Microbial Community Dynamics during Rearing of Black Soldier Fly Larvae (Hermetia illucens) and Impact on Exploitation Potential

Impact on oviposition.While it has been known for decades that H. illucens females prefer decomposing material for oviposition (47), the role of bacteria in this preference only recently came into focus (48). This kind of interkingdom interaction was already reported for several other Diptera. For instance, gravid mosquitoes are attracted to a specific combination of 14 bacterial species (49), and gravid house flies sense volatiles produced by microbes on conspecific eggs to ensure synchronous larval development (50). Other insects can also interfere with each other's interkingdom interactions. For example, the presence of H. illucens larvae inhibits oviposition by the house fly (51). The most likely explanation is that the H. illucens larvae reduce the Escherichia coli numbers in the substrate (52), which is assumed to be a nutritional source for the house fly larvae (53).

Considering H. illucens, the presence of sterile eggs, devoid of microbes, results in less oviposition by other females than when nonsterile eggs are present (48). An assessment of the impact on oviposition of 18 bacterial species, isolated from H. illucens or from competing insects, illustrates the complex interplay between H. illucens and specific microbial species. More specifically, an aliquot of specific bacteria at a concentration of 10⁴, 10⁶, or 10⁸ CFU per ml was put on an agar plug, which was sterilely inserted in the flutes of one of two sterilized cardboard blocks to assess the impact of their presence on the egg deposition preference. Two bacterial species, Ignatzschineria sp. and Acinetobacter sp., isolated from known competitors of H. illucens, the secondary screwworm (Cochliomyia macellaria) and the lesser mealworm (A. diaperinus), respectively, were found to trigger repellency and result in less egg deposition. In contrast, two other bacterial species enhance oviposition, Gordonia sp. (isolated from H. illucens eggs) and Providencia sp. (isolated from the hairy maggot blowfly [Chrysomya rufifacies]). As the latter insect preys on other blowfly species, it makes sense that there is no competition between H. illucens and the hairy maggot blowfly. Intriguingly, this bacterial species is also reported to be present in all life stages of H. illucens, adding weight to the expectation that this species plays a key role in regulating oviposition (48).

Impact on food digestion (enzymatic activities).A biochemical characterization of extracts from the GIT of H. illucens demonstrated high amylase, lipase, and protease activities (54). Most of these activities were very high in the guts of the larvae, while extracts from their salivary glands exhibited less than 10% of the total enzyme activity. Hence, for H. illucens, the gut appears to be the main site for digestion, in contrast to the situation in, for example, insects with primarily extraoral digestion (54). Furthermore, both qualitative and quantitative assays reveal that more kinds of digestive enzymes, often also with higher levels of activity, are present in the gut of H. illucens than in that of the house fly, which explains why the former larvae belong to the most efficient scavengers of all known species of fly (54, 55).

Gut microbes play an important role in aiding their host during the digestion of complex substrates by possessing metabolic properties that the insect lacks. Indeed, studies have shown that a number of metabolic reactions occurring in the insect GIT are encoded by the metagenome of its microbiota rather than by its own genome (56, 57). These microbial functions during digestion could be the basis of the observed substrate-dependent microbial community structure variations, as the bacteria more adapted to digest a specific substrate will have a fitness advantage in the GIT over less-adapted bacteria.

Considering the metabolic capacities of H. illucens larvae and the observation of the high enzymatic activity in their GITs, the bacterial communities in the larval GITs can also be mined to find novel microbial enzymes such as proteases, cellulases, lipases, xylanases and pectinases that degrade organic compounds and might be of particular interest for industrial processes. This potential is demonstrated by the identification of a novel cellulase and an alkaline amylopullulanase from the larval gut microflora using a function-based metagenome screen (24, 25). Particularly, the amylopullulanase has attractive biotechnological characteristics toward exploitation. It is active at lower temperatures than the current industrial enzymes and shows a broad pH tolerance and an increased durability toward several noxious compounds. It can be expected that in the future, more microbial enzymes will be harvested from the GITs of H. illucens larvae, considering their remarkable capacity to digest a wide range of complex substrates, e.g., the digestion of lignocellulose into xylose, which enables biodiesel production (58).

