A successful raid on a fortress requires ingenious strategies in addition to a large number of soldiers. When a microorganism faces a potential host many factors are important, including not only the capacity to proliferate but also the ability to hide, escape or subvert the defence arsenal of the infected organism. This ability confers microbial pathogenicity and relies on complex virulence mechanisms, which are tightly regulated during the course of the infection. The amazing versatility of some microbes that can infect a wide broad of hosts undoubtedly relies on virulence factors intent on fighting evolutionarily conserved innate immune mechanisms. This makes the use of alternative invertebrate models, which are of outstanding interest because they demand less ethical consideration and lower experimental costs, extremely relevant. These simpler organisms are used to analyse genes and mechanisms involved in resistance or tolerance to microorganisms. They can also be used to study bacterial virulence factors that allow proliferation or persistence in the host. In particular, the Drosophila fruit fly has a complex immune response (similar to the mammalian innate immune response) and is particularly appropriate for deciphering many events underlying bacterial pathogenicity from acute virulence to biofilm formation. As highlighted in this review, Drosophila has been notably extensively used to study virulence traits of the opportunistic bacteria Pseudomonas aeruginosa, such as proliferation or persistence, translocation through an epithelial barrier, subversion of the phagocytic machinery, in vivo biofilm formation and enhanced virulence provided by commensal flora or a polymicrobial community. Moreover, these small flies now appear to be a useful system for assaying chemicals with therapeutic potential.
The adaptability of some pathogens to various hosts is a fascinating aspect of the microbiological world, particularly considering that an intimate host–pathogen relationship must occur that can shape the co-evolution of specific virulence factors and host defence mechanisms in the long term. One of the possible reasons underlying this versatility might be a common ancestral defence shared by multicellular organisms. Host invasion requires the ability to cross physical barriers, escape phagocytic machinery and resist the microbicidal activity of secreted proteases, reactive oxygen species or antimicrobial peptides. These inherent protective and defence processes take part in the innate immune system, which is well conserved among eukaryotes (Medzhitov and Janeway, 1998). Consequently, pathogenic microorganisms in humans (bacteria, fungi or viruses) may employ similar set of virulence mechanisms when confronted with mammalian and non-mammalian hosts. While this paradigm is now well accepted, evidence for real host–pathogen interplay in simple model systems arose due to the accumulation of many experimental data through the years (Boman et al., 1972; Mahajan-Miklos et al., 2000; Kurz and Ewbank, 2007). This remarkable observation opened a tremendous number of opportunities for scientific exploration of host–pathogen interactions using invertebrate models such as the amoeba Dictyostelium discoideum, the worm Caenorhabditis elegans or the fruit fly Drosophila melanogaster, to name a few. Using these model systems allows for the infection of large numbers of individuals to conduct genetic screens and efficient statistical analyses with few or none of the practical and ethical restrictions that apply to vertebrate models. Using this approach, Jander and collaborators showed a correlation between Pseudomonas aeruginosa virulence factors used to infect mammals and insects that was stronger than those used to infect worms or plants (Jander et al., 2000). This observation, combined with the deep knowledge of the fly immune response and its similarity to the mammalian innate immune system, makes D. melanogaster a particularly attractive model to investigate the virulence of pathogenic bacteria (Shirasu-Hiza and Schneider, 2007; Apidianakis and Rahme, 2011; Bier and Guichard, 2012).
D. melanogaster: a genetic model possessing a conserved immune response
In their natural environment, all eukaryotes are permanently confined to the microbiological world. For example, D. melanogaster feeds and lays eggs on rotten fruits, which are infested with microbes. Historically, investigation of the biochemical and molecular mechanisms underlying insect resistance to microbes led to identification and purification of inducible antimicrobial peptides that are secreted in the insect haemolymph and the subsequent description of the complex regulatory events controlling their production by the fat body cells in Drosophila (Lemaitre and Hoffmann, 2007).
From the immunologist's point of view, infecting flies with any type of bacteria, fungus or virus, pathogenic or not, has been used as an easy tool to analyse the host defence response. Use of these methods led to the remarkable discovery of the evolutionarily conserved NF-κB like Toll and Imd signalling pathways, which trigger the expression of humoral stress and antimicrobial defence genes (Lemaitre and Hoffmann, 2007). A large number of additional studies established the role of phagocytosis in bacterial clearance by macrophage-like cells derived from the haemocyte lineage, elucidated the role of prophenoloxidase and clotting factors in counteracting microbe invasion and replication, or highlighted the existence of an integrated defence response at the level of epithelial intestinal cells, which includes production of reactive oxygen species and the capacity to renew damaged intestinal cells via stem cell proliferation and differentiation (Lemaitre and Hoffmann, 2007; Ryu et al., 2010; Fauvarque and Williams, 2011; Buchon et al., 2013; Kim and Lee, 2014).
