The present invention relates in general to a method of screening for potential therapeutic agents for use in the treatment or prevention of chronic respiratory diseases, and in particular to the use of a Drosophila model in the screen.

Chronic respiratory diseases such as asthma and chronic obstructive pulmonary disease (COPD) represent major health, economic and societal burdens. The World Health Organisation (WHO) estimates that 235 million people worldwide currently suffer from asthma and has identified the condition as the most common non- communicable disease among children. The WHO further estimates that 65 million people have moderate to severe COPD and predicts that this disease will become the third major cause of death globally by 2030. Although asthma does not kill on the scale of COPD, it creates a substantial burden to individuals and their families and can restrict day-to-day activities for a lifetime.

Asthma and COPD are obstructive airway diseases characterised by chronic inflammation of the large and small airways and structural remodelling involving the epithelium, smooth muscle, microvasculature and extracellular matrix (ECM). Obstruction is usually intermittent and reversible in asthma, but is progressive and irreversible in COPD . However the two diseases may overlap and converge, especially in older people. Asthma management is primarily directed towards suppressing inflammation with inhaled corticosteroids (ICS) and relieving bronchoconstriction with bronchodilators. However, asthma prevention has not been achieved and, once established, there is no cure and there are currently no medications that can alter its natural history. The annual UK health care cost is estimated to be £2.5 billion, with those 10% of patients with severest disease accounting for -50% of the health costs. Bronchodilators are also the mainstay treatment for COPD and maintenance therapy of either a long-acting muscarinic antagonist or 2-agonist (LABA) alone or in combination with an ICS can be provided for those who remain symptomatic despite bronchodilator treatment. However, none of these drugs are able to reduce the progressive decline in lung function, which is the hallmark of COPD. The estimated cost of COPD to the UK economy is around £ 1.2 billion each year. Few new drugs for asthma and COPD have made it to the market in the last 50 years, despite massive investment by the pharmaceutical industry. Traditionally, there has been a high dependency on in vivo animal models of asthma and COPD allowing investigation of the physiological, biochemical and molecular relationships between the dynamics of the inflammatory processes and alterations in airway wall behaviour. A range of species including primates, cats, dogs guinea pigs and rodents have been used to investigate disease mechanisms. These approaches frequently involve procedures of moderate severity such as repeated allergen exposure, smoke-exposure or instillation of proteases (e.g. elastase), chemical irritants, dusts or lipopolysaccharide into the airways of the experimental animal. Although these types of models have contributed to our knowledge of basic mechanisms of lung inflammation/repair and have identified novel pathways and effector molecules, they have not translated into patient benefit. The cost of using such models to screen for potential new therapeutic agents is very high and very time consuming.

Recent technological advances, lower costs and access to large cohorts have resulted in systematic identification of around 200 genes associated with susceptibility to asthma or COPD . This raises the possibility that inheritance of different combinations of disease susceptibility genes underlies the clinical heterogeneity of asthma and COPD. While some susceptibility genes are associated with altered adaptive immune responses (e.g. the Th2 cytokine gene cluster on chromosome 5q3 1 -33 in asthma), many others are postulated to contribute to local tissue susceptibilities including epithelial (dys)function and innate immunity (e.g.ORMDL3, PCDH1 , CDHR3, MYH15, TMEM26, FOXA 1 , CCDC91 , TEKT5, IL-33 and its receptor, IL- 1RL 1) and/or airway remodelling and abnormal lung function (e.g. ADAM33, SGCD, PCDH9 and GALNT 13). However, identifying genetic variants represents just the first step in linking genes and disease . Understanding the mechanisms by which the gene is expressed, how it functions, how it is influenced by other genes, gene products and the environment are crucially important for the development of preventive, diagnostic, and therapeutic strategies. Nonetheless, for asthma and COPD, the explosion in knowledge at the genetic level, has not been matched by understanding of the precise function(s) of these susceptibility genes and/or how their genetic polymorphism contributes to airways disease, or to identification of new therapeutic agents. It is therefore an aim of the present invention to provide a non-mammalian model system, specifically the fruit fly, Drosophila melanogaster, for use in early drug discovery/target identification. It offers many advantages over more complex vertebrate systems, including conservation of basic cellular pathways, unparalleled genetic tractability, in vivo imaging capabilities, low rearing costs and culture conditions that are compatible with large scale screens - allowing high-throughput screening in a physiological context. Drosophila possess a very simple airway system and has been ignored as a model for studying susceptibility genes linked to airways disease due to the lack of an adaptive immune system which is considered important in modelling respiratory disorders, as well as the lack of smooth muscle around the trachea which undergoes structural remodelling in many chronic respiratory diseases . However, the data presented herein surprisingly shows that the innate immune system of Drosophila can reflect key functional characteristics of mammalian airway diseases, and that by expressing the mammalian genes in the gut of Drosophila, muscle remodelling can be observed. Similarly, an increase in epithelial permeability can be observed in the trachea or gut of Drosophila depending on where the mammalian genes are expressed. The observations were both surprising and unexpected, in particular the phenotypic effects observed when mammalian genes associated with respiratory diseases are expressed in the gut of Drosophila.

The present invention provides a genetically modified Drosophila line expressing one or more mammalian genes associated with a chronic respiratory disease .

The invention further provides a genetically modified Drosophila line carrying one or more mammalian genes associated with a chronic respiratory disease, wherein the one or more mammalian genes are capable of being driven to express the mammalian genes, preferably in a tissue specific manner. The genes may not be expressed, but may be capable of being expressed when the line is crossed with an appropriate driver line.

