Cane toads

Bio-control approaches to cane toad control

A.J. Robinson 1, A. Hyatt 3, J. Pallister 3, N.H.R. Hamilton 1 and D.C.T. Halliday 1

1 CSIRO Entomology, GPO Box 1700 , Canberra , ACT 2601

2 CSIRO Livestock Industries, Australian Animal Health Laboratory, Private Bag 24, Geelong , Victoria 3220.

Abstract

 Cane toads (Bufo marinus) were introduced into Australia in 1935 to control native beetles that were damaging sugar cane crops. At the time, many other countries were importing the toads for insect control. In Australia they become a pest in their own right mainly due to their toxicity to native predators but also as competitors for food and refuge sites. A means for their control has been on the agenda for a number of decades but to date no effective methods have been found. The most effective means of control would be a bio-control agent similar to the viruses that were introduced into Australia to control rabbits. Between 1986 and 1996 a determined effort was made to identify cane toad-specific bio-control agents both in Australia and overseas but no suitable agent was discovered. It is now considered unlikely that such off-the-shelf agents will be found and as a consequence other approaches are being explored. One such approach being carried out at the CSIRO is to engineer a virus to interfere with metamorphosis. All amphibians express new proteins during metamorphosis and these are potential “vaccine” antigens or RNAi targets for delivery to tadpoles. The challenge is to ensure that the antigens or RNAi molecules are effective and cane toad specific. Progress toward these goals will be presented.

Introduction

Cane toads (Bufo marinus) were introduced into Australia in 1935 to control native beetles that were damaging sugar cane crops. At the time, many other countries were importing the cane toads for insect control. In Australia , they were unsuccessful in controlling sugar cane beetles and they began their move out of the cane fields to become a pest in their own right. This was mainly due to their toxicity to native predators but also as competitors for food and refuge sites (for a comprehensive review of cane toad introductions see Lever (2001) and for a short review of impacts see Phillips et al. (2003 )).

A means to control cane toad populations has been on the agenda for a number of decades but to date no effective methods have been found. The most effective means of control would be a bio-control agent similar to the viruses that were introduced into Australia to control rabbits. Between 1986 and 1996 a determined effort was made to identify cane toad-specific bio-control agents both in Australia and overseas, however no suitable agent was discovered. It is now considered unlikely that such off-the-shelf agents will be found and as a consequence other approaches are being explored. One approach is the construction of a bio-control agent that would affect metamorphosis or other life stages of the cane toad. Another is the possibility of developing a cane toad-specific toxin.

CSIRO Cane Toad Bio-control Project

In 1969, a paper appeared in the journal, “Science” describing the effects of inoculating bullfrog (Rana catesbiana) tadpoles with haemoglobin extracted from the blood of adult bullfrogs (Maniatis et al. 1969) . The result was the death of a number of tadpoles as they metamorphosed and those that survived produced no adult or tadpole haemoglobin but instead produced a variant haemoglobin that appeared to be neither adult nor tadpole. This was the first demonstration that exposure of tadpoles to adult proteins not expressed until metamorphosis could interfere with gene expression during metamorphosis and causing an auto-immune response. The interference was presumed to be mediated by antibodies. It is now known that a wide range of proteins are up-regulated during metamorphosis (Amano 1998, Brown et al. 1996) and those not expressed until metamorphosis would potentially be treated as foreign antigens by tadpoles.

Our current approach is to see if we can exploit this phenomenon for the control of cane toads. Interference with metamorphosis leading to either death or the weakening of individuals would be one way of reducing the size of cane toad populations to give our native fauna an opportunity to survive the invasion of this highly toxic animal. To be successful, a practical and efficient means of delivering proteins to tadpoles would need to be found. One way to do this would be using a virus that was cane toad specific. This could either be self-disseminating or self-limiting, although a strategy using a self limiting virus would restrict its use to controlling toads on a local scale only. Our strategy is to identify potential target genes that if interrupted would compromise metamorphosis. The particular gene or genes would be inserted into the virus and when expressed in the animal would produce an antibody response that would affect metamorphosis (Figure 1). This is a long term strategy that cannot deliver a rapid solution to the current cane toad problem and we envisage that it will take ten years to develop. We are about to enter the fifth year of that development in partnership with The Australian Department of Environment and Heritage through the Natural Heritage Trust.

