Functional Characterization of Paillotin: An Immune Peptide Regulated by the Imd Pathway with Pathogen-Specific Roles in Drosophila Immunity

Abstract

Insects, such as Drosophila melanogaster, rely on innate immune defenses to combat microbial threats. Antimicrobial peptides (AMPs) play an important role in limiting pathogen entry and colonization. Despite intensive research into the regulation and biochemical properties of AMPs, their exact significance in vivo has remained uncertain due to the challenges of mutating small genes. Fortunately, recent technologies have enabled the mutation of individual AMP genes, overcome previous obstacles, and opened new avenues for research. In this study, we characterized one novel host-defense peptide, Paillotin (IM18, CG33706), using loss-of-function mutants. Paillotin is an ancient host defense peptide of Diptera, regulated by the Imd pathway. Loss of Paillotin does not impact the activity of either the Imd or Toll pathways. Importantly, we found that Paillotin mutants are viable but exhibit increased susceptibility to specific infections, particularly Providencia burhodogranariea. Paillotin was further found to contribute synergistically to defense against P. burhodogranariea when combined with other AMPs. However, we did not detect direct microbicidal activity of Paillotin in vitro in our hands. Taken together, our findings identify Paillotin as a novel host defense peptide acting downstream of Imd signaling, advancing our understanding of the Drosophila antimicrobial response.

Introduction

Insects inhabit environments rich in microbes, necessitating a robust immune system to defend against opportunistic pathogens [1]. Unlike vertebrates, insects rely solely on innate immune defense mechanisms. The conservation of innate immunity across Metazoans and the powerful genetics of Drosophila melanogaster make this insect a powerful model for studying immune responses [2–4].

Flies exhibit behavioral immunity to prevent pathogens from entering their body cavity, such as grooming and avoidance of pathogens [5,6]. Once pathogens breach this barrier, immune responses are activated in different body regions to limit growth. Local immunity occurs at initial pathogen entry sites, like the gut or trachea. Systemic immunity involves cellular and humoral responses that take place in the body cavity upon systemic infection. Cellular responses involve hemocytes (blood cells), while humoral responses involve the production of host defense peptides as well as other effectors secreted by hemocytes and the fat body, an organ analogous to the mammalian liver [1,2]. Two NF-κB signaling pathways, the Toll and Imd pathways, regulate the humoral response. The principal effectors downstream of the Toll and Imd signaling pathways consist of AMPs, which are synthesized by both the fat body and hemocytes and then secreted into the hemolymph. AMPs are positively charged and can bind to the negatively charged membranes of microorganisms, disrupting membrane integrity through pore formation [1,7]. Upon infection, their expression is markedly increased to high levels in Drosophila, notably Drosomycin and Diptericin, which are often used as readouts for Toll and Imd activity, respectively [8]. Besides microbe killing, AMPs have been implicated in multiple processes such as tumor suppression [9,10], microbiota control [11], and neurodegeneration [12,13]. Currently, D. melanogaster is known to possess eight families of inducible AMPs, including antifungal peptides like Drosomycin, Baramicin, and Metchnikowin; Cecropins (four isoforms) exhibiting both antibacterial and antifungal activities in vitro; and Drosocin, Attacins (four isoforms), Defensin, and Diptericins (two isoforms), primarily displaying antibacterial activity [14–21]. Additionally, the Drosophila genome encodes numerous other host defense peptides, such as the Toll-regulated Daisho genes (two genes), Bomanins (twelve genes), peptides encoded by Baramicin A, and the Imd-regulated peptide Buletin (product of the Drosocin gene), whose antimicrobial activity in vitro remains to be fully demonstrated, but functional studies have indicated their importance in vivo for surviving microbial infections [22–26].

Functional studies using single or combined mutations have demonstrated that these host defense peptides largely account for the critical roles of the Imd and Toll pathways in host survival following infection with Gram-negative bacteria, Gram-positive bacteria, or fungi. The antibacterial peptides Diptericin, Cecropin, Attacin, Drosocin, and Buletin account for most of the Imd pathway–mediated resistance to Gram-negative bacterial infection [25,27–29]. Drosomycin, Metchnikowin, Daisho (two genes), Baramicin A mediate much of the Toll-mediated resistance to fungal infection [23,24,26,27,30], and Bomanins (twelve genes) appear to be the most important family of Toll-regulated host defense peptides contributing to resistance against Gram-positive bacteria and fungi [22,27,31]. These host defense peptides display both additive and synergistic effects, along with a high degree of specificity, whereby individual peptides greatly contribute to survival against a defined pathogen [27,29,32]. Some immune-induced peptides regulated by Toll and Imd have regulatory functions. This includes GNBP-like3 and Ibin A/B which have been reported to modulate Toll pathway activity, as well as Edin, which has been implicated in regulating hemocyte sessility [33–35].

Despite significant advances in understanding Drosophila host defense peptides, many peptides remain underexplored. The Drosophila genome encodes several other immune-induced peptides that remain poorly characterized, including notably Listericin, CG4269, and IM18 [1,36,37].

