Isolation and identification Drosophila-associated fungi
A strain of Drosophila-associated fungus was isolated from fly food with mold in accordance with standard protocols. This strain was typically a filamentous fungus with white, velvety-like mycelia and dark grey conidial masses (Fig. 1A and B). The conidiophores were bi-verticillate with smooth-walled stipes, bearing short conidial chains (Fig. 1C). It promptly grew at the optimal temperature of 28°C - 30°C and filled the plates (Φ = 90 mm) within 48 h. To confirm the reliability of the morphological identification, the strain was subjected to molecular identification based on an rDNA ITS sequence analysis. Based on the BLASTn search, it displayed > 99% similarity with a published sequence of Diaporthe sp. (identities = 555/558) and was relatively close to other Diaporthe members. To distinguish our isolate from other strains, it was henceforth termed, Diaporthe FY. For taxonomic reconstruction, the other 12 sequences, including out-group species, were retrieved from GenBank to generate a phylogenetic tree (Fig. 1D). Diaporthe species are among the most frequent endophytes of a wide-range of plants, including grapevines [21, 22]. Due to the saprophytic foraging behavior, flies might ingest Diaporthe from either food or surrounding resources.
Diaporthe FY is a potential pathogen of D. melanogaster
Drosophila frequently encounter a variety of commensal or pathogenic microbes in the wild. We first asked whether Diaporthe FY was beneficial or detrimental to Drosophila. To address this question, we examined the developmental timing and survival rate of the flies challenged with Diaporthe FY. Dechlorinated eggs were used to generate a specific interaction between hosts and specific microbes as previously described [23]. The results presented in Fig. 2A show that the hatched larvae rapidly succumbed to Diaporthe FY infection in vials containing more than 4 × 106 CFU spores. The eclosion duration of the flies infected with Diaporthe FY was extended compared to conventionally reared (CR) flies (Fig. 2A). This finding implies that Diaporthe FY impeded the normal development of Drosophila. Consistent with a previous study [24], Drosophila was susceptible to Aspergillus flavus (Suppl. 1A), suggesting that the pathogenic fungus-induced morbidity of Drosophila could be generated in our laboratory. Intriguingly, the developmental time of adults challenged with less than 4 × 106 CFU spores was shorter than that of germ free (GF) flies (Fig. 2A). This can be partially explained by the fact that the fungi could produce very low concentrations of toxic secondary metabolites during the exponential phase of nutritional growth; thus, its growth-promoting effects could override the inhibition of host development. Additionally, the survival of CR and GF adults fed the Diaporthe FY molds was significantly lower than that of the mock-infected flies (Fig. 2B), suggesting that Diaporthe FY reduced the relative survival of the flies. Moreover, innate immunity-associated genes were significantly triggered in the Diaporthe FY-infected flies compared to their counterparts (Fig. 2C), indicating that the flies developed a robust immune response to this invader. In agreement with a previous study [25], immune deficient PGRP-LC mutant flies were much more susceptible to Diaporthe FY than the wild-type flies (Fig. 2B), indicating that Diaporthe FY was a potentially virulent pathogen towards D. melanogaster. Hence, we subsequently investigated the survival to septic injury by injecting Diaporthe FY spores into the body cavity of flies. Concomitantly, the flies challenged with septically infection were more likely to die compared to uninfected flies (Fig. 2D), indicating that Drosophila was susceptible to Diaporthe FY. Collectively, these results suggest that Diaporthe FY functions as a Drosophila-associated pathogen.
L. plantarum undermines the susceptibility of Drosophila to Diaporthe FY infection
Given that pathogenic fungal infections impose morbidity and mortality upon animals in the wild, it was proposed that the natural host microbiota could promote the survival of flies challenged with Diaporthe FY [26]. Previous studies have shown that L. plantarum, due to its vast metabolic repertoire, fostered host development by accelerating their growth rate [17]. We then examined the antifungal response of L. plantarum against Diaporthe FY by simultaneously inoculating them into sterilized Drosophila GF eggs. Indeed, supplementation with L. plantarum efficiently rescued the lethality of the Diaporthe FY-infected flies, as well as ameliorated the delay of pupa formation and adult eclosion (Fig. 2E and F). This result suggest that L. plantarum mitigated Drosophila susceptibility to Diaporthe FY.
L. plantarum suppresses the growth of Diaporthe FY
To confirm that L. plantarum competes with Diaporthe FY, we tested the inhibition of L. plantarum on the growth of Diaporthe FY in vitro. Our data showed that L. plantarum outcompeted Diaporthe FY in a dose-dependent manner (Fig. 3A). We quantified the suppressive effects by colony growing, mycelia branching, and spore forming assays. First, the colony growth of Diaporthe FY was decreased by L. plantarum compared to the control (Fig. 3B and C). Secondly, there were fewer mycelia in the presence of L. plantarum (Fig. 3D). In addition, the number of spores was dramatically decreased following L. plantarum inoculation (Fig. 3E). Taken together, these results suggest that L. plantarum potently reduced the survivability of Diaporthe FY.
