*3.1. Bioassays*

Mortality of larvae injected with *C. militaris* conidia began at 5–7 days post injection and reached 80–100% after 7–10 days, depending on the dose and temperature (Figure 1A–C). More rapid mortality of larvae at 15 ◦C compared to 25 ◦C was observed following injection of all doses (log rank test, χ2 > 9.6, df = 1, *p* < 0.002). No mortality was registered for larvae injected with saline. Notably, 9–20% of insects infected with low and intermediate doses and maintained at 25 ◦C were able to survive and complete metamorphosis.

**Figure 1.** Mortality dynamics and outcome of infection in wax moth larvae after injection with *C. militaris* conidia and subsequent incubation at 15 ◦C and 25 ◦C. (**A**–**C**)—mortality dynamics after injection of larvae with 1250, 2500 and 5000 conidia per larva (c/L). Different letters indicate significant differences determined by log rank test (χ<sup>2</sup> > 9.6, df = 1, *p* < 0.002). (**D**)—*C. militaris* hyphal bodies (Hb) and cocci (C) in wax moth hemolymph at 5 days after injection with the fungus. Scale bar, 1 μm. (**E**)—portion of mummified and decomposed larvae during the development of mycoses at different temperatures. Different letters indicate significant differences (χ<sup>2</sup> > 8.7, df = 1, *p* < 0.003). (**F**)—*C. militaris* conidial yield on mummified cadavers at 15 ◦C and 25 ◦C. Different letters indicate significant differences determined by *t*-tests (*t* = 6.1, df = 8, *p* < 0.001).

Microscopy observations showed the simultaneous presence of hyphal bodies and cocci in the hemolymph of infected insects maintained at 25 ◦C (Figure 1D), however, these cocci were not observed in the hemolymph at 15 ◦C. At 15 ◦C, mycosis led to the formation of mummified cadavers (94–100%) after all treatment doses (Figures 1E and 2A). However, at 25 ◦C, we documented a large number of bacterially decomposed insects (χ<sup>2</sup> > 8.7, df = 1, *p* < 0.003 compared to 15 ◦C). The bacterioses were identified by symptoms of darkening and liquefaction of the larvae for several hours after death (Figure 2C). The increase in frequency of bacterioses at 25 ◦C was dose-dependent and increased from 36% after injection with the lowest dose and to 83% after injection with the highest dose (Figure 1E). Notably, we registered the formation of abnormally dark mummies in these experiments (Figure 2B). The percent of abnormal mummies was 52% at 25 ◦C and only 12% at 15 ◦C (χ<sup>2</sup> = 8.3, df = 1, *p* = 0.004). Moreover, the production of conidia on mummified cadavers at 25 ◦C decreased 2.5-fold compared to 15 ◦C (*t* = 6.1, df = 8, *p* < 0.001, Figures 1F and 2D,E). Thus, insects were less susceptible to *C. militaris* infection at 25 ◦C, but they were more predisposed to spontaneous bacterial infections compared to those incubated at 15 ◦C.

**Figure 2.** Phenotypes of larvae that died after injection with *C. militaris* conidia. (**A**)—mummification, (**B**)—defective mummies, (**C**)—bacterial decomposition, (**D**)—conidiation at 15 ◦C, (**E**)—conidiation at 25 ◦C.

#### *3.2. CFU Counts in the Hemolymph and Midgut*

In the hemolymph of control larvae, we registered single colonies of enterococci and enterobacteria at both temperatures (Figure 3A). At 15 ◦C, fungal infection did not lead to significant changes in the CFU count (Dunn's test, *p* > 0.17 compared to controls). In contrast, CFU counts of both enterobacteria and enterococci increased in the hemolymph by 39,000–54,000-fold at 25 ◦C in response to *C. militaris* infection (*p* < 0.002 compared to controls). A significant interaction between factors (mycosis × temperature) was found for enterobacteria (H1,19 = 9.4, *p* = 0.002). However, this interaction was not found for enterococci (H1,19 = 1.5, *p* = 0.23) because there was still a slight increase in these bacteria at 15 ◦C in response to *C. militaris* infection.

**Figure 3.** CFU counts in the hemolymph (**A**) and midgut (**B**) of wax moth larvae at 96 h post injection of *C. militaris* (2500 conidia per larva) with subsequent incubation at 15 ◦C and 25 ◦C. Selective media for enterobacteria (Endo agar) and enterococci (Bile esculin agar) was used. Different letters show significant differences within the specified media and tissue (Dunn's test, *p* < 0.05).

In the midgut, we observed an elevation in the enterobacteria and enterococci CFU counts in response to fungal infection at both temperatures (enterobacteria, H1,19 = 4.2, *p* = 0.04; enterococci, H1,19 = 8.5, *p* = 0.004). However, the post hoc tests showed significant elevation only at 25 ◦C (4–7-fold relative to controls, Dunn's test, *p* < 0.04, Figure 3B). E ffects of temperature on CFU counts were not significant, however, a trend toward increased enterococci was observed at 15 ◦C compared to 25 ◦C (H1,23 = 2.7, *p* = 0.09). Notably, uninfected larvae maintained at 15 ◦C were characterized by the highest enterococci CFU counts compared to larvae maintained at 25 ◦C (*p* = 0.03).

