Lack of an antibacterial response defect in Drosophila Toll-9 mutant.

Toll and Toll-like receptors represent families of receptors involved in mediating innate immunity response in insects and mammals. Although Drosophila proteome contains multiple Toll paralogs, Toll-1 is, so far, the only receptor to which an immune role has been attributed. In contrast, every single mammalian TLR is a key membrane receptor upstream of the vertebrate immune signaling cascades. The prevailing view is that TLR-mediated immunity is ancient. Structural analysis reveals that Drosophila Toll-9 is the most closely related to vertebrate TLRs and utilizes similar signaling components as Toll-1. This suggests that Toll-9 could be an ancestor of TLR-like receptors and could have immune function. Consistently, it has been reported that over-expression of Toll-9 in immune tissues is sufficient to induce the expression of some antimicrobial peptides in flies. These results have led to the idea that Toll-9 could be a constitutively active receptor that maintain significant levels of antimicrobial molecules and therefore provide constant basal protection against micro-organisms. To test theses hypotheses, we generated and analyzed phenotypes associated with a complete loss-of-function allele of Toll-9. Our results suggest that Toll-9 is neither required to maintain a basal anti-microbial response nor to mount an efficient immune response to bacterial infection.


Introduction
Innate immunity is a rapid and efficient response that multicellular organisms mount to defend themselves against infection and that has been conserved throughout evolution (reviewed in [1,2]). Upon infection, microbe-specific immune elicitors, known as microbe-associated molecular patterns (MAMPs), are recognized by a set of so-called pathogen recognition receptors (PRRs) [3,4]. Mainly through genetic screens, the Drosophila community has largely contributed to the molecular dissection of the signaling modules that regulate expression of immune genes following infection. Drosophila humoral immune response is under the transcriptional control of two NF-kB signaling cascades, the Immune deficiency (Imd) and the Toll pathways, which closely resemble mammalian TNF-R and Toll-Like Receptor (TLR), respectively (reviewed in [5,6]). The Imd pathway can be activated by Gram-negative bacteria that contain diaminopimelic acid type peptidoglycan (DAP-type PGN) in their cell wall. Binding of DAP-type PGN to the transmembrane receptor Peptidoglycan-Recognition-Receptor-LC (PGRP-LC) induces phosphorylation and proteolytic activation of the NF-kB/Rel transcription factor Relish. This, in turns, activates the expression of several antimicrobial peptides (AMPs), such as Diptericin [2,7,8]. The Toll pathway is triggered by Gram-positive bacteria with Lysine-type peptidoglycan (Lys-type PGN), fungi and yeast, and mediates phosphorylation and subsequent degradation of the IkB-like inhibitor Cactus, freeing the NF-kB transcription factor Dif that translocates into the nuclear compartment where it induces the expression of several genes encoding AMPs, including Drosomycin and Cecropin [9][10][11][12][13].
The Toll (also known as Toll-1) gene was originally isolated for its role in specifying dorso-ventral polarity of the Drosophila embryo [14]. Toll encodes a type I transmembrane receptor composed of three domains: the extracellular leucine-rich repeats (LRR) domain, containing tandem arrays of a short leucine-rich motif in the Nterminal region, a sequence likely to form a single transmembrane helix, and an intracellular C-terminal domain significantly related to the vertebrate interleukin-1 receptor (the Toll-interleukin receptor -TIR-domain), involved in signaling processes [15][16][17]. Although it is well established that vertebrates TLRs function as PRRs for numerous microbial ligands such as LPS, dsRNA, Flagellin, DNA CpG (reviewed in [18,19]), there are so far no indication that Drosophila Toll family members are able to directly recognized microbial motifs. How then is the Toll receptor activated by bacterial PGN in flies? Recognition of bacterial PGN by PGRP-SA triggers a proteolytic cascade whose last substrate is the circulating pro-cytokine Pro-Spätzle [20,21]. This proteolytic cleavage transforms the immature cytokine Spätzle into an active ligand which triggers Toll receptor dimerization [22][23][24]. Receptor dimerization allows the recruitment of an adaptator complex containing three intracellular Death domaincontaining proteins, MyD88, Tube and Pelle [12,25,26]. Then, by a still unknown mechanism, Cactus is phosphorylated, allowing the release of the Dif transcription factor [27,28].