In this article, we have characterized the function of IM18, which we renamed Paillotin. We showed that Paillotin is an ancient host defense peptide regulated by the Imd pathway, which does not affect the regulation of the Imd or Toll pathways. Importantly, we found that Paillotin mutants exhibit increased susceptibility to specific infections, particularly Providencia burhodogranariea. Paillotin was also found to contribute synergistically to defense against P. burhodogranariea when combined with other AMPs, specifically Drosocin, Diptericins, and Attacins. However, we did not detect direct microbicidal activity of Paillotin in vitro. Taken together, our findings identify Paillotin as a novel host defense peptide downstream of Imd signaling, advancing our understanding of the Drosophila antimicrobial response.

Results

Paillotin is a conserved host defense peptide in Diptera

A previous proteomic analysis identified an immune inducible peptide called Immune-induced Molecule 18 (IM18) based on its mass by MALDI-TOF [38]. We rename the IM18 gene (CG33706) as Paillotin (Pai) from André Paillot, who in 1920 discovered the humoral immune response of insects [39,40]. The gene encoding Paillotin was identified as CG33706 in a region annotated as polycistronic with another gene, CG10332, making it challenging to distinguish their expression patterns (Figure 1a); the CG10332-RA transcript is identical to the Pai-RA transcript and contains the full coding sequence of both CG10332 and Pai. Indeed, in FlyAtlas data, IM18 and CG10332 are shown with identical expression levels [41]. However, the characterization of IM18 as an immune-induced peptide pointed to the existence of two distinct gene products, perhaps regulated by the distinct transcripts Pai-RB and CG10332-RA. To confirm this, we monitored the expression of either CG10332-RA alone or the cumulative products of CG10332-RA and Pai-RB upon septic injury with either the Gram-negative bacterium Escherichia coli (E. coli) or the Gram-positive bacterium Micrococcus luteus (M. luteus). The results showed that the mRNA encoding Pai-RB (Figure 2a, b) but not CG10332-RA was strongly induced upon septic injury (Figure 2c, d), confirming that only Pai-RB is immune-induced at the transcriptional level.

Figure 1. The Paillotin gene structure and conservation.

(a) Schematic representation of the Paillotin (Pai) locus. The Pai-RB transcript is annotated as polycistronic, sharing a promoter and transcript with CG10332 (CG10332-RA), indicating both gene CDS regions are co-transcribed. Promoter regions (250 bp upstream) for Pai-RB and CG10332-RA are shown in cyan. Colored triangles indicate putative NF-κB binding motifs: Relish-responsive motifs (GGRDNNHHBS) in teal, Dif/Dorsal-responsive motifs (GGGHNNNDVH) in purple. Red triangles represent sites containing overlapping matches to both motifs. (b) Sequence alignment of Paillotin orthologues across diverse Brachyceran flies revealed a conserved protein structure, including a signal peptide, a dipeptidyl peptidase cleavage site (XA/XP), and the mature Paillotin peptide. Highlighting corresponds to residues with >50% identity across the alignment.

Figure 2. Paillotin is regulated by the Imd pathway.

Transcript levels of Paillotin/CG10332 in w¹¹¹⁸, Relish^E20, spz^rm7, and Relish^E20, spz^rm7 flies collected at 8 h, 16h, and 24h post-E. coli infection (a, c) and M. luteus infection (b, d). (e) Quantified Paillotin peptide levels of w¹¹¹⁸, Relish^E20, and spz^rm7 hemolymph upon either Ecc15 or M. luteus infection (data extracted from (Rommelaere et al., 2025)). Gene Expression was measured by qRT-PCR and normalized with w¹¹¹⁸ UC (unchallenged) flies set as a value of 1. The data shown in the figure are based on three independent experiments, with at least 30 flies per genotype at each time point in each experiment. Error bars represent SEM. Statistical significance was determined using two-way ANOVA (*p<0.05, ** p<0.01, *** p<0.001, n.s. = not significant, p>0.05).

The Paillotin gene encodes a 71-residue precursor peptide composed of a signal peptide, dipeptidylpeptidase motif, and mature peptide (Figure 1b). To investigate the evolutionary conservation of Paillotin, we used a recursive tBLASTn approach to screen outgroup fly species [42]. Orthologues of Paillotin were identified in Diptera families such as Tephritidae and Diopsidae, which share a common ancestor with D. melanogaster approximately 126 million years ago [43] (Figure 1b). The C-terminal domain displayed the highest level of conservation across species, suggesting a critical role in Paillotin function. The presence of conserved features, such as the two-residue dipeptidyl peptidase cleavage site (XA/XP) and a positive charge (+3.89 in D. melanogaster) commonly found in AMPs [24,44], supports the hypothesis that Paillotin plays a key role in host defense and could have microbicidal activities. This evolutionary conservation underscores Paillotin as an ancient host defense peptide in Diptera.