#### *3.3. Bacterial Communities in Midguts and Cadavers*

In the midgut, we registered 168 OTUs (37 ± 5.2 OTUs per sample) with a predominance of two *Enterococcus* OTUs (Figure 4). A BLAST search against sequences in GenBank showed strong similarity with *Enterococcus faecalis* (OTU 1, 100% similarity) and *E. lemanii* (OTU 100, 99.53% similarity). Temperature did not have a significant impact on the relative abundance of di fferent groups and diversity indexes (H1,11 < 0.4, *p* > 0.52, Figure S2). However, trends toward increased diversity indexes in warm conditions were observed for uninfected larvae (Dunn's test, *p* > 0.08, Figure S2). Fungal infection led to a significant decrease in OTU counts and the Chao1 index (H1,11 > 4.7, *p* < 0.03), as well as to shifts in community structure. Under both temperatures, *C. militaris* infection caused a partial displacement of *E. faecalis* by *E. lemanii* (e ffect of the infection: H1,11 = 8.3, *p* = 0.004). In addition, a decrease in the abundance of the subdominant bacteria *Acinetobacter*, *Melaminivora*, *Comamonas*, and *Diaphorobacter* was revealed under the influence of the fungal infection (H1,11 > 5.0, *p* < 0.024). These e ffects were more evident at 25 ◦C (Dunn's test, *p* < 0.013) compared to 15 ◦C (Dunn's test, *p* > 0.17).

In bacterially decomposed cadavers, we registered the lowest bacterial diversity (OTU count, 10 ± 1.9; Chao1, 13 ± 2.6; Shannon, 0.47 ± 0.14). In the cadavers, either the enterococci *E. faecalis* or *Enterobacter* sp. prevailed (Figure 4). Enterobacteriaceae were represented by two predominant OTUs that were also detected in the midgut. One of them, OTU 2, was close in identity to *Enterobacter* sp. (99.53% similarity) which was previously isolated from the midgut of same line of *G. mellonella* [7]. The other, OTU 144, was close to *Cronobacter sakazakii* (99.77% similarity).

#### *3.4. AMP Gene Expression*

We observed a significant upregulation in the expression of the studied AMP genes (except for cecropin) in both the fat body and the midgut under the influence of fungal infection (Figure 5, Table 1). Temperature had a significant impact on the expression of cecropin and lysozyme only. Overall, we observed a stronger expression of antifungal peptide genes in response to infection at 25 ◦C compared to 15 ◦C. In contrast, antibacterial peptide genes trended toward higher expression at 15 ◦C compared to 25 ◦C. For example, expression of the antifungal peptide gene gallerimycin in the fat body was increased by 77-fold at 25 ◦C, but only by 12-fold at 15 ◦C compared to uninfected insects (Dunn's test, *p* < 0.0005 and *p* = 0.10, respectively). The galiomycin gene in the fat body was upregulated by 18-fold at 25 ◦C but only 8-fold at 15 ◦C (*p* = 0.001 and *p* = 0.04 compared to controls, respectively). Gallerimycin and galiomycin gene expression followed the same pattern in the midgut (Figure 5, Table 1).

Unlike the antifungal peptides, expression of the antibacterial peptide gloverin in the fat body in response to fungal infection increased by 55-fold at 15 ◦C (*p* = 0.005 compared to control) and 17-fold at 25 ◦C (*p* = 0.01 compared to control). For the cecropin and lysozyme genes, we observed increased expression in the fat body at 15 ◦C compared to 25 ◦C (e ffect of temperature, H1,19 = 3.9, *p* = 0.05 and H1,23 = 3.2, *p* = 0.07, respectively), and more active expression in response to fungal infection was also observed at low temperature (Figure 5). Changes in the expression of the gloverin, cecropin and lysozyme peptide genes in the midgut were less than in the fat body. The gloverin gene was upregulated by 2.8-3-fold in the midgut in response to fungal infection (H1.19 = 6.2, *p* = 0.01), independent of temperature. Expression of cecropin in the midgut was not significantly changed in response temperature or fungal infection. Expression of the lysozyme gene in the midgut was decreased at 15 ◦C compared to 25 ◦C (H1,23 = 4.6, *p* = 0.03); however, there was a significant upregulation in response to *C. militaris* infection, which occurred only at 15 ◦C (6-fold, *p* = 0.004 compared to control) and not at 25 ◦C (2-fold, *p* = 0.15 compared to control).

**Figure 4.** Bacterial communities (16S rRNA) in the midguts of wax moth larvae during the development of *C. militaris* infection at different temperatures and the communities in the cadavers that decomposed after the infection. Midgut communities were analyzed at 96 h after injection with a dose of 2500 conidia per larva. Decomposed cadavers were analyzed at 6–7 days post injection. Each treatment represents 3 replicates.

