Hermetia illucens larvae used in this study were obtained from a colony established in 2014 at the University of Milan (Milan, Italy) starting from larvae purchased from a local dealer (Redbug, Milan, Italy).

BSF adults were kept at 27 ± 0.5°C, under a 12:12 h light:dark photoperiod, 70 ± 5% relative humidity with water supply and egg traps to promote oviposition. Eggs were collected in a Petri dish (9 cm × 1.5 cm), maintained at 27 ± 0.5°C until hatching, and then subjected to a weaning procedure as described in Pimentel et al. (2017). Lots of 300 larvae were grown on a standard diet for Diptera (Hogsette, 1992), composed of 50% wheat bran, 30% corn meal, and 20% alfalfa meal mixed at a ratio of 1:1 dry matter:water. The larvae were maintained at 27.0 ± 0.5°C, 70 ± 5% relative humidity, in the dark. Fresh diet was added to the feeding substrate every 2 days. All experiments were performed on actively feeding last instar larvae weighing between 180 and 230 mg. For Western blot analysis of phospho-Histone 3, midgut samples collected from larvae molting from third to fourth instar were also used.

Larvae were anesthetized on ice with CO₂. The gut was isolated in Phosphate Buffered Saline (PBS, 137 mM NaCl, 2.7 mM KCl, 8.1 mM Na₂HPO₄, 1.76 mM KH₂PO₄, pH 7.4) at 4°C, and the foregut and the hindgut were removed. The midgut, with the enclosed intestinal content, was subdivided into five regions: anterior midgut (AMG), middle midgut 1 and 2 (MMG1 and MMG2), and posterior midgut 1 and 2 (PMG1 and PMG2) (see Figure 1 and section “General organization of the larval midgut and pH of the midgut lumen” in Results).

For morphological and immunohistochemical analyses, the five midgut regions were fixed as described in the following sections.

For the rest of the analyses, the midgut epithelium was separated from the gut content, lightly blotted on filter paper, placed into cryovials and weighed. Tissues were stored in liquid nitrogen for aminopeptidase N (APN) assay (performed on AMG, MMG2, for simplicity termed middle midgut, and PMG2, for simplicity termed posterior midgut), Western blot (performed on the whole midgut for phospho-Histone 3 and on MMG1 for H⁺ V-ATPase), and qRT-PCR analysis (performed on AMG, MMG1, MMG2, PMG1, and PMG2).

The peritrophic matrix of each midgut region, with the enclosed intestinal content, was isolated, lightly blotted on filter paper and placed into microcentrifuge tubes. Samples were then centrifuged at 15000 × g for 10 min at 4°C. Supernatant, i.e., the midgut juice, was collected and used immediately to measure pH value, or stored at -80°C for the enzymatic assays. For these analyses only AMG, middle midgut 2 (for simplicity termed middle midgut), and posterior midgut 2 (for simplicity termed posterior midgut) were used to avoid contaminations of midgut content between tracts.

After isolation, midgut samples were fixed in 4% glutaraldehyde in 0.1 M Na-cacodylate buffer, pH 7.4, overnight at 4°C. After postfixation in 2% osmium tetroxide for 2 h at room temperature, specimens were dehydrated in ascending ethanol series and embedded in Epon/Araldite 812 mixture resin. Sections were obtained with a Leica Reichert Ultracut S (Leica, Nußloch, Germany). Semi-thin sections were stained with crystal violet and basic fuchsin and then observed with an Eclipse Ni-U microscope (Nikon, Tokyo, Japan) equipped with a TrueChrome II S digital camera (Tucsen Photonics, Fuzhou, China). Thin sections were stained with lead citrate and uranyl acetate and then observed with a JEM-1010 transmission electron microscope (Jeol, Tokyo, Japan) equipped with a Morada digital camera (Olympus, Münster, Germany).