Paillotin is regulated by the Imd pathway

To explore the expression pattern and regulation of Paillotin during systemic immune responses, we analyzed transcript levels in wild-type, Toll, and Imd pathway-deficient flies (Figure 2). Specifically, we challenged w¹¹¹⁸ (wild-type), spz^rm7 (deficient for the Toll pathway), Relish^E20 (deficient for Imd pathway) and Relish^E20, spz^rm7 double mutant flies with either E. coli, an inducer of the Imd pathway (Figure 2a), or M. luteus, an inducer of the Toll pathway (Figure 2b). Expression of Paillotin was monitored at 8h, 16h, and 24h post-infection using qRT-PCR. We observed that Paillotin gene expression was strongly induced after both M. luteus and E. coli infections in w¹¹¹⁸ flies at all time points. However, this induction was completely abolished in Relish^E20 and Relish^E20, spz^rm7 double mutants, but not in spz^rm7 flies, indicating Paillotin is regulated by the Imd pathway (Figure 2a, b). Consistent with this expression profile, we found multiple putative NF-κB responsive motifs within the first 200bp of promoter sequence upstream of gene transcription start sites of Paillotin [24,45] including five motifs uniquely matching the Relish consensus motif sequence (GGRDNNHHBS), one motif matching a Dif/Dorsal (GGGHHNNDVH) binding site, and two κB motifs with putative matches to either Relish and/or Dif/Dorsal binding sites. In contrast, no NF-κB binding motifs were identified in the 250bp upstream region of IM18-RA/CG10332-RA, consistent with its non-immune inducibility.

A recent proteomic analysis of Drosophila hemolymph following bacterial infection [46] indicated that Paillotin peptide was strongly induced in the hemolymph of wild-type flies infected with M. luteus or the Gram-negative bacterium Pectobacterium carotovorum (Ecc15), and this induction was abolished in Relish^E20 flies (Figure 2e). Consistent with its regulation by the Imd pathway, Paillotin was more induced upon infection with Ecc15 than M. luteus. We conclude that Paillotin is a novel host defense peptide regulated at the transcriptional level by the Imd pathway and secreted into the hemolymph.

Identification and characterization of Paillotin mutants

The Drosophila genome nexus represents a global sampling of the genetic diversity of D. melanogaster wild types subjected to genome sequencing [47]. Using a combination of the genome nexus and more detailed variant information in the Drosophila Genetic Reference Panel (DGRP) [48], we identified two strains from Kenya (KR7, KR4N) encoding a premature stop codon in Paillotin (“Gly37∗”, numbering per start codon). Globally, we could assess the Paillotin region in 789 genome-sequenced strains, finding 39 that encode putative loss-of-function mutations (Table S1). The majority of these mutant strains were collected from Winters California and Raleigh North Carolina, which have a high frequency of putative loss-of-function mutations (~26% and ~8% respectively within these populations). We characterized two of the loss-of-function mutations segregating in the DGRP further (Figure 3a). The first mutation (2R_19488505_DEL) is found in 17 genotypes from the DGRP, including DGRP-41, which deletes 61 nucleotides within the Paillotin mature peptide coding region, leading to a frameshift and protein truncation (referred to as PaiΔ41). The second mutation (2R_19488690_SNP), found in DGRP-370 (Pai Δ³⁷⁰), involves a C-T transition that disrupts the Paillotin start codon. We utilized these naturally-occurring mutations to study the role of Paillotin in host defense by backcrossing PaiΔ⁴¹ and PaiΔ³⁷⁰ into the w¹¹¹⁸ iso DrosDel background for seven generations.

Figure 3. Description and validation of PaiΔ⁴¹, PaiΔ³⁷⁰, and PaiΔ^SFHI mutations.

(a) Multiple sequence alignment between the Dmel_R6 reference genome Paillotin and Paillotin peptide variants. Conserved residues are highlighted by color coding according to sequence similarity. (b) MALDI-TOF proteomic analysis of hemolymph from w¹¹¹⁸, PaiΔ⁴¹, PaiΔ³⁷⁰, and PaiΔ^SFHI male flies collected 24 hours after septic injury with Ecc15.

In addition, we used a third mutant referred to as PaiΔ^SFH, a gift from the Guangzhou Drosophila Resource Center (Figure 3a), which was generated by CRISPR-Cas9 and involves a 20-nucleotide deletion that causes a complex frameshift, resulting in a nonsense peptide. The precise nucleotide alterations associated with each Paillotin mutant allele used in this study are presented in Supplementary Figure S1. For PaiΔ^SFHI, only the X chromosome was replaced by the w¹¹¹⁸ iso DrosDel X chromosome.

RT-qPCR confirmed reduced Paillotin gene expression in PaiΔ⁴¹ and PaiΔ³⁷⁰, while Paillotin expression was completely absent in PaiΔ^SFHI (Figure S2). We next performed a MALDI-TOF analysis of hemolymph from adult flies infected with P. c. carotovara Ecc15. The results revealed the absence of the 4625Da Paillotin peaks in PaiΔ⁴¹, PaiΔ³⁷⁰, and PaiΔ^SFHI flies (Figure 3b). All the mutants were viable without morphological defects.

Paillotin does not regulate the Imd and Toll pathways

Some host defense peptides, such as mammalian LL37 or β-Defensins, have been shown to function similarly to cytokines, modulating immune system activity [49,50]. This prompted us to assess whether Paillotin could influence Toll and Imd pathway activity. To address this, we monitored the expression of Diptericin A (a readout of the Imd pathway) in wild-type (w¹¹¹⁸) and Paillotin -deficient flies (PaiΔ⁴¹ and PaiΔ³⁷⁰) at 6- and 12-hours post-infection with Ecc15 and Drosomycin (a readout of the Toll pathway) at 8- and 24-hours post-infection with M. luteus (Figure 4a, b). The expression levels of Diptericin A in PaiΔ⁴¹ and PaiΔ³⁷⁰ mutants were comparable to those in wild-type flies at both time points, indicating no effect of Paillotin on Imd pathway activity. Similarly, the expression of Drosomycin in PaiΔ⁴¹ and PaiΔ³⁷⁰ mutants showed no significant difference compared to w¹¹¹⁸, suggesting that Paillotin is not involved in regulating the Toll pathway. Together, these results demonstrate that Paillotin is not a regulator of the Imd or Toll pathways.

