Biosurfactants Induce Antimicrobial Peptide Production through the Activation of TmSpatzles in Tenebrio molitor

Biosurfactant immunomodulatory activities in mammals, nematodes, and plants have been investigated. However, the immune activation property of biosurfactants in insects has not been reported. Therefore, here, we studied the defense response triggered by lipopeptides (fengycin and iturin A), glycolipids (rhamnolipid), and cyclic polypeptides (bacitracin) in the coleopteran insect, mealworm Tenebrio molitor. The in vitro antimicrobial activities against Gram-positive (Staphylococcus aureus) and Gram-negative (Escherichia coli) bacteria and fungi (Candida albicans) were assessed by mixing these pathogens with the hemolymph of biosurfactant-immune-activated larvae. E. coli growth was remarkably inhibited by this hemolymph. The antimicrobial peptide (AMP) induction results also revealed that all biosurfactants tested induced several AMPs, exclusively in hemocytes. The survivability analysis of T. molitor larvae challenged by E. coli (106 CFU/µL) at 24 h post biosurfactant-immune activation showed that fengycin, iturin A, and rhamnopid significantly increased survivability against E. coli. Biosurfactant-induced TmSpatzles activation was also monitored, and the results showed that TmSpz3 and TmSpz-like were upregulated in the hemocytes of iturin A-injected larvae, while TmSpz4 and TmSpz6 were upregulated in the fat bodies of the fengycin-, iturin A-, and rhamnolipid-injected larvae. Overall, these results suggest that lipopeptide and glycolipid biosurfactants induce the expression of AMPs in T. molitor via the activation of spätzle genes, thereby increasing the survivability of T. molitor against E. coli.


Introduction
Biosurfactants are surface-active biomolecules containing hydrophobic and hydrophilic moieties that are produced by bacteria, fungi, and yeast [1]. Unlike chemically synthesized surfactants, biosurfactants are generally categorized based on their microbial origin and chemical composition [2]. Chemically, biosurfactants are categorized into glycolipids (rhamnolipids) [3], trehalolipids [4], sophorolipids [5], lipopeptides and lipoproteins (surfactin, fengycine, and iturin), fatty acids, phospholipids and neutral lipids [5], polymeric, and particulate biosurfactants [6]. enzymes in mammals to provide either a direct toxic response or indirect actions through modulating antimicrobial peptides [39]. Lipopeptides, the most commonly known biosurfactants, also consist of lipids connected to a peptide backbone containing a mix of L-and D-amino acids [40]. Additionally, rhamnolipids, one of the well-studied glycolipids, are composed of one or more monosaccharide moieties bound by a glycosidic linkage to a hydrophobic moiety [41]. Considering their peptide or sugar backbone compositions, lipopeptides and glycolipids are supposed to induce the innate immune response. Based on these reports of biosurfactant applications in plants and animals, it can be said that biosurfactants have both antimicrobial and immune-eliciting properties. However, the biosurfactant-triggered defense response in insects is yet to be studied.
Therefore, in this study, the induction of TmSpatzle by lipopeptides (fengycin and iturin A), polypeptide (bacitracin), and glycolipid (rhamnolipid) was studied. To the best of our knowledge, this is the first report indicating that lipopeptides (fengycin and iturin A) and glycolipids (rhaminolipds) increase the survivability of the mealworm Tenebrio molitor by activating different antimicrobial peptides through the activation of TmSpz4, TmSpz6, TmSpz3, or TmSpz-like in T. molitor.