From the microbiologist's point of view, Drosophila is being increasingly used to determine bacterial virulence factors which are essential for multiplication and/or persistence in the host, to monitor the degree of pathogenicity of various clinical isolates or mutant strains, to evaluate the mechanism of action of bacterial toxins and to test for bioactive molecules against bacteria (Shirasu-Hiza and Schneider, 2007; Vallet-Gely et al., 2008; Bier and Guichard, 2012).
To add a level of complexity, manipulating both bacterial and fly genomes allows us to assess host–pathogen interaction processes by searching for host genes required to counteract specific bacterial virulence factors. Conversely, we can also identify bacterial virulence factors required to neutralize specific aspects of the immune response. Such studies have been successfully conducted in combinatorial analysis of bacterial and host mutants with entomo-pathogens (which may exceptionally infect humans) such as Pseudomonas entomophila (Opota et al., 2011; Chakrabarti et al., 2012), Erwinia carotovora (Acosta Muniz et al., 2007; Quevillon-Cheruel et al., 2009) and Serratia marcescens (Nehme et al., 2007), and are actively pursued with an increasing number of pathogens causing problematic humans infections, such as P. aeruginosa (D'Argenio et al., 2001), Listeria monocytogenes (Mansfield et al., 2003), Staphylococcus aureus (Needham et al., 2004), Salmonella typhimurium (Brandt et al., 2004), Streptococcus pneumoniae (Pham et al., 2007), Mycobacterium sp. (Dionne et al., 2003), Vibrio cholerae (Blow et al., 2005; Berkey et al., 2009), Burkholderia sp. (Castonguay-Vanier et al., 2010; Pilatova and Dionne, 2012) and Francisella sp. (Vonkavaara et al., 2008; Ahlund et al., 2010; Akimana et al., 2010) (reviewed in Shirasu-Hiza and Schneider, 2007; Vallet-Gely et al., 2008).
Beyond the capacity to infect and potentially kill flies, important aspects to be taken into consideration when analysing the virulence of a microorganism are the type of metabolic disorder and pathogenesis induced in the fly with regard to human-related diseases (Shirasu-Hiza and Schneider, 2007). For instance, Mycobacterium marinum replicates in fly haemocytes in a manner similar to Mycobacterium tuberculosis replication in human phagocytic cells (Dionne et al., 2003), and V. cholerae induces fly weight loss and other characteristic features of human infections, ultimately causing fly death (Blow et al., 2005). Importantly, death may result from the combinatorial effect of toxic bacterial effectors and damage from the host immune response (Ferrandon, 2013). In contrast, host tolerance mechanisms and/or bacterial persistence strategies favour host survival despite increasing the bacterial burden. In a screen specifically designed to identify host tolerance factors to S. typhimurium, Shinzawa and colleagues showed that p38 MAPK overexpression triggers sequestration of the bacteria in enlarged haemocytes in a process called ‘phagocytic encapsulation’ (Shinzawa et al., 2009). Interestingly, a mutated strain (SPI-2) that was impaired in the secretion of type III toxic effectors similarly survived in wild-type haemocytes and caused delayed fly mortality while proliferating at higher rates than the fully virulent parental strain (Brandt et al., 2004; Shinzawa et al., 2009).
Focusing on the use of Drosophila for the discovery and understanding of human bacterial pathogen virulence and persistence, in this review, I wish to emphasize the pioneer work performed using P. aeruginosa, while the reader is referred to other excellent reviews for detailed information on other host–pathogen mechanisms described for various types of bacterial species (Shirasu-Hiza and Schneider, 2007; Vallet-Gely et al., 2008; Bier and Guichard, 2012; Buchon et al., 2013).