Impact on detoxification of substrates.Various xenobiotics, such as antibiotics or pesticides, are complex molecules that can cause problems in the environment and for prolonged periods of time due to their high stability. Not only can H. illucens larvae grow in substrates contaminated with these molecules, but their presence also results in a shorter half-time of these pharmaceuticals and pesticides, e.g., 1.1 days instead of 25 days for trimethoprim (59). Hence, H. illucens could play a significant role not only in the recycling of biological waste but also in its bioremediation into a safer composted material for reuse in the environment. Data suggest that in the house fly, the GIT microbiota aids in the degradation of such molecules by harboring specific metabolic pathways (60). Also, in pest insects, symbioses with microorganisms have been identified that enable the degradation of complex plant allelochemicals and xenobiotic insecticides (61). Therefore, the H. illucens-associated microbiota most likely also facilitates the observed degradation, maybe even in a faster way than the house fly. In the future, it will be interesting to mine the metagenome of this microbiota to identify the xenobiotic degradation mechanisms and exploit those enzymes for bioremediation applications.

A category of natural toxic molecules is the mycotoxins, such as aflatoxin B1 (AFB1), which are produced by fungi that grow on specific crops (such as cereals or nuts). The contaminated crops are often removed from the food chain. However, H. illucens larvae can grow on substrates containing up to 0.415 mg/kg of AFB1 without showing accumulation. In fact, AFB1 appears to be catabolized or at least bound to proteins (62). H. illucens larvae could thus be used to recycle mycotoxin-contaminated substrates. The question remains as to whether the microbiota plays a role in this catabolism.

Improvement of growth performance by adding companion bacteria.For several farm animals, a well-defined set of probiotics has already been developed and is commercially available. Probiotics are live microorganisms which confer a health benefit for the host when administered in appropriate and regular quantities (18). For insects used in industrial rearing, this is virtually uncharted terrain. Nevertheless, Yu et al. (63) reported in 2011 the potential for the use of so-called companion bacteria. Four different Bacillus strains, three isolated from the larval gut (B. subtilis S15, S16m, and S19) and one from the feed (B. subtilis subsp. natto), were shown to enhance the growth of the larvae. This is most likely due to their aid in the digestion of the provided substrate (chicken manure). Treatment with each of the strains resulted in heavier larvae combined with a shortened development time, going from 34.33 ± 3.51 days between hatching and 90% reaching the prepupal stage for the uninoculated sample to 29.00 ± 1.00 days for the samples inoculated with the B. subtilis S15 strain (63). This reduced development time can directly impact the productivity of H. illucens as an insect for industrial rearing. However, with the reported error margins, it remains to be seen if this effect would yield an economic benefit. At the same time, it might be that other bacteria will have a more pronounced effect under specific conditions. Future efforts may aim to identify novel strains with even more pronounced beneficial effects, which can be matched with specific substrates, and to commercialize them for the insect industry.

Reduction of specific pathogens administered in the substrate.To maximize the sustainable nature of H. illucens as an alternative protein source, it is envisioned that in the future, these insects will be industrially grown on various waste streams, probably even in different stages of decay or including animal or human manure, which is currently not allowed in the EU (EG no. 767/2009). Such substrates are an ideal reservoir for many significant foodborne pathogens, including but not limited to Escherichia coli O157:H7 and Salmonella spp. Therefore, it is of key importance to assess how their abundance is affected by the presence of H. illucens larvae to ensure that pathogens do not accumulate in the larvae nor the final product. So far, a few studies focused on the interactions between the larvae and the two mentioned pathogens and found that the bacterial counts in the substrates for both pathogens were reduced (Table 2) and that neither of their numbers accumulated inside the larval body (52, 64, 65).