AMP Activity in the Hemolymph of Biosurfactant-Injected T. molitor Larvae
To determine the in vitro AMP activity against Gram-positive and Gram-negative bacteria and fungi, the hemolymph of biosurfactant-immune activated larvae was mixed with E. coli, S. aureus, and C. albicans. All the tested biosurfactants significantly (p < 0.05) reduced the growth of E. coli compared with PBS ( Figure 1). Conversely, there was no difference between the control group (PBS) and biosurfactants group regarding growth of Gram-positive bacteria and fungi ( Figure 1). modulating antimicrobial peptides [39]. Lipopeptides, the most commonly known biosurfactants, also consist of lipids connected to a peptide backbone containing a mix of L-and D-amino acids [40]. Additionally, rhamnolipids, one of the well-studied glycolipids, are composed of one or more monosaccharide moieties bound by a glycosidic linkage to a hydrophobic moiety [41]. Considering their peptide or sugar backbone compositions, lipopeptides and glycolipids are supposed to induce the innate immune response. Based on these reports of biosurfactant applications in plants and animals, it can be said that biosurfactants have both antimicrobial and immune-eliciting properties. However, the biosurfactant-triggered defense response in insects is yet to be studied. Therefore, in this study, the induction of TmSpatzle by lipopeptides (fengycin and iturin A), polypeptide (bacitracin), and glycolipid (rhamnolipid) was studied. To the best of our knowledge, this is the first report indicating that lipopeptides (fengycin and iturin A) and glycolipids (rhaminolipds) increase the survivability of the mealworm Tenebrio molitor by activating different antimicrobial peptides through the activation of TmSpz4, TmSpz6, TmSpz3, or TmSpz-like in T. molitor.

AMP Activity in the Hemolymph of Biosurfactant-Injected T. molitor Larvae
To determine the in vitro AMP activity against Gram-positive and Gram-negative bacteria and fungi, the hemolymph of biosurfactant-immune activated larvae was mixed with E. coli, S. aureus, and C. albicans. All the tested biosurfactants significantly (p < 0.05) reduced the growth of E. coli compared with PBS ( Figure 1). Conversely, there was no difference between the control group (PBS) and biosurfactants group regarding growth of Gram-positive bacteria and fungi ( Figure 1). Figure 1. The in vitro AMP activity of biosurfactant-immune activated larvae against microbes. The in vitro AMP activities of biosurfactant-immune activated larvae against Gram-positive (S. aureus, Sa), Gram-negative bacteria (E. coli, Ec), and fungi (C. albicans, Ca) were measured by mixing the hemolymph of immune-activated larvae and the microbes. Excitingly, the results revealed that all the tested biosurfactants including bacitracin (BAC), fengycin (FUN), iturin A (ITU), and rhamnolipid (RHA) significantly (p < 0.05) reduced E. coli growth, compared with the PBS-injected control group (red dotted line).* significant difference (p < 0.05) of CFU, compared to the PBS-injected group.

Biosurfactants Increased T. molitor Larvae Survivability against E. coli
To confirm if biosurfactant-activated AMPs could increase survivability against E. coli, a T. molitor larvae survivability assay was conducted by challenging the insect with E. coli (10 6 CFU/µL) at 0, 12, and 24 h post immune activation by biosurfactants. Interestingly, fengycin, iturin A, and rhamnolipids significantly (p < 0.05) increased the survivability in 24 h post immune activated groups ( Figure 3). Fengycin, iturin A, and rhamnolipids increased survivability against E. coli by 33.72, 34.48, and 28.75%, respectively. However, larvae challenged by E. coli at both 0 and 12 h post immune activation by biosurfactant did not show significant survivability when compared with the control group. This may suggest that AMP expression was regulated 24 h post biosurfactant-immune activation. Thus, survivability results confirm the AMP expression observed in this study.

Biosurfactants Increased T. molitor Larvae Survivability against E. coli
To confirm if biosurfactant-activated AMPs could increase survivability against E. coli, a T. molitor larvae survivability assay was conducted by challenging the insect with E. coli (10 6 CFU/µL) at 0, 12, and 24 h post immune activation by biosurfactants. Interestingly, fengycin, iturin A, and rhamnolipids significantly (p < 0.05) increased the survivability in 24 h post immune activated groups ( Figure 3). Fengycin, iturin A, and rhamnolipids increased survivability against E. coli by 33.72, 34.48, and 28.75%, respectively. However, larvae challenged by E. coli at both 0 and 12 h post immune activation by biosurfactant did not show significant survivability when compared with the control group. This may suggest that AMP expression was regulated 24 h post biosurfactant-immune activation. Thus, survivability results confirm the AMP expression observed in this study. Interestingly, fengycin, iturin A, and rhamnolipid significantly increased the survivability against E. coli by 33.72, 34.48, and 28.75%, respectively. The graph indicated by asterisks (*) shows significant difference at p < 0.05 following the Student-Newman-Keuls test.

