobius, and similar looking spores in Australian frogs. To our Myxobolus cottis. However, the development of knowledge no other myxosporeans possess such fine

57 Ashlie Hartigan and others 492 appendages. Appendages such as those described for Spores oval or ellipsoidal with raised sutural Tetrauronema desqualis or Henneguya adiposa are ridge between the valves. The sutural ridge straight thicker and easily observed in light microscopy or curved along the medial axis of the spore. One (Azevedo and Matos, 1996;Griffin et al. 2009). or more longitudinal ridges parallel to the sutural While the FPA of C. axonis sp. n. were observed ridge. Transverse ridges across the valve sur- using light microscopy, they may often remain un- face absent or present. Disc-like plasmodia with noticed. It is likely that Delvinquier (1986) observed spore development in gallbladders of members of C. axonis sp. n. during his extensive frog parasite Amphibia. survey and SEM myxospore studies. Therefore, thorough morphological re-evaluation of parasitolo- Synopsis of Cystodiscus Lutz, 1889 emend. gical specimens is imperative for accurate species identification and classification. Cystodiscus immersus Lutz, 1889 The taxonomy and systematics of Myxosporea is Syn. Cystodiscus immersus Lühe, 1899; Cystodiscus largely based upon the spore morphology and host immersus Cordero, 1919; Myxidium immersum Kudo species (Lom and Dyková, 2006). More recently, and Sprague, 1940 molecular characterization became a complementary part of species descriptions. Such combination of Myxospore morphology: 12·0–14·0×9·0–10·0 μm molecular phylogenetics and morphology is serving with sutural ridge at 45° angle from lateral view, as the basis for improved systematics of Myxosporea but no ridges are mentioned or illustrated (Lutz, (Fiala and Bartošová, 2010). One of the most 1889). Later, Cordero (1919) noted the presence of diverse myxosporean genera is Myxidium Bütschli, transverse ridges. Spores are re-described as “broadly 1882 which is now comprising numerous phylogen- fusiform with bluntly rounded extremities in etically unrelated clades sharing similar spore mor- front view [...] rounded rectangular in side view. phology (Fiala, 2006; Lom and Dyková, 2006). The slightly curved sutural line of the shell-valves Molecular phylogenetics based on SSU rDNA and makes an acute angle with the longitudinal axis of LSU rDNA laid the foundation for several robust the spore and extends to the diagonally opposite clades within Myxidium sensu lato that await points at the extremities” (Kudo and Sprague, 1940). taxonomic reclassification (Fiala, 2006; Fiala and Re-description: 11·8–13·3×7·5–8·6 μm, S-shaped Bartošová, 2010; Hartigan et al. 2011). Here, we re- suture, 1 longitudinal depression encircling the classify one such clade embracing amphibian gall- valve parallel to the sutural ridge, 7–9 transverse bladder species and resurrect the genus Cystodiscus ridges (6–8 transverse depressions) situated toward Lutz, 1889 from the synonymy of Myxidium the smaller end not reaching the larger end of the shell Bütschli, 1882. valve. Polar capsules broadly piriform 3·5–4·2 μm. Lutz (1889)defined the genus Cystodiscus based No FPA. Spores in the gallbladder. on the large disc-like appearance of the gallbladder plasmodia “Für das Genus möchte ich zur Type host: Bufo agua =Cane toad Rhinella marina Bezeichnung der regelmässigen Form und des zelli- Linnaeus, 1758 syn. Bufo marinus (Bufonidae, gen Baus den Namen Cystodiscus und für die Species Anura) (Lutz, 1889). Also described in captive Bufo die Bezeichnung immersus vorschlagen” (=for the agua =Bufo marinus by Lühe (1899). Re-described genus, I would like to propose a name Cystodicus from R. marina (Kudo and Sprague, 1940). referring to its regular form and the cellular Type locality: Brazil. construction, and for the species the name immersus) Type material: Material described by Lutz (1889) of the species Cystodiscus immersus Lutz, 1889 from and the re-description based on material that Cane toads in Brazil. The genus name was later consisted of several sporulating trophozoites “from abandoned and fell into the synonymy of Myxidium the gallbladder of a Bufo marinus from Brazil in Bütschli, 1882 (Kudo and Sprague, 1940). The August 1937” (Kudo and Sprague, 1940). No available molecular phylogenetic data clearly place material deposited. Cystodiscus Lutz, 1889 as a distinct monophyletic Other hosts: Lutz (1889) noted that the same parasite lineage termed “Myxidium-frog clade” (Hartigan was found in Cystignathus ocellatus=Leptodactylus et al. 2011). Due to the polyphyletic nature of the bolivianus Boulenger, 1898 (Leptodactylidae, genus Myxidium we propose to resurrect the genus Anura). The list of frog species infected with with the following emendations and provide a C. immersus by Cordero (1919) is treated with caution synopsis for named species. because Hartigan et al. (2011) demonstrated the presence of multiple distinct genotypes within one morphotype. The species was assumed to be translo- Cystodiscus Lutz, 1889 emend. cated to Australia (Delvinquier, 1986), but now Type species: Cystodiscus immersus Lutz, 1889 all Australian gallbladder Cystodiscus spp. are con- Bivalved myxospores with 2 polar capsules at sidered distinct from this species (Hartigan et al. opposite poles and a central binucleate sporoplasm. 2011).

58 Myxosporea in gallbladders of amphibians 493

Remarks: Kudo and Sprague (1940) wrote about Kudo and Sprague (1940); some spores were seen C. immersus that “when the spore is seen in front with branching transverse ridges and depressions, view, transverse ridges occupy about two-thirds of some with both transverse and longitudinal patterns. the shell valve and the remaining one-third is without The nearly spheroidal polar capsules measure any transverse ridges” and use it to distinguish it 5·0–5·5 μm with 3–5 polar filament coils. No FPA. from C. serotinus. Kudo and Sprague (1940) men- Spores in the gallbladder. tioned that the numbers and arrangement were remarkably constant for spores they had observed. Type host: Northern Leopard Frog Rana pipiens Genetic analysis of spores consistent with Cystodiscus Schreber, 1782 (Ranidae, Anura). cf. immersus from R. marina collected in Brazil Type locality: First seen in a frog in a “class room”,at revealed the presence of 2 distinct genotypes The University of Illinois, USA; an additional (Hartigan et al. 2011). These Cystodiscus cf. immersus survey by Kudo and Sprague (1940) revealed that spores were characterized ultrastructurally and there- “11·3 per cent of a group of 150 Rana pipiens which fore they remain unassigned until morphological were collected from Wisconsin, Iowa, Minnessota, analysis coupled with genetic characterization is etc.”, were infected with the Myxosporea. conducted. Etymology not given. Type material: Material described by Kudo and Sprague (1940). No material deposited. Cystodiscus lyndoyense Carini, 1932 Other hosts: Rana sp. “15·4% of 26 undetermined species of Rana of Louisiana” (Kudo and Sprague, Syn. Myxidium lyndoyense Kudo and Sprague, 1940 1940). In the later study by Kudo (1943) the parasite Myxospore morphology: Broadly oval shaped and was identified in 25/41 (60·9%) of the Southern fusiform in sutural view, 11–12×7·5–8 μm, with toad Bufo terrestris =Anaxyrus terrestris Bonnaterre, rounded polar capsules 4 μm in diameter. The suture 1789, 6/21 (28·6%) of Rana clamitans =Lithobates S-shaped with a sutural ridge and an additional clamitans Latreille, 1801 and 33/88 (37·5%) of longitudinal ridge running parallel present (2 longi- Rana sphenocephala=Lithobates sphenocephalus tudinal depressions). No transverse ridges/ Cope, 1886. The species was considered identical depressions noted (Carini, 1932). Transverse ridges to the one found in the gallbladder of the Two considered present in the revision by Kudo and lined salamander Eurycea bislineata Green, 1818 Sprague (1940). No FPA. Spores in the gallbladder. (Plethodontidae, Caudata) from Cabell and Wayne countries, West Virginia, USA in high (87·9%, 51/58, Type host: Cane toad Rhinella marina Linnaeus, 1758 spring of 1969) prevalence (Clark and Shoemaker, syn. B. marinus (Bufonidae, Anura). 1973). While the illustration (Clark and Shoemaker, Type locality: Lindóia – São Paulo, Brazil. 1973) and SEM (Clark, 1982) of the spores of Type material: Material described by Carini (1932). salamanders appear to show spores to be identical No material deposited. to those seen by Kudo and Sprague (1940), Other hosts: Carini (1932) has recorded this species further work using molecular typing is deemed in at least 4 species of frogs in 3 families (Bufonidae, necessary to confirm identify of these parasites in Hylidae, Leiuperidae). frogs and salamanders. In addition, a large number Remarks: Kudo and Sprague (1940) and all of North American frogs and salamanders have future authors treat this name as a synonym of tentatively been considered to be infected by this C. immersus Lutz, 1889. We treat this name species (McAllister and Trauth, 1995; McAllister and cautiously as nomen dubium/species inquirendae until Bursey, 2005; McAllister et al. 2008, 2009). further data on the diversity of Cystodiscus spp. in Remarks: Kudo and Sprague (1940) distinguished South American frogs is available (see remarks for this species by its size and the large number of ridges C. immersus Lutz, 1889). Species named after the and depressions (covering entire spore surface) from type locality – Lindóia. the only other known myxosporean species they recognized valid in frog gallbladders of the time, Cystodiscus serotinus (Kudo and Sprague, 1940) C. immersus. Etymology not given. comb. n. Cystodiscus haldari (Sarkar, 1982) comb. n. Syn. Myxidium serotinum Kudo and Sprague, 1940 Syn. Myxidium haldari Sarkar, 1982 Myxospore morphology: Ellipsoidal, broad ended, 16·0–18·0×9·0 μm, with 2–4 longitudinal ridges Myxospore morphology: ‘Fusiform or cylindriobico- (3–5 longitudinal depressions) which are parallel nical in sutural view’, spindle to oval shaped, 10·8 to the sutural ridge. The suture curves (S-shape) (10·0–12·0) ×6·7 (6·5–7·0) μm with 8–10 longitudinal before the end of the spore. The valve surface is (end to end) ridges (7–11 longitudinal depressions) covered with 10–13 transverse ridges (9–14 transverse on the surface without transverse ridges. Polar depressions), some ridges are fused. The orientation capsules round, 3·6 (3·0–4·0) μm. No FPA. Spores of ridges and depressions was variable according to in the gallbladder.

59 Ashlie Hartigan and others 494

Type host: European tree frog Hyla arborea Cystodiscus typhonius (Gray, 1993) comb. n. Linnaeus, 1758 (Hylidae, Anura), 20% (1/5) frogs Syn. Myxidium typhonius Gray, 1993 (Sarkar, 1982). As this frog is absent in India, we speculate that it was Jerdon’sTreeFrogHyla Myxospore morphology: Ellipsoidal, 10·9 (9·8– annectans Jerdon, 1870, because it is the only 12·2) ×7·2 (5·7–8·9) μm with a single straight sutural representative of the genus Hyla in the Hyla arborea ridge extending from pole to pole, accompanied with group in India (Faivovich et al. 2005). a sutural depression. The valve surface with 9–11 Type locality: Chakdaha, West Bengal, India. non-branching transverse ridges (8–10 transverse Type material: Specimens on slides (H–M 6) stained depressions) parallel to each other. Polar capsules with Giemsa described by Sarkar (1982)as“to be 3·8 (2·5–5·5)×3·6 (3·3–5·2) μm with 4–5 polar deposited soon” in the National Collection of the ﬁlament coils. No FPA. Spores in the gallbladder. Zoological Survey of India, Calcutta (http://www. envfor.nic.in/zsi/). Type host: Bufo typhonius =Trachycephalus typhonius Remarks: This species is distinct from all other Linnaeus, 1758, 88% (30/34) of adult frogs collected Cystodiscus spp. by the presence of 8–10 longitudinal in December 1989 – January 1990. ridges (9–11 longitudinal depressions) and no trans- Type locality: Puerto Maldonaro, Peru, South verse ridges (Sarkar, 1982). Named after Dr America. D. P. Haldar. Type material: Formalin preserved plasmodia containing spores deposited into the United States Cystodiscus lesminteri (Delvinquier, Markus National Museum, 10th St. and Constitution Ave. and Passmore, 1992) comb. n. NW, Washington, DC 20560, Helminth collection number 81272. Syn. Myxidium lesminteri Delvinquier, Markus Other hosts: A survey of Peruvian amphibians in and Passmore, 1992 December 1989 – January 1990 showed: Rhinella Myxospore morphology: Ovoid, 12·5 (9·5–15·0) ×6·5 marina Linnaeus, 1758, 72·7% (8/11) (Bufonidae, (5·7–8·0) μm with 1 longitudinal depression encir- Anura); Dull Rocket Frog, Colostethus marchesians= cling the valve parallel to the sutural ridge. Allobates marchesianus Melin, 1941 90·9% (10/11) Additional longitudinal depressions (1–2 seen in (Aromobatidae, Anura); Crump Treefrog, Hyla SEM) occupying two-thirds of spore valve, no brevifrons=Dendropsophus brevifrons Duellman and transverse depressions. Polar capsule measurements Crump, 1974 100% (12/12) (Hylidae, Anura), not given. No FPAs. Spores in the gallbladder. Koechlin’s Treefrog Hyla koechlini=Dendropsophus koechlini Duellman and Trueb, 1989, 92·8% (26/28) Type host: Knocking Sand Frog, Tomopterna kruger- (Hylidae, Anura); White leaf frog Hyla leucophyllata ensis Passmore and Carruthers, 1975 (Ranidae, group=Dendropsophus leucophyllatus Beireis, 1783 Anura), 20% (1/5) of adult frogs collected in 71·4% (10/14) (Hylidae, Anura); Dendropsophus February 1989. leali Bokermann, 1964 93·8% (15/16) (Hylidae, Type locality: Naboomspruit District,Transvaal, Anura); Wagner’s White-lipped Frog, Leptodactylus Mosdene Nature Reserve, South Africa. wagneri Peters, 1862 100% (3/3) (Leptodactylidae); Type material: SEM stub with spores and Pointedbelly Frog, Leptodactylus podicipinus Cope, record image deposited in the Queensland 1862 100% (1/1) (Leptodactylidae, Anura); Basin Museum, PO Box 3300, South Brisbane BC, White-lipped Frog, Leptodactylus mystaceus Spix, Queensland 4101, Australia under collection 1824 85·7% (6/7) (Leptodactylidae, Anura); number GL13041 (http://www.qm.qld.gov.au/ Leptodactylus bolivianus Boulenger, 1898 75·0% Collections/Collection+Online) (Delvinquier et al. (9/12) (Leptodactylidae, Anura); Perez’s Snouted 1992). Frog, Edalorhina perezi Jiménez de la Espada, 1870 Other hosts: Eastern Olive Toad Amietophrynus 50·0% (3/6) (Leiuperidae, Anura); Bolivian Bleating garmani=Bufo garmani Meek, 1897 (Bufonidae, Frog, Hamptophryne boliviana Parker, 1927 92·0% Anura); Natal Ghost Frog Heleophryne natalensis (23/25) (Microhylidae, Anura) (Gray, 1993). Hewitt, 1913 (Heleophrynidae, Anura); Common Remarks: The size of C. typhonius spore is similar to Sand Frog Tomopterna cryptotis Boulenger, 1907 the spores of C. immersus, both species are described (McAllister and Freed, 1996). from sympatric toads in South America. Compared Remarks: The only Cystodiscus species described to C. typhonius (entire spore is covered with from Africa. Delvinquier et al. (1992) distinguished depressions and ridges), the spore of C. immersus this species from all known gallbladder spores in has ridges only in 2/3 of the spores. While frogs by the absence of transverse spore surface C. typhonius is not recorded in North America, it ridges. The shape and low number of longitudinal shares similar spore morphology with C. serotinus. ridges distinguishes this species from Cystodiscus The spore of C. serotinus is larger. The spore suture of haldari Sarkar, 1982 which is devoid of transverse C. serotinus is S-curved compared to C. typhonius ridges. Named after Mr Les Minter. straight suture and the ridges on the valve of