In addition to the family founder Toll-1, the Drosophila genome contains 8 additional members: Toll-2 also known as 18-Wheeler, and Toll-3 to Toll-9 [29,30], the function of some of which have been recently studied. Toll-8, also known as Tollo, is required for the induction of neural specific glycosylation in the Drosophila embryo [31,32]. Initially thought to be the membrane receptor upstream of the Imd cascade, Toll-2/18-W is now believed to function in the embryonic epithelium by regulating cell apical constriction via the Rho-GTPase-signaling pathway [30]. More recently, by analyzing double loss-of-function mutants, Yagi et al. showed that 18-W, Toll-7 and Toll-8 may have redundant functions in regulating developmental processes [33]. Although limited to some of the Toll-related proteins, these studies tend to indicate that the implication of Toll family members in immunity is rather the exception (Toll-1) than the rule.
Phylogenetic analysis of TIR domains indicate that, with the exception of Toll-9, Drosophila Tolls are more closely related to each other than to mammalian TLRs [34]. This suggests that these two groups of receptors, Drosophila Tolls and mammalian TLRs, evolved independently to carry distinct functions: putative cellular and developmental roles in insects, and a role in host defense in mammals [34]. However, Drosophila Toll-9 is structurally related to TLRs, both in its ectodomain and in the TIR domain as it is the only Drosophila Toll receptor that does not have N-flanking cysteinrich motifs at the C-terminal end of the ectodomain [35,36]. Additionally, Toll-9 is, with Toll and to a less extend Toll-5, the only Drosophila Toll able to activate immune genes transcription upon over-expression experiments in tissue culture cells [37].
Finally, Toll-9 uses, at least in part, the same intracellular signaling components that Toll-1 to promote Drosomycin expression Figure 1. Knock-out of the Toll-9 locus and Toll-9 expression. (A) A complete deletion of the Toll-9 locus was generated by targeted homologous recombination using 59 (red, II) and 39 (red, I) homology regions; the white cDNA replaced the Toll-9 locus (black box in genomic region). (B) PCR performed on the resulting KO Toll-9 line and Oregon R control. The primers used are localized on the genomic region (A, green). (C) Toll-9 expression is independent of commensal and infectious bacteria. Toll-9 expression measured by quantitative RT-PCR in conventionally reared wildtype (-1-WT CR ), conventionally reared wild-type 4 hours after Erwinia carotovora carotovora 15 ingestion (-2-WT CR inf. Ecc15 -4 h-), conventionally reared wild-type 18 hours after Ecc15 ingestion (-3-WTCR inf. Ecc15 -16 h) and germ-free wild-type (-4-WT GF ), in whole larvae, larval gut, whole adult and adult gut. Toll-9 is enriched in the gut of both larvae and adult, and its expression level is not dependent on commensal and infectious bacteria. Additionally, Toll-9 is no longer expressed in Toll-9 2/2 mutant (-5-). rp49 was used as the experimental expression standard. doi:10.1371/journal.pone.0017470.g001 in vitro (i.e. DmMyD88 and Pelle [37]). All together, these data suggest that Toll-9 could be another Toll family members involved in Drosophila immunity. In order to test the putative implication of Toll-9 in the Drosophila immune response, we generated a Toll-9 complete loss-of-function allele and tested its ability to mount an immune response. The following data suggest that Toll-9 is not required to mount an efficient response to bacterial infection.

Knock-out of the Toll-9 locus
Knock-out of the Toll-9 locus has been realized by ends-out gene targeting as described in [44]. The resulting deletion covers a genomic region starting from 852 bp upstream of the Toll-9 ATG, and up to 41 bp after Toll-9 stop codon. The 39 and 59 homology regions I (2926 bp long) and II (3116 bp long), showed in red in Figure 1, were amplified by PCR from Oregon R genomic DNA. The region I has been amplified using the following primers: F-ACATGCATGCATGTGGAAGCACTCTCGATTCAGC, R-GGGGTACCCCAGAGTTCTAGTCAGTTGTGC and con-tains SphI and KpnI sites respectively, whereas the region II has been amplified using the following primers: F-AGGCGCGCCTT-CATCGGATACCCATTGAGG; R-CCCGTACGGGCGAG-GATTCCGATAGATGCC and contains AscI and BsiWI sites. These two DNA fragments were then cloned into the pW25 vector (Drosophila Genomics Ressource Center), on both sides of the miniwhite gene w hs . This pW25-Toll-9-KO construct was then transformed into y,w embryos using the standard procedure for Pelement [45]. Male donor flies were crossed with y,w; hsp70-FLP, hsp70-I-SceI, noc Sco /CyO females and the resulting progeny were heat-shocked at 37uC for one hour twice a day, at day 2 and 3 after egg laying. Mosaic and white eyes females were then collected from the progeny and crossed to males carrying an eyeless-FLP transgene on X chromosome. In the resulting progeny red eye males were screened. Out of 22.400 males, 26 red eye males were isolated but 7 of them were sterile. Genomic deletion of the Toll-9 locus and replacement with the w hs gene were investigated by PCR, using 5F (59-GCTGTTGACGAAGAGGGAAG-39), 5R (GAATTGAATTGACGCTCCGT-39), 3F (59-GTCCGGTT-GTTTTCGTGCTC-39), 3R (59-GTACACTTCCTTGGCT-GGCG-39), 1 (59-CGTATTAGTATGCCTGTTCC-39) and 21 (59-ACAACTGACTAGAACTCTCC-39) primers as indicated in Figure 1. Among the 19 lines established, only one showed ampification using the primer couples 5F-5R and 3F-3R, but no amplification using the primer couple 1-21 ( Figure 1). This line has been isogenized using y,w flies, and kept as heterozygous in the same vial. Consequently, the control y,w has white eyes, the heterozygous y,w; Toll-9 2 /+ has orange eyes, and the y,w; Toll-9 2 / Toll-9 2 mutant has red eyes.