P. aeruginosa-induced fly pathogenesis
In humans, P. aeruginosa is a Gram-negative opportunistic pathogen and a major cause of nosocomial infections. It mainly infects immune-compromised and cystic fibrosis patients. Its pathogenicity is multifactorial and relies upon surface-associated (fimbriae, flagella, …), secreted virulence factors (toxins, proteases, …) and toxins directly injected into the host cell cytoplasm through the type III secretion system (T3SS) (Galan and Collmer, 1999; Yahr and Wolfgang, 2006). Expression of these virulence factors is tightly regulated and determines whether the bacterium adopts a planktonic (highly virulent) or a sessile (biofilm) lifestyle, which are commonly associated with acute and chronic infections respectively. Many virulence factors are co-ordinated by quorum sensing (QS), a form of bacterial cell-to-cell communication depending on the concentration of diffusible molecules that directly correlates to cell density (Williams and Camara, 2009). The two QS signal inductor and receptor pairs, LasI–LasR and RhlI–RhlR, are interdependent systems based on secretion of N-acyl-homoserine lactone (acyl-HSL) molecules, while the QS LysR-type transcription factor MvfR depends on hydroxy-2 alkylquinolines molecules (HAQs) (Parsek and Greenberg, 2000; Cao et al., 2001).
Drosophila is highly susceptible to P. aeruginosa infections, and as few as 10 to 50 bacteria introduced into a fly can induce death by septicaemia in 1 day (D'Argenio et al., 2001; Fauvarque et al., 2002; Lau et al., 2003). Use of GFP-expressing bacteria revealed that the infection is restricted to the point of inoculation for the first 12 h and is then followed by widespread bacterial invasion causing tissue damage and septicaemia. Alternatively, Drosophila can be orally infected by P. aeruginosa contaminated food, causing death in 2 to 15 days (or even more) depending on the time of exposure and the method of infection: death kinetics typically depends on whether or not flies are subjected to starvation prior to the infection and/or to concomitant food restriction (Erickson et al., 2004; Avet-Rochex et al., 2007; Limmer et al., 2011). When provided orally, wild-type P. aeruginosa rapidly crosses the intestinal barrier to reach and proliferate in the haemolymph compartment, most likely causing death from bacteraemia as reported in three independent studies (Limmer et al., 2011; Mulcahy et al., 2011; de Bentzmann et al., 2012). An intriguing aspect of oral infection described in these studies is that despite high bacterial titres in the haemolymph (eventually reaching more than thousands cells per fly) death occurs in several days. These data imply that the bacteria fail to induce toxic virulence factors or that flies became tolerant to elevated numbers of bacteria in their haemolymph, putatively via a phagocytic encapsulation mechanism as described for Salmonella infections (Shinzawa et al., 2009). Additional work is however needed to clarify this apparent paradox in the case of P. aeruginosa.
Whether haemolymph bacterial invasion is due to gut damage or to active bacterial passage through intact epithelial cells is still an open question. Apidianakis and collaborators (Apidianakis et al., 2009) reported extensive stem cell renewal in the gut (a marker for enterocytes replacement and tissue repair) after ingestion of the P. aeruginosa PA14 human isolate. This proliferative state favours intestinal dysplasia in a sensitized genetic context (Apidianakis et al., 2009). Damage to the epithelial and muscle layers surrounding the crop (a storage organ situated just at the end of the foregut, Fig. 1A) has also been observed after oral ingestion of the laboratory strain PAO1 (Sibley et al., 2008; Mulcahy et al., 2011). Meanwhile, Limmer and collaborators failed to detect any gut damage when feeding flies PA14 (Limmer et al., 2011). Some discrepancies may result from variability among various Pseudomonas strains or differences in the procedure used to infect the flies, including the use of exponential or stationary phase culture or, as mentioned earlier, the time period and methods used (associated or not with starvation) to provide contaminated food to the flies. Finally, various commensal flora in different laboratories may also impact the severity of the outcome of the infection (see also below).