60 Myxosporea in gallbladders of amphibians 495

C. typhonius are transverse and parallel, compared to Type locality: Coleambally, New South Wales, C. serotinus which is noted to have branched ridges Australia (−34°47′60″, 145°46′00″). and varied ridge orientation. Named after the type Type material: G.17758. Australian Museum, host. Sydney, New South Wales, Australia. Other hosts: Green and golden bell frog Litoria aurea Cystodiscus melleni (Jirků, Bolek, Whipps, Lesson, 1826 (Hylidae, Anura); Striped marsh frog Janovy, Kent and Modrý 2006) comb. n. Limnodynastes peronii Duméril and Bibron, 1841 (Limnodynastidae, Anura); Peron’s tree frog Litoria Syn. Myxidium melleni Jirků, Bolek, Whipps, peronii Tschudi, 1838 (Hylidae, Anura); Cane toad Janovy, Kent and Modrý 2006 Rhinella marina Linneaus, 1758 (Bufonidae, Anura). Myxospore morphology: “Mature spores, broadly ellipsoidal with slightly flattened ends in valvular Cystodiscus australis sp. n. ... view [ ] elongate and rhomboidal with rounded Syn. Myxidium immersum Delvinquier, 1986 pro edges in sutural view. Sutural ridge turns at angle of parte; Myxidium immersum Hartigan et al. (2011) pro ” approximately 90° near each pole (suture Z-shaped) parte (Jirků et al. 2006), 12·3 (12·0–13·5) ×7·6 (7·0–9·0) μm with 1 longitudinal depression encircling the valve Myxospore morphology: Ellipsoid to ovoid, straight or parallel to the sutural ridge, 2–5 transverse ridges (1–4 S-shaped suture, 16·0 (15·0–18·0) ×8·7 (8·0–10·0) μm transverse depressions) situated centrally. Polar cap- with 5–11 transverse ridges (5–12 depressions). No sules 5·2 (4·8–5·5) ×4·2 (3·8–4·5) μm with 6–7 polar FPA. filament coils. No FPAs. Spores in the gallbladder. Type host: Striped marsh frog Limnodynastes peronii Type host: Western Chorus Frog Pseudacris triseriata Duméril and Bibron, 1841 (Limnodynastidae, triseriata Wied-Neuwied, 1838 (Hylidae, Anura), Anura). 65% (20/31) frogs during March 2004. Type locality: Riverstone, New South Wales, − ′ ″ ′ ″ Type locality: Pawnee Lake near Lincoln, Lancaster Australia ( 33°41 7 , 150°50 53 ). County, Nebraska, USA (40° 51.18′ N, 96° 53.11′ W). Type material: G.17756. Australian Museum, Type material: Myxospores, SEM stub and slides Sydney, New South Wales, Australia. deposited in the Harold W. Manter Laboratory of Other hosts: Green and golden bell frog Litoria aurea Parasitology collection, University of Nebraska State Lesson, 1826 (Hylidae, Anura); Striped marsh frog Museum (http://manter.unl.edu/hwml/) under col- Limnodynastes peronii Duméril and Bibron, 1841 lection numbers 48167–48171 and phototypes as (Limnodynastidae, Anura); Cane toad Rhinella collection number 48172; SSU rDNA in GenBank: marina Linneaus, 1758 (Bufonidae, Anura). DQ003031. Other hosts: Recorded from 1/9 (11%) Blanchard’s ACKNOWLEDGEMENTS cricket frog, Acris crepitans blanchardi Conant, 1958 We thank Tomáš Scholz and Astrid Holzer (Institute (Hylidae, Anura) and 1/6 (16·7%) Upland chorus frog of Parasitology, Biology Centre of the ASCR, Czech Pseudacris feriarum feriarum Hedges, 1986 (Hylidae, Republic) for their support. We acknowledge the Anura) (Jirků et al. 2006; McAllister et al. 2008). Australian Key Centre for Microscopy and Microanalysis at The University of Sydney and the Electron Microscopy Remarks: This species is distinct from all Cystodiscus unit at the Institute of Parasitology, Biology Centre of the spp. based on its size, centrally positioned transverse ASCR, Czech Republic. A.H. was the recipient of the J.D. ridges and depressions and sutural ridge shape. In Smyth Travel Award from the Australian Society for particular the sympatric but larger C. serotinus whose Parasitology and the ARC/NHMRC Research Network for spore is entirely covered with transverse ridges. See Parasitology. We thank Elaine Chew for expert histopathology processing, Glenn Shea, Lance Jurd, David C. serotinus and C. immersus remarks regarding Newell for their insight and two anonymous reviewers for genetic identification. constructive suggestions. This study was supported by the Australian Academy of Science, Dr William Richards Cystodiscus axonis sp. n. Award in Veterinary Pathology and in part by the Grant Agency of the Academy of Sciences of the Czech Republic Syn. Myxidium immersum Delvinquier, 1986 pro (KJB600960701) and the Grant Agency of the Czech parte; Myxidium immersum Hartigan et al. (2011) pro Republic (204/09/P519). The Australian Registry of parte Wildlife Health acknowledges the Australian Biosecurity Intelligence Network (ABIN) for information management Myxospore morphology: Ellipsoid to ovoid, straight or and communication systems that contributed to this S-shaped suture, 14·5 (13·0–15·0) ×8·7 (8·0–10·0) μm research. ABIN is a project of the National Collaborative Research Infrastructure Strategy (NCRIS). with 4–8 centrally located transverse ridges (5–9 depressions). FPA present. REFERENCES

Type host: Southern bell frog Litoria raniformis Azevedo, C. and Matos, E. (1996). Light and electron microscopic study Keferstein, 1867 (Hylidae, Anura). of a myxosporean, Tetrauronema desaequalis sp. n. (Family

61 Ashlie Hartigan and others 496

Tetrauronematidae), from an Amazonian fish. Journal of Parasitology 82, Kudo, R. R. (1943). Further observations on the protozoan, Myxidium 288–291. doi: 10.2307/3284162. serotinum, inhabiting the gall bladder of North American Salientia. Journal Bartošová, P., Fiala, I. and Hypša, V. (2009). Concatenated SSU and of Morphology 72, 263–277. LSU rDNA data confirm the main evolutionary trends within myxospor- Kudo, R. R. and Sprague, V. (1940). On Myxidium immersum Lutz and M. eans (Myxozoa: Myxosporea) and provide an effective tool for their serotinum n. sp., two Myxosporidian parasites of Salientia of South and molecular phylogenetics. Molecular Phylogenetics and Evolution 53,81–93. North America. Revista de Medicina Tropical Habana 6,65–73. doi: 10.1016/j.ympev.2009.05.018. Lom, J. (1987). Myxosporea a new look at long known parasites of fish. Carini, A. (1932). Myxidium lindoyense n. sp., parasita da vesicula biliar de Parasitology Today 3, 327–332. doi: 10.1016/0169-4758(87)90115-3. batrachios do Brasil. Revista de Biologia e Hygiene Sao Paolo 3,83–84. Lom, J. and Arthur, J. R. (1989). A guideline for the preparation Castresana, J. (2000). Selection of conserved blocks from multiple of species descriptions in Myxosporea. Journal of Fish Diseases 12, 151– alignments for their use in phylogenetic analysis. Molecular Biology and 156. Evolution 17, 540–552. Lom, J. and Dyková, I. (1992). Protozoan Parasites of Fishes, Elsevier, Clark, J. G. (1982). Studies on the biology of Myxidium serotinum Kudo and New York, USA. Sprague, 1940 (Myxozoa), a parasite of amphibian invertebrates, Ph.D. Lom, J. and Dyková, I. (2006). Myxozoan genera: definition and notes thesis. Department of Biological Sciences, College of Arts and Sciences, on taxonomy, life-cycle terminology and pathogenic species. Folia University of Cincinnati, Cincinnati, USA. Parasitologica 53,1–36. Clark, J. G. and Shoemaker, J. (1973). Eurycea bislineata (Green) Two- Lom, J., Feist, S. W., Dyková, I. and Kepr, T. (1989). Brain myxoboliasis lined salamander, a new host of Myxidium serotinum Kudo and Sprague of bullhead, Cottus gobio L., due to Myxobolus jiroveci sp. nov.: light and 1940 (Myxosporidea, Myxidiidae). Journal of Protozoology 20, 365–366. doi: electronmicroscope observations. Journal of Fish Diseases 12,15–27. doi: 10.1111/j.1550-7408.1973.tb00902.x. 10.1111/j.1365-2761.1989.tb01235.x. Cordero, E. H. (1919). Nota sobre ‘Opalina antilliensis’ Metcalf, ciliado Lom, J., Pavlásková, M. and Dyková, I. (1985). Notes on kidney- parasito de los batracios del Uruguay. Physis Seccion A los Oceanos y sus infecting species of the genus Sphaerospora Thélohan (Myxosporea), Organismos 4, 531–535. including a new species of Sphaerospora gobionis sp. nov., and on de Leon, G. P. P. and Nadler, S. A. (2010). What we don’t recognize can Myxosporean life cycle stages in the blood of some freshwater fish. hurt us: a plea for awareness about cryptic species. Journal of Parasitology 96, Journal of Fish Diseases 8, 221–232. doi: 10.1111/j.1365-2761.1985. 453–464. doi: 10.1645/Ge-2260.1. tb01217.x. Delvinquier, B. L. J. (1986). Myxidium immersum (Protozoa, Myxosporea) Lühe, M. (1899). Cystodiscus immersus Lutz. Verhandlungen der Deutschen of the cane toad, Bufo marinus, in Australian anura, with a synopsis of the Zoologischen Gesellschaft 9, 291–293. genus in amphibians. Australian Journal of Zoology 34, 843–853. doi: Lutz, A. (1889). Üeber ein Myxosporidium aus der Gallenblase brasilia- 10.1071/ZO9860843. nischer Batrachier. Centralblatt für Bakteriologie und Parasitenkunde 5, Delvinquier, B. L. J., Markus, M. B. and Passmore, N. I. (1992). 84–88. Myxidium lesminteri sp. n. (Myxosporea, Myxidiidae) from the gallbladder Mann, R. M., Hyne, R. V. and Choung, C. B. (2010). Hormonal in- of 3 Southern African anura. Systematic Parasitology 23,25–30. doi: duction of spermiation, courting behavior and spawning in the Southern 10.1007/BF00008005. bell frog, Litoria raniformis. Zoo Biology 29, 774–782. doi: 10.1002/ Faivovich, J., Haddad, C. F. B., Garcia, P. C. A., Frost, D. R., zoo.20331. Campbell, J. A. and Wheeler, W. C. (2005). Systematic review of the Marquet, E. and Sobel, H. (1970). A protozoon in central nervous system frog family hylidae, with special reference to hylinae: phylogenetic analysis of lungfish Polypterus enlicheri. Journal of Protozoology 17, 71. doi: 10.1111/ and taxonomic revision. Bulletin of the American Museum of Natural j.1550-7408.1970.tb05160.x. History 294,1–240. doi: 10.1206/0003-0090(2005)294[0001:SROTFF]2.0. McAllister, C. T. and Bursey, C. R. (2005). Endoparasites of the eastern CO;2. narrowmouth toad, Gastrophryne carolinensis (Anura: Microhylidae), from Fiala, I. (2006). The phylogeny of Myxosporea (Myxozoa) based on small northwestern Louisiana and northeastern Texas, USA. Comparative subunit ribosomal RNA gene analysis. International Journal for Parasitology Parasitology 72, 124–128. doi: 10.1654/4149. 36, 1521–1534. doi: 10.1016/j.ijpara.2006.06.016. McAllister, C. T., Bursey, C. R. and Trauth, S. E. (2008). New host and Fiala, I. and Bartošová, P. (2010). History of myxozoan character geographic distribution records for some endoparasites (Myxosporea, evolution on the basis of rDNA and EF-2 data. Bmc Evolutionary Biology Trematoda, Cestoidea, Nematoda) of amphibians and reptiles from 10. doi: 22810.1186/1471-2148-10-228. Arkansas and Texas, USA. Comparative Parasitology 75, 241–254. doi: Frost, D. R. (2011). Amphibian Species of the World: an Online Reference. 10.1654/4329.1. Version 5.5 (31 January, 2011). Electronic Database accessible at http:// McAllister, C. T., Crawford, J. A. and Kuhns, A. R. (2009). Myxidium research.amnh.org/vz/herpetology/amphibia/, American Museum of serotinum (Protista: Myxozoa) from a Jefferson salamander Ambystoma Natural History, New York, USA. jeffersonianum, in Illinois. Texas Journal of Science 61,61–64. Gray, M. E. (1993). A new species of Myxidium (Myxozoa: Myxidiidae) McAllister, C. T. and Freed, P. S. (1996). A new host and geographic from the gall bladders of anuran amphibians from Peru. Transactions of the record for Myxidium lesminteri (Protozoa: Myxosporea) from Tomopterna Kansas Academy of Science 96, 152–157. doi: 10.2307/3628327. cryptotis (Amphibia: Ranidae), in Namibia, South-West Africa. Journal of Griffin, M. J., Wise, D. J. and Pote, L. M. (2009). Morphology and small- the Helminthological Society of Washington 63, 123–124. subunit ribosomal DNA sequence of Henneguya adiposa (Myxosporea) from McAllister, C. T. and Trauth, S. E. (1995). New host records for Ictalurus punctatus (Siluriformes). Journal of Parasitology 95, 1076–1085. Myxidium serotinum (Protozoa, Myxosporea) from North American doi: 10.1645/ge-2012.1. amphibians. Journal of Parasitology 81, 485–488. Hartigan, A., Fiala, I., Dyková, I., Jirků, M., Okimoto, B., Rose, K., Sarkar, N. K. (1982). Myxosporidan Myxidium haldari sp. n. (Myxozoa, Phalen, D. N. and Šlapeta, J. (2011). A suspected parasite spill-back of Myxidiidae) from Indian tree frog Hyla arborea. Acta Protozoologica 21, two novel Myxidium spp. (Myxosporea) causing disease in Australian 197–199. endemic frogs found in the invasive Cane Toad. PloS ONE 6, e18871. doi: Schlick-Steiner, B. C., Seifert, B., Stauffer, C., Christian, E., 10.1371/journal.pone.0018871. Crozier, R. H. and Steiner, F. M. (2007). Without morphology, cryptic Hartigan, A., Phalen, D. N. and Šlapeta, J. (2010). Museum material species stay in taxonomic crypsis following discovery. Trends in Ecology & reveals a frog parasite emergence after the invasion of the cane toad in Evolution 22, 391–392. doi: 10.1016/j.tree.2007.05.004. Australia. Parasites & Vectors 3. doi: 10.1186/1756-3305-3-50. Stamatakis, A. (2006). RAxML-VI-HPC: Maximum likelihood- Jirků, M., Bolek, M. G., Whipps, C. M., Janovy, J., Kent, M. L. and based phylogenetic analyses with thousands of taxa and mixed models. Modrý, D. (2006). A new species of Myxidium (Myxosporea: Myxidiidae), Bioinformatics 22, 2688–2690. doi: 10.1093/bioinformatics/btl446. from the western chorus frog, Pseudacris triseriata triseriata, and Stensaas, L. J., Stensaas, S. S. and Sotelo, J. R. (1967). An intra-axonal Blanchard’s cricket frog, Acris crepitans blanchardi (Hylidae), from protozoon in spinal cord of toad Bufo arenarum (Hensel). Journal of Eastern Nebraska: Morphology, phylogeny, and critical comments on Protozoology 14, 585. doi: 10.1111/j.1550-7408.1967.tb02045.x. amphibian Myxidium taxonomy. Journal of Parasitology 92, 611–619. doi: Swofford, D. L., Waddell, P. J., Huelsenbeck, J. P., Foster, P. G., 10.1645/GE-728R.1. Lewis, P. O. and Rogers, J. S. (2001). Bias in phylogenetic estimation Katoh, K., Kuma, K., Toh, H. and Miyata, T. (2005). MAFFT version 5: and its relevance to the choice between parsimony and likelihood improvement in accuracy of multiple sequence alignment. Nucleic Acids methods. Systematic Biology 50, 525–539. doi: 10.1080/ Research 33, 511–518. doi: 10.1093/nar/gki198. 106351501750435086.

62 Supporting Table 4.1: Myxosporean species DNA used for specificity test of Cystodiscus spp. with species specific multiplex PCR

Myxosporea species Host species PCR* Myxidium sp. Haemulon sciurus Neg. Myxidium chelonarum Kachuga smithi Neg. Myxidium cuneiforme Cyprinus carpio Neg. Myxidium paresiluri Pseudogabrus fulvidraco Neg. Sphaeromyxa hellandi Melanogrammus aeglefinus Neg. Sphaeromyxa longa Trisopterus minutus Neg. Myxidium lierberkuehni Esox lucius Neg. Chloromyxum leydigi Torpedo marmorata Neg. Myxobolus lentisuturalis Cyprinus carpio Neg. Myxobolus aeglefini Melanogrammus aeglefinus Neg. Henneguya salmincola Salmo salar Neg. Henneguya zschokkei Salmo salar Neg. Myxobolus sp. Aristichthys nobilis Neg. Myxobolus sp. Carassius gibelio Neg. Chloromyxum cyprini Hypophthalmichthys molitrix Neg. Chloromyxum fluviatile Hypophthalmichthys molitrix Neg. Chloromyxum leydigi Torpedo marmorata Neg. Sphaerospora ranae Rana temporaria Neg. Sphaerospora truttae Salmo salar Neg. Hoferellus anurae Hyperolius kivuensis Neg. Enteromyxum leei Diplodus puntazzo Neg. Kudoa dianae Sphoeroides annulatus Neg. Kudoa crumena Thunnus maccoyii Neg. Sphaerospora sp. Mugil curema Neg. Sphaerospora dicentrarchi Dicentrarchaus labrax Neg. Ceratomyxa appendiculata Lophius piscatorius Neg. Ceratomyxa longipes Melanogrammus aeglefinus Neg. Ceratomyxa cottoidii Clinus cottoides Neg. Palliatus indecorus Alepocephalus grandis Neg. Myxidium gadi Pollachius virens Neg. Myxidium incurvatum Callionymus lyra Neg. Myxidium bergense Heliocolenus dactylopterus Neg. Auerbachia pulchra Coryphaenoides rupestris Neg. Sinuolinea lophii Lophius piscatorius Neg. Parvicapsula sp. Oncorhynchus sp. Neg. Sphaerospora testicularis Dicentrarchus labrax Neg. Ceratomyxa shasta Lophius piscatorius Neg. Soricimyxum fegati Sorex araneus Neg. Tetracapsuloides Oncorhynchus mykiss Neg. bryosalmonae

Note: * All PCR (25 μl) reactions were run with 1 U of Taq-Purple polymerase and a 5x PCR buffer (Top Bio, Czech Republic) with 250 µM of each dNTP (Invitek, Germany). Primer sets (see Material and Methods) were used in combination as species specific multiplex PCR (each 10 pmol concentration). Positive and negative controls were run with each reaction. The PCR cycling condition were initiated with 95°C for 5 mins, then 35 cycles (95°C for 0.5 min, 55°C for 1 min, 72°C for 1 min) and concluded with 72°C for 7 mins. All tested DNA were PCR negative (Neg.). Results were identical with annealing temperature set to 45°C.