Furthermore, it appears that the effectiveness of pathogen reduction depends on the temperature and the composition of the substrate. Indeed, E. coli O157:H7 numbers were reduced more effectively at a temperature between 27°C and 32°C than at 23°C (64). At a higher temperature (35°C), substrate-dependent reductions in bacterial counts were observed, even in the controls not containing H. illucens, as well as 100% mortality of the larvae (52). An example of the effect of the substrate composition is the difference in the reduction of fluorescently labeled E. coli O157:H7 cells in chicken manure compared to that in hog manure. In chicken manure, the presence of H. illucens larvae led to an increased level of reduction in the E. coli O157:H7 population, while in the hog manure, the observed pathogen reduction was the same whether the larvae were present or not. A difference between these two manures that might play a role in the missing antimicrobial activity of the larvae in the hog manure is their pH. The hog manure had an initial pH of 6.0 to 6.2 that decreased over time, while the chicken manure had an initial pH of 7.4 to 8.2 that increased over time (64). The growth of the larvae was not influenced by this lower pH, which matches with the claims of industrial rearers that larvae grow on fermented substrates (personal communications). In fact, larval growth was more than double in the hog manure compared to that in the chicken manure; thus, the reduced growth cannot explain the absence of pathogen reduction. The authors pose as a possible explanation that the antimicrobial compounds present in H. illucens larvae are less stable or active at the lower pH values (64). For example, this is the case for a known insect-harbored antibacterial protein, defensin A, of the flesh fly, Phormia terranovae, which has a maximal stability at a pH of around 8.3 (66).

In conclusion, the antibacterial mechanisms of H. illucens larvae seem to be able to differentiate between organisms, as they do not target all bacteria. As stated, specific strains of Salmonella and E. coli are clearly targeted (64). Indeed, Lalander et al. (65) also observed a quick drop below detection levels of a strain of Salmonella enterica in the GITs of human-manure-fed larvae. Interestingly, the counts of another tested pathogen, Enterococcus spp., were unaffected for more than 8 days by larval presence (65). These observations illustrate that H. illucens larvae are able to reduce the load of specific pathogens in their substrates, under the right conditions, without themselves accumulating these pathogens. However, the studies performed so far do not indicate a complete eradication of pathogens. Therefore, a processing step with a decontaminating effect is still recommended prior to the use of larvae as a feed ingredient.

Use of larvae as a novel feed additive to aid host immunity.If the antibacterial activities observed in H. illucens larvae could persist in the farm animals fed with these larvae, then these larvae could be used as an alternative to the AGPs used in the livestock industry in the past to improve gut health. Unfortunately, the effect on animal health of H. illucens larvae has hardly been investigated. So far, the focus has been on the nutritional impact of feeding insects to other animals (69–72). Recently, Spranghers et al. considered gut health parameters of weaned piglets that were fed with diets containing either defatted or full-fat BSF prepupae (73). They reasoned that the high content of lauric acid (C_12:0), known for its antimicrobial effects against Gram-positive bacteria, in the fat of the larvae could result in antimicrobial effects in the piglet gut and provide added value to H. illucens larvae as a protein source. While they observed a 0.5-log reduction in group D streptococci counts in the gut, the decrease was not statistically significant. A possible explanation for this weak impact is that only 0.03 g C_12:0/100 g of the total 1.58 g C_12:0/100 g present in the feed reached the gut, meaning that most of the C_12:0 was absorbed earlier in the stomach (73). What could boost the efficacy of H. illucens larvae as an antimicrobial feed additive is high levels of other components of the insect humoral response (e.g., AMPs). Unfortunately, the levels of these components were not specifically investigated in this study. However, the authors stated the number of AMPs was expected to be low, and it is also not known if AMPs remain stable during feed processing and are able to reach the gut. In the future, this stability will have to be studied to achieve a successful implementation of whole H. illucens larvae as an antibacterial feed additive.

