P. aeruginosa forms biofilm in the Drosophila crop
Using the fly oral model of infection, 1–3 day old male flies were infected with P. aeruginosa
. As we have been previously published, the predominant site of P. aeruginosa
accumulation at 24 h was the food storage organ known as the crop 
with bacteria moving from the crop into other areas of the gut over time (Figure S1
). The presence of bacterial cells in areas of the gut outside of the crop is consistent with what has been shown by other groups 
We used P. aeruginosa
PAO1pCHAP6656 to visualize P. aeruginosa
colonization in vivo
as these cells produced mCherry as an outer membrane-anchored lipoprotein. PAO1pCHAP6656 was used as it has an easily identifiable membrane-staining pattern 
and can be used to differentiate true bacterial cells from red autofluorescence in the tissue of the Drosophila
. The pCHAP6656 plasmid 
encodes for gentamicin resistance.
To ensure that this was a suitable system for use in vivo, in the absence of antibiotic selection, we monitored plasmid maintenance up to 48 h postinfection. Plating of bacteria recovered from infected flies indicated that pCHAP6656 was maintained up to 48 h postinfection (). At 24 and 48 hours postinfection, flies were sacrificed and crops were surgically removed for microscopic analysis. Imaging of dissected crops indicated that bacteria localized to the periphery of the crop () and that large aggregates (50–250 µM) or microcolonies were visible at 24 h (). These microscopic observations showed P. aeruginosa was present in the Drosophila crop, at high cell densities (~3×107 CFU/crop, ) and organized in microcolonies, indicating the presence of biofilms as early as 24 h postinfection of Drosophila.
PAO1 (pCHAP6656) infection of the Drosophila crop.
In vivo biofilms are composed of P. aeruginosa microcolonies, EPS and DNA, that form a characteristic honeycomb-like shape
grown biofilms are characterized by an extracellular biofilm matrix composed of DNA and EPS 
. To determine if the P. aeruginosa
microcolonies observed in Drosophila
displayed similar characteristics to those of in vitro
biofilms, crops were stained for EPS and DNA. We exploited the red and green autofluorescence of the crop (), while simultaneously visualizing red fluorescent PAO1pCHAP6656 (), green fluorescent EPS () and blue fluorescent DNA (). The gross morphology of uninfected crops was compared to PAO1pCHAP6656-infected crops. Uninfected crops had clearly defined muscular fibers and cellular structures (). PAO1pCHAP6656-infected crops () demonstrated a loss in the musculature, a blurring of the fibers and an overall lack of the organized structure visible in uninfected crops (). FITC conjugated Hippeastrum hybrid Lectin (HHA) 
(green fluorescence) was used to label exopolysaccharide present in the microcolonies (). DAPI (blue fluorescence) was used to visualize DNA, which is present in the nuclei of Drosophila
epithelial cells lining the crop (), in bacterial cells () and as extracellular DNA surrounding bacterial cells and in the biofilm matrix (). PAO1pCHAP6656 (red) was visible as aggregates in the crop (). An overlap between bacteria (red), EPS (green) and DAPI (blue) was visible () suggesting that bacterial aggregates stain positively for EPS and DNA. Red and green autofluorescence of the crop itself was also observable (). DAPI was used as the stain of choice for DNA. Use of other green or red DNA stains such as TOTO-1 or Sytox red, resulted in very high background due to autofluorescence in the crop. These data indicated that at 24 h postinfection, P. aeruginosa
biofilms can be visualized in dissected crops of infected flies and these biofilms stain positively for DNA and EPS, two major characteristics of an in vitro
In vivo P. aeruginosa biofilms stain positively for EPS and DNA.
To show that DNA and EPS were important biofilm components in vivo, infected crops were DNAse- and cellulase-treated prior to EPS and DNA staining () and compared to uninfected () and PAO1pCHAP6656-infected crops without DNAse and cellulase treatment (). Aggregates of bacteria, which stained positively for DNA and EPS were visible in infected crops (). No aggregates were detected in PAO1pCHAP6656-infected crops treated with DNAse and cellulase, indicating that bacterial biofilms were dissolved following enzymatic treatment (). DNAse and cellulase treatment of uninfected crops had no effect on crop structure (data not shown).
