Nutri-Fly^® Bloomington Formulation (Genesee Scientific, USA) media was used to feed the flies, prepared following manufacturer’s instructions at 24 ± 1°C and 70% humidity. After being autoclaved, 2 mL of propionic acid were added to prevent yeast contamination. D. melanogaster Oregon-R-C were raised and maintained in 250 mL polypropylene bottles (57 x 103 mm) with 50 mL of media, while polystyrene tubes (28.5 x 99 mm) with approximately 10 mL of media were used for experimental procedures.

Culture tubes were changed every 3-4 days during experiments, and empty tubes with media were stored at room temperature for up to one month. Flies were kept at 25°C with constant light:dark cycles of 12 hours each and 70% humidity, as Drosophila immunity is influenced by circadian rhythms. To avoid ageing bias during experiments, it was essential that all flies were the same age. To ensure this, 30 males and 30 females were placed into new bottles for 24 hours, allowing the females to lay eggs, after which adults were removed and the empty bottles were kept at 25°C. New flies emerged after 10 days, guaranteeing that all of them were born on the same day.

Two C. albicans strains were used in this study: C. albicans ATCC 90028, a reference strain, and C. albicans 4372, a clinical isolate obtained from a pleural fluid sample. Both present intrinsic resistance to ampicillin, and they were kindly provided by the Microbiology Department of Hospital Universitari Germans Trias i Pujol (HUGTiP). In the case of M. manresensis, it has intrinsic resistance to doxycycline.

To study the effect that oral administration of heat-killed M. manresensis (hkMm) and C. albicans (hkCa) could have on infection progression with the aforementioned yeast isolates, male and female flies in equal proportions were orally exposed to hkMm and hkCa, placed on top of the fly medium, for 24 and 48 hours. For the 48-hour groups, new hkMm and hkCa was added at 24 h. Both treatment regimens were included to assess whether there was a dose-dependent effect on the priming response, and also because female flies did not respond to the treatment during the first 24 hours (data not shown). For the untreated groups, fly medium was supplemented with phosphate-buffered saline (PBS). After treatment, flies were fed with normal medium for 72 h before injection with 13.8 nL of either PBS or C. albicans at a final concentration of 50 CFUs/fly. This was the time point that we identified corresponded to the peak of the priming response. The experimental design for this study is outlined in Figure 1 .

One aliquot of each isolate was defrosted and centrifuged 5 min at 5,000 g. Supernatants were discarded to remove residual glycerol, and pellets were resuspended with 100 μL of sterile PBS. The samples were then inactivated at 80°C during 20 min, after which all microorganisms were dead. 75 μL containing 10⁵ CFUs of each heat-killed sample were then inoculated in fly maintenance tubes, on top of the fly media. Tubes were left to dry overnight before introducing the flies.

To evaluate whether flies were ingesting the hkMm treatment, 5 male and 5 female flies were selected at 24- and 48-hours after the start of the oral treatment. After being anaesthetized, the heads and wings of the selected flies were removed using a razor blade. Flies were then transferred into a glass slide and coated with cold PBS to keep the tissues hydrated. Then, for each fly, one pair of tweezers was used to grasp the thorax, while a second pair was used to gently pull the posterior end of the abdomen, until the entire gut was removed from the abdominal cavity. This process was repeated five times, and the resulting guts were homogenized into 50 μL of PBS. Three 5 μL drops of each preparation were placed into glass slides and further stained using the Ziehl-Neelsen kit (BD BBL™ TB Stain Kit ZN). Slides were observed under the 100X objective of an optical microscope, and treatment was considered successful after confirming the presence of acid-fast bacilli in the preparations ( Figure 2 ).

Systemic infections were performed on 3 to 5-days-old flies. Stock aliquots were defrosted, centrifuged 5 minutes at 5,000 g and resuspended in PBS before diluting them to the desired infective dose. Flies were anaesthetized using CO₂ and injected in the ventrolateral surface of the anterior abdomen with 13.8 nL of the infection mix, using Nanoject II^® (Drummond). All infection solutions consisted in a 50 μL volume mix of either PBS or the designated microorganism and Brilliant Blue. Survival of the control groups was monitored up to 2 days post-infection for wounding control, and flies that died within these days were eliminated from the experiments.