Hemocytes count
For each genotype (Toll-9 2 /TM3 GMR YFP and Toll9 2/2 larvae), 4 wandering third instar larvae were bled onto a coverslip, and the hemocytes collected were incubated for 10 minutes in a humidified chamber to allow them to adhere to the slide, and then fixed in PFA 4% for 5 minutes. The cells were thereafter rinsed 3 times in PBS and mount in Vectashield with DAPI. The results are the mean of 4 different experiments, and each experiment is the mean of 3 counts.

Longevity assays
Assays for longevity were performed on normal food or on food supplemented with an antibiotics cocktail. For each genotype (y,w, Toll-9 2/2 and Toll-9 2 /Def(Toll-9)) at least 80 flies grouped by 10 males in independent vials were tested.

Immunocytochemistry
Immunocytochemistry was performed as previously described (McKearin and Ohlstein, 1995). The mouse anti-Delta (1:20, Developemental Studies Hybridoma Bank), and chicken anti-GFP (1:500, Aves Labs) were used in this study. Fluorescenceconjugated secondary antibodies were purchased from Interchim Jackson (Alexa546 goat anti-mouse IgG antibody) and Molecular probes (Alexa488 goat anti-chicken antibody), and used at a 1:500 dilution. For DNA labeling, tissues were incubated with 1 mg/mL of DAPI (Sigma). All samples were mounted in Vectashield mounting medium (Vectorlabs) and examined using an apotome microscope (Zeiss Axioplan 2). Images were acquired using AxioImager and composite figures were prepared using Adobe Photoshop CS3.

Phenotypic characterization of Toll-9 mutant
In order to study Toll-9 receptor function, we generated a complete loss-of-function allele by targeted homologous recombi- nation [44]. 59 and 39 flanking regions of homology were introduced into the pW25 vector ( Figure 1A) to generate transgenic flies. The Toll-9 gene disruption knock-out candidate flies were screened and verified using PCR (with specific primers in the two flanking regions of Toll-9 -5F and 3R-, within Toll-9 -1 and 21-and within w hs -5R and 3F-, Figure 1A, B) and quantitative PCR (using q5 and q3 primers, Figure 1C, D). Toll-9 2/2 mutants in which both maternal and zygotic contributions have been removed hatched into morphologically normal and fertile adults, although the Toll-9 2/2 flies emerged one day later than the w control. Indeed, the t 50 (time by which 50% of the progeny has emerged) is of 10.7 days for the control and of 11.6 days for Toll-9 2/2 ( Figure 2A). Moreover, Toll-9 2/2 and Toll-9 2 /Def(Toll-9) flies showed a reduced lifespan ( Figure 2B) with the mean lifespan of w control flies being at 42 days versus 32 and 27 days for Toll-9 2/2 and Toll-9 2 /Def(Toll-9) mutant flies, respectively. A similar lifespan shift was observed when flies are grown in axenic conditions ( Figure 2B). Finally, an embryonic lethality was observed, since only one third of the expected Toll-9 2/2 embryos hatched into larvae. The analysis of the dead embryos revealed that the lethality occurs at the end of embryogenesis. In conclusion, loss of Toll-9 2/2 function is associated with partial embryonic lethality, delay in adult flies emergence and reduced lifespan.