Figure 4. Paillotin does not impact the Imd or Toll pathways.

The expression of Diptericin A (a) at 6 or 12h post-infection with Ecc15 and Drosomycin (b) at 8h or 24h post-infection with M. luteus were monitored in w¹¹¹⁸, PaiΔ⁴¹, and PaiΔ³⁷⁰ flies. Relish^E20 and spz^rm7 were used as the positive control. Expression was normalized to w¹¹¹⁸ UC (unchallenged) flies set as a value of 1. The data shown in the figure are based on three independent experiments, with at least 30 flies per genotype at each time point in each experiment. Error bars represent SEM. Statistical analysis was performed using two-way ANOVA. For Diptericin A, significant effects were detected for genotype (F(3,24) = 148.1, p < 0.0001), time (F(2,24) = 332.8, p < 0.0001), and their interaction (F(6,24) = 38.98, p < 0.0001). For Drosomycin, significant effects were also observed for genotype (F(3,24) = 72.82, p < 0.0001), time (F(2,24) = 191.5, p < 0.0001), and interaction (F(6,24) = 22.84, p < 0.0001). However, Tukey’s post hoc comparisons revealed no significant difference between w¹¹¹⁸ and Paillotin mutants at any time point (n.s. = not significant).

Paillotin mutant flies are susceptible to P. burhodogranariea infection

Recent studies have demonstrated that the deletion of a single AMP gene can significantly increase susceptibility to specific pathogens. For instance, the Diptericin A mutation leads to marked susceptibility to Providencia rettgeri, while the Diptericin B mutation greatly reduces resistance to Acetobacter sp. [29,32]. To investigate the role of Paillotin in resistance to infections, we examined its contribution against a range of Gram-negative, Gram-positive bacteria, and fungi. In these experiments, we compared the survival of PaiΔ⁴¹, PaiΔ³⁷⁰, and PaiΔ^SFHI flies to that of the wild-type (w¹¹¹⁸), Relish^E20, Bom^Δ55, and spz^rm7 immune-deficient mutants. Flies were systemically infected by bacteria using a needle dipped in a bacterial pellet, a mode of infection that causes cuticle injury. To exclude a role of Paillotin in resistance to injury, we first monitored the susceptibility of Paillotin mutants to clean injury, and did not detect any susceptibility after clean injury alone (Figure S3). Regarding Gram-negative bacteria, there was no statistically significant difference in survival between wild-type and Paillotin mutants following infections with P. rettgeri, Ecc15, or Enterobacter cloacae (Figure S4a-c). However, all three Paillotin mutant flies (PaiΔ⁴¹, PaiΔ³⁷⁰, PaiΔ^SFHI) exhibited a modest but significant increase in susceptibility to septic injury with Providencia burhodogranariea (P. burhodogranariea) (Figure 5a). This susceptibility was less marked than Relish^E20 flies, indicating that other factors downstream of the Imd pathway contribute to survival upon P. burhodogranariea infection. In addition, we observed higher P. burhodogranariea bacterial load in Paillotin mutants compared to wild-type, indicating that Paillotin suppresses bacterial growth (Figure 5b). For Gram-positive bacteria, Paillotin mutants showed no consistent increased susceptibility to Enterococcus faecalis (Figure 5c), Bacillus subtilis, Listeria monocytogenes, or Staphylococcus aureus infections (Figure S4d-f). Similarly, fungal infections with Beauveria bassiana (Figure 5d), Metarhizium anisopliae, or Metarhizium robertsii did not reveal any increased susceptibility of Paillotin mutants compared to wild-type (Figure S4g-h).

Figure 5. Paillotin mutants are susceptible to P. burhodogranariea infection.

(a) The three Paillotin mutants exhibit increased susceptibility following septic injury with P. burhodogranariea at the indicated bacterial inoculum. (b) Increased bacterial load in Paillotin mutants at 30 hpi. Bacterial load was log₁₀-transformed prior to statistical analysis. One-way ANOVA revealed a significant genotype effect (F(3,30) = 18.95, p < 0.0001). Post hoc Dunnett’s test confirmed that all three mutant lines carried significantly higher bacterial loads than w¹¹¹⁸ (p < 0.0001 for all comparisons). Data points represent the mean bacterial load per fly, derived from individual experiments using pools of four flies per genotype following P. burhodogranariea infection. No consistent effect of the Paillotin mutation on resistance to the Gram-positive bacterium Enterococcus faecalis (c) or the fungus Beauveria bassiana (d). Survival data were analyzed using the Kaplan–Meier method. Statistical significance is indicated relative to the w¹¹¹⁸ (*** p<0.001, n.s. = not significant, p>0.05). N=total number of flies in experiments.