Biosurfactants Activated the Expression of Spätzle Genes in T. molitor
To understand if biosurfactants induce AMP expression via the activation of spätzle genes in T. molitor, biosurfactant-induced activation of T. molitor spätzle was first examined and screened. The results revealed that TmSpz3 and TmSpz-like were upregulated in hemocytes by iturin A, while TmSpz4 and TmSpz6 were upregulated in the fat bodies by fengycin, iturin A, and rhamnolipids ( Figure 4). The other spatzles remained constant or downregulated, compared to the PBS-injected group. In the bacitracin-injected larvae, no spätzle was significantly upregulated. Comparatively, higher fold changes in TmSpz4 and TmSpz6 expressions were observed in rhamnolipid-injected groups. Interestingly, fengycin, iturin A, and rhamnolipid significantly increased the survivability against E. coli by 33.72, 34.48, and 28.75%, respectively. The graph indicated by asterisks (*) shows significant difference at p < 0.05 following the Student-Newman-Keuls test.

Biosurfactants Activated the Expression of Spätzle Genes in T. molitor
To understand if biosurfactants induce AMP expression via the activation of spätzle genes in T. molitor, biosurfactant-induced activation of T. molitor spätzle was first examined and screened. The results revealed that TmSpz3 and TmSpz-like were upregulated in hemocytes by iturin A, while TmSpz4 and TmSpz6 were upregulated in the fat bodies by fengycin, iturin A, and rhamnolipids ( Figure 4). The other spatzles remained constant or downregulated, compared to the PBS-injected group. In the bacitracin-injected larvae, no spätzle was significantly upregulated. Comparatively, higher fold changes in TmSpz4 and TmSpz6 expressions were observed in rhamnolipid-injected groups. . Twenty-four hours post biosurfactant injection, three tissues-hemocytes, fat bodies, and gutwere dissected. The expression of spätzle in each tissue was quantified by qRT-PCR using specific primers. The data are the average of three replicates. * significant difference (p < 0.05) of Spatzles expression, compared to the PBS-injected group.

Effect of TmSpz Gene Silencing on the Expression of AMPs in T. molitor
To study whether activated spatzles have a role in biosurfactant-induced AMP expression in T. molitor, the screened spätzle genes (TmSpz3, TmSpz4, TmSpz6, and TmSpz-like) were silenced. Then, we checked and confirmed the gene silencing levels of TmSpz3 ( Figure 5A), TmSpz-like ( Figure 5C), TmSpz4 ( Figure 6A), and TmSpz6 ( Figure 7A). The biosurfactants were then injected into the genesilenced larvae, respectively. The quantification of AMP expression showed that TmSpz3, TmSpz4, TmSpz6, and TmSpz-like positively regulated the expression of most of the AMPs in the whole body. Specifically, the expressions of TmTen-2, -4, TmAtt-1a, TmAtt-1b, TmAtt-2, TmCol-1, and TmCol-2 were significantly downregulated in the TmSpz3-silenced and iturin A-injected larvae ( Figure 5B). The expressions of TmTen-2, TmCol-1, and -2 were positively regulated by TmSpz-like ( Figure 5D). Twenty-four hours post biosurfactant injection, three tissues-hemocytes, fat bodies, and gut-were dissected. The expression of spätzle in each tissue was quantified by qRT-PCR using specific primers. The data are the average of three replicates. * significant difference (p < 0.05) of Spatzles expression, compared to the PBS-injected group.

T. molitor Larvae Survivability Study
To further understand if the biosurfactants increased the survivability of T. molitor via spätzle gene activation and AMP induction against microbial challenge, TmSpzs were individually silenced ( Figure 8A,C, Figures 9A and 10A), and each biosurfactant was injected into the gene-silenced larvae. Next, 24 h post biosurfactant injection, E. coli was injected. The results revealed that in the TmSpz3 and -like gene-silenced larvae, iturin A injection did not increase the survivability of larvae against E. coli infection, whereas in the dsEGFP injected larval group (control), iturin A significantly increased the survivability (Figure 8B,D). Meanwhile, fengycin and iturin A increased the survivability of larvae injected with dsEGFP, while TmSpz4 gene-silenced larvae were susceptible to E. coli ( Figure 9B,C). Conversely, rhamnolipid showed no significant induction of larval survivability among the TmSpz4 gene-silenced and dsEGFP-injected larvae ( Figure 9D). Similarly, although fengycin and rhamnolipid significantly (p < 0.05) increased control group survivability, the dsTmSpz6-injected larvae were susceptible to E. coli ( Figure 10B-D).