The one or more mammalian genes may be expressed, or be arranged to be expressed, in the trachea and/or gut of Drosophila larvae and/or adult Drosophila.

The one or more genes may be associated with susceptibility to asthma or COPD. The one or more genes may include, but is not limited to, the following: CDHR3, ADAM33, B4GALT5, DLG2, MAGIl, MAGI2, MAML3, PTTGIIP, HHIP, F AMI 3 A, FRMD4A, THSD4, Cl Oorfl l, LIMS1, GSDMA, GSDMB, ZPBP2, CLEC16A, AO AH, ZNF707, CLK3, C3AR1, SYNE I, LINGO 2, ABI3BP, NAF1, CSMD3, DPP 10, STARD3/PGAP3, STOML2, TRIM24, DAD I, FOXB1, FOXA1, ADAMTS9, AD AMI 9, KANSLI, TSEN54, TET2, RBM19/TBX5, NPNT, DEXI, ZBTB10, CTNNA3, SEMA3D, PCDH9, PCDH1, GALNT13, TMEM1 76A, PITPNC1, ACMSD, ZBTBI 6, ODZ3, CDHI 7, SGCD, MYH15, RAB27A, DDX1, COMMD10, EFEMP1, BMP6, PRDM11, WWOX, KCNJ2, PTCH1, SGK493, CA10, AGFG1, SPATS 2L, CELSR1, SOWAHB, TRAPPC9, ENSA, AK097794, ASTN2, LHX3, CCDC91, TBX3, RIN3, TEKT5, LTBP4, MN1, AP1S2, GPR126, GPR61, OLFM4, SEC16B, TNNI3K, CCDC37, MAP1B, TEM26, or a mutated form thereof. The gene may be CDHR3 or a mutated form or variant thereof. Alternatively, or additionally the gene may be ADAM33 or a mutated or variant form thereof. In particular the gene may encode CDHR3 with the mutation C529Y.

The CDHR3 protein may have the following protein sequence (Seq ID No: 1 and 2), the mutation C529Y is identified at position 529 where the mutation to Y is shown in brackets. Seq ID No : 1 is the wild type sequence with a C at position 529 and Seq ID No: 2 is the mutant sequence with a Y at position 529.

Seq ID No: 1 (and 2)

MQEAIILLAL LGAMSGGEAL HLILLPATGN VAENSPPGTS VHKFSVKLSA SLSPVIPGFP QIVNSNPLTE AFRVNWLSGT YFEVVTTGME QLDFETGPNI FDLQIYVKDE VGVTDLQVLT VQVTDVNEPP QFQGNLAEGL HLYIVERANP GFIYQVEAFD PEDTSRNIPL SYFLISPPKS FRMSANGTLF STTELDFEAG HRSFHLIVEV RDSGGLKAST ELQVNIVNLN DEVPRFTSPT RVYTVLEELS PGTIV ANITA EDPDDEGFPS HLLYSITTVS KYFMINQLTG TIQVAQRIDR DAGELRQNPT ISLEVLVKDR PYGGQENRIQ ITFIVEDVND NPATCQKFTF SIMVPERTAK GTLLLDLNKF CFDDDSEAPN NRFNFTMPSG VGSGSRFLQD PAGSGKIVLI GDLDYENPSN LAAGNKYTVI IQVQDVAPPY YKNNVYVYIL TSPENEFPLI FDRPSYVFDV SERRPARTRV GQVRATDKDL PQSSLLYSIS TGGASLQYPN VFWINPKTGE LQLVTKVDC(Y)E TTPIYILRIQ ATNNEDTSSV TVTVNILEEN DEKPICTPNS YFLALPVDLK VGTNIQNFKL TCTDLDSSPR SFRYSIGPGN VNNHFTFSPN AGSNVTRLLL TSRFDYAGGF DKIWDYKLLV YVTDDNLMSD RKKAEALVET GTVTLSIKVI PHPTTIITTT PRPRVTYQVL RKNVYSPSAW YVPFVITLGS ILLLGLLVYL VVLLAKAIHR HCPCKTGKNK EPLTKKGETK TAERDVVVET IQMNTIFDGE AIDPVTGETY EFNSKTGARK WKDPLTQMPK WKESSHQGAA PRRVTAGEGM GSLRSANWEE DELSGKAWAE DAGLGSRNEG GKLGNPKNRN PAFMNRAYPK PHPGK

The gene may encode the soluble form of ADAM33. The amino acid sequence of ADAM33 is detailed below (Seq ID No: 3 and 4) . The soluble form is underlined and the inactive variant containing a mutation from glutamic acid to alanine at position 346 is detailed. Seq ID No : 3 is the wild type sequence with an E at position 346 and Seq ID No: 4 is the mutant sequence with an A at position 346.