Screens of P. aeruginosa mutants, in either injured or orally infected flies, have highlighted the involvement of twitching motility factors (D'Argenio et al., 2001; Potvin et al., 2003), QS regulators (Lau et al., 2003), and a number of either previously known or newly discovered QS regulated genes, in full virulence towards Drosophila (Chugani et al., 2001; Lau et al., 2003; 2005; Erickson et al., 2004; Kim et al., 2008; Lutter et al., 2012). Moreover, we showed the importance of T3SS for full virulence of the cystic fibrosis isolate CHA and identified new regulators of QS or T3SS, which are commonly required for full virulence of PAO1 in amoebae, mice and orally infected flies (Fauvarque et al., 2002; Alibaud et al., 2008). Additional studies have nicely corroborated the essential contribution of Pseudomonas QS in fly killing by using small chemicals or secreted bacterial molecules. Indeed, co-inoculation of bacteria with QS inhibitory molecules slowed down bacterial growth in flies (Cady et al., 2012). Similarly, damping down MvfR through a small volatile molecule reduced P. aeruginosa-induced fly killing while promoting its persistence in the flies (Kesarwani et al., 2011). Actually, while a direct correlation between bacterial growth rate and fly death was observed in T3SS or various QS mutants (Fauvarque et al., 2002; Lau et al., 2003), a twitching motility mutant (chpA) was less virulent but maintained its capacity to proliferate in infected flies (D'Argenio et al., 2001). Such an uncoupling between bacterial growth and fly death had been previously described with SPI-2 mutants of Salmonella (Brandt et al., 2004) and was also observed in an elegant study exploring the protective function of the human lactonase PON1, an enzyme that interferes with QS signalling molecules (Stoltz et al., 2008). In that study, the authors established that when co-injected along with the bacteria, exogenous acyl-HSLs restored virulence to PAO1 mutants deficient in acyl-HSLs production (ΔlasI/rhlI), but not to PAO1 mutants deficient in the receptors (ΔlasR/rhlR). Having demonstrated the activity of the human enzyme when expressed in transgenic flies, the authors further showed that it reduced PAO1-induced lethality to a survival rate similar to that observed in flies infected with ΔlasI/rhlI mutants. This protection was independent of the bacterial growth rate. Thus, the toxicity of Pseudomonas in flies may depend on its capacity to proliferate in the infected organism, or alternatively, it may depend on the expression of QS-dependent genes that encode toxic factors.
Intriguingly, the QS regulator RhlR, but not LasR, was shown to play a major role in PA14 bacterial invasion and proliferation in flies (Limmer et al., 2011) while these two QS interdependent systems were found required for PAO1-induced fly killing in an independent study (Lutter et al., 2012). Because the rhlR mutant induced killing kinetics similar to those in the parental PA14 strain in flies with impaired phagocytosis, it was suggested that RhlR may block haemocyte function through an as yet unknown mechanism (Limmer et al., 2011). Indeed, proper phagocytic function of haemocytes is required to fight P. aeruginosa: blocking phagocytosis by expressing the GTPase Activating Protein (GAP) domain of the ExoS toxin in haemocytes (Avet-Rochex et al., 2005) or by injecting latex beads (Limmer et al., 2011) markedly decreased fly resistance to infection by P. aeruginosa. In addition to cellular immunity, the Imd pathway, largely known for its role in fighting Gram-negative infections, and the Toll pathway are both important immune processes required to resist P. aeruginosa infections. Thus, a role for the Toll pathway in fighting a wider spectrum of pathogens, other than Gram-positive and fungi infections, could be anticipated from the studies of P. aeruginosa fly infection (Lau et al., 2003; Apidianakis et al., 2005).
Exploring in vivo biofilm formation
Many bacteria species have the capacity to sense their environment and respond to stressful stimuli by secreting small signalling molecules and aggregating in multicellular microbial communities to form biofilms (Bordi and de Bentzmann, 2011). In these biofilms, bacteria are protected by a complex extracellular matrix whose composition varies depending upon species but typically contains multiple exopolysaccharides (EPS), proteins, lipids and extracellular DNA (eDNA) (Flemming and Wingender, 2010). Biofilms cause major health problems because they colonize indwelling medical devices such as catheters and prosthetics, and they allow for bacterial persistence in wounds or chronically infected individuals. Furthermore, they confer protection to bacteria, which can escape host immune phagocytic cells, defensive molecules or antibiotic treatment. Biofilms are also suspected to be a source of multidrug-resistant strains due to higher mutations rates and gene transfer by bacterial conjugation. The bacteria P. aeruginosa displays an exceptional capacity to promote the transition between a planktonic to a sessile community lifestyle, a property that is a major cause for morbidity and mortality in cystic fibrosis patients (Bordi and de Bentzmann, 2011).