**Figure 5.** Relative expression of AMP genes in the fat body and midgut of wax moth larvae at 96 h after injection with *C. militaris* (2500 conidia per larva) and subsequent incubation at 15 ◦C and 25 ◦C. Data were normalized to the expression of two reference genes, eEF1a and RBP11. The *Y*-axis shows the fold change relative to uninfected larvae maintained at 25 ◦C. **Gal**—galiomycin, **Glm**—gallerimycin, **Glo**—gloverin, **Cec**—cecropin-like, **Lys**—lysozyme-like. Different letters indicate significant differences between treatments (Dunn's test, *p* < 0.05).

**Table 1.** Two-way effects of *C. militaris* infection and temperature on the expression of AMP genes. Significant effects are highlighted in bold. Arrows show up- or downregulation of genes in response to infection and in response to cooling to 15 degrees. Arrows are shown only for significant (*p* < 0.05) and marginal (*p* = 0.05–0.10) effects.


#### *3.5. Apoptosis, ROS and Stress-Related Gene Expression*

Expression of the IAP gene in the fat body was temperature dependent (Figure 6, Table 2). The gene was upregulated in the fat body in response to fungal infection only at low temperature (Dunn's test, *p* = 0.07 compared to control at 15 ◦C and *p* < 0.01 compared to other treatments). At 25 ◦C, expression of this gene in response to infection was not changed compared to the control (*p* = 0.50). In the midgut, regulation of the IAP gene was not caused by temperature (Table 2), but only by the infection (effect of fungus: H1,23 = 7.7, *p* = 0.006). Significant upregulation in response to *C. militaris* was also registered only at 15 ◦C (*p* = 0.04 compared to control).

**Figure 6.** Relative expression of apoptosis, ROS and stress-related genes in the fat body and midgut of wax moth larvae at 96 h after injection with *C. militaris* (2500 conidia per larva) and subsequent incubation at 15 ◦C and 25 ◦C. Data were normalized to the expression of two reference genes, eEF1a and RBP11. The *Y*-axis shows the fold change relative to uninfected larvae maintained at 25 ◦C. Different letters indicate significant differences between treatments (Dunn's test, *p* < 0.05).

NOX-DUOX domain gene expression was slightly (1.6-fold) upregulated in the fat body in response to fungal infection at both temperatures (effect of infection: H1,23 = 4.2, *p* = 0.04), but the effect of temperature was not significant (Table 2). Scheirer–Ray–Hare test showed a downregulation of this gene in the midgut under the influence of a low temperature (H1,23 = 6.16, *p* = 0.01), however strong downregulation (>2.4-fold) was observed during mycosis development at 15 ◦C only (*p* = 0.08 compared to control at 15 ◦C and *p* < 0.01 compared to other treatments).

HSP70 gene expression in the fat body was increased at a low temperature (Figure 6, Table 2) and fungal infection downregulated its expression at both temperatures (effect of fungus: H1,23 = 4.6, *p* = 0.03). In the midgut, upregulation of HSP70 was also observed at 15 ◦C (H1,23 = 13.7, *p* < 0.001) but fungal infection had no significant effect. The HSP90 gene was slightly and insignificantly upregulated in the fat body in response to infection and independent of temperature. Its expression in the midgut was increased under the influence of low temperature (H1,23 = 8.7, *p* = 0.003) but fungal infection had no significant effect.


**Table 2.** Two-way effects of*C. militaris*infection and temperature on the expression of apoptosis, ROS and stress-related genes. Significant effects highlighted in bold. Arrows show up- or downregulation of genes in response to infection and in response to cooling to 15 degrees. Arrows are shown only for significant (*p* < 0.05) and marginal (*p* = 0.05–0.10) effects.

#### *3.6. Interaction between Fungi and Bacteria In Vitro*

We showed that *E. faecalis* and *Enterobacter* inhibited *C. militaris* more strongly than *M. robertsii* and *B. bassiana*. In particular, *E. faecalis* inhibited *C. militaris* mycelial growth on SDAY medium by 2–2.2-fold more than *M. robertsii* or *B. bassiana* growth (Dunn's test, *p* < 0.012, Figure 7). *Enterobacter* sp. also inhibited *C. militaris* growth more strongly than *M. robertsii* and *B. bassiana*, but the differences were only marginally significant (*p* = 0.06–0.10). None of these fungi inhibited bacterial growth on nutrient agar. However, *M. robertsii* and *B. bassiana* were able to grow on cultures of both bacteria (Figure 7). In contrast, *C. militaris* was not able to grow on cultures of *E. faecalis* or *Enterobacter* sp.

**Figure 7.** Inhibition of fungi by *Enterococcus faecalis* and *Enterobacter sp*. in vitro. (**A**)—zone of mycelial growth inhibition by bacteria on SDAY medium. (**B**)—radial growth of fungi on nutrient agar and this medium with lawns of the bacteria. Different letters indicate significant differences between treatments (Dunn's test, *p* < 0.05).
