Establishment of consensual husbandry protocols for laboratory Xenopus laevis using novel physiological and behavioural techniques.

The african clawed frog, Xenopus laevis, is widely used for scientific research purposes and is maintained in large numbers in laboratories worldwide. Hundreds of thousands of these animals have contributed to an enormous range of scientific research in genetics, developmental, cell and molecular biology since the 1940s. They are famous for their role in early pregnancy testing work, were the first vertebrate animal to be cloned and yet little work has ever examined what the most appropriate care and housing might be for them. This, however, is vital to ensure the best possible conditions for ensuring their (often 15+ years-long) laboratory lives are optimally healthy and happy. Since happy, healthy animals also make the best scientific subjects, it is of paramount importance that this neglected area of animal welfare is addressed. Over the decades so-called 'best practice' (BP) husbandry guidelines have been developed that identify key parameters for appropriate care but little consensus exists on what exactly are the best ways of maintaining these animals. Some sources, for example, suggest tanks should have a water depth of 5cm, other say 'no more than 50cm'; a huge difference to a frog a few cm long. Some authors suggest group housing of up to 10 individuals, others say 100 in a group is fine. Truthfully we have no evidence that either is appropriate. BP guidelines identify that enrichment objects, e.g. providing a plastic tube as a refuge to hide in, should be provided, as for other lab animals. Some keepers suggest this achieves no benefit, however, but others find it reduces bites amongst tank-mates.
A major reason why there is little data in this area is that, until now, there has been no easy way to measure whether particular housing conditions, or the presence of enrichment object, is associated with more or less stress. Amphibians have a 'stress hormone' similar to that released by mammals but it is hard to get samples from them which yield hormones for analysis. Whilst a human can be asked to chew a cotton-bud for saliva, or a monkey trained to urinate into a cup for a sample, frogs are rather less amenable. The few studies that have analysed frog stress have had to resort to taking blood samples, which involve a needle at least, or urine samples, which involve squeezing the animal or rubbing the delicate skin. All of these are likely to be stressful and therefore impact on the stress measures the researchers are trying to assess. Our research has developed a genuinely non-invasive technique for measuring the frog stress hormone, corticosterone. We have already shown it works for a smaller, close relative of Xenopus laevis, and been able to measure the hormone and associated behaviours it shows under more or less stressful conditions. This is the first time this has been done for an amphibian. Our proposed work will develop the technique so we can measure the stress hormone in Xenopus under different housing conditions. We will then be able to establish, conclusively, which conditions produce the least stress to these animals and recommend these to Xenopus keepers in laboratories worldwide. We will also establish a comprehensive, detailed description of all the behaviours shown by each sex in this species (also currently lacking). Once identified we will measure which behaviours, their frequencies and durations, are associated with different stress hormone measurements. In this way, a Behavioural Stress Score (as has been used for cats, and we have developed for horses) can be created which aligns certain levels of behaviour with certain levels of stress. This can be used to regularly monitor behaviour and therefore welfare, without the need for costly biochemical analysis of hormone samples every time. Taken together, the work we propose will significantly improve the welfare of this neglected lab species by comprehensively refining their husbandry recommendations and leading to great reductions in animal use.

The African Clawed frog, Xenopus laevis, is an extremely widely used laboratory animal but comprehensive, consensually developed Best Practice (BP) guidelines for housing and husbandry in this species are lacking. Our work will develop an entirely novel, non-invasive, no-contact biochemical assay for measuring corticosterone (stress hormone) in tank water for X. laevis. We have recently developed and fully validated these methods for Hymenochirus boettgeri (close relative of X. laevis) and used them to detect significantly higher levels of corticosterone release in a more 'stressful' condition compared to a less 'stressful' one. Development of this technique for use with X. laevis will enable non-invasive physiological stress assessment in this species for the first time. The assay would be used to refine the huge variation in a range of key current X. laevis husbandry recommendations by experimentally investigating tank water depth and volume ranges; water change frequency; water temperature and pH ranges and lighting regime to identify conditions associated with lower corticosterone release. A fully comprehensive ethogram of X. laevis behaviour (also lacking) will be compiled. The behavioural responses of animals will then be directly quantified for different enrichments (physical, social and their interactions) and these correlated with concomitant corticosterone measurements to evaluate these for husbandry use. From this we will develop a novel, non-invasive Behavioural Stress Score (BSS) for this species by using our data to outline a putative scale of behaviours known to be associated with varying degrees of corticosterone release. Comprehensive discussion with X. laevis users (through invited workshops) will enable us to test the potential of this BSS for wide use as a welfare monitoring tool. This work will enable direct refinement of husbandry/housing for this animal and reduction in animal use through improved welfare monitoring and increased re-use.

Objective 1: We will develop novel biochemical methods to measure levels of the major stress hormone, corticosterone in the tank water of Xenopus laevis. This non-invasive technique allows us to refine the way in which corticosterone is measured in X. laevis since current methods involve either puncture, whole body homogenisation, cloacal insertion or body squeezing. Since our techniques yield experimental corticosterone titres that are not masked by sampling stress we will produce data sets that are more robust and have less variability. This will reduce the number of animals required for a statistically significant effect. As non-invasive sampling is comparatively stress free, we expect our techniques to increase the research life of a subject, (thus reduce animals required) by allowing repeated sampling from a single individual and reducing the negative health consequences of stress.
Objective 2: Our original non-contact, non-invasive tool for measuring corticosterone enables assessment and optimisation of current husbandry recommendations for Xenopus, for which there is currently little consensus. We will be able to provide, for the first time, a set of husbandry guidelines for X. laevis that are based on robust physiological data. By defining husbandry protocols that minimise stress we are clearly refining the captive environment of this species as well as providing standardised housing conditions (within and between laboratories) which will reduce the variability of experimental data and reduce the numbers of animals required.
Objective 3: The research will construct a comprehensive ethogram for X. laevis so behaviour can be subsequently used as a non-invasive, cost-effective tool to quantify welfare in X. laevis.
Objective 4: There is no consensus on appropriate physical enrichment or group sizes for X. laevis. Behaviour will be combined with physiological indices to assess for the first time, the impact of physical enrichment and group sizes and the interaction between the two. Behaviours found to be significant predictors of corticosterone in the latter studies can be advocated as non-invasive indicators of stress.
Objective 5: We propose hosting two workshops for key Xenopus personnel aimed at discussing our experiments with established expertise and developing a behaviour stress score (BSS) for application to X. laevis, using behaviours shown to be correlates of physiological stress. The BSS can be used to monitor welfare and refine the environment in the scientific, conservation and education arena in the absence of costly, technical biochemical analyses.
Objective 6: Our experimental work will allow, for the first time, compilation of a set of husbandry modifications for X. laevis, based on quantitative, complementary physiological and behavioural data. These will be widely disseminated in a timely fashion to pertinent bodies, as proposed in the 'Communications Plan'. Once implemented, they will result in significant refinement of maintenance protocols for this species. This will have a wide impact by producing physically and psychologically healthy animals for research, education and conservation.
Our quantitative techniques allow us to assign real numbers to the 'extent of refinement' experienced by the individual. We will be able to compute differences in levels of corticosterone produced by different conditions. The predicted reduction in experimental variability can be measured by comparing the coeffecience of variation across our studies. The comprehensive ethogram will enable us to quantify changes in the frequencies of behaviours across conditions. The research is expected to lead to an overall reduction in the numbers of animals required for any one experiment. Our data can be used to compute the scale of this reduction by conducting power analysis to ascertain a sample size required for an effect based on data collected from subjects experiencing optimal versus less optimal conditions.