Modulation of larval immunity to improve their antibacterial properties.Alternatively, the antibacterial efficacy of the H. illucens larvae can be increased by developing methods to boost their immune systems, compensating for the losses in activity during processing. To achieve this, it is necessary to know how the expression of the immunity actors in the larvae is regulated and whether immunity can be directed to target specific pathogens. Recent research, focusing on the immune mechanisms of H. illucens and the modulation thereof, has shown that its immunity is indeed inducible, increasing the feasibility for the exploitation of H. illucens larvae as an antibacterial feed additive (68). Challenging the immunity of H. illucens larvae by introducing bacteria via piercing was found to increase phenoloxidase activity in the hemolymph and to induce the production of AMPs (68). As a result, the antimicrobial activity of the hemolymph increased, even in a pathogen-specific manner. Indeed, larvae challenged with the Gram-positive Micrococcus luteus produced no activity toward the Gram-negative E. coli. On the other hand, larvae challenged with E. coli did generate activity against Gram-negative bacteria and also produced a higher activity against Gram-positive bacteria (68). This suggests a discrimination between Gram-positive and Gram-negative bacteria by the immune system. This distinct immunity response is related to the production of two proteins of approximately 58 kDa and 27 kDa in the hemolymph of the E. coli-challenged larvae, while two other proteins of approximately 3.8 kDa and 43 kDa are expressed after both challenges (68). The authors proposed that by varying the bacteria used for challenging, different sets of antimicrobial compounds will be induced. While the piercing method could improve the antibacterial properties of H. illucens larvae in feed, it is not feasible at an industrial scale.

A more achievable approach was recently demonstrated, which involves the substrate-dependent alteration of H. illucens immunity (74). The manipulation of larval immunity via the substrate composition is an example of the emerging field of nutritional immunology. In this study, larvae were fed a standard diet supplemented with a mixture of bacteria or with different additives of organic waste, such as sulfonated lignin and sunflower oil. Next, the antimicrobial activity of larval extracts and the immunity-related transcriptome were analyzed. The transcriptome analyses revealed that H. illucens harbors a more diverse spectrum of AMP and lysozyme gene family members than most insects, containing 6 attacins, 7 cecropins, 26 defensins, 10 diptericins, and 4 knottin-like peptides (74). Surprisingly, the protein-rich diet (supplemented with brewer's grains) triggered the most differential expression of these genes, followed by the plant oil diet. Both diets induced stronger immune responses than when bacteria were directly added to the diet. Nevertheless, the supplementation of bacteria also triggered a significant upregulation of more than half of the identified candidate AMP genes (74). On the basis of growth inhibition assays, the strongest inhibition of Gram-negative bacteria was observed with the aqueous extracts from the larvae reared on the high-protein and cellulose diet, while for Gram-positive bacteria, the most active extracts were those from the larvae reared on chitin, cellulose, bacteria, and plant oil (Table 3) (74). This diet-dependent expression of AMPs most likely enables H. illucens larvae to adapt the microbiome from the substrate as well as the host-associated core microbiome to enable a flexible digestion of a variety of diets. Future research will have to explore if the expression of specific compounds is responsible for the selection of a unique microbial community in response to environmental factors in H. illucens.

The characterization of the underlying mechanisms behind these diet-dependent expressions would enable the exploitation of this immunity activation and/or modulation, increasing the potential of H. illucens as a viable alternative to antibiotics in feed. On the basis of the substrate the larvae are fed, they can then be directed to express specific immunity genes enabling the control of infectious pathogens in and safe-guarding the health of a wide array of livestock and companion animal species.

Search for antibacterial compounds expressed by the larvae or their microbial community.To circumvent the problems associated with molecules having to withstand processing and digestion, antibacterial molecules can be purified and identified directly from larval extracts, or alternatively, selected immunity-related genes of H. illucens can be expressed in vitro and characterized. Both approaches aim to discover novel molecules that could serve as the inspiration for the design of novel drugs.