Supporting Figure 4.1: Alignment of 18S, ITS1, 5.8S, ITS2 and 28S regions from Cystodiscus axonis and Cystodiscus australis annotated with diagnostic primers used in multiplex species specific PCR

Chapter 5: Comparative pathology and ecological implications of two Myxosporean parasites in native Australian frogs and the invasive Cane toad

65 Comparative Pathology and Ecological Implications of Two Myxosporean Parasites in Native Australian Frogs and the Invasive Cane Toad

Ashlie Hartigan1, Navneet K. Dhand1, Karrie Rose2, Jan Sˇ lapeta1, David N. Phalen1* 1 Faculty of Veterinary Science, University of Sydney, New South Wales, Australia, 2 Australian Registry of Wildlife Health, Taronga Conservation Society Australia, Mosman, New South Wales, Australia

Abstract Myxosporean parasites Cystodiscus axonis and C. australis are pathogens of native and exotic Australian frog species. The pathology and ecological outcomes of infection with these parasites were investigated in this study. Gliosis was correlated to Cystodiscus axonis plasmodia in the brains of (9/60) tadpoles and (3/9) adult endangered Green and golden bell frogs using ordinal regression. Severe host reactions to C. axonis (haemorrhage, necrosis, and vasulitis) were observed in the brains of threatened Southern bell frogs (8/8), critically endangered Booroolong frogs (15/44) and Yellow spotted bell frogs (3/3). Severe brain lesions were associated with behavioural changes, neurological dysfunction, and spontaneous death. Both C. axonis and C. australis develop in the bile ducts of tadpoles, the plasmodia were significantly associated with biliary hyperplasia, inflammation and the loss of hepatocytes in (34/72) Green and golden bell frog tadpoles using ordinal regression. These lesions were so severe that in some cases 70% of the total liver was diseased. Normal liver function in tadpoles is necessary for metamorphosis, metabolism, and immune function. We postulate that this extensive liver damage would have significant host health impacts. Severe hepatic myxosporidiosis was more prevalent in tadpoles examined in autumn and winter (overwintered), suggestive of delayed metamorphosis in infected tadpoles, which would have serious flow-on effects in small populations. We compared the sensitivity of histopathology and species-specific PCR in the detection of C. australis and C. axonis. PCR was determined to be the most sensitive method (detection limit 1 myxospore equivalent of ribosomal DNA). Histology, however, had the advantage of assessing the impact of the parasite on the host. It was concluded that these parasites have the potential for significant ecological impacts, because of their high prevalence of infection and their ability to cause disease in some frogs.

Citation: Hartigan A, Dhand NK, Rose K, Sˇlapeta J, Phalen DN (2012) Comparative Pathology and Ecological Implications of Two Myxosporean Parasites in Native Australian Frogs and the Invasive Cane Toad. PLoS ONE 7(10): e43780. doi:10.1371/journal.pone.0043780 Editor: Matt Hayward, Australian Wildlife Conservancy, Australia Received April 1, 2012; Accepted July 25, 2012; Published October 3, 2012 Copyright: ß 2012 Hartigan et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by the Australian Academy of Science (2009 Margaret Middleton Fund for the Conservation of Endangered Australian vertebrate species award) and The University of Sydney (WP Richards Bequest for Veterinary Pathology). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction the same salmonid range as M. cerebralis and yet causes no host reaction in wild fish [17,18]. Myxosporea are two host metazoan parasites, closely related to Two myxosporean species, Cystodiscus axonis and Cystodiscus jellyfish (Cnidarian) [1–3] with over 2, 100 species of myxosporean australis, have recently been characterized in Australian frogs parasites identified to date [4]. Intermediate hosts are primarily [19,20]. Both Cystodiscus axonis and C. australis have presporogonic aquatic vertebrates including fish, reptiles, and amphibians. stages in bile ducts that are followed by the development of spore Recently however, myxosporea have been identified in shrews producing plasmodia in the gallbladder. Cystodiscus axonis is unique (Sorex araneus) [5,6], a mole (Talpa europaea) [7] and several species of in that it has a presporogonic stage that develops in the central waterfowl [8]. Myxosporean parasites can be either species specific nervous system. Lesions in the central nervous system often or can infect a range of closely related species. Known definitive accompany infection with C. axonis and liver lesions are found with hosts for Myxosporea are annelids (oligochaetes and polychaetes), infection by both of these parasites in some species of frogs, but not however only 34 life cycles have been identified so far [9]. others. The origin of these parasites is not known, but both appear Myxosporea can develop in different tissues of their intermediate to have emerged in Australian frogs in the mid 1960’s [21]. One or hosts, including gills, muscle, intestine, urinary tract, gonad, both of these parasites have now been identified in 7 Australian gallbladder, liver, bone and the central nervous system [9–11]. frogs species (Litoria aurea, Litoria peronii, Litoria raniformis, Litoria The impact of infection on the intermediate host can range from castanea, Litoria booroolongensis, Limnodynastes peronii, Rhinella marina), minimal to severe. Myxobolus cerebralis, a parasite of salmonid fish, representing 3 (Hylidae, Myobatrachidae, Bufonidae) out of 5 frog has caused morbidity and mortality in captive raised and wild fish families in Australia [19,20,22]. Importantly, infections and wherever the parasite has been introduced to susceptible species associated lesions have been found in 4 endangered species (L. [12–16]. Chloromyxum truttae, in contrast, is a common parasite in aurea, L. raniformis, L. booroolongensis and L. castanea). The populations

PLOS ONE | www.plosone.org 1 October 2012 | Volume 7 | Issue 10 | e4378066 Pathology of Cystodiscus spp. in Australian Frogs of each endangered frog in which these parasites have been Cane toad (Rhinella marina) tadpoles (n = 14) and adults (n = 31) identified began declining from the 1970’s and 80’s [23–27] which were collected from areas around Lismore, NSW (228.813, is approximately the time that these Cystodiscus species first 153.279) in February 2009 and 2010, at the southern end of their emerged [21]. distribution in eastern Australia. Infectious diseases are known to be important drivers of population declines and extinctions in amphibians and other Tissue processing species of animals. Disease causing organisms however, are often Tadpoles were bisected along the mid line. One half was fixed found in healthy populations of animals and, proving that in 10% buffered formalin, paraffin embedded, and 5 mm sagittal organisms are causing significant disease and are having delete- sections of entire tadpoles were stained with haematoxylin and rious impacts on a population may be difficult [28]. This process is eosin. Brain, liver, spinal cord, kidney, spleen, and reproductive especially difficult when organisms cannot be cultured and organs from adult frogs were individually fixed, and similarly controlled infection trials cannot be undertaken. sectioned and stained. A section through the cervical, thoracic and It is the purpose of this study to document and statistically lumbar spinal cord was examined for each adult frog. evaluate C. axonis and C. australis as agents of significant disease and Liver sections from adult frogs (n = 3) were stained with Congo their potential to impact the survivability of both tadpoles and red to determine the presence of amyloidosis. For PCR analysis, adult frogs. We compared the sensitivity and specificity of different brain, brainstem or liver were separately removed from the techniques (histology and PCR) that can be used for diagnosis in unfixed half of the tadpoles and adult frogs, or unfixed half of the individual animals and for population surveillance. Lastly, we brain stem and a representative portion of the liver of adult frogs. provide evidence that differing host responses may play a role in Individual tissues were stored at 280uC until processing. the outcome of infection in these two parasites and show how the Many of the tadpoles were very small. In these cases, it was not host-pathogen-environment interactions may have important always possible to get tissue for both histopathology and the PCR ecological implications. assay and therefore, tissues were only examined by histopathology.

Methods Evaluation of disease Ethics Statement Behavioural observations of captive frogs. Southern bell frogs, Booroolong frogs and Yellow spotted bell frogs were Green and golden bell frogs, Striped marsh frogs, Peron’s tree maintained at Taronga Zoo, Sydney, Australia. All animals were frogs and Cane toads were collected under Scientific license subject to zoo staff observations; behavioural abnormalities were numbers 12686 and 12969 (Office of Environment and Heritage, noted in individual pathology and diagnostic reports within the Australia). Animals were euthanized with AQUI-S (AQUI-S, New Australian Registry of Wildlife Health. Zealand) [29] according to University of Sydney Animal Ethics Assessment of lesions. A summary of the species and tissues Licenses (N00/9-2008/3/4855 and N00/9-2009/3/5134). Dead examined to build a syndrome description is provided within Yellow spotted bell frogs, Southern bell frogs and Booroolong frogs Table 1. were examined through the opportunistic sample collection Association of C. axonis infection with microscopic program under the auspices of the Taronga Conservation Society lesions of the central nervous system of the Green and Australia Animal Ethics Committee. golden bell frog. Gliosis was defined as increased numbers of glial cells occurring focally or more diffusely within a section of the Specimens brain or spinal cord. Cystodiscus axonis infection and the presence of Green and golden bell frog (Litoria aurea) tadpoles (n = 97), gliosis and other lesions in sagittal sections of the brain and spinal metamorphs (n = 1) and adults (n = 11); Striped marsh frog cord was recorded from 9 Green and golden bell frog tadpoles. (Limnodynastes peronii) tadpoles (n = 61) and adults (n = 27) and Sagittal sections through the brain and 3 transverse sections Peron’s tree frog (Litoria peronii) tadpoles (n = 18), metamorphs through the spinal cord were also examined in 6 Green and golden (n = 1) and adults (n = 7) were collected from a semi-captive bell frog adults. When an infected tadpole or frog was found, the breeding colony for Green and golden bell frogs in Riverstone, severity of infection was assessed using the area of the central NSW (233.679, 150.860) from May 2008 to December 2010 nervous system where the parasites were in highest concentration. (Table 1). Green and golden bell frogs were bred at this facility in Once this area was identified, parasite load was graded into three multiple man-made pools. Adult Green and golden bell frogs, classes depending on the number of organisms that could be seen Striped marsh frogs and Peron’s tree frogs spawned and moved on 10X magnification: mild (1–5 plasmodia), moderate (6–10 freely in and out of the pools. Both Cystodiscus axonis and C. australis plasmodia) and severe (more than 10 plasmodia). The entire brain have previously been shown to be enzootic to this breeding facility and spinal cord sections were examined for evidence of host [19,20]. response to the parasite. Gliosis was recorded as either present or Booroolong (Litoria booroolongensis) frogs (n = 44) were collected absent and when applicable, lesions such as perivascular lymphoid from two locations, Maragle Creek (n = 25) and Abercrombie cell cuffing, malacia or haemorrhage were recorded. (n = 19) in NSW, Australia between 2007 and 2010. These animals Association of infection with Cystodiscus species and were kept at the Taronga Zoo, Sydney, Australia for variable microscopic lesions in the liver of Green and golden bell lengths of time and were either found dead or were euthanased as frog. To examine the possible relationship between Cystodiscus part of a disease screening program or due to some abnormal species infection and microscopic liver disease, 72 tadpoles, 1 behaviour (e.g. leg dragging). Eight Southern bell frogs (Litoria metamorph (here after included in the analysis for adults), and 11 raniformis) were wild caught in southern NSW, Australia in 2006 adult Green and golden bell frogs were examined. Each liver and kept in captivity until 2009 when they were euthanased due to section was examined for the presence or absence of plasmodial lethargy and lack of breeding. Fourteen Yellow spotted bell frogs stages within the bile ducts, biliary hyperplasia, lymphoplasma- (Litoria castanea) were wild caught from southern NSW in 2010 and cytic inflammation, loss of hepatocytes and biliary fibrosis. Bile held in a captive breeding colony. Three frogs (1 sub-adult and 2 duct plasmodia were recorded as either present or absent. The adults) died spontaneously and were submitted for necropsy. severity of the liver lesions were classified as mild (10–30% of liver

Range in Range in hepatitic Gallbladder Total numberBrains parasite load Livers Liver myxosporidiosis myxospores Fibrosis (F)/Amyloid-like Species Age examined examined Brain positive (brain) examined positive severity positive tissue in liver (A)

Green and golden bell frog Tadpoles 97 60 9 (15.0%) None = 51 72 34 None = 15 0 0 (Litoria aurea) Mild = 6 Mild = 16 Moderate = 3 Moderate = 28 Severe = 13 Adults 12 9 3 (33.3%) None = 6 12 1* None = 12 1 1 – F + A Mild = 1 Moderate = 2 Southern bell frog Adults 8 8 8 (100%) Mild = 2 8 0 None = 7 7 8 – F + A (Litoria raniformis) Moderate = 3 Severe = 2 Booroolong frog Adults 44 44 15 (34.1%) Mild = 9 38 3 None = 38 n.a. 3 – F + A (Litoria booroolongensis) Moderate = 6 Yellow spotted bell frog Adults 3 3 3 (100%) Severe = 3 3 1* None = 2 n.a. 2 – F + A (Litoria castanea) Mild = 1 Peron’s tree frog Tadpoles 18 11 7 (63.6%) Mild = 7 14 12 None = 1 0 0 (Litoria peronii) Mild = 4 Moderate = 7 Severe = 2 Adults 8 6 3 (50.0%) Moderate = 2 8 1* 5 4 – F + A, 1 – F only Severe = 1 Striped marsh frog Tadpoles 61 42 0 None = 42 38 10 None = 29 0 0 ahlg of Pathology (Limnodynastes peronii) Mild = 9 Adults 27 17 2 (11.7%) Moderate = 1 25 0 None = 24 20 12 – F only Severe = 1 Mild = 1 Cane toad Tadpoles 14 12 0 None = 12 14 0 None = 14 0 0 ytdsu spp. Cystodiscus (Rhinella marina) Adults 27 26 2 (7.7%) None = 24 27 0 None = 27 9 6 – F only Mild = 1 Severe = 1

doi:10.1371/journal.pone.0043780.t001 nAsrla Frogs Australian in 68 Pathology of Cystodiscus spp. in Australian Frogs

Table 2. Seasonal prevalence of Cystodiscus axonis plasmodia, Cystodiscus spp. liver plasmodia and hepatic myxosporidiosis in Green and golden bell tadpoles.