T. molitor Larvae Survivability Study
To further understand if the biosurfactants increased the survivability of T. molitor via spätzle gene activation and AMP induction against microbial challenge, TmSpzs were individually silenced ( Figure 8A,C, Figure 9A and Figure 10A), and each biosurfactant was injected into the gene-silenced larvae. Next, 24 h post biosurfactant injection, E. coli was injected. The results revealed that in the TmSpz3 and -like gene-silenced larvae, iturin A injection did not increase the survivability of larvae against E. coli infection, whereas in the dsEGFP injected larval group (control), iturin A significantly increased the survivability (Figure 8B,D). Meanwhile, fengycin and iturin A increased the survivability of larvae injected with dsEGFP, while TmSpz4 gene-silenced larvae were susceptible to E. coli ( Figure 9B,C). Conversely, rhamnolipid showed no significant induction of larval survivability among the TmSpz4 gene-silenced and dsEGFP-injected larvae ( Figure 9D). Similarly, although fengycin and rhamnolipid significantly (p < 0.05) increased control group survivability, the dsTmSpz6-injected larvae were susceptible to E. coli ( Figure 10B-D). coli (E.c) was injected into the gene-silenced larvae. dsEGFP and 1XPBS were injected as control for gene and iturin A-or microbe-injected groups, respectively. The data represent the mean of three biological replicate experiments. *significant differences between dsTmSpz-3, -like, and dsEGFP-treated and PBS-injected larval groups (p < 0.05). Figure 8. Survivability of T. molitor larvae against E. coli after TmSpz-3 or -like gene silencing and iturin A injection. The TmSpz-3 or -like gene were silenced, and the silencing efficiency was measured by qRT-PCR at 3 d post-injection (A,C). Iturin A (It) was then injected into the TmSpz-3 (B) and -like (D) gene-silenced larvae. After 24 h of iturin A injection, E. coli (E.c) was injected into the gene-silenced larvae. dsEGFP and 1XPBS were injected as control for gene and iturin A-or microbe-injected groups, respectively. The data represent the mean of three biological replicate experiments. * significant differences between dsTmSpz-3, -like, and dsEGFP-treated and PBS-injected larval groups (p < 0.05).  biosurfactant injection. The TmSpz4 gene was silenced, and the silencing efficiency of TmSpz4 mRNA was measured by qRT-PCR at 5 d post-injection (A). Then, biosurfactants (BS) including fengycin (B), iturin A (C), and rhamnolipid (D) were separately injected into the gene-silenced larvae. E. coli was then injected into the gene-silenced and biosurfactant-injected larvae. dsEGFP and 1XPBS were injected as control for the gene-and biosurfactant-or microbe-injected groups, respectively. The data represent the mean of three biological replicate experiments. * significant differences between dsTmSpz4-and dsEGFP-treated groups of biosurfactant-and PBS-injected larvae (p < 0.05). , and rhamnolipid (D) were separately injected into gene-silenced larvae. E. coli was then injected to the gene-silenced-and biosurfactant-injected larvae. dsEGFP and 1XPBS were injected as control for the gene-and biosurfactant-or microbe-injected groups, respectively. The data represent the mean of three biological replicate experiments. * significant differences between dsTmSpz6-and dsEGFP-treated groups, biosurfactant-, and PBS-injected larvae (p < 0.05).