On closer inspection of DAPI-stained bacteria in the crop (digital zoom of 4.4X), we observed that the bacteria in the microcolonies were organized into characteristic patterns or clusters (). Cells that we previously observed to be positive for DNA and EPS staining () were organized into a cluster of hexagonal bacterial colonies. These structures were found in two orientations, with bacteria lined up side-to-side and stacked one on top of the other or with the pole to pole length of the bacterial cells visible, and each cell attached side-to-side (). Because of the size of each of the hexagons, approximately 6–8 µM, we predict that each hexagon is made up of approximately 7 to 15 bacterial cells. This is consistent with what has been observed for mature honeycomb structures produced by S. epidermidis 
. These honeycomb-like shaped microcolonies were not firmly attached to the epithelial cell surface but were floating inside the enclosed fly crop (Video S1
Visualization and staining of in vivo microcolonies in the Drosophila crop.
Microbial species, including Sinorhizobium meliloti
, Rhizobium leguminosarum 
, Staphlococcus epidermidis and P. aeruginosa 
were previously shown to form complex biofilm structures, organised in honeycomb- and veil-like patterns. Honeycomb structures are one of the most densely packed structures found in nature and similar to what is observed with bee honeycombs 
, it is predicted that these structures enable close packing together of cells with the least amount of matrix components, including energy-expensive EPS. To our knowledge, bacterial microcolonies that resemble honeycomb structures have not previously been visualized in vivo
in an animal model and highlight the use of Drosophila
as an infection model amenable to microscopic analysis of infected tissues.
P. aeruginosa biofilm infection results in loss of integrity of the fly crop structure
To investigate the potential for detrimental consequences of P. aeruginosa biofilm infections on host tissue we examined the architecture of infected fly crops during a biofilm infection. Excised crops were visualized macroscopically for gross changes in shape and size at 10X magnification (). Brightfield imaging of uninfected () and PAO1-infected () crops indicated significant changes in the size and gross morphology of the crop in response to infection. Infected crops were smaller in size, softer in texture and were more sensitive to breaking apart upon handling (data not shown). This observation is consistent with tissue damage however it is also possible that a smaller softer crop may be as a result of an empty crop as P. aeruginosa infection may interfere with the ability of Drosophila to feed and drink.
Crop integrity in response to P. aeruginosa infection.
To investigate the morphological structure of the musculature of the crop in greater detail in the presence and absence of biofilm infection, F-actin staining using Phallodin 488 was performed. Drosophila nuclei were counterstained with DAPI and crops examined by fluorescence microscopy. The musculature of uninfected crops consisted of wide ribbons of circular muscles covering the crop wall with a intricate network of branched and interconnecting fibers () This architecture was severely compromised or absent in infected crops (). P. aeruginosa (red) localized predominantly to the crop edge, which was where the most disorganised actin staining was detected, including depolymerisation and degradation of actin filaments ().
Expression of EPS is essential for in vivo biofilm formation
We examined in vivo
biofilm phenotypes of P. aeruginosa
mutants known to exhibit altered EPS production and biofilm formation phenotypes in vitro.
is defective for biofilm formation in vitro
, as mutants in the pel
operon are known to have decreased EPS production and biofilm formation 
, while strain PAZHI3 (a mutant in the posttranscriptional regulatory protein RsmA) displayed increased production of both pel
EPS (Figure S2
) and is a hyperbiofilm former 
. Flies were infected with PAO1, pelB::lux
or PAZHI3, all carrying pCHAP6656, and 24 h postinfection crops were excised and examined for the presence of microcolonies. Microscopy was performed on PAO1pCHAP6656, pelB::lux
pCHAP6656, and PAZH13pCHAP6656-infected crops. Twelve fields of view were captured, from a minimum of 3 crops infected with each strain (), for quantification of microcolony formation (). Image analysis (ImageJ) was performed to differentiate and count the frequency of individual cells, as well as small and large microcolonies () (See materials and methods
for additional information). Large microcolonies (>20 cells, ) were present only in PAO1- or PAZHI3-infected crops and were absent from pelB::lux
-infected crops (). PAZHI3-infected crops had more microcolonies (n
15) present than those seen in PAO1-infected crops (n
9) (). Furthermore, the large microcolonies (categorized as those microcolonies consisting of >20 cells) observed in PAZH13-infected crops were significantly larger (p<0.001) in size (approx 17-fold) that those microcolonies observed in PAO1-infected crops indicating the hyper-biofilm features of PAZHI3 detectable in vitro
were also observed in vivo
. No large microcolonies were detected in pelB::lux
-infected flies ().
The role of Pel EPS during in vivo biofilm formation in Drosophila.