Brain plasmodia Spring Summer Autumn Winter n

Plasmodia absent 6 (75.0%) 14 (82.3%) 7 (63.6%) 24 (100%) 51 Plasmodia present 2 (25.0%) 3 (17.6%) 4 (36.4%) 0 (0%) 9 Total 8 17 11 24 60

Liver plasmodia Spring Summer Autumn Winter n

Plasmodia absent 7 (77.8%) 8 (57.1%) 7 (35.0%) 16 (55.1%) 38 Plasmodia present 2 (22.2%) 6 (42.9%) 13 (65.0%) 13 (44.8%) 34 Total 9 14 20 29 72

Severity of hepatic myxosporidiosis Spring Summer Autumn Winter n

None 5 (55.6%) 6 (42.9%) 3 (15.0%) 3 (10.3%) 17 Mild 3 (33.3%) 4 (28.6%) 9 (45.0%) 10 (34.5%) 26 Moderate 1 (11.1%) 2 (14.3%) 4 (20.0%) 10 (34.5%) 17 Severe 0 2 (14.3%) 4 (20.0%) 6 (20.7%) 12 Total 9 14 20 29 72

doi:10.1371/journal.pone.0043780.t002 affected), moderate (30–60% of the liver affected) and severe (60% used for all analyses. Odds ratios are reported with 95% or more of the liver affected) when a representative section of the confidence intervals (CI) and 2-sided p-values. liver was viewed under 10X magnification. It has been previously Prevalence data for Cystodiscus plasmodia in Green and golden shown that C. axonis and C. australis are morphologically bell frog tadpole livers and C. axonis in brains was compared to indistinguishable in liver sections and both infections can occur other species of tadpoles (Striped marsh and Peron’s tree frog simultaneously [20]. In this study, it was not possible to determine tadpoles) that were collected from the same location (Riverstone, which parasites were present in the liver. Thus the observed lesions NSW) in the same season. The overall differences in prevalence could be associated with C. axonis, C. australis or a combined were evaluated with x2 tests or Fisher’s exact test if the cell infection. frequencies were low. If the overall differences were significant, Statistical association of microscopic lesions with pair-wise comparisons were made using Z-test for comparing Cystodiscus species infection in Green and golden bell proportions or Fisher’s exact test, as appropriate. tadpoles and frogs. Associations were initially investigated by creating contingency tables of the presence of infection with Assessment of the impact of season on the presence and presence of lesions and lesion severity. Univariable binary logistic severity of lesions in the Green and golden bell frog regression analyses were then conducted to determine association between infection, age and season and the binary outcome tadpoles and adult frogs variables – presence or absence of lesions in the brain (Supple- Three histological parameters were mapped in Green and mentary Table 1A) and liver (Table S1D) for the tadpoles and golden bell frogs across seasons between May 2008 and December adult Green and golden bell frog. Similar binary logistic regression 2010. Seasons were defined as autumn (March, April, May), models were constructed to evaluate the association of age and winter (June, July, August), spring (September, October, Novem- season with presence or absence of infection in brain and liver. For ber) and summer (December, January, February). The presence or ordinal outcome variables (e.g. severity, graded with none, mild, absence of Cystodiscus spp. infection of the liver and the degree of moderate and severe) ordinal logistic regression analyses were severity of associated lesions in the liver seasonally were examined conducted to determine their association with infection, age and in tadpoles only (n = 72). Similarly, the presence or absence of season. Assumption of proportional odds for ordinal models was Cystodiscus axonis infection in the brain was examined in tadpoles evaluated using Score test. Multivariable logistic regression only (n = 60) (Table 1). analyses were then conducted (binary or ordinal, as appropriate) Due to the small amount of tissue available in a tadpole, not all to evaluate the association of infection, age and season after sagittal sections contained both brain and liver. Therefore the adjusting for each other. First order interaction terms were finally numbers of animals that were included in the liver and brain tested by adding to the multivariable model and retained if seasonality study were not always the same (Table 1 and 2). significant. Fisher’s exact test was used to evaluate associations Association between the lesion of interest and season was analysed where logistic regression model did not converge because some of with ordinal logistic regression and when necessary (brain the cell frequencies were zero. plasmodia) Fisher’s exact test in SAS (Table S1A and D). SAS statistical software (release 9.3, 2002–10, SAS Institute Inc., Cary, NC, USA) and UniLogistic macro (Dhand, 2010) were

PLOS ONE | www.plosone.org 4 October 2012 | Volume 7 | Issue 10 | e4378069 Pathology of Cystodiscus spp. in Australian Frogs

Comparison of histology and multiplex PCR to detect gliosis was often in close proximity to plasmodia (Fig. 1), but in Cystodiscus species infection in the brain and liver some circumstances gliosis could be seen in areas of the nervous A minimum of twelve serial sections were obtained for a sub- system where there were no plasmodia. Gliosis was observed in the sample of the frog specimens (not all samples qualified due to small brain (Fig. 1E, F), spinal cord (Fig. 1C), cranial nerves, spinal tissue size available). Depending on the size of the tissue, the nerves and dorsal root ganglia (Fig. 1A–C). Gliosis was often the number of 5 mm sections ranged from 12 to 34 in liver (n = 17) and most severe in nerve roots and the dorsal root ganglia. A mild to 12 to 48 in brain (n = 15). Each section was screened individually moderate degree of malacia occurred concurrently in the brain, and presence of Cystodiscus spp. was recorded. The liver specimens spinal cord and ganglia (Fig. 1A, B) of some specimens, but it was were all from tadpoles of the Green and golden bell frogs. The rare. Haemorrhage was not seen in the Green and golden bell frog brain specimens were from the Green and golden bell frogs specimens. (tadpoles, n = 5; adults, n = 1), Peron’s tree frog (adult, n = 1) and Gliosis is a nonspecific reaction and can be caused by multiple Striped marsh frog (tadpoles, n = 5; adults, n = 4). infectious agents [30], it was significantly associated with the For a subset of specimens for which a tissue sample was presence of C. axonis plasmodia (p,0.001, odds ratio = 42.5) available in 220uCor280uC, we isolated DNA and used a (Table S1B). Gliosis was found in 78% of the (7/9) of infected multiplex species specific PCR according to a previously published tadpoles and 100% of the (3/3) of infected adults. Gliosis was protocol [20]. The liver specimens were from Green and golden either seen as a diffuse and extensive (40%, 4/10) or multifocal bell frogs (tadpoles, n = 11) and Striped marsh frog (tadpoles, and locally extensive (50%, 5/10) with the exception of one animal n = 8). The brain specimens were from the Green and golden bell with a single focus of gliosis (10%, 1/10). Glial activity was found frogs (tadpoles, n = 2), Peron’s tree frog (adult, n = 1) and Striped in brains of 10% (6/60) tadpoles, for which plasmodia were not marsh frog (tadpoles, n = 4; adults, n = 4). identified. Moderate parasite loads (6–10 cysts per view) were The sensitivity of single histological section to demonstrate significantly associated with the presence of gliosis (43%, 5/12, presence of the parasites in tissue compared with serial section and p#0.001, Table S1B). PCR and vice versa was calculated using a standard equation (sensitivity of a test = # of true positives/# of true positives + # Absence of Cystodiscus axonis plasmodia in brains of of false negatives). The true positives were considered those overwintered Green and golden bell frog tadpoles returning positive outcomes in histology or PCR. Cystodiscus axonis plasmodia were found in the nervous system of To determine the detection limit of the multiplex species specific Green and golden bell frog tadpoles in the spring, summer and PCR [20] we tested samples with a known number of C. axonis and autumn (Table 2A). The prevalence of C. axonis infection in the C. australis myxospores (in PBS counted by haemocytometer). The nervous system was lowest in the summer (18%, 3/17), then spring DNA isolation and PCR reactions were performed according to (25%, 2/8) and was significantly higher (36%, 4/11) in the autumn Hartigan et al. [19], all DNA samples were eluted in 50 ml buffer. (Table S1A, p = 0.003). No tadpoles with brain stages of C. axonis Extracted DNA was diluted 10-fold to the equivalent approxi- were identified in winter (n = 25) (Table 2A). mately ,0.01 myxospore per ml (calculated myxospores in extraction/eluted volume). PCR reactions (25 ml) contained 1 ml Severe host reactions and increased prevalence of of template DNA and were run in triplicate, with positive and negative control in each PCR batch. Results were visualised on a Cystodiscus axonis in threatened frog species compared 2% agarose gel stained with GelRed (Biotium, Australia). to common frog species Tadpoles. Overall chi-sq analysis for brain plasmodia pres- Results ence in tadpoles was found to be significant (p = ,0.001). Prevalence of infection in the Peron’s tree frog tadpoles (64%, Distribution of Cystodiscus axonis in nervous system and 7/11) was significantly higher than that observed in the Green and association with gliosis in Green and golden bell frogs golden bell frog tadpoles (Table 1 and Table S1B, x2 p = 0.0004). and tadpoles None of the Striped marsh frog tadpoles were found to be infected 2 Cystodiscus axonis produce 5–25 mm multinucleated non-sporo- (0% 0/42) (Table 1 and Table S1B, 0% x p = 0.008). The gonic producing plasmodia within axons [19,20] (Fig. 1). Plasmo- numbers of organisms found in the brains of the Peron’s tree frog dia were found in 15% (9/60) of the Green and golden bell frog tadpoles was low (100% mild infection load) and while gliosis was tadpoles and 33% (3/9) of adults (Table 1). Plasmodia were not seen in these tadpoles, the severity was less than that seen in consistently seen in all portions of the brain in all specimens. infected Green and golden bell frog tadpoles. Haemorrhage within Organisms were rarely found the diencephalon (8%, 1/12) at the the nervous system was not present in any of the Peron’s tree frog level of the optic lobe, but were more common caudally in the tadpoles. mesencephalon (34%, 4/12) and the rhombencephalon (34%, 4/ Adult frogs. The prevalence of C. axonis in other species of 12) (Fig. 1E). Plasmodia were seen in one specimen in a cranial frogs varied significantly. The highest was the Southern bell frogs nerve as it left the brain and in a section of the trigeminal nerve as (100%, 8/8) and Yellow spotted bell frogs (100%, 3/3). Moderate it passed through muscle. Plasmodia were never seen in the prevalences were seen in the Peron’s tree frogs (50%, 3/6) and the prosencephalon, they were typically more abundant in the spinal Booroolong frogs (34.1%, 15/44). Low prevalences were observed cord (67%, 8/12) and nerve roots (50%, 6/12) (Fig. 1A, B) and if in Striped marsh frogs (12%, 2/17) and Cane toads (8%, 2/26) found in one nerve root, were generally found in multiple other (Table 1). nerve roots. In one frog plasmodia were found in a peripheral Varying degrees of gliosis were found within the spinal cord nerve. There was no association between age (adult or tadpole) (Fig. 2A–D), brain (Fig. 2E) and nerve root (Fig. 2F) sections of and the presence of plasmodia (p = 0.21) (Table S1A). infected adults. Multifocal to locally extensive haemorrhage was Inflammatory changes were observed in 83% of infected seen in the adult Yellow spotted bell frogs (100%, 3/3) and animals (10/12); changes were characterised by increased num- Southern bell frogs (38%, 3/8). Haemorrhagic lesions were most bers of glial cells (gliosis) and the response varied between focal to severe and widespread in the Yellow spotted bell frogs (Fig. 2C and multifocal and locally extensive in different intensities (Fig. 1). The D). Vasculitis characterized by perivascular lymphoid cell cuffing,

PLOS ONE | www.plosone.org 5 October 2012 | Volume 7 | Issue 10 | e4378070 Pathology of Cystodiscus spp. in Australian Frogs

Figure 1. Cystodiscus axonis lesions in brains of Green and golden bell frogs and tadpoles. A. Root nerve of a Green and golden bell tadpole showing severe inflammation, loss of nerve ganglia and numerous C. axonis plasmodia (arrowheads), scale 200 mm; B. Higher magnification (20X view) of root nerve (A), infiltrate has completely replaced nerve cells, scattered foci of necrosis are seen throughout the nerve (circle), scale 100 mm; C. Higher magnification view of C. axonis plasmodia within the same nerve root (A) (40X), scale 50 mm; D. Posterior end of Green and golden bell metamorph spinal cord showing moderate glial activity in response to moderate plasmodia load (arrowheads indicate plasmodia). This animal had motor dysfunction possibly due to the high load of plasmodia in its spinal cord and dorsal nerve roots, scale 200 mm; E. Moderate load of plasmodia (arrows) in diencephalon and mesencephalon of tadpole, scale 200 mm; F. Higher magnification (20X) view of plasmodia in diencephalon of tadpole (E) showing increased glial activity in response to plasmodia presence, scale 40 mm. doi:10.1371/journal.pone.0043780.g001 necrosis of the vascular endothelium and hylanization of the vessel Behavioural abnormalities in adult frogs with Cystodiscus walls was found in both Yellow spotted bell frogs and one axonis Booroolong frog. Some degree of necrosis and malacia (Fig. 2C A Green and golden bell frog metamorph was captured because and E) was present in Yellow spotted bell frogs, Southern bell frogs it could not use its back legs. This animal had many organisms and and Booroolong frogs. associated gliosis in the dorsal root ganglia and spinal nerves (Fig. 1D). Three of the infected Booroolong frogs (20%, 3/15) had

PLOS ONE | www.plosone.org 6 October 2012 | Volume 7 | Issue 10 | e4378071 Pathology of Cystodiscus spp. in Australian Frogs

Figure 2. Severe reactions to Cystodiscus axonis plasmodia in other species of Australian frogs. A. Spinal cord of Southern bell frog showing multifocal and locally extensive gliosis in reaction to moderate load of C. axonis plasmodia (arrows), individual was noted to be lethargic and inappetent, scale 200 mm; B. Spinal cord of Southern bell frog showing congestion and severe haemorrhage in association with high numbers of C. axonis plasmodia, individual was found dead, scale 50 mm; C. Diffuse haemorrhage, multifocal melacia (grey arrowheads) and gliosis and C. axonis plasmodia (black arrowheads) in spinal cord of Yellow spotted bell frog, scale 100 mm; D. C. axonis plasmodia associated with blood vessel epithelial cell in Yellow spotted bell frog spinal cord, scale 20 mm; E. Cross section of Booroolong frog infected with C. axonis, showing severe haemorrhage, extensive multifocal gliosis and spongiform change, scale 100 mm; F. Nerve root of Striped marsh frog showing mild focal gliosis in association with C. axonis plasmodia (arrow), scale 50 mm. doi:10.1371/journal.pone.0043780.g002 a history of neurologic disease. The animals were dragging their Hepatic myxosporidiosis in tadpoles characterised by the back legs and one could not right itself or react to simple association between biliary hyperplasia, inflammation, stimuli. Seven of the fifteen (46.7%) Booroolong frogs were loss of hepatocytes and the presence of Cystodiscus spp. found dead. Seven of the (87.5%, 7/8) Southern bell frogs were plasmodia reported to be lethargic and unresponsive to stimuli prior to The plasmodia of C. axonis and C. australis develop within the euthanasia, the eighth animal was found dead. All of Yellow bile ducts as multinucleated intraluminal spherical to tubular spottedbellfrogs(3/3)diedspontaneouslywithnoobserved structures distributed uniformly across the liver [19,20]. Plasmodia behavioural abnormalities. (Fig. 3) were found in 47% (34/72) of the Green and golden bell

PLOS ONE | www.plosone.org 7 October 2012 | Volume 7 | Issue 10 | e4378072 Pathology of Cystodiscus spp. in Australian Frogs frog tadpoles examined (Table 1). Associated lesions consisted of The presence of liver plasmodia, age (tadpole) and season were all biliary hyperplasia, loss of hepatocytes, and lymphoplasmacytic significantly associated with the following hepatic lesions within inflammation at the margins of the expanding lesions (Fig. 3). multivariable logistic regression models: biliary hyperplasia (odds Biliary hyperplasia was multifocal to focally extensive with lesions ratio = 10.8), loss of hepatocytes (odds ratio = 5.1) and severe coalescing as they became more advanced. The amount of liver hepatic myxosporidiosis (combination of lesions) (odds ratio = 7.9) affected ranged from less than 10% to in excess of 70%. No (Table S1D). evidence of regeneration was noted in any of the sections. Lesions Plasmodia in the gallbladder were observed in 14% (10/74) of were found in all (n = 34) infected tadpoles. Hepatic lesions were Green and golden bell frog tadpoles (confirmed by PCR). found in 33% (24/72) of tadpoles where plasmodia were not Organisms were not attached to the epithelium, and were identified. vacuolated with numerous vegetative nuclei in the cytoplasm, as The presence of hepatic plasmodia was significantly associated described by Hartigan et al. [20]. The immature forms were never with age (tadpole) (p = 0.006) and biliary inflammation observed with myxospores and were distinct from biliary (p = ,0.001) in univariable logistic regression models (Table S1C). plasmodia based on size and vacuolated appearance. When

Figure 3. Myxosporiodiosis and developing Cystodiscus spp. plasmodia in tadpoles. A. Coalescing extensive biliary lesions in Green and golden bell tadpole, scale 500 mm; B. Green and golden bell tadpole showing loss of hepatocytes surrounding hyperactive biliary tree and lymphoplasmacytic inflammation, moderate grade of hepatic myxosporidiosis, scale 200 mm; C. Mild inflammation and biliary hyperplasia seen in Striped marsh tadpole, scale 100 mm; D. Higher magnification of Cystodiscus spp. (either C. axonis or C. australis) intra-biliary plasmodium, scale 25 mm; E. Peron’s tree frog tadpole with moderate hepatic myxosporidiosis and plasmodia development in gallbladder (free floating forms), scale 200 mm; F. higher magnification view of plasmodia in (E) (40X), plasmodia appear vacuolated and with scattered nuclei in cytoplasm, these are immature forms (presporogonic) of Cystodiscus spp. that develop in the gallbladder, scale 40 mm. doi:10.1371/journal.pone.0043780.g003