Toll-9 expression is highly spatially and temporally restricted during embryogenesis [46]. It starts at stage 5 in the vitellophages and is, from stages 7 to 11, limited to a group of cells in the head region, located at the site where the hemocytes progenitors are specified. Later in embryogenesis, Toll-9 transcripts are no longer detectable [46]. Knowing that hemocytes are essential for normal embryogenesis to take place and are playing essential roles in immunity [42,47,48], we checked whether the embryonic hemocyte population was affected in Toll-9 2/2 mutants. For that purpose, we monitored the expression of two hematopoietic markers, Serpent, a GATA transcription factor expressed in hemocytes precursors which will differentiate into plasmatocytes and crystal cells, and Lozenge, an AML-1/Runx1 homologue required for crystal cell differentiation and expressed in a subset of prohemocytes [49,50]. As showed in Figure 3, both factors were expressed normally in Toll-9 2/2 mutant embryos (even at the end of embryogenesis, data not shown), suggesting that the embryonic hematopoiesis is not affected by the absence of Toll-9. This was further confirmed by counting the number of larval hemocytes. We found that the number of larval hemocytes was slightly increased in Toll9 2/2 mutants. Indeed, we found 17756238 hemocytes per mm 2 (n = 11420) in Toll-9 2/2 mutant larvae, compared to 12426138 hemocytes per mm 2 (n = 7990) in Toll-9 2 /TM3 GMR YFP control larvae (less than a 2-fold increase). It is however unlikely that this increase of embryonic hemocytes could explain the partial embryonic lethality observed in Toll-9 2/2 stocks. Toll-9 function is not required for basal and inducible AMP production It has been shown that Toll-9 overexpression in vitro and in vivo is sufficient to induce the expression of antimicrobial peptides genes known to be regulated by the Toll pathway, such as Drosomycin or Cecropin ( [37,38] and Figure 4A, A'), but not Diptericin, an Imd pathway regulated gene ( Figure 4B, B'). This work had led to the idea that Toll-9 functions as a constitutively active signaling receptor that will maintain a substantial level of antimicrobial peptides required to ward off continuous challenge of microorganisms.
We then wanted to test whether Toll-9 expression was dependent on the presence of either commensal or pathogenic bacteria. We therefore compared Toll-9 transcript levels between flies grown in axenic conditions, conventionally-reared flies, and flies infected with the phytopathogenic bacteria Erwinia carotovora carotovora 15 (Ecc15). Our results indicate that Toll-9 expression was neither modified under axenic conditions nor after immune challenge by Ecc15 (Figure 1C, D). This is in agreement with previously published results showing that Toll-9 transcript levels did not seem to change in the whole larvae and adult following septic injury with Escherichia coli [37].
Among immune-competent tissues, the gut is permanently in contact with a large number of well-tolerated commensal bacteria, which under normal condition do not elicit an immune response. Our results, together with microarray analysis (http://www. flyatlas.org), indicate that Toll-9 is highly and specifically enriched in larval and adult guts (3.3 times in the larval midgut compared to whole larvae, Figure 1C and 13.8 times in the adult midgut, compared to whole fly, Figure 1D). It was therefore tempting to speculate that Toll-9 may play a role in localized epithelial defense by maintaining a basal level of antimicrobial peptides in the gut, a tissue in constant contact with microorganisms. We then tested whether Toll-9 expression was modified specifically in the gut in axenic conditions or following oral infection with Ecc15. Contrary to Bombyx mori Toll-9, whose expression is significantly increased in the gut by microbial challenge [51], Drosophila Toll-9 expression levels was not significantly different in conventionally reared flies, flies grown in axenic conditions or in immune challenged flies ( Figure 1C, D).
To further test the hypothesis that Toll-9 is required to maintain a basal level of antimicrobial peptides, we measured the levels of several peptides in Toll-9 2/2 mutants, in the whole animal to quantify the systemic response, and specifically in the gut to measure the local response. The following antimicrobial peptides were tested: Drosomycin and Defensin, regulated by the Toll pathway, Diptericin, regulated by the Imd pathway, and Drosomycin3, a gut expressed gene whose transcription is JAK/STAT pathway dependent [52]. Toll-9 2/2 mutant larvae and adults were always compared to Toll-9 2 /TM3 control larvae and +/+ control adults, respectively, from the same vial, to avoid variations due to external factors. Basal expression levels of Diptericin, Drosomycin, Defensin ( Figure 4C, D) in whole animals and of Diptericin, Drosomycin3 and Defensin ( Figure 4E, F) in guts were not significantly different in controls and Toll-9 2/2 mutants (for each experiment, P$0.06). We confirmed this result using antimicrobial peptide reporter lines. Indeed, uninfected wild-type and Toll-9 2/2 mutants show identical Diptericin-Cherry, Drosomycin-GFP and Cecropin-GFP expression pattern in both larvae and adult (constitutive Cecropin-GFP expression in the adult is shown in Figure 5A, B).