We further speculated that after dominating the niche, L. plantarum could thwart Diaporthe FY colonization. To this end, we pre-incubated the food with L. plantarum for different lengths of time (24 h, or 48 h) and added Diaporthe FY to the ‘‘modified’’ diet. In agreement with the simultaneous competition, the growth of Diaporthe FY was also hindered when it was pre-inoculated into the food with L. plantarum. Of note, the longest incubation period completely inhibited the growth of Diaporthe FY (Fig. 3F). This inhibitory effect was further fortified by the decreased number of mycelia and spores (Fig. 3G and H). Taken together, these findings support the L. plantarum-mediated inhibition of Diaporthe FY growth and dispersal.
Lactic acid inhibits the growth of Diaporthe FY
To further characterize the mechanism involved in L. plantarum-mediated inhibition of Diaporthe FY, we next sought to identify candidate inhibitory factors derived from L. plantarum metabolites. Lactic acid is generated by many lactic acid bacteria and exerts its antimicrobial effects by disrupting the cytoplasmic membrane or reducing the intracellular pH [27]. Since the strain of L. plantarum used in this study typically produced more than 75 mM (approximately 0.7% w/v) L-lactate at the end of fermentation (Suppl. 2), we therefore focused on the role of L-lactate on inhibiting the growth of Diaporthe FY. To determine whether lactic acid could inhibit the growth of Diaporthe FY, we scored the fungal growth on a medium supplemented with different concentrations of L-lactate. The results showed that the growth of Diaporthe FY was inhibited in a L-lactate dose-dependent manner (Fig. 4A). Diaporthe FY was modestly inhibited by 0.5% lactic acid and completely inhibited by 1% or higher doses of lactic acid. It was unlikely that this antifungal property was derived from the lower pH value, since the comparable pH values adjusted with HCl were unable to inhibit the growth of Diaporthe FY (Suppl. 3). These data indicate that the inhibition of Diaporthe FY was partly attributed to the properties of lactic acid. The data further showed that the colony growth of lactic acid-treated Diaporthe FY was prominently decreased compared to the mock-infected flies (Fig. 4B). Likewise, the number of mycelia and hyphae were considerably lowered in the presence of lactic acid (Fig. 4C and D). Overall, our data suggest that lactate was an important factor that could inhibit the growth of Diaporthe FY.
The synergism between Drosophila and L. plantarum to combat Diaporthe FY infection
Upon pathogenic infection, Drosophila initiates an innate immune response through the production of reactive oxygen species and antimicrobial peptides. It was assumed that collaboration between the host and its commensals could more efficiently resist pathogenic fungi than either alone. To the end, the early third-instar larvae were seeded into the fly diet with Diaporthe FY and L. plantarum. Our data revealed that the colony growth of Diaporthe FY was significantly obstructed in the presence of larvae compared to that in the absence of larvae (Fig. 5A and C). This result indicated that Drosophila and commensals collaborated to antagonize pathogens. Similarly, the number of branching mycelia was reduced in the presence of larvae compared to jn the absence of larvae (Fig. 5B and C). Intriguingly, Diaporthe FY did not form any spores in the case of larvae, partly due to the disrupted configuration of the hypha. These results demonstrate that Drosophila synergized with L. plantarum to suppress the growth of Diaporthe FY, which was critical for host survival against infection.
L. plantarum reverses ovipositional avoidance to Diaporthe FY
Using various sensory modalities, animals are able to swiftly respond to certain stimuli in their surrounding environment. To enhance the survival and fitness of their offspring, Drosophila females select favorable sites to deposit their eggs [28, 29]. Our previous work showed that commensals (e.g., L. plantarum), elicited an oviposition preference of Drosophila using the two-choice assay [23]. Since Diaporthe FY imposed morbidity on both larval and adult Drosophila (Fig. 2), it would be reasonable to hypothesize that Drosophila could sense the presence of Diaporthe FY in potential egg-laying sites. As expected, female adults overwhelmingly avoided egg-laying on the food treated with Diaporthe FY (Fig. 6A). Many molds produce an extraordinary range of secondary metabolites that repel insects [30, 31]. Indeed, the flies were robustly repulsed to laying their eggs on the surface of the halves containing metabolites of Diaporthe FY (Fig. 6B), which indicated that secondary metabolites of Diaporthe FY alerted the flies to the presence of toxic molds. We next wondered whether L. plantarum could alter the ovipositional repulsion of females to Diaporthe FY. As expected, the addition of L. plantarum dose-dependently increased the oviposition index of the females, and even switched to laying eggs in fermented food with a predominance of L. plantarum (Fig. 6C), indicating that L. plantarum attenuated ovipositional avoidance to Diaporthe FY. We further asked whether L. plantarum could abolish the ovipositional aversion to Diaporthe FY when it had dominated the community. The diet was pre-incubated with L. plantarum for different lengths of time and then exposed to Diaporthe FY. We found that although the flies were adverse to ovipositing in fermented food pre-incubated with L. plantarum for 24 h, this response was over-ridden in fermented food pre-incubated with L. plantarum for 48 h (Fig. 6D). Hence, our results demonstrat that commensals, if dominating the niche, significantly reversed the oviposition avoidance to pathogenic fungi.