To further confirm the importance of EPS during oral Drosophila
infection, qRT-PCR was used to measure expression of pel
during infection. Psl
expression was significantly induced, approximately 150 fold (p<0.01). Pel
expression was highest during oral infection of flies, induced approximately 2200 fold (p<0.001), relative to acute infection of flies (). These data indicate that Pel may play a more important role during oral infection of Drosophila
, since it is more highly expressed. These data highlight the importance of Pel EPS as a biofilm matrix component for the establishment and/or maintenance of biofilms in vivo
, in addition to its well-characterized importance for attachment and maturation, during the early and later stages of biofilm formation in vitro 
Non-biofilm forming strains disseminate at a faster rate than biofilm forming strains following oral infection
We hypothesized that biofilm forming and non-biofilm forming strains would differ in their ability and/or timing to disseminate and that ultimately the kinetics of bacterial dissemination may play a role in fly survival. Initial experiments were performed to compare in vivo
localization of PAO1, pelB::lux
and PAZHI3 strains in infected flies. Results of viable plate counts indicated a slightly lower bacterial load was recovered from the GI tract of pelB::lux
-infected flies (3.8×105
CFU/fly; mean ± SEM), compared to that of PAO1-infected flies (4.8×105
CFU/fly). There was a corresponding increase in the number of viable pelB::lux
CFU/fly), recovered from the fly body, excluding the GI system, compared to that of PAO1-infected flies (1.03×103
CFU/fly) 5 days postinfection (). Similar numbers of bacteria were isolated from the GI system or fly body of PAZH13-infected flies compared to PAO1-infected flies. To provide evidence of altered dissemination between biofilm and non-biofilm forming strains, hemolymph was recovered from infected flies at day 2 and day 5 postinfection. The pelB::lux
mutant was present in the hemolymph at significantly higher numbers than PAO1 or PAZH13 at two days postinfection, while no significant difference in dissemination was observed five days postinfection (). Previous studies have shown that pelA
mutants demonstrated increased rates of swarming motility 
, which in combination with reduced biofilm formation and may contribute to the increased rate of dissemination observed during infection of Drosophila
2 days postinfection. The fact that significantly increased numbers of pelB::lux
are observed in the hemolymph 2 days postinfection relative to PAO1, while no significant difference is observed 5 days postinfection, suggests that upon detection by the host immune system in the hemolymph, pelB::lux
bacteria are unable to persist or are cleared by the immune system.
In vivo localization and antibiotic resistance profiling of biofilm and non-biofilm infections.
P. aeruginosa recovered from biofilm infections in vivo have increased resistance to antimicrobials
Resistance to antimicrobials is a general feature of all biofilms. We hypothesized that PAO1 recovered from a biofilm infection of Drosophila would display increased resistance to antimicrobials. Antimicrobial sensitivities were compared in PAO1 directly recovered from flies relative to PAO1 planktonic cultures or PAO1 planktonic cultures exposed to pulverized fly tissues, termed “mock-infected” PAO1. Identical bacterial inocula from these three conditions were swabbed onto Pseudomonas Isolation Agar (PIA) and antimicrobial sensitivity was measured by disk diffusion. Antimicrobial resistance of PAO1 recovered from mock-infected cultures was not significantly different relative to the resistance profiles of PAO1 recovered from planktonic cultures (data not shown). PAO1 directly recovered from infected flies had significantly increased resistance to polymyxin B, colistin, and ciprofloxacin, but not to gentamicin or ceftazidime relative to planktonic PAO1 cultures (). PAZHI3 recovered from infected flies was also significantly more resistant to polymyxin B, colistin, and ciprofloxacin than planktonic PAZHI3 (). In contrast, antimicrobial resistance profiles of the pelB::lux mutant (which failed to form biofilms in vivo) were not significantly different in cells directly recovered from infected flies relative to planktonic or mock-infected cells ().