Seq ID No: 3 (and 4)

MGWRPRRARG TPLLLLLLLL LLWPVPGAGV LQGHIPGQPV TPHWVLDGQP WRTVSLEEPV SKPDMGLVAL EAEGQELLLE LEKNHRLLAP GYIETHYGPD GQPVVLAPNH TDHCHYQGRV RGFPDSWVVL CTCSGMSGLI TLSRNASYYL RPWPPRGSKD FSTHEIFRME QLLTWKGTCG HRDPGNKAGM TSLPGGPQSR GRREARRTRK YLELYIVADH TLFLTRHRNL NHTKQRLLEV ANYVDQLLRT LDIQVALTGL EVWTERDRSR VTQDANATLW AFLQWRRGLW AQRPHDSAQL LTGRAFQGAT VGLAPVEGMC RAESSGGVST DHSELPIGAA ATMAHE(A)IGHS LGLSHDPDGC CVEAAAESGG CVMAAATGHP FPRVFSACSR RQLRAFFRKG GGACLSNAPD PGLPVPPALC GNGFVEAGEE CDCGPGQECR DLCCFAHNCS LRPGAQCAHG DCCVRCLLKP AGALCRQAMG DCDLPEFCTG TSSHCPPDVY LLDGSPCARG SGYCWDGACP TLEQQCQQLW GPGSHPAPEA CFQVVNSAGD AHGNCGQDSE GHFLPCAGRD ALCGKLQCQG GKPSLLAPHM VPVDSTVHLD GQEVTCRGAL ALPSAQLDLL GLGLVEPGTQ CGPRMVCQSR RCRKNAFQEL QRCLTACHSH GVCNSNHNCH CAPGWAPPFC DKPGFGGSMD SGPVQAENHD TFLLAMLLSV LLPLLPGAGL AWCCYRLPGA HLQRCSWGCR RDPACSGPKD GPHRDHPLGG VHPMELGPTA TGQPWPLDPE NSHEPSSHPE KPLPAVSPDP QADQVQMPRS CLW

A genetically modified Drosophila line according to the invention may be used in an assay to screen agents that may be useful in the treatment of a chronic respiratory disease. A genetically modified Drosophila line according to the invention may be a

Drosophila melanogaster line.

The invention provides a method of screening for agents useful in the treatment of a chronic respiratory disease, the method comprising;

i) obtaining genetically modified Drosophila carrying one or more mammalian genes (transgenes) associated with a chronic respiratory disease; ii) administering an agent to the Drosophila larvae and/or adults;

iii) screening for one or more of:

a) a change in the innate immune response;

b) a change in epithelial permeability; and

c) a change in intestinal muscle structure or organisation; iv) identifying a compound or agent that affects one or more of a), b) or c) in step iii).

The mammalian gene may be referred to as a transgene once integrated into the Drosophila genome.

Preferably the mammalian gene/transgene is carried such that it may be expressed in a tissue specific manner. Preferably the Drosophila used express one or more mammalian gene/transgene in a tissue specific manner.

Preferably one or more mammalian genes associated with a chronic respiratory disease are expressed in the gut of the Drosophila. Alternatively, or additionally, the one or more mammalian genes associated with a chronic respiratory disease may be expressed in the trachea of Drosophila.

For initial large scale screens, Drosophila larvae may be employed as their size and transparency allows for plate-based assays utilising large compound libraries which may be added to the larval food. To allow visualization of responses in intact living animals, transgenic Drosophila lines carrying the mammalian genes may express the transgene or be crossed with driver lines to express the transgene. The driver line may also endogenously express reporters for antimicrobial peptide expression such as drosomycin-GFP (drs-GFP) to allow visualisation of cellular induction of immune response, or YFP-tagged collagen (Viking-YFP) to visualise hemocyte-mediated repair. To eliminate off target responses from this assay, positive hits may be tested in assays that directly measure epithelial barrier function (such as. Smurf assays using adult Drosophila or by immunostaining using antibodies against septate junction proteins in larvae where expression of the mammalian transgene is driven in the gut or trachea respectively).

The one or more genes may be associated with susceptibility to asthma or COPD . The one or more genes may be selected from, but is not limited to, the list comprising

CDHR3, ADAM33, B4GALT5, DLG2, MAGI1, MAGI2, MAML3, PTTG1IP, HHIP, F AM 13 A, FRMD4A, THSD4, Cl Oorfl l, LIMS1, GSDMA, GSDMB, ZPBP2, CLEC16A, AO AH, ZNF707, CLK3, C3AR1, SYNE1, LINGO 2, ABI3BP, NAF1, CSMD3, DPP 10, STARD3/PGAP3, STOML2, TRIM24, DAD1, FOXB1, FOXA1, ADAMTS9, AD AMI 9, KANSL1, TSEN54, TET2, RBM19/TBX5, NPNT, DEXI, ZBTB10, CTNNA3, SEMA3D, PCDH9, PCDH1, GALNT13, TMEM1 76A, PITPNC1, ACMSD, ZBTB16, ODZ3, CDH1 7, SGCD, MYH15, RAB27A, DDX1, COMMD10, EFEMP1, BMP6, PRDM11, WWOX, KCNJ2, PTCH1, SGK493, CA10, AGFG1, SPATS 2L, CELSR1, SOWAHB, TRAPPC9, ENSA, AK097794, ASTN2, LHX3, CCDC91, TBX3, RIN3, TEKT5, LTBP4, MN1, AP1S2, GPR126, GPR61, OLFM4, SEC16B, TNNI3K, CCDC37, MAP1B, TEM26, or a mutated form thereof. The gene may be CDHR3 or a variant or mutated form thereof and/or ADAM33 or a variant or mutated form thereof. In particular the gene may encode CDHR3 with the mutation C529Y. The gene may encode sADAM33 the soluble form of ADAM33.

Preferably the Drosophila is a Drosophila melanogaster.