No in vitro activity of Paillotin peptide

Cecropins were the first inducible AMPs identified in moths [14]. In vitro studies have shown that Cecropins are highly effective against a wide range of Gram-negative bacteria at concentrations lower than those induced in insects during infection (25–50 µM) [51]. To investigate the antimicrobial potential of Paillotin in vitro, we produced the peptide and evaluated its activity against P. burhodogranariea at 100µM alone or in combination with sublethal concentrations of Cecropin A. However, Paillotin exhibited no detectable microbicidal activity against P. burhodogranariea in vitro (Figure S5a). It also failed to inhibit the growth of other Gram-negative bacteria, including Ecc15 and E. coli (Figure S5b, c). Thus, we found no activity of Paillotin in vitro under these conditions.

Paillotin synergistically contributes to the defense against P. burhodogranariea with other AMPs

Recent studies have demonstrated synergistic or additive cooperation among AMPs, notably against P. burhodogranaria. While single mutations in Drosocin, Attacins, or Diptericins have minimal overall effect, a compound mutant referred to as the “Group B” AMP mutant lacking Drosocin, Attacins A, B, C and D, and Diptericins A and B [full genotype: w¹¹¹⁸; AttC^Mi, Dro-AttAB^SK2, Dpt^SK1; AttD^SK1], displays marked susceptibility to P. burhodogranariea infection [27]. To investigate potential interactions between Paillotin and these AMPs, we used genetic crosses to combine the Paillotin mutation with Group B mutants. We initially infected Group B mutants, Paillotin mutants, and Group B, Paillotin compound mutants [full genotype: w¹¹¹⁸; AttC^Mi, Dro-AttAB^SK2, Dpt^SK1, Pai] with P. burhodogranariea. However, we observed no synergistic or additive effects, as the introduction of Paillotin mutations into Group B mutants did not further increase susceptibility. This lack of an observable effect may be due to the already high susceptibility of Group B AMP mutants to this bacterium, even at a low inoculum (OD₆₀₀=0.1), potentially masking any additional contribution from the Paillotin mutation (Figure 6, dashed curves). To circumvent this, we assessed the impact of Paillotin homozygous deficiency in flies heterozygous for Group B AMP mutations (Group B/+), which depresses the inducibility of AMPs compared to homozygotes [27]. Notably, in this sensitized genetic background, the absence of Paillotin, regardless of the allele used, significantly increased susceptibility of Group B/+ flies to P. burhodogranariea (Figure 6a-c, solid curves). This co-occurring loss of resistance reflects a synergistic effect on the Hazard Ratio (HR), evidenced by increased mortality beyond additive effects (Paillotin alone, HR =-0.17, p =0.01; Group B alone, HR =-1.06, p<0.001; Paillotin*Group B interaction, HR =-0.15, p=0.02). In addition, a marginally significant trend consistent with a synergistic effect between Paillotin and Group B mutants was also seen for infections with the Gram-negative bacteria Providencia rustigianii and Providencia vermicola, but not E. coli (Figure S6). This supports the notion that Paillotin confers a fairly specific resistance to P. burhodogranariea, perhaps also extending to other Providencia species.

Figure 6. Paillotin synergistically contributes to defense against P. burhodogranariea with other Imd-regulated AMPs.

(a-c) Survival curves showing the synergistic effects between different Paillotin mutants (PaiΔ⁴¹, PaiΔ³⁷⁰, PaiΔ^SFHI) and Group B mutations (Drosocin, Attacin A, B, C, and Diptericin A, B) in the heterozygous background (Group B/+). Survival curves were analyzed using the Cox proportional hazards (CoxPH) model in R version 3.6.3. (** p<0.01, *** p<0.001, n.s. = not significant, p>0.05). N=total number of flies in experiments.

Overexpression of Paillotin restores resistance in Paillotin-deficient flies against P. burhodogranariea

The observation that three distinct null mutations in Paillotin are associated with a higher susceptibility and the synergistic effect between Paillotin and other AMPs against infection provide strong evidence for a role of this host defense peptide in resisting P. burhodogranariea. To reinforce this finding, we performed a genetic rescue experiment by combining the Paillotin mutation with a UAS-Pai transgene. We initially observed that overexpression of Paillotin in a wild-type background was not sufficient to enhance resistance to P. burhodogranariea (Figure S7). In contrast, overexpression of Paillotin using either the C564-Gal4 (a Gal4 driver with strong expression in the fat body) or Actin-Gal4 (ubiquitous) driver restored pathogen resistance to near wild-type levels in the PaiΔ⁴¹ (Figure 7a) and PaiΔ³⁷⁰(Figure 7b) mutant backgrounds. The observation that three distinct Pai alleles confer susceptibility, together with the results of the rescue experiment, collectively demonstrates that Paillotin acts as a pathogen-specific host defense peptide essential for restricting P. burhodogranariea proliferation.

Figure 7. Over-expression of Paillotin rescues the resistance of Paillotin-deficient flies against P. burhodogranariea.

(a-b) Overexpression of Paillotin via a combination of the C564-Gal4 or Actin-Gal4 and UAS-Pai constructs enhances the resistance of the Paillotin mutant to P. burhodogranariea infection. (c) Overexpression of Paillotin in the Relish^E20 background failed to restore resistance. Survival curves were analyzed using the Kaplan-Meier method. (*** p<0.001, n.s. = not significant, p>0.05). N=total number of flies in experiments.