The ability of P. aeruginosa to form bacterial aggregates and dense microcolonies associated with EPS and eDNA (resembling biofilms) was first detected in the dissected crop of flies that were fed PAO1 (Mulcahy et al., 2011). Biofilm formation was strictly dependent on EPS secretion. Indeed, the pelB mutant, which displays decreased secretion of EPS and poor ability to form biofilm in vitro, did not colonize the crop of infected flies while the PAZHI3 strain, displaying increased production of EPS, had the capacity to form an enhanced number and larger microcolonies compared to PAO1. Such biofilm-like structures surrounded by eDNA were also observed with PAO1 and a PprBK mutant that overexpresses the Pseudomonas membrane permeability regulator (PprB), positively regulating biofilm formation (de Bentzmann et al., 2012). This mutant is characterized by its ability to rapidly form unconventional biofilm containing, among other molecules, the BapA adhesin and eDNA, but no EPS. In this study, the parental PAO1 strain, the PprBK mutant insertion and the PprBK::ΔBapA double mutant (impaired for the secretion of the BapA adhesin), were all found to colonize not only the crop but also the proventriculus, a complex structure that connects the foregut to the midgut (Fig. 1). Interestingly, the plague bacillus Yersinia pestis colonizes and persists in the foregut proventriculus of its flea vector under an obligatory biofilm state (Jarrett et al., 2004). Moreover, the crop and proventriculus are prominently colonized by commensal enterococci in the natural fly populations (Cox and Gilmore, 2007). All these observations suggest that the crop and the proventriculus are particularly permissive for pathogenic or commensal bacterial persistence. Additionally, PAO1 and PprBK biofilms were observed in the midgut, a place where food is subjected to continuous flux (Fig. 1). This implies that efficient attachment of the bacteria to either the peritrophic membrane and/or the gut epithelial cells occurs, an aspect that remains to be further elucidated. Both hyperbiofilm strains PAZHI3 (Mulcahy et al., 2011) and PprBK (de Bentzmann et al., 2012) displayed a reduced capacity to cross the epithelial barrier and invade the haemolymph together with reduced virulence in both acute and chronic models of fly infection. This observation highlights the fact that host tolerance is greater for bacteria embedded in a biofilm protective matrix. In other words, bacteria survival and persistence in a protective biofilm matrix are harmless to the infected organism compared to dissemination of acute virulent planktonic bacteria.
Reproducing polymicrobial community interaction
An infection protocol with a single pathogen failed to mirror the presence of polymicrobial communities observed in chronic wounds or in clinical sputum in which several microorganisms coexist, typically including variable bacterial species in combination with various fungi pathogens (Wolcott et al., 2013). In such communities, either mutualistic relationships or competitive behaviours favour bacterial persistence or enhanced virulence and ultimately aggravates the outcome of the infection. A better understanding of molecular and cellular mechanisms underlying multiple colonization of a living host may help improve medical treatment of polymicrobial infections. Sibley and colleagues used Drosophila to assay potential cooperation between P. aeruginosa and a number of strains isolated from the oropharyngeal flora, which are commonly found in association with P. aeruginosa in the lungs of cystic fibrosis patients (Sibley et al., 2008). Interestingly, some of them were not pathogenic to Drosophila in the feeding assay but dramatically reduced fly survival in combination with PAO1. Two isolates in particular, belonging to Gram-positive Staphylococcus sp. and Streptococcus sp., clearly modulate a number of P. aeruginosa virulence factors in vivo, a property which may account for increased PAO1 virulence (Sibley et al., 2008). Alternatively, it is also possible that gut damage induced by PAO1 allowed for translocation of these isolates in the haemocoel where they may contribute to fast fly killing.
An independent study aimed at exploring synergy mechanisms during P. aeruginosa co-infection with Gram-positive bacteria provided particularly interesting information on how virulence enhancement is achieved at the molecular level (Korgaonkar et al., 2013). Starting from the fact that P. aeruginosa exposure to cell wall fragments from Gram-positive bacteria enhances expression of the virulence factor pyocyanin, specifically due to the presence of N-acetyl glycosamine (GLcNAc), these authors isolated a P. aeruginosa mutant unable to respond to peptidoglycan. This mutant showed a reduced capacity to kill flies following oral infection, suggesting that peptidoglycan from native commensal flora contributes to P. aeruginosa-induced pathogenesis. Indeed, eliminating Gram-positive commensal bacteria through specific antibiotic treatment resulted in decreased virulence of the wild-type P. aeruginosa strain (but not the mutant strain), which was restored by feeding flies peptidoglycan. Enhanced fly killing upon peptidoglycan addition was not due to an increase in P. aeruginosa growth, which indicates that Gram-positive peptidoglycan enhances production of a toxic virulence factor, as indeed illustrated by P. aeruginosa pqsA virulence gene expression analysis, rather than promoting bacterial proliferation. Moreover, the number of Gram-positive bacteria within P. aeruginosa-infected crops was significantly lower than in uninfected crops. Therefore, sensing the peptidoglycan from the commensal flora likely enhances P. aeruginosa virulence and induces killing of that commensal flora (Korgaonkar et al., 2013).