The method of the invention may further comprise the step of driving expression of the mammalian genes in the Drosophila prior to administering the agent. Expression of the mammalian genes may be controlled by linking the mammalian transgenes to promoters that can be activated in a tissue specific manner. For example, expression of the mammalian genes may be limited to one or more of the trachea, the gut, the reproductive tract or the wings, preferably the trachea and/or the gut. Expression may be driven by tissue specific expression of an activator of the promoter linked to the mammalian gene. Tissue specific expression may be controlled by crossing a Drosophila line carrying the mammalian transgene with a Drosophila driver line that will drive expression of the transgene in a particular tissue . In step iii) the agent may be an organic compound, a protein, a peptide, an antibody, an oligonucleotide, an siRNA, an RNAi or a microRNA. An organic compound may be a C I to C I 00 compound, or a C I to C60 compound, or a C5 to C60 compound, or a C I to C50 compound, or a C5 to C50 compound, or a C I O to C60 compound, or a C I O to C50 compound. An organic compound for use in the method of the invention may contain less than 100 carbon atoms. An organic compound for use in the method of the invention may contain more than 10 carbon atoms.

The agent may be administered to the Drosophila in its food.

Preferably in step iii) a change in any of the characteristics can be readily observed. Examples of how a change may be screened for are given below.

Where one or more mammalian genes associated with a chronic respiratory disease are expressed in the gut of the adult Drosophila, a change in epithelial permeability may be determined by using a Smurf assay (Tricire and Rera, PLoS One. 2015 Nov 3 ; 10( 1 1)). This assay works where the presence of a mammalian gene associated with a chronic respiratory disease results in an increase in epithelial permeability, as occurs in adult Drosophila expressing CDHR3 C529Y in the gut. In such Drosophila, epithelial permeability is increased in the gut such that if the Drosophila is fed a coloured food stuff, for example a blue dye added to the food stuff, the dye leaks into the body cavity producing a Drosophila with a coloured, for example blue, body. The number of blue flies (ie. those with intestinal leakage) can be assessed by direct counting. By looking for changes in the extent of epithelial permeability, agents that may reduce this permeability may be identified. Such agents may then be studied further as possible potential therapeutic agents for use in the treatment of a chronic respiratory disease, such as asthma and/or COPD .

Where one or more mammalian genes associated with a chronic respiratory disease are expressed in the trachea of a Drosophila larva a change in epithelial permeability can be determined by screening for a change in the extent of disruption of tracheal epithelial septate junctions. Fluorescently labelled anti-coracle antibodies may be used to define the septate junctions, and allow junctional integrity or disruption to be observed by microscopy. By looking for a reduction in the extent of epithelial septate junction disruption, agents that may reduce the permeability of the epithelial barrier may be identified. Such agents may then be studied further as possible potential therapeutic agents for use in the treatment of chronic respiratory disease, such as asthma and/or COPD.

Where one or more mammalian genes associated with a chronic respiratory disease are expressed in the gut and/or in the trachea of Drosophila larva, a change in the innate response may be monitored by looking for a change in the level of hemocyte recruitment or induction of an innate immune response typified by induction of antimicrobial peptide (AMP) expression. These responses may be measured in a number of ways, including visualization in intact living larvae after crossing transgenic Drosophila lines carrying the mammalian genes with Drosophila driver lines. Examples of driver lines include those which express PPK-GAL4 or NP 1 -GAL4 which can drive expression in the gut or trachea respectively. The driver lines may also endogenously express (a) ds-Red-tagged hemocytes (eater-Ds-Red) to enable visualisation of hemocyte recruitment, or (b) drosomycin-GFP (drs-GFP) an antimicrobial peptide (AMP) that allows visualisation of cellular induction of immune response. Induction of AMP expression can be quantified in a plate based assay, whereas changes in hemocyte recruitment can be measured in a larval chip which immobilises the living organism for microscopy. The skilled person will readily appreciate that the visualisation marker (for example, DS-red and GFP) may be any suitable visualisation marker. By looking for a reduction in hemocyte recruitment or antimicrobial peptide expression, agents that may reduce the effects of or prevent respiratory diseases may be identified. Such agents may then be studied further as possible potential therapeutic agents for use in the treatment of a chronic respiratory disease, such as asthma and/or COPD . Where one or more mammalian genes associated with a chronic respiratory disease are expressed in the gut of the Drosophila larvae, a change in mesenchymal tissue remodelling may be determined by looking at the thickness and structural organisation of the gut visceral muscle. Expression of mammalian genes associated with a chronic respiratory disease may cause thickening and disorganisation of the Drosophila visceral muscle . If an agent can reduce or prevent this thickening or disorganising of the visceral muscle then it may be useful as potential therapeutic agent for use in the treatment of a chronic respiratory disease, such as asthma and/or COPD, and may accordingly be a target for further study. Induction of mesenchymal responses may be screened in intact living larvae after crossing transgenic Drosophila lines with appropriate driver lines, for example lines expressing NP 1 -GAL4. These lines may also express YFP-tagged collagen (Viking-YFP)) to visualise transgene-induced repair and remodelling or GFP-tagged visceral muscle (MHC-GFP) to assess the muscle thickness and organisation. Gut muscle remodelling may also be visualised and quantified after staining with phalloidin.

When the mammalian genes are expressed in the gut of Drosophila they may be expressed in gut enterocytes.

Any compound or agent identified in iv) may be a potential therapeutic agent for use in the treatment of a chronic respiratory disease. The agent may then be identified as a "lead" compound for further study and characterisation.

The method of the invention may be performed as a high throughput screen, with optional follow-up in functional assays.