The Imd and Toll signaling pathways are essential for insect immunity, with double pathway-deficient flies exhibiting extreme susceptibility to infection due to their inability to upregulate hundreds of immune effector genes, particularly AMPs [52]. Previous work has demonstrated that transgenic overexpression of individual AMPs in some cases is sufficient to restore resistance in Imd, Toll-deficient flies [53]. Building on this paradigm, we generated flies overexpressing Paillotin using Actin-Gal4 or C564-Gal4 drivers in Relish^E20 mutant flies (Figure 7c). However, Paillotin overexpression failed to confer protection against P. burhodogranariea infection. While we cannot exclude that Paillotin was not expressed at a sufficient level to overcome the extreme susceptibility of Relish mutants, this suggests that Paillotin alone is insufficient and the host requires the presence of other AMPs for the defense response to function effectively, consistent with the synergistic interactions observed with Group B/+ heterozygous mutants.

Discussion

In this study, we characterized Paillotin, an evolutionarily conserved host-defense peptide regulated by the Imd pathway. Use of three Paillotin mutants reveals a role of Paillotin in the resistance to P. burhodogranariea infection, and perhaps other moderately virulent Providencia species, without altering survival to other tested bacterial or fungal species. Furthermore, the Paillotin mutation enhances the susceptibility to P. burhodogranariea in flies heterozygous for mutations in Drosocin, Diptericins A and B, and Attacins A, B, and C. Taken together, the consistent susceptibility of three independent Paillotin alleles to P. burhodogranariea, accompanied by elevated bacterial loads, the resistance of Paillotin-deficient flies to other tested pathogens, and the successful rescue of the phenotype by overexpression, all provide strong evidence that Paillotin functions as a host defense peptide specifically effective against this bacterium. A high degree of specificity has been observed for other host defense peptides, whose inactivation frequently causes pathogen-specific immune deficiency. This has been illustrated not only for Diptericin A and B as described previously [29,32], but also for Drosocin, which provides significant protection against E. cloacae [25,27]. The existence of such specificity between a defined host defense peptide and particular pathogens suggests that these genes have been evolutionarily maintained under selective pressure imposed by specific microbes encountered in natural environments [54].

Previous studies have shown that resistance to P. burhodogranariea involves multiple antimicrobial peptides, including Buletin (encoded by the Drosocin gene), Diptericins A and B, and Attacins A, B, C, and D. While single AMP mutations in many genes do not greatly impact survival following infection with P. burhodogranariea, compound mutants lacking Drosocin, Diptericins and Attacins exhibit increased susceptibility, suggesting the presence of a synergistic effect among these peptides [27]. More recently, the absence of Buletin—a peptide generated by furin cleavage of the Drosocin precursor—results in notable survival differences following P. burhodogranariea infection [25]. Thus, host resistance to P. burhodogranariea is mediated by the collective action of multiple host defense peptides, and our study now adds Paillotin to this list. Future studies may further reveal additional roles of Paillotin, either alone or in combination with other host defense peptides, in resistance to other pathogens.

We were unable to detect any direct in vitro microbicidal activity of Paillotin against P. burhodogranariea and other bacteria, either alone or in combination with Cecropin. This reflects a common mismatch between in vitro and in vivo activities. Despite demonstrating effective antimicrobial activity in vivo in relevant animal models, many AMPs show minimal or no activity in standard minimal inhibitory concentration (MIC) or minimal microbicidal concentration (MMC) assays when tested under physiological salt concentrations or in serum-containing environments [55–58]. Several factors could explain this lack of detectable activity: i) Paillotin may depend on host-specific factors for its function, such as the requirement for a cofactor or modulation by environmental cues (e.g., pH or ionic strength), or ii) synergistic interactions with other immune molecules that are only present in vivo. Similar synergistic mechanisms have been observed in other systems. For instance, in Galleria mellonella (the greater wax moth), an anionic antimicrobial peptide (anionic peptide 2) enhances lysozyme-mediated membrane perforation, facilitating the elimination of E. coli [59]. Furthermore, AMPs often enhance each other’s efficacy through complementary mechanisms of action [21,60]. For instance, the bumblebee linear peptides hymenoptaecin and abaecin exhibit a synergistic interaction against E. coli. Although abaecin alone shows no detectable activity at concentrations up to 200 μM, hymenoptaecin disrupts the bacterial membrane, allowing abaecin to enhance bactericidal effects at doses as low as 1.25 μM [61]. In Tenebrio molitor, RNAi-mediated knock-down of the defensin Tenecin-1 did not affect host survival but reduced bacterial load following S. aureus infection. However, simultaneous knock-down of Tenecin-1 with other AMPs, such as the coleoptericin Tenecin-2 or the attacin Tenecin-4, resulted in increased mortality and higher bacterial loads, highlighting functional complementarity among AMPs during infection [62]. In vitro experiments showed that Attacin, an antibacterial protein that accumulates in the hemolymph of the giant silk moth Hyalophora cecropia, compromises the permeability barrier function of the E. coli outer membrane. Pretreatment with Attacin significantly increases the sensitivity of E. coli to Cecropin B, another antimicrobial protein present in the hemolymph of H. cecropia [63]. In addition, recent studies have pointed to the role of host defense peptides in resilience mechanisms, including tolerance to host immune molecules [46] and pathogen toxins [26,31]. However, we did not detect a role for Paillotin in resilience to the fungal toxins Destruxin A or Verruculogen (Figure S8). Interestingly, in contrast to recent studies [26,31], we did not detect a susceptibility of Toll-deficient spz^rm7 flies to these two toxins (Figure S8). Of note, the description of DIMs has been done largely by collecting peptides in Trifluoroacetic acid, which is known to modify peptides by, for example, deamidating NG motifs [24,64]. It is possible that the endogenous peptide products of the host immune response differ from the peptides observed in proteomic analysis and synthetic peptides used in recent studies, in unknown ways. Early studies on AMPs typically found striking activity in vitro using peptide extracts purified from endogenous hemolymph or cell line supernatant [14,17,19,21,65]. Future studies may benefit from a return to such approaches, including gel overlay experiments that can simplify the isolation of the activity of single peptides from endogenous samples [66] or collecting crude peptide extracts from overexpressing cells or larval hemolymph. However, future studies might also explore whether Paillotin protects against virulence factors specific to P. burhodogranariea. For now, despite displaying many features of AMPs, we cannot conclusively classify Paillotin as a bona fide AMP. It should be classified as a host defense peptide until its molecular mode of action is fully determined.