After isolation, midguts were fixed in 4% paraformaldehyde in PBS overnight at 4°C. Specimens were then dehydrated in ascending ethanol series and embedded in paraffin. Sections (7 μm-thick), obtained with microtome Jung Multicut 2045, were deparaffinized, rehydrated, and pre-incubated for 30 min with PBS containing 2% bovine serum albumin (BSA) before incubation with anti-H⁺ V-ATPase antibody (Ab 353-2, polyclonal antibody raised against the highly purified V₁ complex of the H⁺ V-ATPase from Manduca sexta (Huss, 2001), a gift by Prof. H. Wieczorek, University of Osnabrück, Germany, dilution 1:10000 in 2% PBS/BSA), for 1 h at room temperature. After washing with PBS, sections were incubated for 1 h at room temperature with an anti-guinea pig Cy2-conjugated secondary antibody (dilution 1:200 in 2% PBS/BSA, Jackson ImmunoResearch, West Grove, PA, United States). After washes with PBS, sections were incubated with DAPI (100 ng/ml in PBS) for nuclear staining and then washed. Slides were mounted in Citifluor (Citifluor Ltd, London, United Kingdom) with coverslips and analyzed with an Eclipse Ni-U microscope (Nikon) equipped with TrueChrome II S digital camera (Tucsen Photonics). The primary antibody was omitted in controls.

Tissues were homogenized with T10 basic ULTRA-TURRAX (IKA, Staufen im Breisgau, Germany) in RIPA buffer (150 mM NaCl, 2% NP-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), and 50 mM Tris, pH 8.0) (1 ml/0.14 g of tissue), to which protease inhibitor cocktail was added (final concentrations: 1 mM AEBSF, 800 nM Aprotinin, 50 μM Bestatin, 15 μM E-64, 20 μM Leupeptin, 10 μM Pepstatin A) and phosphatase inhibitors (final concentrations: 1 mM sodium orthovanadate and 5 mM sodium fluoride) (Thermo Fisher Scientific, Waltham, MA, United States). The homogenate was centrifuged at 15000 × g for 15 min at 4°C and the pellet discarded. Bradford assay (Bradford, 1976) was used to determine the protein concentration of the supernatant. Clarified homogenates were denatured at 98°C in 4 × gel loading buffer for 5 min and loaded on 8% or 12% Tris-glycine acrylamide gel for SDS-PAGE analysis (60 μg protein/lane). Proteins were then transferred to a 0.45 μm nitrocellulose membrane (GE Healthcare Life Sciences, Little Chalfont, United Kingdom). The membranes were blocked with 5% milk in Tris Buffered Saline (TBS) (50 mM Tris-HCl, 150 mM NaCl, pH 7.5) for 2 h at room temperature, and subsequently incubated for 1 h at room temperature with the following antibodies: anti-phospho-Histone 3 (polyclonal antibody able to react with human and mouse phospho-Histone 3, but previously tested on different insect species (Barton et al., 2014; Apiz and Salecker, 2015; Choi et al., 2015; Franzetti et al., 2015; Franzetti et al., 2016)) (#06-570, Merck-Millipore, Burlington, MA, United States), diluted 1:1000 in TBS to which 2% milk was added, anti-H⁺ V-ATPase (Ab 353-2, see above for detail, section “Immunohistochemistry”), diluted 1:10000 in TBS to which 5% milk was added, or anti-GAPDH (Proteintech, Rosemont, IL, United States), diluted 1:2500 in TBS to which 5% milk was added. After three washes with TBS to which 0.1% Tween 20 was added, antigens of phospho-Histone 3 and GAPDH were detected with an anti-rabbit (dilution 1:7500 in TBS to which 5% milk was added; Jackson ImmunoResearch Laboratories), and antigen of H⁺ V-ATPase with an anti-guinea pig (dilution 1:10000 in TBS to which 5% milk was added, Jackson ImmunoResearch Laboratories) HRP-conjugated secondary antibodies. After three washes with TBS to which 0.1% Tween 20 was added, immunoreactivity was detected by SuperSignal chemiluminescent substrate (Thermo Fisher Scientific).