The data above suggest that contrary to what has been proposed earlier, Toll-9 is not required to maintain a basal level of AMPs production in vivo. We then tested whether Toll-9 is necessary to regulate immune gene induction after microbial challenge. As Toll-9 is mainly expressed in the gut, we used the Ecc15 oral infection model. First, we observed that the survival of Toll-9 2/2 mutants was not affected by Ecc15 oral infection ( Figure  S1). We next checked induction levels of Diptericin and Drosomycin in whole larvae and adult, and Diptericin and Drosomycin3 expression in the adult gut ( Figure 6). Since antimicrobial peptides gut induction has been shown to depend essentially on the Imd pathway, we used relish E20 as a control. Upon Ecc15 oral infection, Diptericin expression increased up to 120-fold in whole larvae ( Figure 6A), 25-fold in whole adults ( Figure 6C) and 6 times in adult guts ( Figure 6E). Similar induction levels were obtained in Toll-9 2/2 mutant background ( Figure 6A, C, E, for each experiment, P$0.25), whereas, as expected, Diptericin induction was abolished in a relish E20 mutant ( Figure 6C, E). Similarly, no significant variation was observed for Drosomycin and Drosomycin3 induction between wild-type and Toll-9 2/2 mutants ( Figure 6B, D, F). These results were confirmed by using reporter lines for Cecropin, Drosomycin and Diptericin (Cec-GFP, Drs-GFP and Dpt-cherry, respectively), both in the larvae ( Figure 5C-F) and in the adult (data not shown). To complete this analysis, we performed oral infection using other bacteria species such as Pseudomonas entomophila, Serratia marcescens and Lactobacillus plantarum. Twenty hours following oral ingestion of these bacteria, Diptericin mRNAs induction was not significantly different in Toll9 2/2 versus control midguts ( Figure S2, for each experiment, P$0.1). From these experiments we concluded that the basal and oral infection- induced antimicrobial gene levels are unaffected by Toll-9 2/2 inactivation. This suggests that Toll-9 does not participate in maintaining basal expression levels of antimicrobial peptides genes in normal conditions or after oral immune challenge.
To complete these experiments, we analyzed AMP production and survival of Toll-9 2/2 flies infected by septic injury. Our results showed that Toll-9 2/2 mutant flies resist as well as control flies to clean injury and to septic injuries performed with either Grampositive or Gram-negative bacteria (Figure 7). We next wanted to test whether the humoral antimicrobial response was affected in Toll-9 2/2 infected mutant. Diptericin mRNA induction upon Ecc15 infection was similar in Toll-9 2/2 mutant and in controls ( Figure  S3), indicating that Imd dependent AMP production response was not affected by the absence of Toll-9.

Gut homeostasis is normal in Toll-9 mutant
Recent studies have revealed that bacterial infection can trigger a gut immune response that involves the production of Reactive Oxygen Species and antimicrobial peptides [52][53][54][55]. Associated with this mechanism of defense, the gut epithelium triggers the proliferation and differentiation of intestinal stem cells (ISCs, stained with Delta (Dl) and Escargot (Esg), [52]. Upon stimulation, the ISCs give rise to two intestinal cell types: enterocytes (young enterocytes still express Esg) and enteroendocrine cells [56]. It has been shown that oral infection with some bacteria species can stimulate epithelium renewal by inducing ISCs proliferation and therefore production of enterocytes [52,53]. We then tested whether the gut homeostasis was modified in a Toll-9 2/2 mutant. As shown in Figure 8, the numbers of ISCs (Esg-positive and Dl-positive cells) and young enterocytes (Esg-positive and Dl-negative cells) were similar in unaffected Toll-9 2/2 mutants and in control guts ( Figure 8A-A'' and B-B''). Upon infection, the epithelium renewal occurred normally in a Toll-9 2/2 mutant midguts ( Figure 8C-C'' and D-D''). These results indicate that neither the basal gut renewal in non-infected guts nor the epithelium renewal triggered by bacterial infection is dependent on Toll-9 function.