The use of compound module-deficient mutants has revealed the major role of the Imd pathway, and to a lesser extent, melanization, but not the cellular response to survive P. burhodogranariea infection [67]. Since Paillotin does not affect Toll and Imd pathway activities, we cannot exclude a role of this peptide in the melanization reaction. However, this role would have to be subtle as Paillotin mutants display a wild-type blackening reaction upon clean injury (Figure S9a). Previous studies have shown that flies lacking 14 AMPs exhibit increased microbiota abundance and altered composition in old flies, suggesting that AMPs are involved not only in pathogen resistance but also in shaping the environmental microbiota [11]. However, we did not detect any significant differences in gut microbiota load between wild-type flies and Paillotin mutants (Figure S9b).

Interestingly, there is a relatively high frequency of loss-of-function mutations segregating in wild Drosophila populations (Table S1), suggesting Paillotin is not integral to the generic host defense. This observation is consistent with our results showing that the Paillotin mutation affects defense against only a few bacteria out of the pathogens tested, suggesting it is dispensable in defense against many microbes, but useful against some microbes like Providencia sp. whose presence is stochastic in wild-caught flies [68].

To conclude, our study characterizes a new Drosophila host defense peptide, Paillotin, which plays a specific role against P. burhodogranariea. IM18/Paillotin is the last gene encoding an immune molecule identified by MALDI-TOF in the 1998 study [38] to be characterized, following the successful characterization of Bomanins [30], Daisho [23], Buletin [25], and Baramicin [24,26]. Together, these studies and the present article provide important insights into the Drosophila humoral immune response, revealing a more complex picture than initially anticipated. Using the power of Drosophila genetics to dissect how immune effectors individually or collectively contribute to host defense provides a general principle on the architecture of innate immune systems.

Materials and Methods

Sequence Comparisons

Genomic sequences were obtained from GenBank default reference assemblies [69], from the DPGP via the PopFly website (www.popfly.uab.cat)[47], and DGRP sequence data from the DGRP database [48] (http://dgrp2.gnets.ncsu.edu/). Sequence comparisons and alignment figures were prepared using Geneious R10 [70]. Putative NF-κB binding sites were annotated using the Relish motif “GGRDNNHHBS” as described in Copley et al [45], along with a manually curated amalgam motif “GGGHHNNDVH” described in Hanson et al [24].

Paillotin annotation across species was performed as described in Hanson et al [42]. In brief, a recursive tBLASTn approach was used with an artificially high E-value (E < 1000) to compensate for the short length of the AMP query sequence. All the best BLAST hits were assessed manually. If no Paillotin was found in a genome or lineage, only the conserved C-terminus was used as a query in a repeat attempt (improving sensitivity). No Paillotin candidates were discovered outside Brachycera, which included screens of the sequenced nematoceran genomes of various mosquitoes, Chironomus riparius, Clogmia albipunctata, Tipula unca, Lutzomyia longipalpis, Phlebotomus papatasi, Sylvicola cinctus, and Bibio marci.

Fly Strains

Flies were maintained at 25°C on standard cornmeal-agar media. w¹¹¹⁸ flies were used as wild-type controls. The positive controls for infection assays for Gram-positive and Gram-negative infections were spz^rm7 and Relish^E20. The PaiΔ⁴¹ and PaiΔ³⁷⁰ mutations are loss-of-function variants in the IM18 gene within the DGRP, and PaiΔ^SFHI was generated via CRISPR/Cas9. The guide RNAs that targeted Pai were cloned into the pUAST-attB plasmids before injecting into the y [1] M {RFP [3xP3.PB] GFP[E.3xP3] =vas-int.Dm}ZH-2A w [*]; M{3xP3-RFP.attP} ZH-86Fb (BL24749) to get gRNA transgenic flies. The gRNA flies were then crossed with flies containing nos-Cas9 to get mutants. The UAS-Pai constructs were prepared using the TOPO pENTR entry vector and cloned into the pTW system. Genetic variants were backcrossed into the DrosDel isogenic background over seven generations as previously described [71] using the following primers that are specific to an indel in the Pai locus common to both the PaiΔ⁴¹ and PaiΔ³⁷⁰, but which is absent in the DrosDel background: Pai_41F (ACGTACGGTCAAAAGCAACTACTTCG) and Pai_R3mut1 (CCAATCCCAAGGGCACTCGTATATG), amplified with a 66°C annealing temperature. All the fly strains used in the paper are listed in Table S2.