Figure 1. Schematic diagram of bio-control approach for cane toads

Progress to date has been good. We have repeated the bullfrog experiments in cane toads using the β-chain of adult cane toad haemoglobin. In this case, the animals survived through metamorphosis but the interesting and surprising result was the significant reduction in adult haemoglobin mRNA detected in the young toads and a persistence of larval haemoglobin mRNA (N. Siddon, S. Tarmo & A.J. Robinson, unpublished data). One would expect to see a reduction in protein levels but not necessarily mRNA. This suggests some form of feedback at the transcriptional level. We have yet to measure the antibody response and the levels of proteins in these animals.

Although haemoglobin is a good model protein to test our ideas, high conservation of the molecule across amphibia raises concerns of species specificity and thus would not be useful in a bio-control approach. Ultimately, we are looking for a protein or part of a protein that is cane toad specific. In parallel to the haemoglobin testing we are searching for other genes that are expressed for the first time during metamorphosis. We have identified 14 such proteins, four of which have been identified by others and the remaining ten by us using a powerful technique called micro-array technology. Using this technology one can look for differences in gene expression across a large number of genes on a single glass slide. Using another technique called semi-quantitative PCR, our results confirm that the 10 identified genes have little or no mRNA expression in tadpoles (Figure 2). The next stage is to see if these genes exist in native frogs and to see if they differ enough to be candidate target genes. Those that look promising will be tested for their effects against cane toads.

Figure 2. Expression analysis of cane toad genes up-regulated during metamorphosis. Semi-quantitative RT-PCR results show mRNA expression levels within four developmental stages (A = tadpole; B = tadpole with developed hind legs; C = toadlet with tail; D = toadlet). The early activation of a number of genes (I), in comparison to the late activation of others (II) is clearly illustrated. The third grouping (III) reveals those genes with a faint signal detected in the early tadpole stage (A). Cane toad b -actin was used as the house-keeping control gene.

On the virus front we have chosen a ranavirus to test the concept. Ranaviruses will infect a range of amphibians, and even fish, so unless such viruses can be attenuated to cause no disease and ensuring that the inserted anti-metamorphosis gene is species-specific, they would not be acceptable bio-control agents. However, they are useful to see if delivery of the anti-metamorphosis gene by a virus will cause the same effect as inoculation with protein alone. We have been successful in attenuating the virus for this purpose and also have developed protocols to enable genes to be inserted into the viral genome. We have inserted the cane toad haemoglobin gene into the virus and are currently testing for expression of the protein. If expression can be detected we will then infect cane toad tadpoles under contained laboratory conditions to see if we can alter the course of metamorphosis (J. Pallister & A. Hyatt, unpublished data).

Ultimately, to build a bio-control agent that is species-specific, one would ideally need to use a virus that is species-specific and not just rely on the species-specificity of the target gene. Species-specific viruses of vertebrates that can be engineered to contain genes without losing their ability to replicate on their own are confined to the large DNA virus families. Poxviruses, adenoviruses and herpesviruses have all been engineered and many are species-specific or at least have a very limited host range. Of these three, only one virus from each of the adenovirus and herpesvirus families have been isolated from amphibians and both of these have been found in the leopard frog (Rana pipiens). Genomic sequence is available for both viruses (Davison et al., 1998, Davison et al., 2000) , which provides a basis for PCR tests in a search for adenoviruses or herpesviruses in cane toads. A herpesvirus (Bennati et al. 1994) has been identified by electron microscopy of lesion material in the agile frog (Rana dalmatina) from the UK, however no virus isolation has been reported. These findings suggest that members of these viral families may be present in other amphibia including cane toads.