Discussion
Biosurfactants have been repeatedly shown to exhibit antimicrobial activity [42][43][44]. In addition, the immune response modulators of biosurfactants have been investigated in mammalian and nematode C. elegans [26][27][28]. Therefore, based on the immunomodulatory reports, it could be hypothesized that biosurfactants could activate the immune response of T. molitor against microorganisms. Thus, to test this postulation, the in vitro growth inhibitions of Gram-positive and Gram-negative bacteria and fungi were performed by mixing microbial cells with the hemolymph of biosurfactant-immune activated larvae. Excitingly, the results revealed that all the tested biosurfactants significantly reduced the CFU of the Gram-negative bacteria E. coli, with no difference observed between the control and biosurfactants against Gram-positive bacteria and fungi. These results suggest that the tested biosurfactants specifically inhibit the growth of Gram-negative bacteria via activating the production of AMPs in T. molitor hemolymph.
The inhibition of E. coli CFU by the hemolymph of biosurfactant-immune triggered T. molitor larvae was a call for concern. As explained earlier, AMPs form the main immune response produced during infection or immune elicitation. Therefore, AMP expression levels were determined by injecting biosurfactants into T. molitor larvae. Various AMPs were induced by the injection of biosurfactants, suggesting that biosurfactants could regulate the innate immunity of T. molitor through the induction of AMPs in hemocytes. However, using a different mechanism, B. licheniformis-derived [58] and rhamnolipid biosurfactants [59] have also demonstrated the ability for modulation of immune responses in Labeo rohita. Moreover, rhamnolipid displayed immunostimulatory properties against human monocyte cells by improving the secretion of the cytokine tumor necrosis factor [60].
The major concern here is how the host recognizes these lipopetides or glycolipids to activate the immune response. Even though the known pattern recognition receptors of D-amino acids have not been identified, D-amino acids are asymmetrically recognized by some other receptors and enzymes in mammals to provide either direct toxic responses or indirect actions through the modulation of antimicrobial peptides [39]. Additionally, Li Hao and his colleagues described that the presence of D-amino acids in the peptide structure can play a key role in mammalian receptor recognition [61]. Based on the composition of the different L-and D-amino acids, including D-amino acids (Tyr, Orn, and Thr, Ala) in fengycin and the D-amino acids (Tyr, Asn, and Asn) of the cyclic lipopeptide iturin A [62], the recognition of these biosurfactants could be attributable to certain recognition receptors that activate AMPs.
Toll receptors have been identified and known to recognize evolutionarily conserved bacterial components such as LPS, LP, LPT, and PGN in mammals [31,63]. Moreover, it has been reported that CD14 facilitates LP recognition by toll-like receptors 2 and 1 in humans [64][65][66]. Similarly, LPS binding protein (LBP) is a serum factor known to have the ability to bind pathogens and initiate the innate immune response [67]. Moreover, it has been reported that LP could enhance TLR2-mediated activation, a promising strategy for vaccine development [68].
Contrarily, in insects, during both embryogenesis and immunity, inactive spätzle protein is synthesized and secreted from the cell, containing a pro domain and a C-terminal region [69,70]. During infection, microbe-associated molecular patterns, such as LPS and lipoteichoic acid (LTA), are recognized by PGN recognition protein and GNBP1 [33], leading to the initiation of the proteolytic cascade, serine protease, and SPE [36]. Therefore, this process finally activates the initially inactive spätzle. Extrapolating to the mammalian and insect immune activation systems, the bacillus LP (fengycin and iturin A) and glycolipid (rhamnolipid) are found activating the TmSpatzle genes in T. molitor. It should be noticed that the recognition receptor for these specific biosurfactants is not well known.

Insect Culture
The coleopteran insect mealworm, Tenebrio molitor, was maintained at 27 ± 1 • C, 60% ± 5% relative humidity, under dark conditions, with an artificial diet (170 g of whole-wheat flour, 20 g of fried bean powder, 10 g of soy protein, 100 g of wheat bran, 200 mL of sterile water, 0.5 g of chloramphenicol, 0.5 g of sorbic acid, and 0.5 mL of propionic acid). Larvae at the 10th to 12th instar were used for experiments [76].

Source of Biosurfactants
The biosurfactants (bacitracin, fengycin, iturin A, and rhamnolipids) used in this study were purchased from Sigma Aldrich (St. Louis, MO, USA) ( Table 1). All the biosurfactants were in powder form.

Preparation of Microorganisms
Gram-negative bacteria (E. coli K12), Gram-positive bacteria (S. aureus RN4220), and fungi (Candida albicans) were used in this study. These microorganisms were cultured in Luria-Bertani (LB; E. coli and S. aureus) and Sabouraud dextrose (C. albicans) broths at 37 • C overnight, subcultured at 37 • C for 3 h. Then, the microorganisms were harvested and washed two times by centrifugation at 3500 rpm for 10 min in phosphate-buffered saline (PBS; pH 7.0). They were then suspended in PBS, and their concentrations were measured at OD600. Finally, 10 6 cells/µL of E. coli, S. aureus, and 5 × 10 4 cells/µL of C. albicans were injected separately [77].