The skilled person will understand that optional features of one embodiment or aspect of the invention may be applicable, where appropriate, to other embodiments or aspects of the invention. There now follows by way of example only a detailed description of the present invention with reference to the accompanying drawings, in which:

Figure 1- shows a schematic representation of human CDHR3 (green/upper representation) and Drosophila DE-cadherin (yellow/lower representation) showing the position of an asthma susceptibility mutation in human CDHR3

(Cys529Tyr) and the analogous residue in Drosophila DE-cadherin (Tyr586). KEY: EC: extracellular cadherin domain, CYTO: cytoplasmic domain, LMG: laminin G domain, EGFL: EGF-like domain, CPS : Cadherin proteolytic site domain. Figure 2A - E - Figure 2A shows the trachea from wt Drosophila larvae . Figure 2B shows the trachea from Drosophila larvae expressing CDHR3 WT (PPK4-GAL4>UAS-CDHR3-WT). Figure 2C shows the trachea from Drosophila larvae expressing CDHR3 C529Y (PPK4-GAL4>UAS-CDHR3- C529Y). In Figures 2A, B and C the tracheae were immunostained with anti- coracle antibodies to define the septate junctions (green) and cell nuclei stained with 4,6-Diamidino-2-phenylindole, dihydrochloride (DAPI, blue). The arrowheads illustrate junctional integrity and the disruption caused by the C529Y variant; the arrows identify increased numbers of cells identified as stem cells (see figure 4) and tubular narrowing caused by CDHR3 C529Y.

Figure 2D shows quantitation of stem cells. Figure 2E compares epithelial cells cultured from normal human airway and cells cultured from volunteers with asthma. The cells are stained with labelled antibodies for the tight junction proteins Occludin (green) and labelled antibodies for ZO- 1 (red) . From the lack of continuity of the staining in the asthma-derived culture, it is clear that the tight junctions are disrupted in the cells cultured from the airway of an asthma patient.

Figure 3 - shows the time to go from egg to pupae (the left end of the bar in the graph), and from pupae to adult Drosophila (the right end of the bar in the graph) in three scenarios, one expressing CDHR3 C529Y (PPK- GAL4>CDHR3 -mutant), one expressing WT CDHR3 (PPK-GAL4>CDHR3- WT 1) and finally a WT Drosophila (PPK-GAL4>W1 1 18). The data presented clearly shows that expression of WT or mutant CDHR3 has no effect on Drosophila development times or survival rate .

Figure 4 - shows stem cell expansion around the tracheal branches when the mutant CDHR3 C259Y (PPK->CDHR3 mutant) or WT CDHR3 (PPK-GAL4 CDHR3 WT) are expressed, the results are compared with control Drosophila (PPK>W1 1 18). Drosophila larval tracheal preparations were fluorescently stained to identify hemocytes (red) or stem cells (green). The stem cell compartment around the tracheal trunk and branches located between dentical bands 3 and 4 of the larvae were imaged and scored as small, medium or large. The data presented clearly show the presence of hemocytes and that expression of mutant CDHR3 dramatically increases stem cell number as a response to injury.

Figure 5 - shows the effects of WT or mutant CDHR3 expression in the gut of larval or adult Drosophila. In the top panel, RNA extracted from gut tissue expressing WT CDHR3 (NP- 1 -GAL4>UAS-CDHR3 WT) or mutant CDHR3(NP- 1 -GAL4>UAS-CDHR3 mutant) shows equivalent levels of expression whereas in control flies no CDHR3 is detected. In the middle panel, control (wild type Dah) adult Drosophila or those expressing CDHR3 WT or CDHR3 mutant were fed a diet containing the blue dye Erioglaucine

(2.5% wt/vol) for 9 hours each week and the time taken for development of the blue Smurf phenotype measured. Note, that in non-Smurf flies the dye is contained within the intestinal tract whereas the Smurf flies have a blue body. The data presented clearly shows that expression of mutant CDHR3 dramatically increases intestinal permeability as evidenced by leakage of the blue dye into the body of the flies from day 27. In the lower panel, the induction of the anti-microbial peptide drosomycin (DRS) in the gut of third instar larvae was assessed by RT-qPCR. Expression of DRS is doubled in flies expressing mutant CDHR3 suggesting increased permeability of the gut barrier, even at the larval stage.

Figure 6 - (A) shows human ADAM33 proteins (full length and the soluble sADAM33 construct where the letters indicate the component exons). Alignment of ADAM33 with Drosophila orthologues showing low sequence homology. (B) shows the effect of expressing sADAD33 in Drosophila larval gut. The visceral muscle in gut tissue from third instar larvae were stained fluorescently using phalloidin (red). In the WT flies, the circular and longitudinal fibres are clearly organised in a regular fashion whereas sADAM33 causes the circular muscles to become disorganised and pulled together in regions where the longitudinal muscles are disrupted.

Figure 7 shows visceral muscle in gut tissue from third instar larvae stained fluorescently using phalloidin (red). Upper panel shows gut from a WT larva (A) or with ectopic expression of PGRP-LC (B). Lower panel shows tissue from larvae expressing CDHR3 WT (C) or C529Y (D). Transgene expression was driven in the gut enterocytes using NP 1 -GAL and actin in larval visceral muscle labelled with phalloidin (red). In the WT flies or those expressing CDHR3 WT, the circular and longitudinal fibres are clearly organised in a regular fashion whereas the circular muscle is thicker and more closely spaced in the gut of the larvae expressing PGRP-LC (in which innate immune signalling is constitutively activated) or mutant CDHR3.