Polymyxin B and colistin are cationic AMPs: short, amphipathic peptides that bind to and disrupt both the outer and cytoplasmic membranes resulting in bacterial cell death 
. Ciprofloxacin is a member of the fluoroquinolone drug class which inhibits DNA gyrase and hence DNA replication. The increased antibiotic resistance phenotype of P. aeruginosa
recovered from the flies compared to planktonic cultures, is analogous to the increased resistance observed in in vitro
biofilm populations compared to planktonic cultures 
Drosophila infected with biofilm and non-biofilm forming P. aeruginosa have altered survival kinetics
To assess the comparative abilities of biofilm and non-biofilm forming P. aeruginosa
strains for their ability to cause disease in Drosophila
, we monitored fly survival over 14 days in response to oral infection. The non-biofilm forming pelB::lux
mutant was significantly more virulent compared to PAO1, having a significantly increased rate of Drosophila
killing (). In contrast, hyperbiofilm-forming PAZHI3 demonstrated significantly reduced virulence compared to PAO1, as indicated by a greater survival of infected flies up to 14 days postinfection (). There was no difference in the bacterial load (CFU) in biofilm and non-biofilm infected flies (data not shown). PAO1 mutants in psl
showed similar killing kinetics to PAO1-infected flies (Figure S3
) indicating that Pel EPS contributes to pathogenesis during infection of Drosophila
while Psl EPS does not. Previous in vitro
studies have indicated that both Pel and Psl are important in P. aeruginosa
biofilm formation 
and that Pel EPS also contributes to antibiotic resistance 
. Our data highlight a unique role for Pel EPS in P. aeruginosa
biofilm formation in vivo
, as well as a role in dissemination and virulence.
Kaplan-Meier survival curves post P. aeruginosa infection.
The production of Pel EPS and biofilm formation inversely correlated with virulence and the ability of P. aeruginosa
to cause death in Drosophila
after oral infection (). To determine if EPS production also affected the outcome of acute P. aeruginosa
killing kinetics were compared in male flies nicked in the thoracic region, with the relevant P. aeruginosa
strains, up to 36 h. Pel production was not found to be important factor during acute infection as PAO1 and pelB::lux
infections resulted in similar killing kinetics in acutely-infected Drosophila
up to 36 h postinfection (). PAZHI3 was attenuated for virulence during acute infection, similar to what was seen for oral infection (). Reduced killing of Drosophila
by PAZHI3 is similar to reduced virulence previously observed for PAZHI3 in a mouse model of acute infection 
Biofilm infections induce AMP gene expression
To monitor the AMP response to biofilm and non-biofilm infections in Drosophila, we assessed the expression of the AMP genes cecropin A1, diptericin and drosomycin using qRT-PCR during oral infection with PAO1, pelB::lux and PAZHI3 (). As no difference in killing kinetics were observed between flies acutely infected with biofilm forming PAO1 and non-biofilm forming pelB::lux, AMP gene expression was not monitored following acute infection.
Biofilm infections induce antimicrobial peptide gene expression in Drosophila.
PAO1 oral infection induced the expression of cecropin A1, diptericin and drosomycin between 4- and 36-fold relative to uninfected flies (). Increased gene expression was also detected in PAZHI3-infected flies at levels between 72- and 446-fold. While PAZH13 is hyperbiofilm former in vitro 
and in vivo
(), it is also a pleiotrophic mutant 
. Thus, while there is a correlation between biofilm formation and increased AMP expression, we cannot rule out the possibility that the higher levels of AMP induction seen in response to PAZH13 infection may not be solely attributable to increased biofilm formation. In response to pelB::lux
infection, we observed lower expression of all three AMP genes, between 1.6- and 5-fold, compared to uninfected flies. Suppression of AMP gene expression is thought to be one of the main mechanisms whereby commensal bacteria fail to elicit an immune response in the host 
. However, virulent strains of P. aeruginosa
have also been documented to suppress AMP gene expression and the Drosophila
immune response during an acute infection 
. In this study, decreased expression of AMP gene expression by the pelB::lux
mutant () appears to be associated with increased fly death following oral infection as flies die at a significantly faster rate compared to those infected with the biofilm forming Pel positive strains PAO1 or PAZH13 (). While this study does not demonstrate active suppression, it is possible that increased fly mortality post oral infection resulted from decreased expression of AMP gene expression in the fly and/or a more rapid (within 2 days) dissemination of pelB::lux
to the hemolymph, resulting in systemic infection and fly death. However in addition to difference in localization of pelB::lux
(), it may also be that pelB::lux
is more toxic, eliciting pathological changes in Drosophila
resulting in more rapid death.