AMPs Activity Study
Each biosurfactant was dissolved by vortexing in the 1x PBS at 1µg/µL (biosurfactant/PBS). Then, the stock solution of each biosurfactant was diluted to the targeted concentration (100, 200, 300, and 1000 ng/µL). Late instars larvae of T. molitor were injected with 300 ng/µL of each biosurfactant. During each injection, the solution was well vortexed to insure uniform distribution of the particle. This concentration was selected based on the previous preliminary experiment. Briefly, we conducted the preliminary experiment with 100, 200, 300, and 1000 ng/µL biosurfactants injection, and the highest AMPs activity was achieved at 300 ng/µL. 1X PBS (0.1 mM sodium phosphate, 2.5 mM potassium chloride, pH 7.2) was injected as control. The injections were done using disposable needles mounted onto a micro applicator (Picospiritzer III micro dispense system, Parker Hannifin, Hollis, NH). The injected larvae were maintained on artificial diet under standard rearing conditions. Twenty-four hours post injection, the hemolymph was extracted in to 1X PBS on ice. Then, the hemolymph was centrifuged at 15,000 rpm and 4 • C for 10 min. The supernatant was retrieved and boiled at 100 • C for 10 min, then centrifuged at 15,000 rpm and 4 • C for 10 min, and then the supernatant transferred to a new tube. Finally, peptide concentration was determined by Ephoch at OD220. The hemolymph sample was diluted by 1X PBS and 100 µg mixed with E. coli (10 6 CFU). The total 200 µL mixture (10 µL hemolymph sample, 10 µL E. coli, and 180 µL PBS) was incubated at 37 • C for 1 h under shaking conditions (200 rpm). Then, 10 −5 serial dilution was done, and 100 µL of the mixture was plated on LB agar media and incubated at 37 • C for 16 h. The assay was done in triplicate for each biosurfactant, and the colony per plate for each biosurfactant was counted.

RNA Extraction, cDNA Synthesis, and Expression of AMPs
Each biosurfactant was injected at 300 ng/µL in to late larvae of T. molitor. 1X PBS was injected into the control group larvae. Twenty-four hours after biosurfactant injection, the whole-body samples from 20 larvae were injected with each biosurfactant to 500 µL of guanidine thiocyanate RNA lysis buffer, which was prepared from 0.5 M EDTA, 1 M MES buffer, 3 M guanidine thiocyanate, 200 mM sodium chloride, phenol red (40 µL), 25 µL of Tween-80, 250 µL of acetic acid glacial (for pH 5.5), and 500 µL of isoamyl alcohol. Samples were homogenized using a homogenizer machine (Bertin Technologies, France) for 20 s. RNA extraction was done following the protocol [76].

Biosurfactant-Triggered Immune Survivability of T. molitor
Thirty young instars (9-12 instars) larvae of T. molitor were injected with 300 ng/µL of each biosurfactant, and 1X PBS as control. At different time points (0, 12, and 24 h) post injection, E. coli (10 6 CFU) were injected in to the immune-triggered and PBS-injected larvae. In the 0 h group, biosurfactants and E. coli were injected at the same time. The experiment was performed in triplicate (10 larvae/replication). Mortality data were counted daily for 10 d.

Screening of Biosurfactant-Activated T. molitor Spatzle Genes
Biosurfactants were injected in to T. molitor larvae, and 24 h post injection, three tissues-the hemocytes, fat bodies and guts-were dissected, RNA extracted, and cDNA synthesized following the abovementioned protocol. Using gene-specific primers, the expressions of nine T. molitor Spätzle genes following biosurfactant injection were analyzed. The expression was quantified by qRT-PCR.