MALDI-TOF Proteomic Analysis

Hemolymph samples were collected from Ecc15-infected flies (OD600 = 200, 24 hours post-infection, 50 flies/sample) using a nanoinjector (Nanoject, Drummond Scientific, Broomall, PA) for MALDI-TOF analysis. Trifluoroacetic acid (0.1% TFA) was added to the hemolymph before mixing with the acetonitrile matrix. Samples are processed by SPE (Zip-Tip C18) before MALDI. Peaks were identified based on corresponding m/z values from previous studies [24,72] in ISIC’s Mass Spectrometry facility of EPFL. Spectra were visualized using Origin software, with figures prepared in Prism 10.

Gene Expression Analysis

Total RNA was extracted from pooled samples of ten flies using TRIzol. cDNA was synthesized using Takara Reverse Transcriptase. Quantitative PCR (qPCR) was performed using PowerUP SYBR Green Master Mix with the primers. Results represent the average of three independent experiments. Primers used in this study were: Drosomycin - F 5’-CGTGAGAACCTTTTCCAATATGAT-3’ Drosomycin - R 5’-TCCCAGGACCACCAGCAT-3’ Diptericin A - F: 5’-GCTGCGCAATCGCTTCTACT-3’ Diptericin A - R:5’-TGGTGGAGTGGGCTTCATG-3’ Rpl32-F: 5’-GACGCTTCAAGGGACAGTATCTG-3’ Rpl32 - R: 5’-AAACGCGGTTCTGCATGAG-3’ Pai - F: 5’-GATCCGTTTCCTGGTCCAGTGTG-3’ Pai - R: 5’-CCATGAAGCTGATCGCATTGTGC-3’ CG10332 - F: 5’-TCAGAGGCTGCACCATCAGAAGT-3’ CG10332 - R: 5’-CCGCTTCAGAGGCTGCACC-3’

Microbial Culturing Conditions

Micrococcus luteus, Erwinia carotovora carotovora (Ecc15), and Providencia burhodogranariea strain B were cultured overnight in LB broth at 29°C. Enterobacter cloacae, Escherichia coli strain 1106, Providencia rustigianii, Providencia vermicola, Providencia rettgeri, and Staphylococcus aureus were cultured overnight in LB at 37°C. Enterococcus faecalis, Bacillus subtilis, and Listeria monocytogenes were cultured overnight in BHI at 37°C. Metarhizium anisopliae and Metarhizium robertisii were cultured on malt agar plates at 25°C for around two weeks until sporulation was observed. Beauveria bassiana strain R444 commercial spores were produced by Andermatt Biocontrol, products: BB-PROTEC. Bacteria were pelleted at 4000 rpm for 15 minutes at 4°C, resuspended in PBS, and diluted to the desired OD600 value indicated in the figure legends. The fungal material was filtered through glass wool using 0.05% Tween-80 to collect spores and diluted to the desired concentration (in spores/ml). All microbial strains used in this study are listed in Table S3.

Microbiota Load Analysis

Flies were collected 30 hours post-infection and surface-sterilized with 70% ethanol. Whole-fly homogenization was performed in LB broth using a bead beater at 7200 rpm for 30 seconds. Serial dilutions of the homogenates were plated (100 µL per plate) on LB agar and incubated at 29 °C for 24 hours. Colony-forming units (CFUs) were then manually counted.

For gut microbiota load analysis, 18-day-old conventionally reared flies were collected and processed using the same procedure, except that MRS medium was used for culturing. Plates were incubated at 29 °C for 48 hours prior to CFU quantification.

In Vitro Antibacterial Activity

The Paillotin peptide (45 residues) was synthesized by GenicBio with >95% purity, and Cecropin A (from silk moth) was obtained from Sigma-Aldrich (≥97% purity). Peptides were dissolved in water. Bacteria were cultured overnight and diluted to OD600 = 0.005 in LB broth, then incubated for 2 hours. Paillotin peptide and Cecropin A were diluted in LB broth to the desired concentration. Bacterial cultures (1 µL) were added to peptide solutions in a 96-well plate, and OD600 was recorded every 10 minutes using a TECAN plate reader.

Survival Assays

Male flies (3-5 days old) were used for survival assays. The thorax of each fly was pricked with a needle (diameter ~0.2mm) dipped in a bacterial suspension. For the fungi infection, flies were rolled in the plates of A. fumigatus or the commercial spores of B. bassiana, or were dipped into a spore solution of M. anisopliae or M. robertisii. Mortality was recorded daily. Infected flies were maintained at 29°C or 25°C and transferred to fresh vials three times per week.

Toxin Injection

Destruxin A (Sigma) was dissolved in high-purity dimethyl sulfoxide (DMSO) and subsequently diluted with PBS to a final concentration of 8 mM. A volume of 9.2 nL of this solution was administered to flies using the Nanoject III microinjector (Drummond) [26]. Verruculogen (Abcam) was prepared in DMSO at a concentration of 1 mg/mL, and 4.6 nL of this solution was injected into the flies [31].

Wounding experiment

Clean injury was performed with a needle sterilized (diameter ~0.1mm) with an EtOH and PBS wash. Cuticle blackening was categorized into normal, weak, or no melanization 24h post-pricking at the injury site.

Data accessibility

All data and code used in this study are publicly available at Dryad [73].