Deciphering the activity of a bacterial toxin using transgenic flies
As illustrated above, the toxicity of a bacterial infection is not only dependent on simple bacterial proliferation but also relies on the production of toxic factors. An interesting strategy to decipher the molecular mechanism of such toxins in a living organism, without dealing with the complexity of the pathogenesis or manipulating dangerous microorganisms, is to create transgenic animals expressing a single bacterial toxin in a cell- or tissue-specific manner and elucidating the host target proteins mediating putative phenotypic defects caused by the toxin (Fig. 2). Using this method, it was shown that expression of the GAP domain of the ExoS toxin of P. aeruginosa in the eye induced developmental defects due to inhibition of Rho GTPase functions (Avet-Rochex et al., 2005), which could be predicted from previous studies in mammalian cells (Galan and Collmer, 1999). Interestingly, expressing this toxin in the haemocyte lineage, but not in the fat body cells, was sufficient to increase fly susceptibility to P. aeruginosa and other types of bacterial infections. This susceptibility could be rescued by coexpressing the Rho GTPase Rac2 (Avet-Rochex et al., 2005; 2007). These data demonstrated both the role of fly phagocytes and the in vivo efficiency of the ExoS toxin in counteracting cellular immunity. A similar strategy was successfully undertaken to understand the toxicity and identify previously unidentified molecular targets of the anthrax lethal and oedema factors of Bacillus anthracis (Guichard et al., 2006; 2010), as well as the CagA virulence factor of Helicobacter pylori (Botham et al., 2008; Muyskens and Guillemin, 2011; Reid et al., 2012) (extensively reviewed in Bier and Guichard, 2012). More recently, two YopJ-related toxins, the Aeromonas salmonicida-secreted protein AopP (Jones et al., 2012) and the Y. pestis YopJ (Paquette et al., 2012), were also expressed in transgenic flies. These types of toxins typically inhibit MAPKs and NF-κB pathways and induce rapid apoptotic death in infected macrophages (Orth et al., 1999). In particular, YopJ possesses acetyl transferase activity on MAP2K, which blocks phosphorylation, thus preventing activation and downstream signal transduction (Mittal et al., 2006; Mukherjee et al., 2006). In Drosophila, expression of AopP inhibited both Imd- and Toll NFκB-like pathways and induced apoptosis in the imaginal discs and the haemocyte lineage through a yet unknown mechanism. Unlike AopP, YopJ uniquely inhibited the Imd pathway through serine/threonine acetylation of the MAP3K, dTAK1 (Paquette et al., 2012). Importantly, YopJ displays a similar inhibitory activity against human TAK1 (Meinzer et al., 2012). Such enzymatic activity on evolutionary conserved targets may contribute to the ability of pathogens to infect multiple hosts, in this last particular case, fleas, rats and humans.
Concluding remarks: using flies for therapeutic issues
A large number of studies describe how human pathogens infect flies and reproduce in many cases some features of human pathology. They provide new insights in bacterial virulence or persistence factors on one side and host defences or tolerance mechanisms on the other side (Shirasu-Hiza and Schneider, 2007; Ferrandon, 2013). An imperative with regard to this large amount of experimental data is to apply these various pathogenesis models to assay drug activity on infected flies, possibly taking advantage of ‘humanized’ flies by transgenic-mediated expression of human proteins (Stoltz et al., 2008). The fact that small QS signalling molecules can interfere with P. aeruginosa toxicity, either when injected in the body cavity or when fed to flies, and that oral antibiotic treatment can clear bacteria in various cases of flies infection, provides proof-of-concept that such an approach is feasible and deserves to gain in intensity (Kesarwani et al., 2011; Cady et al., 2012). Presently, infectious disease treatment faces the problematic emergence of antibiotic resistance. A novel strategy to circumvent resistance is targeting virulence mechanisms rather than bacterial vital proteins. Along this line, using simple genetic models allows for screening of compounds targeting virulence factors in the context of host defence mechanisms in action. While very large screens can be more easily performed on very simple organisms such as the amoeba D. discoideum or the worm C. elegans (Cosson and Soldati, 2008; Ewbank and Zugasti, 2011), the Drosophila fruit fly can be advantageously used to further analyse the efficacy and mechanism of action of selected drugs in a living organism possessing a complex immune response.
The author thanks Dr Didier Grunwald for confocal imaging and Dr Emmanuel Taillebourg for critical reading of the manuscript and apologizes for additional work that could not be cited due to space restriction and focus on Pseudomonas-related studies in living flies. Research work by the group in the infectiology field has received financial support from Region Rhône-Alpes and the FINOVI foundation.