As EPS can be a cell-surface or secreted product, we hypothesized that co-infection of Drosophila
with a 1-1 mixture of P. aeruginosa
wildtype and pelB::lux
would restore AMP gene expression and killing, similar to levels observed in orally PAO1-infected flies. In these experiments, PAO1::p16Slux 
was used instead of PAO1 as the wildtype strain as bacterial load and AMP gene expression did not differ significantly in PAO1::p16Slux
-infected flies compared to PAO1-infected flies (data not shown). Use of PAO1::p16Slux
allowed us to differentiate between wildtype and mutant strains for quantitative bacteriology using erythromycin resistance in PAO1::p16Slux
as the differentiating marker. Relative to uninfected flies, AMP gene expression was measured in flies co-infected with PAO1::p16Slux
and PAO1, pelB::lux
or PAZHI3. There was no significant difference in AMP gene expression following co-infection with PAO1::p16Slux
and PAO1 (). Co-infection with PAO1::p16Slux
resulted in induction of 6.3-, 4.3-, and 23-fold for cecropin A1, diptericin and drosomycin, respectively (), induction levels similar to those observed for wildtype infections. For PAO1::p16Slux
and PAZHI3 co-infected flies, AMP genes were induced at levels between 62 to 97 fold (). These data indicated that co-infection of pelB::lux
and PAO1 restored AMP gene expression to levels similar to those observed in PAO1-infected flies. In all co-infection experiments, quantitative bacteriology was performed at T0
(hours) to ensure that the bacterial load was at a ratio of approximately 1-1 at the initial stage of infection (T0
), at the time of RNA extraction (T24
), and at later time points during infection (T120
) (Figure S4
). No significant differences were observed in the growth of different bacterial strains in Drosophila
following co-infection at any of the time points investigated, indicating that pelB::lux
or PAZH13 mutants were not altered in their ability to compete with PAO1 for colonization during infection of Drosophila
Drosophila survival was also monitored following co-infection experiments. Co-infection of flies with pelB::lux and PAO1::p16Slux resulted in significantly increased fly survival relative to pelB::lux-infected flies, increasing fly survival to levels similar to those seen during wildtype infection (PAO1 and PAO1::p16Slux) (). Co-infection of flies with PAZH13 and PAO1::p16Slux had no significant effect on fly survival with flies dying at similar rates regardless of whether they were infected with PAZH13 alone or co-infected with PAZH13 and PAO1::p16Slux. Single infection with PAO1::p16Slux or PAO1, or co-infection with PAO1 and PAO1::p16Slux, had similar killing kinetics (data not shown).
Prior to this study, it was not known if the fly immune system responded differently to biofilm and non-biofilm forming bacteria. Drosomycin expression is regulated through the Toll pathway 
; Diptericin is regulated via the Imd pathway 
and both pathways overlap to regulate cecropin A1 expression 
. Our data indicates that both of the central immune pathways in Drosophila
are activated in response to biofilms. In addition this data indicates that it is Pel positive biofilms, and possibly Pel EPS itself, that may act as a specific host immune signal inducing AMP gene expression in Drosophila
as a psl
mutant has no effect on Drosophila
killing (Figure S3
). Future work will focus on identifying the specific bacterial components involved in AMP gene expression and other host signalling pathways in response to Pel and Psl positive biofilms and non-biofilm P. aeruginosa
In the Drosophila oral infection model, our data suggests that Pel positive biofilms induced AMP gene expression in the fly. Although biofilm infections induce AMP gene expression (), biofilm-forming bacteria isolated from fly crops postinfection are more resistant to the AMPs polymyxin B and colistin than those recovered from planktonic cultures (). Bacterial Pel EPS may be a cue to the host to increase AMP gene expression thus serving to slow dissemination of the bacteria, and in this way slow systemic infection which would rapidly kill the host. On the other hand, EPS may also induce inflammation in the crop/GI system resulting in a localized damage to the host. Strains incapable of forming Pel positive biofilms in vivo resulted in a decreased AMP response but disseminated earlier, resulting in a systemic infection associated with faster host killing. These interpretations are supported by the Drosophila survival data obtained from co-infection experiments, where co-infection of flies with pelB::lux and PAO1 significantly increases Drosophila survival compared to infection with pelB::lux alone.
