Use of Drosophila Models to Explore the Function of Asthma Susceptibility Genes

Asthma is a disorder of the airways (the tubes that carry air into the lungs), whose incidence is increasing worldwide. In the UK, asthma affects 1 in 11 children and 1 in 12 adults (Asthma UK). Asthma symptoms include wheeze, cough and shortness of breath caused by airways that narrow too much and too easily in response to harmless stimuli such as dust, cold air, air pollutants and pet dander. There is no cure for asthma. For many sufferers, current therapies help manage the symptoms of the disease, but for approximately 10% of patients with the severest disease current therapies are ineffective. There is an urgent need to identify the underlying basis of asthma, and to develop appropriate new treatment strategies.
The causes of asthma are not fully understood. While exposure to environmental triggers can bring on an asthma attack, it is known that asthma tends to run in families and may be inherited. Over the last 10-15 years, considerable progress has been made in identifying small changes in a number of genes that have been associated with inheritance of asthma. However, the functions of these genes (ie. how they work inside a cell) and/or how the small changes in their genetic code contribute to changes in their function that contribute to development of asthma remain poorly understood. Traditionally, scientists try to understand how a gene works by using 'animal models' (usually mice or rats) which allow them to investigate disease states in ways that would be inaccessible in a human patient. For example, they can deliberately delete or mutate the gene of interest and look at the consequence for the animal. However, such approaches are costly, time consuming and may not give any useful information because on the one hand, the animal may not show any difference or may give information that does not tell scientists about humans, and on the other hand, it may die as a result of the genetic modification. Therefore, there is a need to develop alternative approaches that replace the use of animals and reliably speed up the screening process, as well as offer potential for introducing several disease-genes into the same organism. This will allow assessment of how they interact together to cause the disease and how they influence responses to environmental triggers.
The fruit fly has proved to be a useful genetic tool for studying a number of common human diseases including neurodegenerative diseases like Alzheimer's, cancer, diabetes and cardiovascular disease. Insects possess a very simple airway system which shares many features with the human airway. Beside these characteristics, there are very efficient and easy methods available for altering genes in Drosophila, making this model so unique. This comprehensive toolbox of methods allows directed manipulation of any gene of interest (including human genes) with complete control in any cell type and at any time. Crucially, fly models can be developed and assessed in a relatively short period (within a month), so if the first approach doesn't work we will be able to make changes and study their effects very quickly. Therefore, we wish to determine whether 'fly models' can be used to replace animal models of asthma to study asthma susceptibility genes. We will demonstrate proof-of-principle by targeting two susceptibility genes called CDHR3 and ADAM33 where we have prior experience in animal and human cell systems, so enabling direct comparison with the fly system. By allowing screening of the function of key asthma genes, as well as providing a model for high through-put drug screening, we believe that this approach will greatly accelerate our understanding of the underlying genetic causes of asthma and help to develop new therapies that can control better the disease, or ultimately, prevent development of asthma.

We wish to test the hypothesis that Drosophila models can replace, augment and inform vertebrate asthma models by allowing first-pass elucidation of key asthma genes, as well as providing an in vivo model for rapid high through-put drug screening. We will demonstrate proof-of-principle by targeting two susceptibility genes, ADAM33 and CDHR3, where we have prior experience in mammalian systems, so enabling direct comparison with the invertebrate system.
We will use the GAL4/UAS (driver/effector) system to analyse gene function in Drosophila. Novel transgenic flies will be generated which express UAS-human sADAM33 and UAS-human sADAM33(E346A) (inactive control). These lines will be crossed with tub-GAL80ts : bin-GAL4 or tubGal80ts: ppk4-GAL4 to drive conditional expression of ADAM33 selectively in either hind-gut visceral muscle or tracheal branches, respectively of third instar larvae. The remodelling effect of sADAM33 on the visceral muscles will be assessed using high resolution light-sheet and confocal microscopy in anaesthetized living intact larvae. Other readouts will include assessment of epithelial barrier immunity, muscle mass around the airways or gut, collagen deposition and oxygen sensitivity.
In Drosophila DE-cadherin, Tyr586 corresponds to the human CDHR3 susceptibility mutation Cys529Tyr; therefore we will create a Tyr586Cys mutation in DE-cadherin in order to compare the two versions of the protein. We will also create transgenic lines expressing human CDHR3 Cys529 or its allelic variant Tyr529 in the larval trachea or gut, as described for ADAM33. Effects on epithelial barrier, immune responses and oxygen sensitivity will be assessed at baseline and after challenge with cigarette smoke or bacteria.
If tractable, such a system will facilitate studies of remodelling and barrier immunity in asthma. As we are developing small molecule inhibitors of ADAM33 this model would allow high throughput screening of the compounds.

The proposed project brings together asthma researchers (Davies, Haitchi, Collins) with experts in genetic manipulation of the fruit fly, Drosophila melanogaster, for modelling human disease (Mudher, Roeder). We aim to use our combined expertise to evaluate the potential of fruit flies to elucidate the function of two asthma susceptibility genes. We expect these models to help REPLACE the use of vertebrate models for asthma research and to ACCELERATE drug discovery and development programmes. By way of illustration of the bottleneck in understanding the function of asthma susceptibility genes, since its discovery in 2002, it has taken 12 years to dissect the functions of the asthma gene, ADAM33, and begin to assess its contribution to airway remodelling. Several factors affected our progress, especially the number of alternatively spliced variants of ADAM33 and a negative study from a US group in 2006 reporting that the Adam33 knock-out mouse lacked a phenotype. Our most recent studies have involved use of various murine models of allergic inflammation and development of transgenic (Tg) mice. However, our first our first Tg mice expressed high levels of ADAM33 mRNA but no protein, so we had to make new mice. We now have several ADAM33 transgenic lines (with differing expression levels of sADAM33 and with different isoforms of ADAM33), as well as the knock-out mouse (where our studies suggest that the original studies were not sufficiently powered to see significant differences in the KO mice). To date, we have used 1300 mice for our studies (founders, breeding, experiments). Going forward, we are proposing more mechanistic studies, as well as a drug screening programme which will involve around 1200 mice. Considering that more than 100 genes have been significantly associated with the pathogenesis of asthma, similar studies for every gene would involve more than 250,000 mice. These numbers would be based on comparison of a wild type and disease variant, although in many cases (as with ADAM33) multiple variants may need to be assessed and are likely to be studied independently in two or three labs, conservatively quadrupling our estimates to one million mice; furthermore, creation of double or triple transgenics to study gene-gene interactions will multiply to this number further. The latter two problems are much more tractable in flies where a variety of genetic tools are readily available for these purposes.
In making our proposal, we acknowledge that there are significant differences between the physiologies of flies and mammals. For this reason, we are taking a conservative approach, focusing on two genes for which data are available to allow validation of the fly models. The first of these genes, ADAM33 has been the focus of our research for more than 10 years and we were the first to describe its function in vitro and in vivo. CDHR3, which we were involved in identification last year, is the topic of ongoing research in our laboratory. As a consequence, we will be able to cross reference our discoveries in mammalian and fly systems and demonstrate the added value of the high through put genetic analyses using the invertebrate system. Assuming the models are successful, we would expect to disseminate comparative results through publication and by interaction with industry within 1-2 years. Wider uptake of this approach may be slower, as it will require expertise in development of fly models. However, this can be achieved through collaboration and training. We are involved in several asthma networks (eg. EU-COST) which could facilitate a Europe-wide asthma genetics - fly consortium.