Total RNA was extracted from 15 to 40 mg of frozen tissue by using TRIzol reagent (Thermo Fisher Scientific), according to the manufacturer’s instructions. DNA contamination was removed by using TURBO DNA-free Kit (Thermo Fisher Scientific), then the purity of RNA was assessed by quantification and the integrity of RNA was tested by electrophoresis on 1% agarose gel. RNA was retrotranscribed with M-MLV reverse transcriptase (Thermo Fisher Scientific). qRT-PCR was performed with iTaq Universal SYBR Green Supermix (Bio-Rad, Hercules, CA, United States) using a 96-well CFX Connect Real-Time PCR Detection System (Bio-Rad). To calculate the relative expression of genes of interest, the 2^-ΔΔCt method was used, with HiRPL5 (Hermetia illucens Ribosomal Protein L5) as a housekeeping gene. The primers used for HiTrypsin (accession number HQ424575) were: F: ATCAAGGTCTCCCAGGTC and R: GGCAAGAGCAATAAGTTGGAT; for HiChymotrypsin (accession number HQ424574) were: F: AGAATGGAGGAAAGTTGGAGA and R: CAATCGGTGTAAGCAGAGACA. The primers used for HiRPL5 (F: AGTCAGTCTTTCCCTCACGA and R: GCGTCAACTCGGATGCTA) were designed on conserved regions of RPL5 in other insect species and the sequence was checked by sequencing the PCR product.

The pH of the midgut juice isolated from midgut samples as described above was measured by universal indicator strips with a resolution of 0.5 pH unit (Hydrion Brilliant pH Dip Sticks, Sigma-Aldrich, Milan, Italy). Samples of midgut juice of the different regions were extracted from at least 15 larvae and the experiment was repeated on 14 independent batches of larvae.

Last instar larvae were transferred to standard diet prepared with a water solution containing 4 mM cupric chloride (CuCl₂) and allowed to feed for 6, 12, 24, and 48 h. Control larvae were fed on diet prepared with water without CuCl₂. At the indicated time points, 15 larvae were collected, anesthetized on ice with CO₂ and dissected. The whole gut was removed and placed in ice-cold PBS. The midgut was divided into five regions as described above and fixed overnight at 4°C in 3.7% formaldehyde in PBS. Tissue samples were then rinsed in PBS and mounted on microscope slides with mounting medium (2:1 glycerol:PBS). Copper-dependent fluorescence was observed with a BX50 fluorescence microscope (Olympus) at an excitation wavelength of 365 nm, analyzing the emitted fluorescence between 585 and 620 nm.

At the same time points (6, 12, 24, and 48 h), midgut juice from 15 larvae was collected from both control and copper-fed larvae, and the luminal pH of anterior, middle, and posterior midgut region was measured as indicated above.

Frozen samples of midgut juice were thawed at 4°C and protein concentration was determined by the method of Bradford (1976), using BSA as standard. Assays were performed under conditions in which product formation was linearly associated with enzyme concentration.

The activity of APN was assayed using L-leucine p-nitroanilide (Sigma-Aldrich) as substrate (Franzetti et al., 2015) and measuring its degradation by release of pNA. After thawing, midgut samples were homogenized with a microtube pestle in 50 mM Tris-HCl, pH 7.5 (1 ml/100 mg tissue). Different volumes of homogenate were diluted to 800 μl with the same buffer and then 200 μl of 20 mM L-leucine p-nitroanilide were added. The diluted samples were subjected to continuous absorbance readings at 410 nm at 45°C. One unit/mg (U/mg) of APN activity was defined as the amount of enzyme that releases 1 μmol of pNA per min per mg of proteins.

Statistical analyses were performed with R-statistical software (ver. 3.3.2). The following analyses were performed: one-way analysis of variance (ANOVA) followed by Tukey’s test, paired and unpaired t-tests. Statistical differences between groups were considered significant at p-value ≤ 0.05. The statistical analysis performed for each experiment and the p-values are reported in the figure legends.