Discussion
The findings that TLRs are implicated in the immune response in mammals and that Toll participates in the Drosophila host response has led to the idea that TLR-mediated immunity is originating from ancestor of bilaterian (review in [18]). If such, one might expect multiple mammalian TLRs and Drosophila Toll family members to have immune function. If it is now clearly demonstrated that every single mammalian TLR family member is implicated in an immune mechanism, it is however not the case for Drosophila for which evidence for immune role has only been gathered for the Toll-1 protein itself [12,57,58]. In contrary, there are some reasons to think that the immune implication of TLRs and Toll have been co-opted independently in insects and mammals [36,59,60]. This hypothesis is based on functional differences between Toll and TLR implication in immunity, the most obvious being the difference in their mode of activation. Whereas TLR receptors are directly activated by versatile microbial ligands, Drosophila Toll-1 receptor activation is dependent of the maturation of its Spä tzle ligand by proteolytic cleavage [18,61,62]. The goal of this study was to test whether another member of the Toll family other than Toll-1 itself could have an immune function in flies. We focused our attention on Toll-9 for two reasons. 1) Its extracellular domain organization resembles that of the TLRs [35], and 2) published gain-of-function experiments have previously suggested that Toll-9 could function as a constitutively active receptor triggering basal AMP production [37]. However, the phenotypic characterization of a complete lossof-function mutant presented here tends to indicate that Toll-9 is not implicated in the antibacterial arm of the immune response. Indeed, Toll-9 is neither required for basal and immune inducible AMP production nor for intestinal response to oral bacterial infection. In addition, our results rules out the hypothesis that Toll-9 is triggering a constant baseline level of AMPs required to eliminate pathogens when they enter the body cavity. We can however not exclude that a functional Toll-9 receptor is not of importance to mediate other aspects of the immune response such as anti-viral or and-parasitic immunities. It is of note that the Toll signaling pathway control anti-dengue virus in Aedes aegypti [63]. It is also possible that Toll-9 is functionally redundant with other Toll proteins, although the rather unique extracellular domain organization of Toll-9 let us believe that this is unlikely. What is then the function of Toll-9? We have identified an incompletely penetrant embryonic lethality associated with the loss of Toll-9. It is however difficult to pinpoint the exact cause of this lethality. The only clear phenotype associated with Toll-9 elimination is an increase of blood cell number. It is however well know that the number of circulating hemocytes is highly dependent on genetic background and can strongly fluctuate even in between wild type stocks. We therefore do not favor the hypothesis than the high hemocytes figures in Toll-9 2/2 mutants are responsible for embryonic lethality. Further work will be needed to dissect the precise role of Toll-9 during embryogenesis. It should however be noted that both TLRs and Drosophila Toll-1 have been shown to be important for non-immune function. Whereas Toll-1 is also involved in early axis determination and in specifying neuromuscular innervation pattern, the number of manuscripts describing TLR expression and function in the central nervous system (CNS) has been increasing steadily and expanding beyond their traditional roles in infectious diseases to neurodegenerative disorders and injury [14,[64][65][66][67][68][69][70][71]. This might be a route to follow for finding Toll-9 function! Supporting Information Figure S1 Toll-9 2/2 mutant survival is not affected by Ecc15 oral infection. 7 days after oral ingestion of the Ecc15 bacteria, Toll-9 2/2 mutant flies survive as well as the control. (TIF) Figure S2 Gut Diptericin expression after bacteria feeding is not modified in Toll-9 2/2 mutant. Quantitative RT-PCR analysis of Diptericin expression in Toll-9 2/2 mutant and control guts after oral challenge with Ecc15, Lactobacillus plantarum, Pseudomonas entomophila and Serratia marcescens. Relative Diptericin/rp49 ratios of unchallenged controls were set to 1 to indicate fold induction. (TIF) Figure S3 Diptericin mRNA levels after Ecc15 septic injury are comparable in wild-type Toll-9 2/2 adults. Quantitative RT-qPCR analysis of Diptericin mRNA 6 and 24 hours after septic injury with Ecc15 in wild-type and Toll-9 2/2 mutant adults. Diptericin/rp49 ratio at 6 hours post challenged was set as 100%. (TIF)

Author Contributions
Conceived and designed the experiments: KNR BC JR. Performed the experiments: KNR BC. Analyzed the data: KNR. Contributed reagents/ materials/analysis tools: KNR BC. Wrote the paper: KNR JR.