Materials and Methods Fly husbandry and Transgenics

Drosophila melanogaster was be reared on standard agar media and housed in environmentally controlled chambers at 25 ^° C, under a 12h light/dark cycle as described in Cowan, C. M. et al. Scientific reports 5 , 17191 , 2015. Females expressing either the tracheal epithelial-specific driver PPK-GAL4 or gut epithelial-specific driver NPY-GAL4 (Bloomington Stock Centre) were crossed with males bearing attp- insertions of CDHR3 wild-type or CDHR3-disease variant allele (CDHR3 C529Y) . Transgenic Drosophila were generated using the insertion site-specific attp methodology and standard Drosophila germ line transformation transgenesis techniques. The CDHR3 wild-type and disease variant stocks can be recombined with a line harbouring a Drosmycin: : GFP reporter insertion to enable tracking of induction of immune response as described in Wagner, C et al FASEB journal 23, 2045-2054, 2009.

Barrier dysfunction assay: Smurf assay

Following the NP 1 driver targeted expression of the CHDR3 WT or mutant variant to the Drosophila gut, adult females were reared on a diet containing a blue dye (Erioglaucine disodium salt purchased from Sigma Aldrich). Dyed medium was prepared using standard medium with dye added at a concentration of 2.5% (wt/vol) as described in Rera et al, PNAS, 109, 21528-21533, 2012. At weekly time points, adult females were exposed to dyed media for 9h before being returned to their normal media. The escape of blue dye from the gut into the body cavities, after return to normal diet, was taken to indicate breakdown of gut barrier integrity as described in Rera et al, PNAS, 109, 21528-21533, 2012. Barrier dysfunction leading to gut remodelling

Gut remodelling was investigated by assessing organisation, architecture and thickness of smooth muscle tissue in hind gut after targeted expression of sADAM33 or CDHR3 wt and the disease-variant in adult flies. This was ascertained by immunostaining using anti-actin (phalloidin) as described in Buchon et al, BMC Biol 8, 152, 2010. The sADAM33 or CDHR3 wt and disease variant lines can also be recombined with a transgenic line expressing GFP-tagged visceral muscles (called MHC-GFP - Wang et al, Biology open 4, 1753- 1761 , 2015) to allow for in vivo investigations of muscle remodelling.

Immunohistoschemical investigation of barrier disruption and consequent induction of immune response

Following targeted expression of CDHR3 wt or disease associated variant to trachea or gut, epithelial barrier integrity in these tissues was ascertained using immunofluorescence microscopy to characterise epithelial barrier proteins such as coracle and discs large as described in Buchon et al, BMC Biol 8, 152, 2010. To ascertain whether expression of the transgenic human proteins causes induction of immune responses, RNA was extracted from isolated larval trachea or gut tissue and expression of Drosomycin measured by RT-qPCR using primers specific for this AMP (Thermofisher, product code DmO 1822006 $ 1 ). Alternatively, a drosmycin: :GFP reporter line can be used to monitor expression in these tissues in situ. In this case, the disease variants should be recombined to this line prior to gut or tracheal targeted expression. Induction of GFP signal can be ascertained in situ in larval tracheal or gut tissue as described in Wagner et al FASEB journal 23, 2045-2054, 2009. Stem cell and hemocyte infiltration into the tracheal branches will be characterised by immunohistochemical staining (phosphohistone H3 or NIMC l respectively) or by recombination with a hemocyte reporter line (eater-ds-Red).

In vivo larval studies

Gut smooth tissue remodelling and induction of GFP-drosomycin can be conducted in situ in wandering third instar larvae . Wandering third instar larvae are anaesthetised in etherised chambers for 10 mins and then immobilised in agarose coverslips. Their tracheal branches or gut organisation can be visualised using fluorescence microscopy as described in Quraishe et al, Molecular psychiatry 18, 834-842, 2013. Results

In order to demonstrate the ability of Drosophila to act as a model organism for use in a screen for potential therapeutic agents for use in the treatment of a chronic respiratory disease, such as asthma or COPD, a variety of different genetically modified Drosophila lines were produced and analysed.

Epithelial permeability

In one embodiment Drosophila were produced which expressed the mammalian CDHR3 gene in the trachea (see the materials and methods above). Drosophila lines were produced that expressed either WT CDHR3 or mutant CDHR3 which carries the mutation C529Y. Details of the protein are given in Figure 1 , and the differences when compared to the naturally occurring Drosophila protein DE-cadherin are highlighted. Figure 1 shows the presence of a tyrosine at 586 equivalent to tyrosine 529 of the asthma variant of CDHR3. However, it should be noted DE-cadherin is much bigger, having four additional extracellular domains that are not present in CDHR3. Therefore it would be anticipated that the human gene may not function in the fly. CDHR3 was identified in 2014 as a susceptibility locus for early childhood asthma with severe exacerbations; however its normal biological function is unknown. CDHR3 belongs to the cadherin superfamily. The prototypic and best-characterized member of the cadherin superfamily is epithelial E-cadherin, which has an extracellular domain that is involved Ca ²⁺ -dependent cell-cell adhesion, a single transmembrane domain and a cytoplasmic domain that interacts with the underlying actin cytoskeleton. E-cadherin is localised in adherens junctions and these are required for the formation of tight junctions. Like E-cadherin, CDHR3 is strongly expressed in the airway epithelium which has essential barrier functions and plays a key role in innate immunity. Immunohistochemical analysis of bronchial biopsies has shown altered localisation of E-cadherin and tight junction proteins in the airway epithelium of asthmatic individuals. E-cadherin is down-regulated by TNFa and other cytokines including IL- Ι β leading to barrier disruption and release of pro- allergic mediators including TSLP and activation of NFkB . However, in studies using fully differentiated bronchial epithelial cultures in vitro, untreated cultures from asthmatic donors retain features of the in vivo phenotype including reduced tight junction formation (see Figure 2E which shows that tight junctions are disrupted in asthma sufferers), suggesting a heritable trait. Furthermore, functional analyses of these cultures have revealed increased permeability and sensitivity to environmental challenges suggesting that deficient barrier function may contribute to asthma pathogenesis.