Three regions could be clearly recognized in the alimentary canal isolated from H. illucens larvae, i.e., the foregut, the midgut, and the hindgut (Figure 1A). The organization of the gut into three regions of different embryonic origin characterizes all insects (Nation, 2008) and allows the sequential function of the gut, i.e., food ingestion, digestion and absorption of nutrients, and elimination of frass. At variance with other Brachycera (Terra et al., 1988; Dubreuil, 2004), the larval gut of H. illucens did not present gastric caeca (Figure 1A). The midgut, the region involved in the production and secretion of digestive enzymes and in the absorption of nutrients, represented the intermediate and the longest part of the digestive system (Figure 1B). Midgut gross morphology and pH values of the lumen content (i.e., midgut juice) revealed the presence of distinct regions in the midgut of H. illucens larvae. The AMG (Figure 1C), with an acidic lumen content (Table 1), was formed by a deeply infolded epithelium. The middle region was characterized by a narrow and short tract (middle midgut 1, MMG1) that continued in a segment characterized by a larger diameter (middle midgut 2, MMG2) and absence of infolding (Figure 1D); the lumen of the middle region was strongly acidic (Table 1). The posterior midgut began downstream of a constriction at the end of the middle tract and was characterized by an alkaline luminal pH (Table 1). It was the longest region of the midgut and showed a short first tract (posterior midgut 1, PMG1) followed by a second one darker in color (posterior midgut 2, PMG2) (Figure 1E).

A detailed morphological characterization of the midgut epithelium was performed for all five regions described above (see “General organization of the larval midgut and pH of the midgut lumen” and Figure 1). Three main cell types were found along the whole length of the midgut, i.e., columnar, endocrine, and regenerative cells. Columnar cells were the most representative cell type of the epithelium and were characterized by the presence of microvilli on the apical membrane and basal infolding (Figure 2A). The endocrine cells were localized in the basal region of the epithelium and presented electron-dense granules in the cytoplasm (Figure 2B). Finally, regenerative cells were randomly distributed at the base of the epithelium (Figure 2C). These cells were able to proliferate during larva-larva molt (Figure 2D,E), as confirmed by Western blot analysis of phospho-Histone 3: a 17-kDa band (the expected molecular weight of the protein) revealed a significant presence of this mitotic marker during the molting phase, which was absent in the intermolt period (actively feeding last instar larvae) (Figure 2F and Supplementary Figure S1).

In addition to the three cell types described above, every midgut district presented typical features, indicating a marked regional differentiation of this organ. The AMG showed a thick epithelium (Figure 3A) formed by columnar cells with a developed basal infolding (Figure 3B). These cells were characterized by abundant rough endoplasmic reticulum (RER) (Figure 3C), and numerous electron-dense granules and mitochondria were present in the apical region under the microvilli (Figure 3D). The epithelium of MMG1 contained specialized cells, i.e., copper cells, characterized by a cup shape (Figure 4A), a large central nucleus (Figure 4B), and long microvilli (Figure 4B). A peculiar trait of these cells was the presence of several and elongated mitochondria inside each microvillus (Figure 4C,D). On the apical membrane of the cells portasome-like particles could be observed (Figure 4E). The presence of H⁺ V-ATPase, previously shown to be associated to portasomes (Shanbhag and Tripathi, 2009; Zhuang et al., 1999), was investigated by immunohistochemistry and Western blot analysis. Immunostaining performed on different midgut regions confirmed the presence of H⁺ V-ATPase only in MMG1, specifically associated with the apical membrane of copper cells (Figure 4F–J). To confirm that the antibody specifically recognized H⁺ V-ATPase in this midgut district, MMG1 was isolated and analyzed by Western blotting. Seven bands with a molecular mass of 13, 14, 27, 34, 55, 56, and 67-kDa, corresponding to most of the subunits of the cytoplasmic V₁ complex of H⁺ V-ATPase, were recognized by the antibody (Figure 4K). The ability of these cells to acidify the middle midgut lumen, thanks to the secretion of proton into the lumen via H⁺ V-ATPase, was evaluated. In copper-fed larvae of Drosophila melanogaster Meigen, 1830, copper cells have been shown to acquire an orange fluorescence signal due to the formation of a copper ions-metallothionein complex and, in turn, the acid-secreting activity of these cells is reduced (McNulty et al., 2001). To verify whether copper cells in H. illucens midgut showed similar features, larvae fed on diet containing CuCl₂ for different time periods (6, 14, 24, and 48 h) were examined. MMG1, the region containing copper cells, from control larvae reared without CuCl₂ in the diet did not exhibit orange fluorescence when examined under UV excitation for any of the time periods considered (Figure 5A and data not shown). Conversely, the same midgut tract of larvae fed on copper-containing diet showed a fluorescence signal starting from 14 h that increased significantly over time (Figure 5B–E). Midgut cells of the other regions showed an orange fluorescence only in larvae fed on copper-containing diet for 48 h (data not shown). We also examined the relationship between copper-dependent fluorescence and midgut lumen pH. Midgut juice samples from anterior, middle, and posterior region of the midgut were extracted from larvae fed for 24 h with copper-containing diet and control diet, and the pH was measured using universal indicator strips. As indicated in Figure 5F, copper feeding reduced larval midgut acidification in the middle midgut, whereas no variation was recorded in the anterior and posterior midgut.