Toxin Approach

Toxicity

When thinking about how one might control a pest species, an early stage in the process is to develop an understanding of how the pest operates and survives in its environment, and then see if there are any aspects of its behaviour or physiology that could make it vulnerable. An obvious strategy for survival for the cane toad is its ability to produce a powerful toxin that is capable of killing any predator that consumes it. All members of the Bufo genus produce toxin but in different amounts in different species (Meyer & Linde 1971) . Of all the species, Bufo marinus produces one of the highest amounts of toxin per gram of body weight. The toxins produced by Bufo are synthesised de novo rather than being derived from their diet and are chemically related to digitalis.

Cardiac effect

Digitalis is the name given to extracts of the leaves of the foxglove Digitalis purpurea. The medicinal and toxic properties of digitalis have been known for centuries, however it is only in relatively recent times that the precise chemical nature of digitalis has been determined (Molero et al. 2000) . It turns out that digitalis is a complex mixture of chemicals known as cardiac glycosides or cardiotonic steroids (Figure 3). As the name suggests these compounds act on the heart causing increased strength of heart beat (inotropic effect) in therapeutic doses, but cardiac fibrillation and death in toxic doses. Many other plant species contain such compounds (Molero et al. 2000) .

The cane toad and other Bufo species produce venoms that contain over fifty cardiotonic steroids grouped into two categories; bufadienilides and bufotoxins (Chen and Kovarikova 1967, Meyer & Linde 1971) (Figures 3B, C & D). The bufadienolides, also known as bufogins, are the more toxic per weight of compound and lack the suberylarginine group of the bufotoxins. Homologues of the bufadienilides are found in plants where they are usually conjugated with carbohydrate moieties. Bufadienolides have also been found in fireflies (Photinus sp.) and snakes (Rhabdophis sp.) (Steyn & van Heerden 1998) .

Figure 3. Basic chemical structures of cardiotonic steroids. A. Digitalis-like compounds. B and C. Bufadienilides. D. Bufotoxins. (From {Steyn, 1998 #47}). (These images were drawn using WINPLT v. 7.1.11. PLT is free shareware. Password request submitted to reich@chem.wisc.edu 21/06/05 ).

The cardiotonic steroids act by binding to the Na +,K +-ATPase, the so-called sodium pump. The Na +,K +-ATPase is a protein composed of two sub-units present in the membrane of all cells and controls the levels of intracellular K + and Na + and, secondarily, Ca 2+. The sub-units are designated a and β, and there are different isoforms of each chain. Various combinations of the a and β chain isoforms are associated with different tissues. A γ subunit has also been described that can influence the function of the pump (McDonough et al. 2002, Yu 2003) .

The main region that influences binding of the cardiotonic steroids to the Na +,K +-ATPase has been determined using the sheep ATPase as a model (Price & Lingrel 1988) . It is the first extracellular loop (designated (M1-M2) of the a chain. The a chain is approximately 1,000 amino acids in length and is predicted to have ten transmembrane domains (Kaplan 2002, Lingrel & Kuntzweiler 1994) (Figure 4). The region is 12 amino acids in length and the amino acids at each end of this domain are critical for binding. The most significant change in resistance recorded in mutagenesis studies of the Na +, K +-ATPase in the sheep was the alteration of glutamine to an aspartate in the first amino acid position and an asparagine to an arginine in the last. This conferred a 50,000 fold resistance (Price et al. 1990) . Other amino acid changes in the a chain can also affect binding in the range of 2 to 1,000 fold.