Dead flies were collected from the maintenance tubes and transferred to individual eppendorf tubes. Each fly was then washed with 70% ethanol, rinsed with PBS, and homogenized into 200 μL of sterile PBS. Serial dilutions were then performed and 5 μL of each dilution were spotted in Luria-Bertani (LB) plates supplemented with ampicillin to select for C. albicans. Plates were incubated 24 h at 37°C, and viable bacteria were quantified by counting colony forming units (CFUs) after incubation time. It is worth noting that this technique carries an intrinsic detection limit of 40 CFUs due to the low sample volume cultured.

Data were graphed and analyzed using GraphPad Prism [version 9.5.1 (733)]. The ROUT test was applied in all groups to identify and subsequently reject one or multiple outliers. CFU counts and gene expression data were checked for normality and lognormality with Shapiro-Wilk normality test. Multiple comparisons for data with Gaussian distributions were conducted using one-way ANOVA, while Mann-Whitney tests were applied for data with non-Gaussian distributions. On the other hand, for two-group comparisons, unpaired t-tests were conducted, while Kruskal-Wallis tests were used to analyze non-normally distributed groups. For the principal component analysis (PCA), RStudio (version 2023.12.1 + 402) was used to analyze the multivariate relationships within the datasets. An initial outlier test was applied to identify and subsequently reject outliers, and the ‘mt’ and ‘ggplot2’ packages was then used to analyze and visualize the resulting data, respectively. GraphPad Prism was used to graph variable contributions and to perform statistical analysis of the principal component coordinates. The p-values ≤ 0.05 (*), ≤ 0.01 (**), ≤ 0.001 (***), and ≤ 0.0001 (****) were considered significant (same for #).

The dissection protocol confirmed the presence of M. manresensis in the gut of treated flies ( Figure 2 ), and hereby demonstrates the success of the both 24-hour and 48-hour oral treatments with hkMm. However, this same methodology could not be applied to flies treated with hkCa, due to the fact that the intestinal microbiota of D. melanogaster is already rich in yeast species.

We evaluated the effect of the oral treatment in the expression levels of two different NADPH oxidase enzymes (i.e., Nox and Duox), which are responsible for the production of reactive oxygen species (ROS) in D. melanogaster. Results for the 48-hour regimens are shown in Figure 3 , while results for the 24-hour regimens in Supplementary Figure S1 . None of the tested treatments induced significative NADPH oxidase enzyme production at the end of their respective regimens. However, both the 24-hour and the 48-hour hkMm administrations lead to an increase in nox expression 72 hours after the end of the treatments (72 h EOT) in males and females, although only male flies treated with hkCa displayed increased levels of nox induction in the 72 h EOT time-point. Regarding duox, no significative increase in expression was detected in flies treated with hkMm, while a significative reduction was detected in both hkCa treatment groups.

To assess whether response to oral treatment with hkMm and hkCa showed sexual dimorphism in flies, a principal component analysis (PCA) was performed with the relative expression levels of treated flies. Results are shown in Supplementary Figure S2 . Differences related to sex were mainly driven by the differential nox and drosomycin expression in the 72 h EOT time-point.

In males, an increase in diptericin, drosomycin, and upd3 expression was observed at 72 h EOT in all treatment regimens in a dose-dependent manner. Results for the 48-hour regimens are shown in Figure 3 , while results for the 24-hour regimens in Supplementary Figure S1 . Even when the treatment was given orally and thus only the Imd pathway should be stimulated, data show both Imd (diptericin) and Toll (drosomycin) stimulation, together with Upd3, while females showed only stimulation of diptericin 72h EOT time-point.