MMG2 showed a wide lumen surrounded by a thin epithelium (Figure 6A) formed by large flat cells with a large elongated nucleus (Figure 6B) and a very short brush border (Figure 6C).

The posterior midgut displayed peculiar morphological features. In this region, the epithelium was thick (Figure 7A) with a well-developed brush border (Figure 7B). In particular, columnar cells in PMG1 differed from those of PMG2 for the numerous electron-dense granules localized under the microvilli (Figure 7B). Similar to the AMG, abundant RER (Figure 7C) was observed. Columnar cells of PMG2 showed very long microvilli (Figure 7D).

A previous study indicated that salivary glands of H. illucens larvae are only a minor source of digestive enzymes and that digestion is mainly accomplished by the midgut (Kim et al., 2011b). To evaluate whether the morphological differences observed in the different midgut regions as well as the strong variations in luminal pH along this organ corresponded to a functional regionalization, we examined the activity of enzymes involved in protein and carbohydrate digestion in each midgut region.

We first measured the total proteolytic activity in the midgut juice isolated from the anterior, middle, and posterior regions using azocasein as substrate at pH values as close as possible to those present in the lumen of each region. As reported in Figure 8A, the posterior tract showed the highest activity, which was more than forty- and twenty-fold higher than in the anterior and middle region, respectively. Since the posterior midgut may have a major role in protein digestion, a detailed characterization of the total proteolytic activity in this region was performed. First, we evaluated the dependence of the total proteolytic activity on temperature (Figure 8B). The highest activity was recorded at 45°C, while beyond this temperature the residual activity declined, reaching approximately 10% at 10 and 70°C. Since the lumen of the posterior midgut had an alkaline pH (Table 1), we wanted to determine whether the total proteolytic activity measured in this region could be ascribed to serine proteases, the major family of endopeptidases that shows a significant activity at alkaline pH values (Terra and Ferreira, 1994, 2005). When the proteolytic activity was measured at acidic pH (pH 5.0), a highly significant decrease in the enzymatic activity (approximately 13-fold reduction) was observed compared to that recorded at pH 8.5 (Figure 8C). Moreover, PMSF, a rather specific, competitive, and irreversible inhibitor of serine proteases, caused a significant reduction in the total proteolytic activity compared to controls, with an inhibition of 56.5 ± 1.4% (Figure 8D). In contrast, no inhibition was observed with E-64, an irreversible inhibitor of a wide range of cysteine proteases (Figure 8D).