PLOS ONE | www.plosone.org 8 October 2012 | Volume 7 | Issue 10 | e4378073 Pathology of Cystodiscus spp. in Australian Frogs sectioned, the plasmodia ranged in size from 5–20 mm wide and like eosinophilic fibrillar to amorphous material (Fig. 4C). 50–100 mm long. Immature plasmodia that had not yet begun to produce spores were found in the gallbladder. Six adult frogs (75%, 6/8) had some Overwintered Green and golden bell frog tadpoles are degree of biliary hyperplasia. This varied from focal to multifocal significantly more likely to have severe hepatic mild lesions to a moderate locally extensive lesion. At no time did myxosporidiosis these lesions represent more than 20% of the liver section. Amyloid-like deposits were present in 4 of 5 of the lesions in adult Cystodiscus spp. plasmodia were found in the livers of Green and frogs. There was only a moderate deposit of amyloid-like material golden bell frog tadpoles all year round. The prevalence of in 3 frogs but larger deposits were present in the other frog. The Cystodiscus spp. liver plasmodia in Green and golden bell frog prevalence of these lesions (80%, 4/5) was significantly higher than tadpoles was lowest in the spring (22.2%, 2/9), increased in the Green and golden bell frog (0%) (x2 p = 0.002). Five of the summer (42.9%, 6/8) and winter (44.8%, 13/29). The highest frogs (63%, 5/8) with liver lesions had spore producing plasmodia prevalence in the autumn (65.0%, 13/20) (Table 2B), however, the in their gallbladders. prevalence did not vary significantly between seasons. Striped marsh frogs (Limnodynastes peronii). Twenty Severe hepatic myxosporidiosis was more prevalent in autumn seven adult Striped marsh frog adults were examined and 20 (20.0%, 4/20) and winter (20.1%, 6/29) than spring (0%, 0/9) and (74%, 20/27) were found to be infected and producing spores. summer (14.0%, 2/14) (Table 2C). There was a significant Mild degrees of biliary hyperplasia were found in 12 (60%) of the association between the season of collection and the level of disease 20 infected frogs. The prevalence of lesions in the liver (44%, 12/ seen (p = 0.01, Table S1D). Animals collected in winter were 7.9 27) was significantly different from those in the Green and golden times more likely to have severe hepatitis than those collected in bell frogs (x2 p = 0.027). The degree of biliary hyperplasia, spring (Table S1D). Severity of hepatic myxosporidiosis was however, was mild resulting in a two to three fold increase in associated with the presence of Cystodiscus spp. plasmodia, age bile duct cross sections and typically only affected scattered portal and season of collection in a multivariable model (Table S1D). triads. A mild degree of peribiliary fibrosis was present in 60% (12/20) of infected frogs, but these changes rarely impacted the Variation in prevalence of hepatic myxosporidiosis in the surrounding hepatocytes. Striped marsh frog and Peron’s tree frog tadpoles Booroolong frogs (Litoria booroolongensis). Liver lesions The description of the lesions seen in the Peron’s tree frog and found in the 15 infected adult Booroolong frogs with brain stages Striped marsh frog tadpoles is identical to those described above in included amyloid-like deposits in 20% (3/15), mild periportal Green and golden bell frog tadpoles. Overall chi-sq analysis for fibrosis in 6.7% (1/15) and mild biliary hyperplasia in 20% (3/15). liver plasmodia presence in tadpoles was found to be significant Presporogonic plasmodia were also present in bile ducts of the 3 (p = 0.0004). Prevalence of Peron’s tree frog tadpoles with hepatic adult Booroolong frogs (Fig. 4E) with biliary hyperplasia. This is plasmodia (86%, 12/14) was significantly higher than seen in the the only time that this stage of development was found in an adult Green and golden bell frog tadpoles (47%, 34/72; x2 p = 0.009) frog. (Table S1E). The majority of infected Peron’s tree frog tadpoles Southern bell frogs (Litoria raniformis). Liver lesions (64%, 9/14) had moderate to severe lesions. The prevalence of were found in all 8 frogs. These lesions varied from mild biliary infection in the Striped marsh frog tadpoles (26%, 10/38) was also hyperplasia with a minimal degree of fibrosis (Fig. 4B), to locally significantly lower than in Green and golden bell frog tadpoles (x2 extensive changes with accompanying amyloid-like material and p = 0.033) and only mild lesions were seen in the livers of these disruption of the adjacent architecture, as seen in the most tadpoles (Table 1). Early development of gallbladder plasmodia advanced lesions observed in the Peron’s tree frogs. Spore was observed in 36% (5/14) of Peron’s tree and 11% (4/38) of production in the gallbladder was occurring in 7 of 8 frogs. Striped marsh frog tadpoles, as described in the Green and golden Yellow spotted bell frog (Litoria castanea). Two adults bell frog tadpoles (Fig. 3). and a sub-adult were examined. The sub-adult had a moderate degree of biliary hyperplasia, with accompanying inflammation Cystodiscus infection associated with fibrosis of biliary and associated loss of adjacent hepatocytes (Fig. 4D). The sub- tracts in adult frogs adult and one adult were found to have an amyloid-like material Both C. axonis and C. australis develop spore producing around the biliary ducts. No frogs were observed to have spores in plasmodia in the gallbladder of adult frogs [19,20]. In histological the gallbladder. sections, plasmodia appear as elongated discs containing numer- Cane toad (Rhinella marina). Six of the nine infected adults ous nuclei and myxospores scattered throughout the cytoplasm examined had fibrosis associated with the portal tract. No biliary and freely in the bile. The oval spores contain characteristic hyperplasia or inflammation was observed in any of the livers seen. circular polar capsules at distal ends (Fig. 4F). Large numbers of myxospores were observed in the gallbladders of Green and golden bell frogs (Litoria aurea). Twelve adults (Fig. 4F). Green and golden bell frogs (11 adults and 1 metamorph) were examined in this study for evidence of Cystodiscus spp. infection. Sensitivity of histology to detect Cystodiscus species Biliary plasmodia, biliary hyperplasia and associated loss of infection in the brain and liver hepatocytes, lesions found in tadpole livers, were only found in Throughout this study we used the microscopic examination of the metamorph (an animal with 4 legs, but had yet to reabsorb its single brain and liver histological sections to determine presence or tail). Myxospores were found in the gallbladder of one adult frog absence of Cystodiscus spp. To calculate sensitivity of this approach with no biliary lesions were seen. we compared single histological section outcome with (i) multiple Peron’s tree frogs (Litoria peronii). Eight (7 adults and 1 serial histological sectioning and (ii) used a previously developed metamorph) Peron’s tree frogs were examined. The single multiplex PCR outcome [20] on a selection of samples (Table 3). metamorph had biliary plasmodia, biliary hyperplasia and The sensitivity of single vs. serial histological sections for both associated loss of hepatocytes. Surrounding the proliferating bile brain and liver samples was 100%, however within the liver ducts within the interstitium were multifocal deposits of amyloid- samples, there were four overall positive samples that contained

PLOS ONE | www.plosone.org 9 October 2012 | Volume 7 | Issue 10 | e4378074 Pathology of Cystodiscus spp. in Australian Frogs

Figure 4. Cystodiscus spp. liver and gallbladder development in adult frogs. A. Normal portal triad in healthy adult Striped marsh frog, scale 50 mm; B. Moderate hepatic lesions in Southern bell frog adult with C. axonis myxospores in gallbladder, moderate biliary hyperplasia, fibrosis and inflammation. No intra-biliary plasmodia were seen, scale 100 mm; C. Severe biliary fibrosis in adult Peron’s tree frog infected with C. axonis myxospores in gallbladder. No plasmodia within bile ducts, calcified deposits observed within biliary tree (also seen in tadpoles), scale 500 mm; D. Intrabiliary C. axonis plasmodia in sub-adult Yellow spotted bell frog, bile duct extremely distended, surrounded by lymphoplamacytic infiltration and some fibrosis, scale 100 mm; E. Cystodiscus axonis plasmodia in adult Booroolong frog bile duct, mild biliary hyperplasia and inflammation accompanied the infection, scale 20 mm; F. Myxospores (arrowheads) in the gallbladder of adult Cane toad, spores identified by circular polar capsules at distal ends, found throughout large plasmodium cytoplasm and freely in bile, scale 40 mm. doi:10.1371/journal.pone.0043780.g004 sections in which parasites were not found (false negatives) (5/16; equivalent of 0.1 myxospore DNA did not return species specific 4/16; 2/30 and 2/34). Importantly, observing all serial sections PCR amplicon for either Cystodiscus spp. including the false negative sections revealed histopathological lesions characteristic of hepatic myxosporidiosis (biliary hyperpla- Discussion sia, inflammation or loss of hepatocytes). Cystodiscus was detected histologically in 43% of brains, and 33% Ecological implications of C. axonis and C. australis for of livers which tested positive by multiplex PCR. PCR was 100% Australian frogs sensitive in detecting brains that were microscopically Cystodiscus Cystodiscus axonis and C. australis are emerging myxosporean was observed, but since it did not detect infection in 2/19 liver parasites of multiple species of Australian frogs. In addition to samples that were positive on histology, sensitivity of PCR in liver characterizing the prevalence and pathology associated with C. is 66% (Table 3). axonis and C. australis infection in several amphibian populations, The limit of detection using 10-fold serial dilutions of C. axonis the question that we address in this paper is whether these and C. australis myxospores DNA (Figure S1) was equivalent to 1 parasites have an ecological impact. Several biological character- myxospore, as confirmed by three independent PCR reactions. An istics are necessary for an emerging pathogen to have a significant

PLOS ONE | www.plosone.org 10 October 2012 | Volume 7 | Issue 10 | e4378075 Pathology of Cystodiscus spp. in Australian Frogs

Table 3. Comparison of detection method results (single histological section vs. serial histological sections) and species specific multiplex PCR.

Serial histological sections

Single histological section Positive Negative

Tadpole liver samples Positive 12 0 Negative 0 5 Frog and tadpole brain samples Positive 3 0 Negative 0 12

Multiplex PCR

Single histological section Positive Negative

Tadpole liver samples Positive 4 2 Negative 8 5 Frog and tadpole brain samples Positive 3 0 Negative 4 4

N.B Sensitivity for both methods is 100%. Lower 95% confidence Intervals for liver = 77.9% and for brain = 36.8%; Upper confidence levels could not be determined. N.B Sensitivity for liver = 33.3% (95% CI: 9.9%–65.1%) and for brain = 42.9% (95% CI: 9.9%–81.6%). doi:10.1371/journal.pone.0043780.t003 and persistent ecological impact. The pathogen must be able to disease [35], and it is possible that the stress or advanced age (in infect a high percentage of the susceptible population, it must be some cases) could have made these frogs more prone to disease. able to cause disease in a significant number of the animals that it Both C. axonis and C. australis were associated with severe disease infects, the disease that it generates must be sufficiently severe that in the liver of tadpoles of susceptible species. A healthy tadpole it will directly or indirectly interfere with recruitment of juvenile liver is essential for the hormonal cascades that bring about animals into the breeding population or kill mature breeding metamorphosis [36], in addition the liver is the site for adults, and the pathogen needs to persist in the environment [31]. haematopoesis, immunoglobulin production [37], and is an In this study, we show that C. axonis and C. australis exhibit these important organ for energy and protein metabolism [38]. It is characteristics in our study populations and thus have a significant likely that the extent of the lesions seen in the livers of the Green potential for ecosystem impact. and golden bell frog tadpoles and the Peron’s tree frog tadpoles Our findings show that C. axonis and C. australis caused high would have significant metabolic and physiological implications. prevalences of infection in multiple species of frogs including the Our seasonal data also suggests that the liver disease resulted in a endangered Green and golden bell frog where prevalences of delayed metamorphosis and overwintering in some Green and infection in tadpoles were as high as 65%. They also produced golden bell frog tadpoles. significant disease that directly or indirectly would have impacted Overwintering usually occurs when tadpoles do not have the their host in the common and endangered species of frog resources to metamorphose and provides the advantage of extra examined in this study. Neurological dysfunction is associated time to feed and grow before spring [39]. However, overwintering with similar haemorrhagic lesions in fish infected with myxosporea also increases the risk of desiccation [40], predation and increased in the brain (Myxobolus balantiochelois and Myxobolus cerebralis) [32– exposure to pathogens and parasites [39]. It is unclear if infection 34]. The inflammatory and haemorrhagic lesions associated with with Cystodiscus spp. predisposes tadpoles to overwinter or if C. axonis described here and possibly the presence of the parasite overwintered tadpoles are just infected for longer and therefore within axons would be expected to significantly interfere with have a severe (chronic) host response. Regardless, any negative neurological function and this was indeed the case in some frogs impact on the ability to metamorphose exposes tadpoles to more that were observed prior to death. The ecological impact of the risks and removes one reproductive season from the life history of nervous system form of C. axonis may be increased by its ability to the tadpole/frog. cause disease in both tadpoles and adult frogs. Significant brain The ability of C. axonis and C. australis to have spread so widely lesions were seen in a high percentage in adults of both the in 50 years could suggest an ability to persist in the environment. Southern bell frog and the Yellow spotted bell frog. However, This could be explained by the resilience of myxospores in the whether these types of lesions occur in the wild is unknown. In this water column as described for other myxosporea species [41], or study, all the adult frogs with C. axonis lesions in the nervous system sustained infection in other (vertebrate and invertebrate) hosts. where held in captivity for months to years before their death. The Subclinical but productive infections were observed in the invasive stress of captivity can exacerbate the impacts of pathogens and Cane toad and at least one species of widely distributed native

PLOS ONE | www.plosone.org 11 October 2012 | Volume 7 | Issue 10 | e4378076 Pathology of Cystodiscus spp. in Australian Frogs frog, the Striped marsh frog [19]. Neither of these species tadpoles of specific ages this could be valuable information for developed significant disease as either tadpoles or adults, but both establishing captive bred populations that are free of disease and species produced myxospores in their gallbladders and would have strategies for translocation of tadpoles. been expected to shed the parasites during their adult life. Similar The presence of presporogonic plasmodia in a small number of differences in susceptibility of native species to an invasive adult Booroolong frogs infected with C. axonis was unexpected pathogen was been shown in studies of the chytrid fungus because these stages were never seen in adult frogs of any other (Batrachochytrium dendrobatidis), where Striped marsh frogs were species in this study. These animals were euthanized immediately shown to have a higher tolerance for infection than the Green and after capture from the wild and thus could have been recently golden bell frogs [42]. infected. Alternately, the liver could have been colonized by reactivation of a latent infection in the brain or in other tissue. Is presporogonic plasmodial development in the nervous Why the bile duct environment could support presporogonic system an amplification strategy? plasmodia in this species and not in others remains unknown and Presporogonic development was observed to be a common would require controlled infection trials to determine. finding in the nervous system of tadpoles and adult frogs infected with C. axonis. The question then arises as to whether it is a PCR and histology are valuable surveillance tools for necessary part of the parasites life cycle? Based on the myxosporea Cystodiscus infections in frogs of fish, the nervous system can act as a route of infection and Screening frogs and tadpoles for infectious diseases is necessary conduit to other tissues, as occurs with M. cerebralis in salmonid fish to prevent the introduction of disease into captive breeding [43], or as site of spore production with spores being released programs, preventing disease release into new areas from captive- when the host dies or is eaten as occurs in several other species of raised frogs, and determining the geographic distribution of the fish [9,10,44–46]. Evidence that the nervous system is used by C. disease. This study shows that both PCR and histopathology have axonis for amplification is based on the finding that the nervous advantages as screening tools for Cystodiscus infections and that a system was infected by C. axonis in all species of frogs examined, combination of both is optimum. arguing that it was a targeted tissue. Additionally, finding the The PCR assay used in this study amplified an rDNA sequence organisms in cranial and peripheral nerves suggests that C. axonis for which as many as 10,000 copies may be present in may use the nervous system as an entry way to the tadpole. Lastly, mxyosporeans [53]. It was found to be very sensitive (sensitivity finding organisms in close association with vessels could explain limit one organism) and was more sensitive than histology in how the parasite moved from the central nervous system to other detecting infection. This is consistent with a similar study, where tissues as has been suggested for other myxosporea species [47,48]. PCR was found to be up to 3 times more sensitive than histology If C. axonis does use the nervous system as a site for amplification for detecting infection with the B. dendrobatidis [54]. Our PCR was within the host it is likely that in its natural host it causes minimal also able to detect and distinguish between infection with one or or no disease as was seen in the Cane toad and the Striped marsh both species of Cystodiscus, and the added advantage of PCR is that frog. Adverse impacts, as seen in Green and golden bell frog it can be used with pooled samples. The only disadvantage of PCR tadpoles and Booroolong frogs, and adults of the Southern bell would be is susceptibility to contamination. Proving that animals frog and Yellow spotted bell frogs appear to represent reaction of were actually infected and that samples were not contaminated by the host’s immune system to the parasite, and would be those environmental organisms can be difficult as has been shown for expected in non-parasite adapted species. Ultimately, to determine other amphibian pathogens in aquatic environments [54]. if the nervous system stages of C. axonis represent an important Histology as a screening tool has the advantages that it can be developmental stage controlled infection trials where the sequence used to detect an array of infectious agents, in addition to of the development of C. axonis can be followed will be necessary. Cystodiscus spp., and that it can determine if infection is associated with disease. Combining these screening techniques or using Possible window of vulnerability and management histopathology as a confirmatory tool for populations that have implications been shown to be PCR positive would provide the most diagnostic From a conservation perspective, it would be valuable to know if power. Whichever technique is used, screening tadpoles in Cystodiscus infections occur only in the tadpole or at any time summer or autumn is optimal to detect infected populations as during the tadpole and frog life cycles. Based on studies of this is when infection prevalence is highest. In addition, myxosporean infections in fish, age of the host and seasonal factors development of an in situ hybridisation protocol for these two would be expected to impact susceptibility to infection and disease species could prove extremely informative for future research in [49–51]. Tadpoles have also been shown to exhibit different levels these parasites. of disease and resistance to other parasites, for example the This study shows that pathologists should consider C. axonis in trematode (Ribeiroia ondatrae). In this case the type and severity Australian frogs (tadpoles and adults) if they are exhibiting gliosis of limb malformations varied according to tadpole their size and and infection with either C. axonis or C. australis or both in tadpoles developmental stage [52]. with biliary hyperplasia, particularly if there is minimal associated The difference in prevalence between frog and tadpoles for both inflammation. Differential diagnoses for the lesions caused by C. Cystodiscus species raises questions about the timing of infection. axonis in the nervous system would be those caused by the rare The low rate of infection in adult Green and golden bell frogs, extension of systemic bacterial and fungal infections to the nervous compared to tadpoles suggests that infected tadpoles are either system and these would typically be accompanied by a neutro- capable of ridding themselves of infection, or infected tadpoles philic or granulomatous inflammation. Liver disease caused by have reduced survival rates compared with non-infected animals, infectious agents (viruses, bacteria, other parasites and fungi), or that adults of this species may be resistant to infection. Whether however, is common in frogs and tadpoles [55–58]. These diseases this represents infection in adult animals or infection of tadpoles differ from the lesions caused by Cystodiscus spp. by the level of with delayed development will require controlled experiments to necrosis observed, multiple organs affected and the physical determine. If Cystodiscus spp. are only able to infect tadpoles or only presence of the pathogen, e.g. fungal hyphae or nematode larvae.