Biofilm infections do not alter kinetics of subsequent acute infection but modify fly survival in response to subsequent oral challenge
It has previously been shown that P. aeruginosa
eludes host defenses by suppressing AMP gene expression in a Drosophila
model of acute infection 
. This study also demonstrated that infection with a less virulent P. aeruginosa
strain resulted in immune potentiation and protected flies from subsequent acute infection with a more virulent P. aeruginosa
. To determine if oral infection, biofilm formation and induction of AMPs in Drosophila
could alter the kinetics of fly survival following subsequent acute infection, we performed the following experiment. Male flies were orally infected with PAO1 (biofilm, AMP induction), pelB::lux
(non-biofilm, AMP repression) or PAZHI3 (hyperbiofilm, AMP induction) for 24 h. After 24 h, orally infected flies from each of the three groups above and uninfected flies were nicked with PAO1 (acute infection), LB (sterile nicking) or not treated. Oral infection with PAO1, pelB::lux
or PAZHI3 had no significant effect on the rate of fly survival during subsequent acute infection (nicking) with PAO1 ().
Kaplan-Meier survival curves of Drosophila orally infected (feeding) for 24h followed by subsequent acute (nicking) or secondary oral infection.
To determine if oral PAO1 or PAZH13 biofilm infections altered Drosophila survival following subsequent oral infection with pelB::lux, the following experiment was performed. Drosophila were allowed to feed on PAO1, PAZH13, pelB::lux or a sucrose control for 24 h (primary infection), which is sufficient for biofilm formation to occur in the crop (). After 24 h, all flies were transferred to new vials containing pelB::lux as the food source (secondary infection). Survival was monitored up to 14 days after the primary infection. Primary infection with PAO1 or PAZH13, followed by secondary infection with pelB::lux significantly increased fly survival compared to flies who were infected with pelB::lux for both the primary and secondary infection. Increased Drosophila survival following primary infection with PAO1 or PAZH13 was not due to failure of the secondary infecting pelB::lux strain to infect Drosophila, as pelB::lux tetracycline resistant colonies (the antibiotic marker of the lux transposon) were recovered (at ≥3.8×106 CFU/fly or 76–99% of total bacterial load) from all secondary pelB::lux infected flies 5 days postinfection.
Primary oral infection with a biofilm-forming strain protected Drosophila from secondary oral infection with pelB::lux. Oral infection with a biofilm forming strain induced AMP gene expression, which may explain why increased fly survival was observed against secondary oral infection with pelB::lux. However the AMPs induced following oral infection may not be sufficient to alter Drosophila survival against subsequent acute infection. A possible reason for this is that AMP induction following biofilm infection is localized to the gut and does not protect Drosophila from death as a result of pricking and acute systemic infection. It is also possible that the pathology resulting from tissue damage following oral infection () may prevent Drosophila from responding to and coping with subsequent acute infection.
infections are associated with the highest case fatality rate of all Gram-negative infections 
. This is partly due to the ability of P. aeruginosa
to resist antimicrobial therapy. One of the main evasion strategies used by P. aeruginosa
, and other microbes, is the formation of multicellular, dense aggregates called biofilms. We have shown that specific antibiotic resistance mechanisms are induced in P. aeruginosa
. Biofilm infections are estimated to account for 65% of all bacterial infections 
. While some studies have investigated the host response to P. aeruginosa
, little is known regarding the bacterial and/or host factors involved in the pathogenesis of biofilm infections. The aim of this research was to develop a Drosophila
infection model that enables biofilms to be intricately studied in vivo
In this work we present evidence that oral infection of Drosophila by P. aeruginosa PAO1 resulted in biofilm formation in the Drosophila crop (). We demonstrated that biofilms formed in vivo retain the typical characteristics of in vitro grown biofilms, including DNA and EPS staining () and increased resistance to antibiotics (). We also showed that biofilm infections resulted in significantly decreased numbers of bacteria disseminating to the hemolymph 2 days postinfection, and contributed to increased AMP gene expression in the fly (, ). Non-biofilm forming pelB::lux infections, on the other hand, resulted in decreased AMP gene expression in the fly, significantly increased numbers of bacteria disseminating to the hemolymph 2 days postinfection, as well as early and increased fly mortality (–). The increased virulence of the pelB::lux mutant was attenuated by co-infection of Drosophila with biofilm-forming and AMP-inducing strains PAO1 or PAZH13 (). Furthermore, primary infection with either of these AMP-inducing strains altered the survival kinetics of Drosophila from secondary oral infection with the more virulent pelB::lux but not from subsequent acute infection (). In summary, we have developed a novel P. aeruginosa biofilm model of infection that can be used for studying both the bacterial and host response during infection. This model has the potential to significantly increase our understanding of the relationship between biofilms and the host during infection and also to tease out fundamental differences between the host response to biofilm and non-biofilm P. aeruginosa infections.