Trypsin- and chymotrypsin-like enzymes are the major serine proteases involved in protein digestion in most insects (Terra and Ferreira, 1994, 2005). To verify the involvement of these enzymes in digestion processes, we first measured mRNA levels of two isoforms of these enzymes (Kim et al., 2011a) in the different midgut districts by qRT-PCR. Both genes were highly expressed in the posterior midgut, mRNA levels being significantly higher in PMG1 than in PMG2 for HiTrypsin (Figure 9A,B). By using specific substrates, we measured the tryptic and chymotryptic activity in the different midgut regions. Figure 10A reports trypsin-like activity in the midgut juice from anterior, middle, and posterior region of the midgut using BApNA as substrate. No activity was recorded in the anterior and middle region, while a significant trypsin-like activity was present in the posterior region. To better characterize the tryptic activity, we evaluated its dependence on temperature and pH. The optimal temperature, as for the total proteolytic activity (Figure 8B), was 45°C (Figure 10B). The evaluation of pH effect on trypsin activity at a temperature of 45°C showed that the highest BApNA hydrolysis occurred at pH values from 7.0 to 10.5, with an optimum at pH 8.5, and decreased significantly at acidic pH (Figure 10C). We also measured the activity of chymotrypsin-like proteases in the midgut juice extracted from the three regions of the midgut using SAAPPpNA as substrate (Figure 10D). The highest activity was recorded in the lumen content of the posterior midgut, confirming that this district is the main site for protein digestion by endopeptidases belonging to the serine proteases family.

To define whether the posterior midgut was also responsible for the final phase of protein digestion in which single amino acids are produced, we measured the activity of APN, one of the abundant exopeptidase families present in the midgut brush border of insect larvae (Terra and Ferreira, 1994). APN activity was recorded only in tissue homogenates prepared from the posterior midgut (Figure 11).

α-amylase activity in midgut juice samples from anterior, middle, and posterior regions of the midgut was also determined. The highest activity was recorded in the AMG, although starch was also digested to a significant degree in the posterior region; conversely, no activity was present in the middle midgut (Figure 12A). As the pH of the lumen in the three midgut regions was rather different in H. illucens larvae (Table 1), we performed a detailed evaluation of pH dependence in the AMG to assess how this chemical parameter influenced α-amylase activity. High activity was recorded at pH values ranging from 5.0 to 8.0, with a maximum between pH 6.0 and 7.0. At pH values lower than 5.0 and higher than 8.0 α-amylase activity significantly decreased and dropped to zero at pH 3.0 and 10.0 (Figure 12B). Therefore, the luminal pH value of the anterior and posterior midgut (Table 1), the regions where hydrolysis of internal α-1,4-glycosidic bonds of polysaccharides (i.e., starch and glycogen) occurs (Figure 12A), fits with the optimum pH range of α-amylase activity (Figure 12B).

We also measured lipase activity in midgut juice samples from anterior, middle, and posterior regions of the midgut. A significant activity was recorded in the anterior (4.51 ± 0.24 U/mg, mean ± SEM of 4 experiments) and in the posterior midgut (6.86 ± 0.62 U/mg, mean ± SEM of 6 experiments), in contrast, no lipase activity was detected in the middle region (Supplementary Figure S2).

Finally, since the midgut can be involved in the enzymatic clearance of ingested microorganisms (Lemos and Terra, 1991a; Lemos et al., 1993; Terra and Ferreira, 1994; Padilha et al., 2009), we measured the activity of lysozyme, which catalyzes the hydrolysis of 1,4-β-glycosidic bonds between N-acetylmuramic acid and N-acetyl-D-glucosamine residues in the peptidoglycan of the cell wall of many bacteria, thus compromising the integrity of the structure and causing the lysis of the cell. Lysozyme activity was measured in midgut juice samples extracted from anterior, middle, and posterior regions of the midgut. The highest activity was present in the middle midgut, whereas the other two regions showed significantly lower activities, especially the posterior region (Figure 13).