PLOS ONE | www.plosone.org 12 October 2012 | Volume 7 | Issue 10 | e4378077 Pathology of Cystodiscus spp. in Australian Frogs

Conclusion Cystodiscus axonis and C. australis using Cystodiscus internal tran- This study shows that both Cystodiscus axonis and Cystodiscus scribed spacer rDNA specific primers. Reactions were run in australis have the potential to infect and cause disease in common triplicate and visualised on a 2% agarose gel stained with Gel Red and endangered Australian frog species. They were also associated (Biotium, Australia). with spontaneous deaths in semi captive and captive populations (TIF) of frogs. If these infection dynamics occur in wild frogs then they Table S1 Summary of statistical analyses results for associating would also be expected to significantly reduce tadpole recruitment Cystodiscus axonis and Cystodiscus australis with lesions of disease. and adult frog longevity, potentially making them a key (DOC) threatening process. Additionally, this study also provides evidence of certain host developmental stages being more susceptible to Acknowledgments infection, which may have significant implications for release of captive bred animals. Lastly, this study provides investigators the We thank the Jane Hall, Cheryl Sangster, Peter Harlow and Michael necessary tools for screening wild and captive populations of frogs McFadden for their specimen collections and advice, and Elaine Chew for for the presence of these parasites and these tools may prove her expert histopathology processing. We acknowledge the Australian instrumental in determining the distribution of these parasites in Biosecurity Intelligence Network for information management and communication systems that contributed to this research. Australia and establishing specific pathogen free captive breeding colonies. Author Contributions Supporting Information Conceived and designed the experiments: AH JSˇ DNP. Performed the experiments: AH KR DNP. Analyzed the data: AH ND KR JSˇ DNP. Figure S1 Detection limit for multiplex species specific Contributed reagents/materials/analysis tools: AH KR JSˇ ND. Wrote the PCR. Ten fold dilutions of quantified myxospore samples for paper: AH JSˇ DNP.

References 1. Siddall ME, Martin DS, Bridge D, Desser SS, Cone DK (1995) The demise of a 20. Hartigan A, Fiala I, DykovaÍ, Rose K, Phalen DN, et al. (2012) New species of phylum of protists: Phylogeny of myxozoa and other parasitic cnidaria. Journal Myxosporea from frogs and resurrection of the genus Cystodiscus Lutz, 1889 for of Parasitology 81: 961–967. species with myxospores in gallbladders of amphibians. Parasitology 19: 1–19. 2. Lom J, Dykova I (1997) Ultrastructural features of the actinosporean phase of 21. Hartigan A, Phalen DN, Sˇlapeta J (2010) Museum material reveals a frog myxosporea (Phylum myxozoa): A comparative study. Acta Protozoologica 36: parasite emergence after the invasion of the cane toad in Australia. Parasites & 83–103. Vectors 3: 50. 3. Jimeńez-Guri E, Philippe H, Okamura B, Holland PWH (2007) Buddenbrockia is a 22. Hartigan A, Sangster C, Rose K, Phalen DN, Sˇlapeta J (2012) Myxozoan Cnidarian worm. Science 317: 116–118. parasite in brain of critically endangered frog. Emerging Infectious Diseases 4. Lom J, DykovaÍ (2006) Myxozoan genera: definition and notes on taxonomy, 18(4): 693–695. life-cycle terminology and pathogenic species. Folia Parasitologica 53: 1–36. 23. Gillespie GR, Hines HB, editors (1999) Status of temperate riverine frogs in 5. Prunescu C, Prunescu P, Pucek Z, Lom J (2007) The first finding of south-eastern Australia. Canberra: Environment Australia. 109–130 p. myxosporean development from plasmodia to spores in terrestrial mammals: 24. White AW, Pyke GH (1996) Distribution and conservation status of the Green Soricimyxum fegati gen. et sp n. (Myxozoa) from Sorex araneus (Soricomorpha). Folia and Golden Bell Frog Litoria aurea in New South Wales. Australian Zoologist 30: Parasitologica 54: 159–164. 177–189. 6. Dykova´ I, Tyml T, Fiala I, Lom J (2007) New data on Soricimyxum fegati 25. Pyke GH, White AW (2001) A review of the biology of the Green and Golden (Myxozoa) including analysis of its phylogenetic position inferred from the SSU Bell frog Litoria aurea. Australian Journal of Zoology 31: 563–598. rRNA gene sequence. Folia Parasitologica 54: 272–276. 26. Wassens S (2008) Review of the past distribution and decline of the Southern Bell Frog Litoria raniformis in New South Wales. Australian Zoologist 34: 446– 7. Friedrich C, Ingolic E, Freitag B, Kastberger G, Hohmann V, et al. (2000) A 452. myxozoan-like parasite causing xenomas in the brain of the mole, Talpa europaea 27. Hamer AJ, Lane SJ, Mahony MJ (2010) Using probabilistic models to investigate L., 1758 (Vertebrata, Mammalia). Parasitology 121: 483–492. the disappearance of a widespread frog-species complex in high-altitude regions 8. Bartholomew JL, Atkinson SD, Hallett SL, Lowenstine LJ, Garner MM, et al. of south-eastern Australia. Animal Conservation 13: 275–285. (2008) Myxozoan parasitism in waterfowl. International Journal for Parasitology 28. Smith KF, Sax DF, Lafferty KD (2006) Evidence for the role of infectious disease 38: 1199–1207. in species extinction and endangerment. Conservation Biology 20: 1349–1357. 9. Feist SW, Longshaw M (2006) Phylum: Myxozoa. Fish Diseases and Disorders, 29. Rose K (2005) Wildlife health investigation manual. Sydney, NSW: The Vol 1: Protozoan and Metazoan Infections, 2nd Edition. Wallingford: Cabi Australian Registry of Wildlife Health. Publishing. 230–296. 30. Ransohoff RM (2007) Microgliosis: the questions shape the answers. Nat 10. Lom J, Dykova´ I (1992) Protozoan parasites of fish. Waltham, MA: Elsevier Neurosci 10: 1507–1509. Science Publishers. 31. Smith KF, Acevedo-Whitehouse K, Pedersen AB (2009) The role of infectious 11. Canning EU, Okamura B (2004) Biodiversity and evolution of the Myxozoa. diseases in biological conservation. Animal Conservation 12: 1–12. Advances in Parasitology 56: 43–131. 32. Levsen A, Alvik T, Grotmol S (2004) Neurological symptoms in tricolor 12. Bartholomew JL (2002) The history and dissemination of whirling disease. 3–24. sharkminnow Balantiocheilos melanopterus associated with Myxobolus balantiocheili n. 13. Boustead NC (1993) Detection and New Zealand distribution of Myxobolus sp. infecting the central nervous system. Diseases of Aquatic Organisms 59: 135– cerebralis, the cause of whirling disease of salmonids. New Zealand Journal of 140. Marine and Freshwater Research 27: 431–436. 33. Frasca S, Poynton SL, West AB, Van Kruiningen HJ (1998) Epizootiology, 14. Hoffman GL (1990) Myxobolus cerebralis, a worldwide cause of Salmonid Whirling pathology, and ultrastructure of the myxosporean associated with parasitic disease. Journal of Aquatic Animal Health 2: 30–37. encephalitis of farmed Atlantic salmon Salmo salar in Ireland. Diseases of aquatic 15. Halliday MM (1976) The biology of Myxosoma cerebralis the causative organism of organisms 32: 211–225. Whirling disease of Salmonids. Journal of Fish Biology 9: 339–357. 34. Rose JD, Marrs GS, Lewis C, Schisler G (2000) Whirling disease behavior and 16. Hedrick RP, El-Matbouli M, Adkison MA, Macconnell E (1998) Whirling its relation to pathology of brain stem and spinal cord in Rainbow trout. Journal disease: Re-emergence among wild trout. Immunological Reviews 166: 365– of Aquatic Animal Health 12: 107–118. 376. 35. Munson L, Terio KA, Worley M, Jago M, Bagot-Smith A, et al. (2005) Extrinsic 17. A´ lvarez Pellitero MP, Pereira Bueno JM, Gonza´lez Lanza C (1982) On the factors significantly affect patterns of disease in free-ranging and captive cheetah presence of Chloromyxum truttae Leger, 1906 in Salmo trutta fario from Leoń (Duero (Acinonyx jubatus) populations. Journal of Wildlife Diseases 41: 542–548. Basin, NW Spain). Bulletin of the European Association of Fish Pathologists 2 36. Shi Y-B, Wong J, Puzianowska-Kuznicka M, Stolow MA (1996) Tadpole (1): 4–7. competence and tissue-specific temporal regulation of amphibian metamorpho- 18. Lom J, DykovaÍ (1992) Protozoan parasites of fishes. New York: Elsevier. 315 p. sis: Roles of thyroid hormone and its receptors. BioEssays 18: 391–399. 19. Hartigan A, Fiala I, DykovaÍ, Jirku˚ M, Okimoto B, et al. (2011) A suspected 37. Rollins-Smith LA (1998) Metamorphosis and the amphibian immune system. parasite spill-back of two novel Myxidium spp. (Myxosporea) causing disease in Immunological Reviews 166: 221–230. Australian endemic frogs found in the invasive cane toad. Plos One 6(4): e18871. 38. Frieden E, Mathews H (1958) Biochemistry of amphibian metamorphosis. III. doi:10.1371/journal.pone.0018871. Liver and tail phosphatases. Archives of Biochemistry and Biophysics 73: 107– 119.

PLOS ONE | www.plosone.org 13 October 2012 | Volume 7 | Issue 10 | e4378078 Pathology of Cystodiscus spp. in Australian Frogs

39. Anstis M (2002) Tadpoles of south-eastern Australia: a guide with keys. Frenchs 49. Raffel T, Lloyd-Smith J, Sessions S, Hudson P, Rohr J (2011) Does the early frog Forest: New Holland Publishers Australia Pty Ltd. catch the worm? Disentangling potential drivers of a parasite age–intensity 40. Newman RA (1992) Adaptive plasticity in amphibian metamorphosis. relationship in tadpoles. Oecologia 165: 1031–1042. BioScience 42: 671–678. 50. Holzer AS, Sommerville C, Wootten R (2006) Molecular studies on the seasonal 41. Hedrick RP, McDowell TS, Mukkatira K, Macconnell E, Petri B (2008) Effects occurrence and development of five myxozoans in farmed Salmo trutta L. of freezing, drying, ultraviolet irradiation, chlorine, and quaternary ammonium Parasitology 132(Pt 2): 193–205. treatments on the infectivity of myxospores of Myxobolus cerebralis for Tubifex 51. Mun˜oz G, Randhawa H (2011) Monthly variation in the parasite communities tubifex. Journal of Aquatic Animal Health 20: 116–125. of the intertidal fish Scartichthys viridis (Blenniidae) from central Chile: are there 42. Stockwell MP, Clulow J, Mahony MJ (2010) Host species determines whether seasonal patterns? Parasitology Research 109: 53–62. infection load increases beyond disease-causing thresholds following exposure to 52. Johnson PTJ, Kellermanns E, Bowerman J (2011) Critical windows of disease the amphibian chytrid fungus. Animal Conservation 13: 62–71. risk: amphibian pathology driven by developmental changes in host resistance 43. El–Matbouli M, Hoffmann RW, Mandok C (1995) Light and electron and tolerance. Functional Ecology 25: 726–734. microscopic observations on the route of the triactinomyxon sporoplasm of 53. Hallett SL, Bartholomew JL (2009) Development and application of a duplex Myxobolus cerebralis from epidermis into rainbow trout cartilage. Journal of Fish QPCR for river water samples to monitor the myxozoan parasite Parvicapsula Biology 46: 919–935. minibicornis. Diseases of Aquatic Organisms 86: 39–50. 44. Kent ML, Hoffman GL (1984) Two new species of Myxozoa: Myxobolus inaequus 54. Skerratt LF, Mendez D, McDonald KR, Garland S, Livingstone J, et al. (2011) New species and Henneguya yheca New species from the brain of a South American Validation of diagnostic tests in wildlife: The case of chytridiomycosis in wild Knife fish Eigenmannia virescens. Journal of Protozoology 31: 91–94. 45. Dykova´ I, Lom J, Cirkovich M (1986) Brain myxoboliasis of common carp amphibians. Journal of Herpetology 45: 444–450. (Cyprinus carpio) due to Myxobolus encephalicus. Bulletin of the European Association 55. Berger L, Speare R, Humphrey J (1997) Mucormycosis in a free-ranging green of Fish Pathologists 6: 10–11. tree frog from Australia. Journal of Wildlife Diseases 33: 903–907. 46. Dzulinsky K, Cone DK, Faulkner GT, Cusack R (1994) Development of 56. Cullen BR, Owens L (2002) Experimental challenge and clinical cases of Bohle Myxobolus neurophilus (Guildford, 1963) (Myxosporea) in the brain of Yellow perch iridovirus (BIV) in native Australian anurans. Diseases of Aquatic Organisms 49: (Perca flavescens) in Vinegar Lake, Nova Scotia. Canadian Journal of Zoology 72: 83–92. 1180–1185. 57. Davis AK, Yabsley MJ, Keel MK, Maerz JC (2007) Discovery of a novel 47. Lom J (1987) Myxosporea: a new look at long known parasites of fish. alveolate pathogen affecting southern leopard frogs in Georgia: Description of Parasitology Today 3: 327–332. the disease and host effects. EcoHealth 4: 310–317. 48. Hedrick RP, MacConnell E, de Kinkelin P (1993) Proliferative kidney disease of 58. Crawshaw GJ, Weinkle TK (2000) Clinical and pathological aspects of the salmonid fish. Annual Review of Fish Diseases 3: 277–290. amphibian liver. Seminars in Avian and Exotic Pet Medicine 9: 165–173.

PLOS ONE | www.plosone.org 14 October 2012 | Volume 7 | Issue 10 | e4378079 Supporting Information: Table S1. Summary of statistical analyses results for associating Cystodiscus axonis and Cystodiscus australis with lesions of disease.

80 Supplementary Table 1 Summary of statistical analyses results for associating Cystodiscus axonis and Cystodiscus australis with lesions of disease A Brain histopathology (GGBF) Univariable binary logistic

regression Odds Outcome Explanatory variables b s.e. 95% C.I p Ratio Age (Adult versus Tadpoles) 1.04 0.79 2.83 -0.64, 2.57 0.21 Presence of brain plasmodia Season 0.003* Presence of brain plasmodia** 3.75 0.89 42.50 8.83, 325.64 <0.001 Gliosis Parasitic load <0.001* *Fisher’s exact test

B Chi-sq test p Striped marsh frog tadpoles versus 0.0086 Presence of brain plasmodia GGBF Peron's tree frog tadpoles versus GGBF 0.0004

C Liver histopathology (GGBF) Univariable binary logistic regression Outcomes Explanatory variables b s.e. Odds Ratio 95% C.I p Presence of Liver plasmodia Age (Adult versus Tadpoles) -2.29 1.07 0.10 0.005, 0.564 0.006 <.000 Biliary inflammation Presence of liver plasmodia 2.245 0.481 9.438 3.78, 25.32 1

81 D Multivariable logistic regression Outcomes Explanatory variables b s.e. Odds Ratio 95% C.I p Biliary hyperplasia Intercept -0.8315 0.7414 Presence of liver plasmodia 2.5827 0.8454 13.233 3.0, 95.4 0.0023 Age (Adults versus Tadpoles) -2.2683 1.0627 0.103 0.009, 0.68 0.0328 Season 0.0285 Summer vs. Spring 0.167 0.9708 1.182 0.17, 8.38 0.86 Autumn vs. Spring 1.5394 0.9679 4.662 0.76, 36.42 0.11 Winter vs. Spring 2.3841 0.968 10.849 1.79, 85.49 0.01 Grade of biliary hyperplasia Intercept 3 -4.48 0.87 Intercept 2 -1.85 0.71 Intercept 1 -0.76 0.68 Presence of liver plasmodia 2.87 0.58 17.70 5.98, 61.83 <0.001 Age (Adults versus Tadpoles) -2.25 0.92 0.11 0.01, 0.56 0.015 Season 0.057 Summer vs. Spring 0.43 0.84 1.53 0.28, 9.40 0.61 Autumn vs. Spring 1.60 0.80 4.96 1.05, 28.20 0.044 Winter vs. Spring 1.62 0.76 5.06 1.16, 26.67 0.034 Loss of hepatocytes Intercept -1.95 0.92 Presence of liver plasmodia 2.12 0.68 8.33 2.39, 36.37 0.002 Age (Adults versus Tadpoles) -2.69 1.28 0.07 0.003, 0.61 0.036 Season 0.002 Summer vs. Spring -0.58 1.15 0.56 0.06, 6.05 0.62 Autumn vs. Spring 2.28 1.06 9.75 1.41, 100.2 0.032 Winter vs. Spring 2.43 1.00 11.35 1.86, 106.76 0.016 Hepatic myxosporidiosis severity Intercept 3 -4.59 0.84 Intercept 2 -2.90 0.75 Intercept 1 -0.73 0.67 Presence of liver plasmodia 2.42 0.52 11.22 4.21, 32.41 <0.001 Age (Adults versus Tadpoles) -2.04 0.94 0.13 0.01, 0.695 0.03 Season 0.01 Summer vs. Spring 0.46 0.83 1.58 0.30, 8.99 0.58 Autumn vs. Spring 1.08 0.78 2.95 0.67, 15.23 0.16 Winter vs. Spring 2.07 0.77 7.91 1.88, 36.65 0.007