Several studies examined the bioactive effects of different larval extracts (74–77). These studies show various antimicrobial activities against different bacteria that depend on the environment in which the larvae were reared, as well as the extraction method used (see Table S1 in the supplemental material). Their activity was also able to hinder the early stages of biofilm formation, which is a logical consequence of their antibacterial activity against planktonic cells (76). So far, no one has investigated whether these extracts also can eradicate mature biofilms. To identify bioactive molecules from the crude extracts, further separation methods (e.g., Sep-Pak C₁₈ or high-pressure liquid chromatography [HPLC]) can be applied until a pure fraction of the active compound is obtained for identification. Using this strategy, Park et al. (77) showed that the compound responsible for activity against E. coli was strongly hydrophilic and that several compounds with strong activity against methicillin-resistant Staphylococcus aureus (MRSA) act in synergy. Furthermore, these compounds are highly robust, as they remained active after several freeze-thaw cycles and after lyophilization. Such parameters are key properties for a stable pharmaceutical product. The identified molecule was named DLP4, a defensin-like peptide with a potent activity against Gram-positive bacteria (78). Using a similar approach, hexanedioic acid was also identified as a new substance with antibacterial activity against various pathogenic bacteria, including MRSA (79).

Recently, the technological advances in genomics and transcriptomics have enabled a more direct approach for the identification and characterization of immunity genes. For AMPs, several online databases have emerged, which contain over 2,000 molecules. Comparing new sequences to such databases enables the rapid identification of candidate AMPs in organisms. Using a reference set of insect-derived AMPs and lysozymes, a total of 53 genes were annotated as putative AMPs in H. illucens (74). The challenge now is to express such genes and characterize their antibacterial properties. So far, a total of 9 AMPs are reported to be cloned, expressed, and structurally analyzed. For three of these, the MIC values against various strains were determined (Table 4) (80, 81). The first is stomoxynZH1, an AMP of the class of the cecropins, which consists of a linear cysteine-free peptide of 60 amino acids (aa) with an alpha helix-like structure. Studies with homologous stomoxyns showed that this molecule has a broad activity spectrum, with low cytotoxicity and a rapid activity upon exposure. The H. illucens-derived molecule, fused to thioredoxin, also had both antibacterial and antifungal activities. In the future, the stability of this peptide without the fusion needs to be investigated (80).

Two other peptides that were characterized are DLP2 and DLP4. These homologous defensin-like peptides have high sequence similarities (60 to 65%) to other known insect defensins: sapecin, insect defensin A, and lucifensin. These molecules are potent antibacterials against Gram-positive bacteria, with DLP4 killing bacteria after just 0.5 h of exposure. On the other hand, they have almost no impact on Gram-negative bacteria. DLP2 and DLP4 retain 85% and 93%, respectively, of their activity after an incubation of 1 h at 80°C, which shows a high thermal stability. Furthermore, the observed cytotoxicity toward eukaryotic cells is low for both molecules. MRSA bacteria also fail to develop additional resistance during 30 serial passages, while they do gain resistance toward other commercially used antibiotics in that time frame. To summarize, in vitro, these two peptides entail a high stability, have no impact on eukaryotic cells, and trigger low rates of bacterial resistance development, which makes them promising drug candidates. However, these in vitro conditions obviously differ greatly from those in clinical use. To this end, 6-week-old female BALB/c mice were injected intraperitoneally with MRSA on one side and two doses of the DLP on the other side to assess the in vivo potential of these novel antibacterial compounds. In mice, both peptides could not only prevent mortality, but they also triggered immunomodulatory effects that resulted in an improved balance between proinflammatory and anti-inflammatory cytokines upon bacterial infection (81). To date, this is the most comprehensive evaluation of an H. illucens-derived AMP toward antibiotic development. The results clearly show their potential as an innovative new class of antibacterial compounds. It can be expected that in the future, all H. illucens AMPs will be systematically characterized using a streamlined methodology to rapidly compare and score antibacterial potential.