Effect of Tm Spätzle Gene Silencing on the Expression of AMPs
To synthesize the double-stranded RNA of the TmSpz genes, forward and reverse primers containing the T7 promoter sequence at their 5 ends were designed using the SnapDragon-Long dsRNA design software (Harvard Medical School, Boston, MA, USA, https://www.flyrnai.org/cgibin/ RNAi_find_primers.pl). The PCR product was amplified using AccuPower ® Pfu PCR PreMix with TmSpz_Fw and TmSpz_Rv under the following cycling conditions: an initial denaturation step at 94 • C for 2 min, followed by 35 cycles of denaturation at 94 • C for 30 s, annealing at 53 • C for 30 s, and extension at 72 • C for 30 s, with a final extension step at 72 • C for 5 min. The PCR products were purified using the AccuPrep PCR Purification Kit (Bioneer, Daejeon, Korea), and dsRNA was synthesized using the Ampliscribe TM T7-Flash TM Transcription Kit (Epicentre Biotechnologies, Madison, WI, USA), according to the manufacturer's instructions. After synthesis, the dsRNA was purified by precipitation with 5 M ammonium acetate and 80% ethanol. Subsequently, it was quantified using an Epoch spectrophotometer (BioTek Instruments, Inc., Winooski, VT, USA). As a control, the dsRNA of enhanced green fluorescent protein (dsEGFP) was synthesized, and all samples were stored at −20 • C until use.
To study whether the biosurfactants induced AMPs expression in T. molitor via the activation of Spätzle genes, TmSpz genes-those expressed by biosurfactants (T. molitor spätzle 3 (TmSpz3), spätzle 4 (TmSpz4), spätzle 6 (TmSpz6), and spätzle like (TmSpz-like))-were silenced by injecting the synthesized dsRNA (1 µg/larvae) of each gene into the larvae hemocoel. dsEGFP was injected to a separate set of larvae that acted as negative control. The gene expression knockdown (>85%) was achieved at 3 d (TmSpz3, TmSpz6, and TmSpz-like) and 5 d (TmSpz4) post injection of each dsRNA. The respective biosurfactants per specific day were then injected in the gene-silenced larvae. 1XPBS was injected as control group. Twenty-four hours post biosurfactant injection, RNA was extracted from the whole body. These time points were selected based on the preliminary experiment of AMP expression. Subsequently, cDNA synthesis was conducted following the previously described protocol. qRT-PCR was then conducted using specific primers to analyze the temporal expression patterns of the fourteen AMPs.

Effect of TmSpz Genes Silencing on T. molitor Survivability
The survivability study was investigated after the TmSpz3, TmSpz4, TmSpz6, and TmSpz-like genes were silenced. After gene silencing, the following biosurfactants were injected: iturin A in TmSpz3 and TmSpz-like gene-silenced larvae; fengycin, iturin, and bacitracin in TmSpz4 and TmSpz6 gene-silenced larvae. This grouping was based on the preliminary screening of TmSpz genes against biosurfactants. Twenty-four hours after biosurfactant injection, E. coli (10 6 CFU) were injected in to the immune triggered larvae. The experiment was done in triplicate. Mortality data were counted daily for 10 days.

Data Analyses
The survivability data were subjected to analysis of variance (ANOVA) using SAS 9.4. Differences from the control were assessed using Tukey's multiple range test (p < 0.05). Comparative AMP gene expression was calculated using delta delta Ct (∆∆Ct). The fold change from the internal control (TmL27a) and external control (PBS) was calculated using the formula 2 -(∆∆Ct) .

Conclusions
The antimicrobial properties of different biosurfactants from bacteria and fungi have been well studied. However, our current study revealed that fengycin and iturin A produced by Bacillus subtilis and rhamnolipids produced by Pseudomonas aeruginosa modulate the immunity of the yellow mealworm T. molitor. Generally, this study suggests that biosurfactants such as lipopeptides and glycolipids induce the expression of AMPs through activation of spätzle genes in hemocytes, thereby increasing the survivability of T. molitor larvae against E. coli ( Figure 11). Based on our current results, these biosurfactants could be used as safe immune elicitors in the mass rearing of T. molitor.

Conclusions
The antimicrobial properties of different biosurfactants from bacteria and fungi have been well studied. However, our current study revealed that fengycin and iturin A produced by Bacillus subtilis and rhamnolipids produced by Pseudomonas aeruginosa modulate the immunity of the yellow mealworm T. molitor. Generally, this study suggests that biosurfactants such as lipopeptides and glycolipids induce the expression of AMPs through activation of spätzle genes in hemocytes, thereby increasing the survivability of T. molitor larvae against E. coli ( Figure 11). Based on our current results, these biosurfactants could be used as safe immune elicitors in the mass rearing of T. molitor.