E Chi-sq test p GGBF vs. Striped marsh frog tadpoles 0.03331 Liver plasmodia GGBF vs. Peron's tree frog tadpoles 0.00824 GGBF vs. Striped marsh frog adults 0.02725 Biliary fibrosis GGBF vs. Peron's tree frog adults 0.0022

Chapter 6: Final discussion, future directions and conclusions

83 6.1 A Dark Horse: Myxosporea in Australian amphibians

Prior to this research, it was thought that the only myxosporea infections of Australian native frogs and tadpoles were those caused by Myxobolus hylae (Johnston, 1918) and Myxobolus fallax

(Browne et al., 2002) both infecting the urogenital tract and Cystodiscus immersus Lutz 1889, which infects the gallbladders of Australian frogs (Delvinquier, 1986). Myxobolus hylae and M. fallax had been found to be pathogenic in two Litoria spp., and C. immersus were assumed to have been introduced to Australia via the exotic Cane toad (Bufo marinus; Rhinella marina) in

1935. Delvinquier put forward the hypothesis of exotic introduction in 1986, after a survey of

Australian frog gall bladders showed similar looking myxospores to C. immersus from South

America. Species identification at this time was based on line drawings for C. immersus of

Brazilian specimens in 1889 and 1940. Delvinquier’s theory held until 2011 when myxosporea was investigated in Australian frogs. Identification of the parasite seen in native Australian frogs and tadpoles was a primary aim of this thesis. The parasite was examined using histological and ultrastructural morphology, molecular sequencing and retrospective analysis of the pathogen’s origins. Molecular identification of gallbladder myxospores and infected brain and liver tissue from seven host species (Figure 1) showed there were in fact two distinct genotypes present in some individuals (Figure 1). This distinction would have been very difficult to establish based on light microscopy alone. Direct and cloned sequencing of multiple ribosomal DNA regions offered the opportunity to compare the two Australian species with the Brazilian myxospore material thought to be C. immersus. This work showed that two distinct species were simultaneously infecting Australian frog species. It was also shown that these were in fact not related to the genotypes found in South America. Therefore due to a lack of type material for

84 genetic sequencing, there would be great difficulty determining which South American genotype

(species) was the true Cystodiscus immersus.

Figure 1. Amphibian hosts identified with Cystodiscus parasite infection

A: International Union for the Conservation of Nature (IUCN) Conservation status; B: Host species identified throughout this research; C: Host species range; D. Cystodiscus axonis development seen in brain tissue; E: Liver lesions associated with either Cystodiscus spp.; F: Cystodiscus species identified in host species, either Cystodiscus axonis or Cystodiscus australis

6.2 The origins of gallbladder myxosporea in Australian frog species

In addition to distinguishing these parasites from the South American species, the current study also aimed to determine the origin of Australian gallbladder myxosporea species. This included determining any role of the Cane toad in translocating any myxosporea into Australia as suggested by Delvinquier (1986) as well as evaluating if these pathogens were an emerging infection. Examination of the translocation route of this supposed vector of gallbladder myxosporea revealed an additional piece of the puzzle. The Cane toad was moved from its native

Central and South America and actively bred and translocated through the Caribbean before being moved to the Hawaiian island Oahu in 1932 (Easteal, 1981). Our studies looked for

85 gallbladder myxosporea in the Cane toads of Hawaii in 2010 and could find no positives in 261 individuals and therefore it can be concluded that C. immersus did not translocate with the Cane toad. This was confirmed by investigation of museum material which showed that gallbladder myxosporea were present in Australian frogs from 1966 in NSW, 31 years after the introduction of the Cane toad and also in areas outside the distribution of the Cane toad. This retrospective analysis of the prevalence and distribution of these parasites has identified the emergence of these pathogens within the last 40 years in Australia, which was the third aim of the work outlined in this thesis. It is now obvious that the Cane toad is not the vector of introduction of these parasites. Attention has turned to speculation about the ecological role of the Cane toad in relation to the gallbladder myxosporea that had been found in Australian frog species.

Introduced species are often thought to be competitors, new predators and vectors of foreign pathogens into new environments full of naïve hosts (Robinson, 2012c, Prenter et al., 2004,

Anderson and May, 1986). Introduced species can cause the decimation of the endemic population (Robinson, 2012a, Holdich and Reeve, 1991) and can also suffer complex effects on themselves. Some host species lose their parasites during their translocation (enemy release hypothesis) (Torchin et al., 2001) only to become infected with parasites from the new environment (Torchin et al., 2003). We hypothesised that not only was the Cane toad a victim of infection to the gallbladder myxosporea encountered in Australia, but it also acted as an amplifying host for these parasites – a “spill back” host. A spill back host can become infected with native pathogens and act as an additional host and source of infection for native species

(Tompkins and Poulin, 2006). This could explain the case of the Cane toad and gallbladder myxosporea in Australia. It has been suggested that the majority of spill back events that occur are unknown due to poor background about native parasites (Kelly et al., 2009). This is indeed

86 the case for myxosporea in Australian frogs, which has been significantly improved by the data provided in this thesis.

6.3 The more you look, the more you find: cryptic diversity in gallbladder myxosporea

The two Cystodiscus species identified by our research were only distinguished after significant scrutiny of molecular differences, ultrastructural features (filiform polar appendages of C. axonis) and histopathology of pre-sporogonic development. The genetic distinction and subtle morphological differences between these Cystodiscus species which overlapped in distribution and host range have implications for the diversity of other myxosporea infecting amphibian gallbladders around the world. We have summarised the described species, their morphology and host ranges from all over the world, yet only five species have been molecularly identified

(Figure 2). Genotyping has increased knowledge of diversity in all taxonomic groups, with the number of cryptic species complexes identified increasing when using PCR and molecular sequencing as part of taxonomical descriptions (Bickford et al., 2007). The group of gallbladder myxosporea described in amphibians around the world share a similar morphology, all produce spores in the gallbladder, and some even overlap in host species (e.g. Cystodiscus melleni and

Cystodiscus serotinum) (Figure 1). It is reasonable to assume that as more species in this complex are investigated there will be more species descriptions based on genetic differences. As well as genetic differences, our data indicates that morphological differences can still be indicative of species status. Another feature that we have shown to be a diagnostic tool in distinguishing between species is the location of the site of presporogonic development.

Cystodiscus axonis and C. australis can be differentiated on the basis of plasmodia presence in the nervous tissue of hosts (indicating C. axonis). No other myxosporean species found in the

87 gallbladder has been examined for histological lesions associated with infection and there are no data on the infection status of these Cystodiscus species in other species of tadpoles. If the tissue tropism can be used as a diagnostic feature between C. axonis and C. australis, then there could be value in examining hosts of other Cystodiscus species around the world.

Figure 2. Cystodiscus species infecting amphibian gallbladders around the world

All of the reported myxosporean species infecting frog gallbladders with adequate species identification (i.e. line drawing, host record). Host species identified and genetic evidence for species identification. Note that Cystodiscus Brazil 1 and 2 have not been formally described and therefore have not been included in this figure, they would infect Bufo marinus in Brazil. Lines on the spore surface indicating pattern of ridges and depressions used previously as species diagnostic characters.

6.4 The impact of Cystodiscus species in Australian frogs and tadpoles

The final aim of this thesis was to establish the host impacts of myxosporean parasites in

Australian frogs and tadpoles. The results of this study showed that both species cause significant pathology (Figure 1). During their development in the liver, Cystodiscus axonis and

C. australis were associated with hepatic lesions that affected the tadpole host’s health. The brain

88 lesions, including intra-axonal plasmodia, gliosis, necrosis and haemorrhage, that are associated with C. axonis can potentially cause significant behavioural changes and contribute to host mortality. Whilst the described pathology in affected hosts certainly poses a significant risk to individual tadpole and frog health, the wider population impacts for wild populations of

Australian frogs remain unknown. Some pathogens elicit a more significant host response in naïve host populations, for example Myxobolus cerebralis in Rainbow trout (Feist and

Longshaw, 2006) or Chytrid fungus (Batrochochytrium dendrobatidis) in amphibians.(Daszak et al., 1999) This may be the case in some Australian frog populations such as the Southern bell frogs and Yellow spotted bell frogs. Using museum material we established the emergence of these parasites in NSW populations between 1966 and 2009. An emerging pathogen could explain the host reactions we documented in Green and golden bell frogs, Booroolong frogs,

Southern bell frogs and Yellow spotted bell frogs. These host species are characterised by small threatened populations with a southern distribution, and C. axonis infection was found, with significant lesions in brain tissue. The severe host reactions seen in these frog species, due to C. axonis at least, could be an example of Cystodiscus spp. recently moving into their species range.

Cystodiscus species have been listed as important threats to captive breeding populations

(Murray and Skerratt, 2012), requiring greater research and management attention. The emergence of one or both of these parasites should be a concern for conservationists as they have the potential to be important pathogens to endangered species.

6.5 Future Research Possibilities

This study has increased knowledge and understanding of two parasites of significance to the conservation of endangered frog species. The identification of these parasites, as well as

89 establishing whether they are exotic or native pathogens, and establishing host impact, have been important steps in the potential success of future control programs. Further studies into the biology of these parasites are required in order to answer some important questions, such as: Is there an invertebrate host? Can infection of tadpoles and frogs be controlled? What Australian frog populations are going to be affected by Cystodiscus parasites now and in the future?

To date, no amphibian-myxozoa life cycle has been determined (Eiras, 2005). Without knowledge and understanding of an established life cycle, any mitigation or control of infection either in laboratory or natural conditions will be impossible. All of the life cycles identified in fish hosts use annelids as definitive hosts (Feist and Longshaw, 2006), and whilst it could be possible that there is no alternate host for Cystodiscus spp., both hypotheses should be tested.

Freshwater oligochaetes are a natural focus for amphibian life cycle studies, given that all of the known invertebrate hosts for myxozoa thus far have been oligochaetes or polychaetes worms

(Feist and Longshaw, 2006). The discovery of known alternate hosts for other myxozoan species has relied on the meticulous search of sediment material for tubificid worms, and the subsequent isolation of actinospores from these candidate worms (Xiao and Desser, 1998, Hallett et al.,

1999, McGeorge et al., 1997). The same approach was adopted for the search for Cystodiscus spp. during this study; however no alternate host was identified.

The impact of Cystodiscus parasites will be most important for captive bred and endangered populations (Murray et al., 2011). These groups are most vulnerable to extinction risks from pathogens due to small population size, genetic bottlenecks and stress (Altizer et al., 2003,

Smith et al., 2006, Daszak et al., 1999). It would therefore be beneficial for their management to establish a controlled transmission experiment of these parasites. An experimental infection

90 study for C. axonis and C. australis would establish if direct transmission can occur from myxospores in adult frog gallbladders to infect tadpoles (Figure 3). If direct transmission is not possible (Figure 3A) and no invertebrate hosts identified, studies will have to rely on sediment as a source of infective material. There are a number of sites where the prevalence of infection in frog hosts is high enough to secure a high number of infected invertebrates for infecting tadpoles.

At sites such as the Riverstone property (where the Green and golden bell frogs used in this research were collected) both Cystodiscus spp. were present and in high prevalence. Naïve or

‘clean’ spawn could be obtained from the Taronga Conservation Society amphibian breeding program. Allowing the tadpoles to develop side by side, with the substrate being the only variable, and sampling a small number of individuals from each species at regular time periods would allow the comparison of infection onset and prevalence of the parasites over a tadpole’s life cycle. This would establish if tadpoles can be infected at all the age stages, including spawn, tadpole and metamorph (Figure 3B). It would also be important to determine the effects of infection during metamorphosis (Figure 3C) as these answers will provide valuable data for conservation managers who are assessing translocation risks or managing captive breeding programs.

91 Figure 3. Representation of Cystodiscus spp. life cycle in Australian frogs

A controlled experiment would primarily examine the important aspects of the parasite-host relationship, such as A: the infectious agent (myxospore or actinospore) for frogs or tadpoles, B: the host development stages susceptible to infection, and C: if infection does inhibit metamorphosis (Figure 3C) as suggested in chapter 5.

Finally, the presence of the invertebrate and vertebrate hosts does not necessarily mean disease will develop, or that there are any negative ecological consequences associated with the presence of infection. Therefore it would be advantageous to develop a predictive tool for use in population risk management. An extremely innovative model was developed for assessing infection risk factors for chytrid fungus in Australian amphibians, which could be used in this case to predict populations that would be prone to myxospore or actinospore infection (Murray and Skerratt, 2012). This model was primarily created for populations with ‘data deficient’ assessments of chytrid infection. The aim was to focus surveillance and management activities

(Murray and Skerratt, 2012) based on the level of risk. A predictive model like this was based on host records (current and retrospective) (Green and Dodd, 2007) and the suitability of the

92 environment for Batrochochytrium dendrobatidis (Murray and Hose, 2005, Murray et al., 2011) presence and persistence. A similar approach would be possible with Cystodiscus parasites in

Australian frog populations, once the life cycle has been clarified. Some similarities exist between chytrid fungus and Cystodiscus parasites in relation to their impacts and the factors influencing infection dynamics in frog populations. Like chytrid fungus, C. axonis and C. australis cause host impacts which vary between species. There is also some indication of environmental persistence either in tadpole populations, or possibly the situation of alternate hosts, and both scenarios require significant surveillance and management. In the case of

Cystodiscus species, the use of retrospective host records such as museum material (Hartigan et al., 2010) and wildlife health records from university and groups such as the Australian Registry of Wildlife Health (Hartigan et al., 2011) has previously been demonstrated (Chapter 2,

Supplementary material). The distribution of the invertebrate host would provide an estimate of suitable environments, and therefore key frog populations could be prioritised (based on IUCN conservation status) for further surveillance, or if a known host of Cystodiscus spp., for possible captive breeding programs.

6.6 Conclusion

This thesis refuted a hypothesis of exotic myxosporean parasite introduction in Australian frogs that had held for 25 years. Key gaps in our knowledge of amphibian myxosporea have been identified, as well as an unprecedented biodiversity in frog myxosporea, by this study of

Cystodiscus myxosporidiosis in Australian frogs and tadpoles. This research has provided insight into importance of detecting amphibian pathogens, the need for multifaceted approaches in

93 species descriptions and has identified an emerging threat to endangered Australian amphibians that requires surveillance and further research.

References

Altizer, S., Harvell, D. & Friedle, E. 2003. Rapid evolutionary dynamics and disease threats to biodiversity. Trends in ecology & evolution, 18, 589-596. Anderson, R. M. & May, R. M. 1986. The invasion, persistence and spread of infectious diseases within animal and plant communities. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 314, 533-570. Bickford, D., Lohman, D. J., Sodhi, N. S., Ng, P. K. L., Meier, R., Winker, K., Ingram, K. K. & Das, I. 2007. Cryptic species as a window on diversity and conservation. Trends in ecology & evolution, 22, 148-155. Browne, R. K., Scheltinga, D. M., Pomering, M. & Mahony, M. 2002. Testicular myxosporidiasis in anurans, with a description of Myxobolus fallax n. sp. Systematic Parasitology, 52, 97-110. Daszak, P., Berger, L., Cunningham, A. A., Hyatt, A. D., Green, D. E. & Speare, R. 1999. Emerging infectious diseases and amphibian population declines. Emerging Infectious Diseases, 5, 735-748. Delvinquier, B. L. J. 1986. Myxidium immersum (Protozoa: Myxosporea) of the Cane toad Bufo marinus in Australian anura with a synopsis of the genus in amphibians. Australian Journal of Zoology, 34, 843-854. Easteal, S. 1981. The history of introductions of Bufo marinus (Amphibia: Anura); a natural experiment in evolution. Biological Journal of the Linnean Society, 16, 93-113. Eiras, J. C. 2005. An overview on the myxosporean parasites in amphibians and reptiles. Acta Parasitologica, 50, 267-275. Feist, S. W. & Longshaw, M. 2006. Phylum: Myxozoa. Fish Diseases and Disorders, Vol 1: Protozoan and Metazoan Infections, 2nd Edition. Wallingford Cabi Publishing. Green, D. E. & Dodd, C. K. 2007. Presence of amphibian chytrid fungus Batrochochytrium dendrobatidis and other amphibian pathogens at warm water fish hatcheries in Southeastern North America. Herpetological Conservation and Biology, 2, 43-47. Hallett, S., Erséus, C. & Lester, R. 1999. Actinosporeans (Myxozoa) from marine oligochaetes of the Great Barrier Reef. Systematic Parasitology, 44, 51-59. Hartigan, A., Fiala, I., Dyková, I., Jirků, M., Okimoto, B., Rose, K., Phalen, D. N. & Šlapeta, J. 2011. A suspected parasite spill-back of two novel Myxidium spp. (Myxosporea) causing disease in Australian endemic frogs found in the invasive cane toad. Plos One, 6. Hartigan, A., Phalen, D. N. & Šlapeta, J. 2010. Museum material reveals a frog parasite emergence after the invasion of the cane toad in Australia. Parasites & Vectors, 3. Holdich, D. M. & Reeve, I. D. 1991. Distribution of fresh water crayfish in the British Isles, with particular reference to crayfish plague, alien introductions and water quality. Aquatic conservation marine and freshwater ecosystems, 1, 139-158. Johnston, T. H. 1918. A parasite, Myxobolus hylae sp. nov. of the reproductive organs of the golden swamp frog, Hyla aurea. Australian zoologist, 1, 171.