Figure 4. Schematic diagram of the a subunit of the sheep Na +,K +–ATPase (Redrawn from {Lingrel, 1994 #60})

The amino acid sequence of the cane toad Na +,K +-ATPase derived from bladder epithelium has been determined (Jaisser et al. 1992) . Not surprisingly, the cane toad molecule (albeit with eight transmembrane domains predicted rather than ten) has a modified M1-M2 domain with the first amino acid being an arginine and the second a lysine. This confers a 500 fold resistance compared to, for example, the unmodified sheep Na +,K +-ATPase (Jaisser et al., 1992) . The cane toad is therefore resistant to its own toxins. Other species that have Na +,K +-ATPases found to be resistant to the cardiotonic steroids are certain leaf eating beetles that feed on plants containing cardiac glycosides (Labeyrie & Dobler 2004) and the rat (Price & Lingrel 1988) .

There have also been studies to determine what components of the cardiotonic steroid are involved in the binding to the sodium pump (Farr et al. 2002, Molero et al. 2000, Thomas et al. 1990) .

Identification and cane toad control

So what is the connection between these observations and cane toad control? The first question one could ask is can we exploit these findings to discover or design a toxin that would bind to the cane toad Na +,K +-ATPase but not that of other species? One possibility would be to identify a source of digitalis-like compounds and screen them for activity against cane toads. Due to their potential value in human medicine, a wide range of compounds with digitalis-like activity have been screened for their inotropic activity. If one could identify a repository of such compounds, a screening strategy, probably using a cell culture system could be devised. Another approach could be to model the potential binding of modified digitalis-like structures to the cane toad Na +,K +-ATPase binding domain.

A third approach could be to identify a naturally occurring endogenous ligand of the cane toad Na +,K +-ATPase and exploit this as a toxin. The Na +,K +-ATPase binding site is highly conserved across vertebrate species except for those species in which it is modified to render it resistant. This has given rise to the suggestion that endogenous ligands exist for the site and act as a sodium pump regulatory mechanism (see comment by Mahon & McKenna 2000 ). Other endogenous ligand/receptor pathways with regulatory roles have been postulated, and later discovered, by observing the effects of plant compounds on humans and later discovered, opioids being a well known example. Searches for such endogenous ligands of the Na +,K +-ATPase in humans have identified cardiotonic steroids similar or identical to ouabain (a digitalis-like compound) and marinobufagenin (a compound identical to one produced by the cane toad) (Manunta & Ferrandi 2004) . It is postulated that these compounds are produced in the adrenal cortex and are sodium pump regulators but there is still debate about the significance of these findings.

If indeed such regulatory molecules exist in humans and other animals, then how does the cane toad, or any other animal with a resistant sodium pump for that matter, regulate its Na +,K +-ATPase? Either it produces a regulatory molecule with a different structure that binds to the modified M1-M2 region or it has dispensed with this regulatory mechanism altogether and relies on other mechanisms of control. Other means of Na +,K +-ATPase regulation have been demonstrated (summarised in Yu (2003) ) and it is likely that for such an important molecule there will be redundancy in the control system. However, if the cane toad has a modified M1-M2 ligand then the discovery of that molecule could the basis for a species-specific toxin.

Acknowledgements

The cane toad bio-control project is funded in part by the Natural Heritage Trust through the Australian Department of Environment and Heritage. We acknowledge the earlier major contribution to this project by Dr Nicole Siddon and Mr James Alderman and also the invaluable technical assistance of Suze Tarmo, Daryl Venables, John Bray, Sarah Goldie, Rhonda Voysey and Veronica Olsen.

References

Amano T. (1998). Isolation of genes involved in intestinal remodelling during anuran metamorphosis. Wound Repair and Regeneration 6: 302-313.

Bennati R., Bonetti M., Lavazza A. & Gelmetti D. (1994). Skin lesions associated with herpesvirus-like particles in frogs (Rana dalmatina). Veterinary Record 135: 625-626.

Brown D.D., Wang Z., Furlow J.D., Kanamori A., Schwartzman R.A., Remo B.F. & Pinder A. (1996). The thyroid hormone-induced tail resorption program during Xenopus laevis metamorphosis. Proceedings of the National Academy of Sciences USA93: 1924-1929.

Chen K.K. & Kovarikova A. (1967). Pharmacology and toxicology of toad venom. Journal of Pharmaceutical Sciences 56: 1535-1541.