Adherens junctions, with homologous cadherins, are expressed in both Drosophila trachea and human airways. From sequence alignments, endogenous Drosophila DE- cadherin has 46% similarity with human CDHR3 and a tyrosine at amino acid 586 corresponding to the human CDHR3 susceptibility mutation at position 529 (Figure 1 ) . Using a Drosophila model in which CDHR3 C529Y is expressed in the trachea of Drosophila larvae, disruption of septate junction integrity can be observed (Figure 2) . Expression of CDHR3 C529Y induces a number of morphological changes including narrowing of the tracheal tube structure, disruption of the localisation of coracle (cor) and discs-large (dig) proteins in epithelial septate junctions and expansion of stem cells in the tracheae (cells responsible for repairing and regenerating damaged tissue (Figure 2C,D). This is not due to non-specific effects of expressing a human gene because expression of wild-type CDHR3 had little effect on barrier homeostasis (Figure 2B). Coracle and dig proteins are components of septate junctions, structures that serve as selective-permeability barriers, separating the apical from the basal regions in sheets of epithelial cells in Drosophila. Septate junctions restrict paracellular solute movement and are the invertebrate functional counterparts of tight junctions in vertebrate epithelia. As it has been shown that epithelial tight junctions are disrupted in asthma (Xiao et al, J Allergy Clin Immunol 201 1 ; 128 :549-556), the findings using Drosophila may provide a genetic explanation for the phenotypic observations in asthma. The data also suggest that CDHR3 C529Y has functional effects on the epithelium that are relevant for asthma, especially the activation of stem cells involved in responses to epithelial injury. Similar epithelial injury and repair responses are evident in asthma (Loxham M, Davies DE. J Allergy Clin Immunol. 2017).

By screening for agents that affect the disruption of septate junctions caused by the expression of CDHR3 C529Y in the trachea of Drosophila larvae, potential therapeutic agents may be identified for further analysis. Preferably potential therapeutic agents will reduce the disruption to tight junctions.

In an alternative embodiment, CDHR3 C529Y is expressed in the gut of Drosophila (flies are produced as described above in the materials and methods). In this case, expression of CDHR3 C529Y in the larval gut causes an increase in antimicrobial peptide expression suggestive of an increase in gut barrier permeability (Fig 5, lower panel) . This was confirmed in adult flies in which gut epithelial barrier permeability was measured in a Smurf assay where the flies are given food containing a blue dye . If gut barrier permeability is increased, the dye can pass from the gut into the body cavity, causing the flies to become blue, as illustrated with CDHR3 C529Y (Figure 5 middle panel). This assay provides direct evidence that CDHR3 C529Y affects epithelial barrier integrity. Innate Immunity

In flies, the respiratory organ is purely epithelial and these barrier cells have the capacity to mount an immune response to pathogens involving a single pathway converging on NF-κΒ activation, a key transcription factor in airways disease. Drosophila is also a valuable model for studying sterile inflammation as a consequence of cellular damage with recruitment of immune cells involved in wound healing. This also involves activation of conserved pathways including caspase, JNK and JAK/STAT with relevance for airways disease . Mechanistic studies in Drosophila have been accelerated through combining live-imaging and genetic tools allowing quantitative 4D studies in vivo that are impractical in mammals. These tools may be used to study genes (such as CDHR3 and sADAM33) associated with mammalian respiratory diseases in Drosophila.