Interestingly, PCA analysis ( Figure 4 ) revealed that nox expression in males was the key differentiating component for the generation of two separate clusters: one containing both hkCa treatments, and the other grouping hkMm and PBS (i.e., absence of treatment) ( Figure 4A ). The same two clusters described in males could be distinguished in females, although in this case duox induction also played an important part in differentiating said clusters ( Figure 4B ). In this case, since no significant differences were observed between the 24- and 48-hour treatment regimens of each treatment group ( Supplementary Figure S3 ), results were pooled to simplify data visualization.

To determine whether oral administration of hkCa and hkMm confers specific and non-specific protection to D. melanogaster against subsequent C. albicans infection -respectively-, flies were systemically infected with 50 CFUs of C. albicans 72 h after the end of the oral treatments. Results for the 48-hour regimens are shown in Figure 5 , while results for the 24-hour regimens are displayed in Supplementary Figure S4 .

Infection with C. albicans ATCC and shame presented similar survival curves, while flies infected with C. albicans 4372 displayed with a far more aggressive profile, with an average lifespan of 1.79 days post-infection ( Figure 5A ). Systemically infected flies with C. albicans 4372 after the end of the tested treatments did not display significant differences in survival compared to untreated flies. However, a drastic decrease in pathogen load was detected in males and females of all tested treatment regimens, with a greater impact in the hkMm 48 h group ( Figure 5B ). Thus no relation between pathogen load and survival could be demonstrated.

Similarly, systemically infected flies with C. albicans ATCC presented no significant differences in survival between treatment groups. Regarding CFU counts of dead flies, no differences were observed between groups (data not shown). Thus, multiple time-points were established to monitor the pathogen load of live flies. In this regard, significant differences between treated and un-treated groups were observed at 1-, 4-, and 10-days post-infection in both males and females, with no differences between 24- and 48-hour treatment regimens (data not shown). At 1-day post-injection, infection was only established in the untreated (i.e., PBS) group, and CFUs in these flies remained fairly stable during the following time-points. Regarding treated flies, infection progressed at 4-days post-infection only in the hkMm group in males, while remaining undetectable in hkCa ATCC. Finally, at 10-days post-infection, a few CFUs were detected in treated male flies, while females continued to exhibit absence of C. albicans infection.

In light of the high mortality detected in the C. albicans 4372 group, we injected 50 CFUs of hkCa 4372 into the flies’ abdomens in a preliminary evaluation with only 30 males and 30 females to ascertain if the observed rapid host death was caused by either the inflammatory response of the flies or a structural component of the pathogen or an exoproduct secreted by C. albicans. Results are displayed in Supplementary Figure S5 . Survival was monitored up to 10 days post-infection, and fly lifespan remained comparable to the control group, indicating that C. albicans 4372 needs to be viable to cause the previously observed fly death.

To evaluate D. melanogaster’s innate immune response to infection according to treatment, patterns of gene expression of male and female flies were analyzed at 24-hours post-infection. Results for the 48-hour regimens are shown in Figure 6 , while results for the 24-hour regimens in Supplementary Figure S6 .

Interestingly, the inhibition of the pathogen growth caused by the hk treatment triggered the same AMP production that the control group in which C. albicans 4372 experiments an extraordinary growth. This response also presents sexual dimorphism. On one hand, male response is Toll-based, further reinforcing the trend prompted by hk priming. On the other, females appear to prioritize the Imd response. Regarding upd3, there is not such a difference, and in both cases, it appears to be enhanced in the group primed with hkCa 4372. In contrast, infection with C. albicans ATCC has no impact in the innate response. No significant differences were detected in duox expression levels (data not shown).

Moreover, PCA analysis ( Figures 7A, C, D ) shows a clear pattern, in which survival is the main factor that clusters the control and C. albicans ATCC-infected groups in one corner of the graph, while flies infected with C. albicans 4372 are placed in the opposed corner of the graph. Dimension 2, on the other hand, emphasizes the sexual dimorphism upon C. albicans 4372 infection, marked by drosomycin expression in males, and diptericin and upd3 in females. This difference might account for the slightly better survival rate in males ( Figure 7B ). Since no significant differences were observed between the 24- and 48-hour treatment regimens of each treatment group ( Supplementary Figure S3 ), results were pooled to simplify data visualization.