Davison A.J., Sauerbier W., Dolan A., Addison C. & McKinnell R.G. (1998). Genomic studies of the Lucke tumor herpesvirus (RaHV-1). GenBank Accession number AF110004. Date accessed 15/6/05.

Davison A.J., Wright K.M. & Harrach B. (2000). DNA sequence of frog adenovirus. Journal of General Virology81: 2431-2439.

Farr C.D., Burd C., Tabet M.R., Wang X., Welsh W.J. & Ball W.J. Jr. (2002). Three-dimensional quantitative structure-activity relationship study of the inhibition of Na(+),K(+)-ATPase by cardiotonic steroids using comparative molecular field analysis. Biochemistry 41: 1137-1148.

Jaisser F., Canessa C.M., Horisberger J.D. & Rossier B.C. (1992). Primary sequence and functional expression of a novel ouabain-resistant Na, K-ATPase. The beta subunit modulates potassium activation of the Na, K-pump. Journal of Biological Chemistry267: 16895-16903.

Kaplan J.H. (2002). Biochemistry of Na, K-ATPase. Annual Reviews of Biochemistry 71: 511-535.

Labeyrie E. & Dobler S. (2004). Molecular adaptation of Chrysochus leaf beetles to toxic compounds in their food plants. Molecular Biology and Evolution 21: 218-221.

Lever C. (2001). "The Cane Toad: The History of A Successful Colonist". Otley, West Yorshire, Westbury Publishing.

Lingrel J.B. & Kuntzweiler T. (1994). Na+,K(+)-ATPase. Journal of Biological Chemistry269:19659-19662.

Mahon N. & McKenna W.J. (2000). Digoxin-like immunoreactive substances in hypertrophic cardiomyopathy. European Heart Journal 21: 1034-1036.

Maniatis G.M., Steiner L.A. & Ingram V.M. (1969). Tadpole antibodies against frog haemoglobin and their effect on development. Science 165: 67-69.

Manunta P. & Ferrandi M. (2004). Different effects of marinobufagenin and endogenous ouabain. Journal of Hypertension22:257-259.

McDonough A.A., Velotta J.B., Schwinger R.H., Philipson K.D. & Farley R.A. (2002). The cardiac sodium pump: structure and function. Basic Research in Cardiology97 (suppl. 1): 19-24.

Meyer K. & Linde H. (1971). Collection of toad venoms and chemistry of the toad venom steroids. [In] W. Bücherl & E.E. Buckley (eds.) "Venomous animals and their venoms". New York, Academic Press. pp. 521-556.

Molero C.P., Medardi M. & San Feliciano A. (2000). A short review on cardiotonic steroids and their aminoguanidine analogues. Molecules5: 51-81.

Phillips B.L., Brown G.P. & Shine R. (2003). Assessing the potential impact of cane toads on Australian snakes. Conservation Biology 17: 1738-1747.

Price E.M. & Lingrel J.B. (1988). Structure-function relationships in the Na, K-ATPase alpha subunit: site-directed mutagenesis of glutamine-111 to arginine and asparagine-122 to aspartic acid generates an ouabain-resistant enzyme. Biochemistry27: 8400-8408.

Price E.M., Rice D.A. & Lingrel J.B. (1990). Structure-function studies of Na, K-ATPase. Site-directed mutagenesis of the border residues from the H1-H2 extracellular domain of the alpha subunit. Journal of Biological Chemistry 265: 6638-6641.

Steyn P.S. & van Heerden F.R. (1998). Bufadienolides of plant and animal origin. Natural Product Reports 15: 397-413.

Thomas R., Gray P. & Andrews J. (1990). Digitalis: its mode of action, receptor, and structure-activity relationships. Advances in Drug Research 19: 311-562.

Yu S.P. (2003). Na+, K+ -ATPase: the new face of an old player in pathogenesis and apoptotic/hybrid cell death. Biochemical Pharmacology 66: 1601-1609.