As described above with reference to Figure 2, expression of CDHR3 C529Y in Drosophila trachea cells causes disruption of tracheal epithelial septate junctions and focal activation of stem cells in the trachea. It is believed that CDHR3 C529Y expression has disturbed the functional integrity of the epithelium and has triggered a local inflammatory wound healing/tissue repair response, reflecting the abnormal injury /repair phenotype observed in airway epithelium of asthmatic subjects. Transgenic Drosophila expressing UAS-CDHR3 (WT and C529Y) and ADAM33 (WT and E346A) genes have been produced. These lines are viable and have been crossed with UAS/GAL4 driver lines PPK-GAL4 and NP 1 -GAL4 to target expression of the genes to larval tracheal branches (PPK-GAL4) or gut enterocytes (NP 1 -GAL4) respectively. By crossing these lines with GAL80ts (TARGET) or GENESWITCH driver variants to incorporate temperature or drug-induced spatial control of CDHR3 or sADAM33 expression the effect of CDHR3 (WT and C529Y) and sADAM33 (WT and E346A) on the structure and function of the tracheal network and on wound healing responses may be observed. To allow this to be visualized in real time in intact living animals, transgenic lines are crossed with one of three recombined PPK- GAL4 tracheal driver lines that also endogenously express either a) ds-Red-tagged hemocytes (eater-Ds-Red) to enable visualisation of hemocyte recruitment, b) drosomycin-GFP (drs-GFP) to allow visualisation of cellular induction of immune response, or YFP-tagged collagen (Viking-YFP) to visualise hemocyte-mediated repair. For in vivo and ex vivo imaging, wandering third instar larvae are anaesthetised and mounted in agarose chambers or larval chips to reveal their tracheal branches. Following induction of CDHR3 or sADAM33 transgenes, the temporal sequence of immune signalling (drs-GFP), infiltration of hemocytes (eater-Ds-Red) and ECM deposition (Viking-YFP) around these structures will be visualised in vivo in real time using high-resolution light microscopy and confocal microscopy techniques established in the Mudher lab (Sinadinos C et al. Neurobiol Dis 2009; 34:389-395 ; Quraishe S et al. Mol Psychiatry 2013 ; 18 : 834-842). The architecture of the trachea (thickness, ECM and actin organisation, basement membrane changes) is analysed comprehensively using immunohistochemistry and high resolution confocal microscopy. These readouts are linked to routinely used functional assays that reflect compromised tracheal or gut barrier function including reduced tracheal gas filling, burrowing behaviour or gut "leakiness" measured in a ' Smurf assay (Tricoire H et al. PLoS ONE 2015) which are conducted in embryonic-stage, third instar larvae or adults. These studies will provide behavioural readouts for the morphological changes induced by the expression of the asthma-susceptibility gene, thus enabling dissection of pathogenic effects from molecular, to cellular and organismal level.

Mesenchymal Remodelling

As Drosophila do not have muscle around the tracheal tubes, the potential to study mesenchymal remodelling responses in the larval gut was studied. The Drosophila gut consists of a simple epithelium surrounded by visceral muscles, nerves and tracheal branches. In preliminary studies, the use of this system was validated by expressing sADAM33 which causes an increase in bronchial smooth muscle in mice (Davies et al JCI Insight 2016 Jul 21 : 1 ( 1 1)) and a hypercontractile phenotype in rats (Duan et al Exp Cell Res 2016 Nov 15 : 349( 1): 100- 1 18). In the Drosophila larval gut, sADAM33 expression caused the muscle to become disorganised with the circular muscle appearing to be pulled together in regions where the longitudinal muscles are disrupted, suggesting a contractile phenotype. As inflammation can lead to airway remodelling in lung disease, the use of the system was also assessed by activating the immune deficiency (IMD) pathway involving NF-κΒ by targeted overexpression of the peptidoglycan receptor gene, PGRP-LC in larval enterocytes using the gut specific NP 1 -GAL4 driver. Preliminary experiments indicated that ectopic activation of the IMD pathway results in thickening (or contraction) of the visceral muscle and disordering of its organisation compared with control, as determined by actin staining with phalloidin staining (Figure 7B vs 7A). Similarly, when CDHR3 C529Y was expressed in the enterocytes, a thickening and disordering of the visceral muscle was observed (Figure 7D), but there was little effect of WT CDHR3 (Figure 7C). Together these data point to a specific effect of the CDHR3 C259Y asthma variant on the visceral muscle, possibly due to gut epithelial barrier disruption driving activation of the IMD pathway.

To characterise in detail the effects of CDHR3, and other asthma genes such as sADAM33, on visceral smooth muscle responses and ECM deposition, transgenic CDHR3 or sADAM33 lines will be crossed with one of two recombined PPK-GAL4 tracheal or NP 1 -GAL4 or 5053A-GAL4 gut driver lines that also endogenously express GFP-tagged visceral muscles (called MHC-GFP) or YFP-tagged collagen (Viking-YFP) (Bunt S et al. Dev Cell 2010; 19:296-306; Altincicek B et al. Dev Comp Immunol 2008). This enables visualisation of visceral muscle responses in real time, using in vivo and ex vivo imaging approaches. For example, wandering third instar larvae can be anaesthetised and then immobilised in agarose coverslips so that their tracheal branches or gut organisation can be visualised using fluorescence microscopy. Contraction or migration of muscle cells around gut or trachea may be visualised in vivo following induction of CDHR3 and the involvement of hemocytes analysed either by staining for them in dissected tissue (as in Figure 2), or by visualising the real time infiltration of RFP-tagged cells. The proliferation and architecture (length, thickness and actin organisation) of the visceral smooth muscle and ECM may be analysed comprehensively using immunohistochemistry and high resolution confocal microscopy. Involvement of the JAK/STAT pathway in the visceral muscle response may be investigated using a JAK/STAT reporter gene, STAT-GFP58.

By screening for agents that affect the thickening and disordering of the visceral muscle caused by the expression of CDHR3 C529Y in the gut of Drosophila larvae, potential therapeutic agents may be identified for further analysis. Preferably potential therapeutic agents will reduce the thickening and disordering of the visceral muscle.

Application of Drosophila Models for Compound Screening

Many of the screens used in the present the invention are capable of being reproducibly and quantitatively undertaken in a 96 well assay - and are capable of developing in a high throughput format.

Conclusion

In summary, the data presented herein demonstrates that it is possible to express genes associated with human airway diseases in Drosophila and to observe disease-relevant effects. As this invertebrate model is amenable to quantifiable dissection of pathogenic mechanisms at molecular, cellular and functional (behavioural) levels, it has the potential to substantially replace rodent models in screening for potential therapeutic agents for use in the treatment or prevention of chronic respiratory diseases, such as asthma or COPD.