94 Kelly, D. W., Paterson, R. A., Townsend, C. R., Poulin, R. & Tompkins, D. M. 2009. Parasite spillback: A neglected concept in invasion ecology? Ecology (Washington D C), 90, 2047-2056. McGeorge, J., Sommerville, C. & Wootten, R. 1997. Studies of actinosporean myxozoan stages parasitic in oligochaetes from the sediments of a hatchery where Atlantic salmon harbour Sphaerospora truttae infection. Diseases of Aquatic Organisms, 30, 107-119. Murray, B. R. & Hose, G. C. 2005. Life-history and ecological correlates of decline and extinction in the endemic Australian frog fauna. Austral Ecology, 30, 564-571. Murray, K., Skeratt, L., Marantelli, G., Berger, L., Hunter, D., Mahony, M. & Hines, H. 2011. Guidelines for minimising disease risks associated with captive breeding, raising and restocking programs for Australian frogs. A report for the Australian Government Department of Sustainability, Environment, Water, Population and Communities. A report for the Australian Government Department of Sustainability, Environment, Water, Population and Communities. Murray, K. A. & Skerratt, L. F. 2012. Predicting wild hosts for amphibian chytridiomycosis: integrating host life-history traits with pathogen environmental requirements. Human and Ecological Risk Assessment: An International Journal, 18, 200-224. Prenter, J., MacNeil, C., Dick, J. T. A. & Dunn, A. M. 2004. Roles of parasites in animal invasions. Trends in ecology & evolution, 19, 385-390. Smith, K. F., Sax, D. F. & Lafferty, K. D. 2006. Evidence for the role of infectious disease in species extinction and endangerment. Conservation Biology, 20, 1349-1357. Tompkins, D. M. & Poulin, R. 2006. Parasites and Biological Invasions Biological Invasions in New Zealand. In: Allen, R. B. & Lee, W. G. (eds.). Springer Berlin Heidelberg. Torchin, M. E., Lafferty, K. D., Dobson, A. P., McKenzie, V. J. & Kuris, A. M. 2003. Introduced species and their missing parasites. Nature, 421, 628-630. Torchin, M. E., Lafferty, K. D. & Kuris, A. M. 2001. Release from parasites as natural enemies: Increased performance of a globally introduced marine crab. Biological Invasions, 3, 333-345. Xiao, C. & Desser, S. S. 1998. The oligochaetes and their actinosporean parasites in Lake Sasajewun, Algonquin park, Ontario. The Journal of Parasitology, 84, 1020-1026.

Appendices

96 Appendix 1A

The University of Sydney Faculty of Veterinary Science, NSW Australia 2006

To Whom It May Concern,

We the undersigned are writing this letter to stipulate the role of Ashlie Hartigan in the preparation and submission of the following manuscript.

Hartigan. A, Phalen. D and Šlapeta J, (2010) Museum material reveals a frog parasite emergence after the invasion of the cane toad in Australia, Parasites & Vectors, 3:50, doi: 10.1186/1756-3305-3-50

Ashlie Hartigan during her PhD candidature was responsible for formulating the research questions, conducting statistical analyses, writing the draft manuscript, responding to reviewer’s reports and coordinating submission and publication of the manuscript.

Task Role of Co-author Collection and analysis of samples AH Statistical analysis AH and JŠ Design, writing and editing of manuscript AH, DP and JŠ

Yours faithfully,

Jan Šlapeta David Phalen The University of Sydney, The University of Sydney, Faculty of Veterinary Science Faculty of Veterinary Science Ph: +61293512025 Ph: +61293511798 [email protected] [email protected]

97 Appendix 1B

Faculty of Veterinary Science, University of Sydney, NSW Australia 2006

To Whom It May Concern,

We the undersigned are writing this letter to stipulate the role of Ashlie Hartigan in the preparation and submission of the following manuscript.

Hartigan. A, Fiala. I, Dyková. I, Jirků. M, Okimoto. B, Rose. K, Phalen. D and Šlapeta. J, (2011), A Suspected Parasite Spill-Back of Two Novel Myxidium spp. (Myxosporea) Causing Disease in Australian Endemic Frogs Found in the Invasive Cane Toad, PLoS One, vol.6 (4)

Ashlie Hartigan collected and performed all extractions and preparation of samples used in this manuscript during her PhD candidature.

Contribution to this manuscript by all authors is as follows: Task Role of Co-author Amplication, cloning and sequencing of genetic markers used for species identification AH, JŠ and IF Collection and submission of samples AH, MJ and BO Analysis of transmission and scanning electron microscopy material AH, IF and ID Examination of histological material AH, KR, DP Phylogenetic analyses AH, IF and JŠ Barcoding analysis JŠ Preparation, design and writing of the manuscript AH, DP and JŠ Response to reviewers and submission of manuscript AH and JŠ

Yours faithfully,

98 Appendix 1C

The University of Sydney, Faculty of Veterinary Science, NSW Australia, 2006

To Whom It May Concern,

We the undersigned are writing this letter to stipulate the role of Ashlie Hartigan in the preparation and submission of the following manuscript.

Ashlie Hartigan, during her PhD candidature, was responsible for collecting data, conducting analyses, writing the draft manuscript, responding to reviewer’s reports and coordinating submission and publication of the original research paper.

The individual roles of co-authors are listed below: Task Role of Co-author Collection and preparation of samples AH, KR and IF Scanning and transmission electron microscopy AH, ID, and IF AH, ID, KR and Review of histological material DNP DNA isolation and PCR AH and IF Phylogenetic analysis IF Preparation of figures used in manuscript AH, IF and JŠ Preparation, submission and editing of the manuscript AH, IF, DNP, JŠ

Yours faithfully,

99 Appendix 1D

The University of Sydney, Faculty of Veterinary Science, NSW Australia, 2006

To Whom It May Concern,

We the undersigned are writing this letter to stipulate the role of Ashlie Hartigan in the preparation and submission of the following manuscript.

Hartigan. A, Dhand. N, Rose. K, Šlapeta. J and Phalen. D.N, (2012), Comparative pathology and ecological implications of two myxosporean parasites in native Australian frogs and the invasive Cane toad, PLoS One, in press.

Ashlie Hartigan, during her PhD candidature, was responsible for collecting and examining material, performing the experiments, conducting analyses, writing the draft manuscript, responding to reviewer’s reports and coordinating submission and publication of the original research paper.

The individual roles of co-authors are listed below: Task Role of Co-author Collection and examination of histological material AH, KR and DP Diagnostic PCR experiments AH Statistical analyses AH and ND Design, preparation and editing of manuscript AH, ND, KR, JŠ and DNP

Yours faithfully,

Karrie Rose Australian Registry of Wildlife Health Navneet Dhand NSW, Australia The University of Sydney Ph: +612 9978 4749 NSW, Australia [email protected] Ph: +61 2 9351 1669 [email protected]

David Phalen Jan Šlapeta The University of Sydney, The University of Sydney, Faculty of Veterinary Science Faculty of Veterinary Science Ph: +61293511798 Ph: +61293512025 [email protected] [email protected]

100 Appendix 2: Summary of Genbank accession numbers

Accession Animal Tissue Organism number Species code isolated Region Clone I.D. name HQ822149 Rhinella marina 45.7 Myxospores partial 18S CTAUSC18 HQ822150 Rhinella marina 45.7 Myxospores partial 18S CTAUSC15 HQ822151 Litoria aurea 5A Liver partial 18S Direct seq. Cystodiscus HQ822152 Limnodynastes peronii 20F Myxospores partial 18S Direct seq. australis HQ822153 Litoria aurea 5A Liver partial 18S GGBFLI267 HQ822154 Litoria aurea 5A Liver partial 18S GGBFLI268 HQ822174 Litoria aurea 5A Liver partial LSU 7A04 HQ822175 Litoria aurea 5A Liver partial LSU 7A05 HQ822176 Litoria aurea 5A Liver partial LSU 7A07 Cystodiscus HQ822177 Litoria aurea 5A Liver partial LSU 7A06 australis HQ822178 Limnodynastes peronii 20F Myxospores partial LSU 7A12 HQ822179 Limnodynastes peronii 20F Myxospores partial LSU 7A13 HQ822180 Rhinella marina 45.7 Myxospores partial LSU CTAUS7A17 HQ822193 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFAH1 HQ822194 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMF20FP7 HQ822195 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMF20FP3 HQ822196 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMF20FP4 HQ822197 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB10 HQ822198 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB7 HQ822199 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB3 HQ822200 Litoria aurea 5A Liver p18S, ITS, 5.8, p28S GGBFIFE6 HQ822201 Bufo marinus 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB5 HQ822202 Litoria aurea 5A Liver p18S, ITS, 5.8, p28S GGBFIFE5 HQ822203 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFIFF6 HQ822204 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB9 HQ822205 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB2 HQ822206 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB1 HQ822207 Litoria aurea 5A Liver p18S, ITS, 5.8, p28S GGBFIF1 Cystodiscus HQ822208 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFITS4 australis HQ822209 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB4 HQ822210 Litoria aurea 5A Liver p18S, ITS, 5.8, p28S GGBFIFE3 HQ822211 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFIF2 HQ822212 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFIF10 HQ822213 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB8 HQ822214 Litoria aurea 5A Liver p18S, ITS, 5.8, p28S GGBFIFE4 HQ822215 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFIF5 HQ822216 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFIF3 HQ822217 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFB6 HQ822218 Litoria aurea 5A Liver p18S, ITS, 5.8, p28S GGBFIFE7 HQ822219 Litoria aurea 5A Liver p18S, ITS, 5.8, p28S GGBFIFE2 HQ822220 Litoria aurea 5A Liver p18S, ITS, 5.8, p28S GGBFIFE9 HQ822221 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFIF9 HQ822222 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFAH3

101 HQ822223 Limnodynastes peronii 20F Myxospores p18S, ITS, 5.8, p28S SMFAH2 HQ822163 Rhinella marina 45.7 Myxospores partial 18S CTAUSC14 HQ822164 Rhinella marina 45.7 Myxospores partial 18S CTAUSC16 HQ822165 Rhinella marina 45.7 Myxospores partial 18S CTAUSC20 HQ822166 Rhinella marina 45.7 Myxospores partial 18S CTAUSC17 HQ822167 Litoria peronii 29A Brain tissue partial 18S PTFP1 Cystodiscus HQ822168 Litoria raniformis 6985.6 Myxospores partial 18S SBFP3 axonis HQ822169 Litoria raniformis 6985.6 Myxospores partial 18S SBFP5 HQ822170 Litoria raniformis 6985.2 Brain tissue partial 18S SBFB1 HQ822171 Litoria raniformis 6985.6 Myxospores partial 18S SBFP1 HQ822172 Litoria raniformis 6985.2 Brain tissue partial 18S SBFB5 HQ822173 Litoria aurea 5A Brain tissue partial 18S direct JN977609 Litoria castanea 7868.1 Brain tissue partial 18S direct HQ822186 Litoria aurea 5A Brain tissue partial LSU GGBF7A02 HQ822187 Litoria aurea 5A Brain tissue partial LSU GGBF7A03 HQ822188 Litoria peronii 29A Brain tissue partial LSU PTF7A21 Cystodiscus HQ822189 Litoria peronii 29A Brain tissue partial LSU PTF7A18 axonis HQ822190 Litoria peronii 29A Brain tissue partial LSU PTF7A19 HQ822191 Rhinella marina 45.7 Myxospores partial LSU CTAUSP3 HQ822192 Rhinella marina 45.7 Myxospores partial LSU CTAUSP4 HQ822234 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFA6 HQ822235 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF7 HQ822236 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF9 HQ822237 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF8 HQ822238 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF5 HQ822239 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFA1 HQ822240 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF10 HQ822241 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFA3 HQ822242 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFA10 HQ822243 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF1 HQ822244 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFA5 HQ822245 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFIF9 HQ822246 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFIF7 Cystodiscus HQ822247 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFAHH1 axonis HQ822248 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFA4 HQ822249 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF6 HQ822250 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF2 HQ822251 Rhinella marina 45.7 Myxospores p18S, ITS, 5.8, p28S CTAUSIFA7 HQ822252 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF3 HQ822253 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFIF1 HQ822254 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFIF5 HQ822255 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFIF2 HQ822256 Litoria aurea 5A Brain tissue p18S, ITS, 5.8, p28S GGBFIFF4 HQ822257 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFIF4 HQ822258 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFIF10 HQ822259 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFIF6

102 HQ822260 Litoria peronii 29A Brain tissue p18S, ITS, 5.8, p28S PTFIF8 JN603234 Litoria raniformis 6985.2 Myxospores p18S, ITS, 5.8, p28S direct JN603235 Litoria raniformis 6985.6 Myxospores p18S, ITS, 5.8, p28S direct JN977605 Litoria castanea 7868.1 Liver p18S, ITS, 5.8, p28S direct JN977606 Litoria castanea 7868.1 Brain tissue p18S, ITS, 5.8, p28S direct JN977607 Litoria castanea 7908.1 Liver p18S, ITS, 5.8, p28S direct JN977608 Litoria booroolongensis 6175.1 Liver p18S, ITS, 5.8, p28S direct HQ822160 Bufo marinus M. immersum Myxospores partial 18S Mi1 Cystodiscus sp. HQ822161 Bufo marinus M. immersum Myxospores partial 18S Mi4 Brazil 1 HQ822162 Bufo marinus M. immersum Myxospores partial 18S Mi7 HQ822181 Bufo marinus M. immersum Myxospores partial LSU CTBRP2 Cystodiscus sp. HQ822182 Bufo marinus M. immersum Myxospores partial LSU CTBR9A01 Brazil 1 HQ822224 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ8 HQ822225 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ3 Cystodiscus sp. HQ822226 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ2 Brazil 1 HQ822227 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ7 HQ822155 Bufo marinus M. immersum Myxospores partial 18S Mi8 HQ822156 Bufo marinus M. immersum Myxospores partial 18S Mi3 Cystodiscus sp. HQ822157 Bufo marinus M. immersum Myxospores partial 18S 9A03 Brazil 2 HQ822158 Bufo marinus M. immersum Myxospores partial 18S 7A28 HQ822159 Bufo marinus M. immersum Myxospores partial 18S Mi6 HQ822183 Bufo marinus M. immersum Myxospores partial LSU CTBR7A26 Cystodiscus sp. HQ822184 Bufo marinus M. immersum Myxospores partial LSU CTBR7A27 Brazil 2 HQ822185 Bufo marinus M. immersum Myxospores partial LSU CTBRP5 HQ822228 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ5 HQ822229 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ4 HQ822230 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ9 Cystodiscus sp. HQ822231 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ1 Brazil 2 HQ822232 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ10 HQ822233 Bufo marinus M. immersum Myxospores p18S, ITS, 5.8, p28S CTBRIFJ6

103 Appendix 3: List of frog species identified with Cystodiscus spp. infection

Species Locality References Bufonidae Rhinella marina QLD and NSW Delvinquier (1986); Hartigan (2010;2011;2012) Myobatrachidae Limnodynastes ornatus QLD Delvinquier (1986) Limnodynastes peronii QLD and NSW Delvinquier (1986); Hartigan (2010;2011;2012) Limnodynastes tasmaniensis QLD Delvinquier (1986) Limnodynastes terraereginae QLD Delvinquier (1986) Mixophyes fasciolatus QLD Delvinquier (1986) Ranidella parinsignifera QLD Delvinquier (1986) Uperoleia laevigata QLD Delvinquier (1986) Hylidae Litoria caerulea QLD and NSW Delvinquier (1986); Hill (1997) Litoria dentata QLD Delvinquier (1986) Litoria fallax QLD Delvinquier (1986) Litoria latopalmatu QLD and NSW Delvinquier (1986) Litoriu lesueuri QLD and NSW Delvinquier (1986) Litoria nasuta QLD and NSW Delvinquier (1986) Litoria nyakalensis QLD Delvinquier (1986) Litoria peronii QLD and NSW Delvinquier (1986); Hartigan (2010;2011;2012a) Litoria revelata QLD Delvinquier (1986) Litoria rubella QLD Delvinquier (1986) Litoria tyleri QLD Delvinquier (1986) Litoria verreauxii QLD Delvinquier (1986) Litoria aurea NSW Hartigan (2010;2011;2012a, b, c) Litoria raniformis NSW and VIC Hartigan (2010;2011;2012a, d) Litoria booroolongensis NSW Hartigan (2010;2011;2012a, d) Litoria castanea NSW and VIC Hartigan (2012b, d)

Litoria gracilenta QLD and NSW Hartigan (2012d)

104