*2.9. Statistical Analyses*

Microbiological data were subjected to one-way analysis of variance (ANOVA). Pair comparison of treatment means was done using Duncan procedure at *p* < 0.05. Differences between *T. molitor* and *A. domesticus* powders were evaluated using the generalized linear model (GLM) procedure. The linear dose-dependent relationship between the scavenging properties of the DPPH radical and the various concentrations tested has been tested through the linear regression test. The statistical analysis was done with Statistical Analysis System 9.2 software (SAS Institute, Cary, NC, USA).

#### **3. Results and Discussion**

#### *3.1. Microbial Loads of Insect Flours*

Insect powders have been recently investigated for their microbiological/safety aspects by different research groups, but so far, very little is known about their characteristics at the expiry date. The microbiological investigation of *T. molitor* and *A. domesticus* powders in this work was carried out after 12 months of storage. Mealworm powder hosted 7 microbial populations, while house cricket

powder was characterized by a higher microbial diversity, since the *Pseudomonas* group was also detected, forming a total of 8 microbial populations (Figure 1).

**Figure 1.** Microbial loads (Log CFU/g) of insect powders. Abbreviation: PCA, plate count agar for detection of total mesophilic microorganism; MRS, de Man-Rogosa-Sharpe agar for detection of mesophilic rod LAB; M17, medium 17 agar for detection of mesophilic coccus LAB; KAA, kanamycin esculin azide agar for detection of enterococci; PAB, *Pseudomonas* agar base for detection of pseudomonads; VRBGA, violet red bile glucose agar for detection of *Enterobacteriaceae*; CPS, coagulase-positive staphylococci; NA, nutrient agar for detection of spore-forming aerobic bacteria; MA, malt agar for detection of yeasts and moulds incubated for 48 h and 7 days, respectively. Results indicate mean values and standard deviation of three plate counts. Different lowercase letters indicate significant differences on microbial concentrations performed according to Duncan test between insect powders for *p* < 0.05 while different uppercase letters indicate significant differences on microbial concentrations between different growth media for *p* < 0.05.

Statistically significant differences were observed for the levels of LAB cocci, enterococci, pseudomonads, CPS and members of the Bacillaceae family between the powders of *T. molitor* and *A. domesticus*. Yeasts and moulds were undetectable for both matrices, while only mealworm powder showed pseudomonads below the detection level. The levels of TMM in both powders were a little lower than 106 CFU/g. The highest levels in mealworm powder were reached by LAB cocci (5.95 log CFU/g) followed by CPS, members of Bacillaceae family and LAB rods, which were all at cell densities above 105 CFU/g, while house cricket powder showed only LAB cocci at these levels. In general, *A. domesticus* powder showed lower levels of all microbial groups than *T. molitor* powder except the members of *Enterobacteriaceae* family.

The microbiology of insect powders soon after production has been studied by other authors considering different microbial groups. Bußler et al. [14] evaluated the total viable counts (on PCA) of *T. molitor* powder and observed levels of 7.72 log CFU/g of dry matter. Klunder et al. [11] analyzed the microbial loads of *T. molitor* powder after the insect were subjected to boiling and crushing reporting levels of 4.8 log CFU/g. The same authors also analyzed house crickets after boiling and stir-frying showing levels of 2.7 log CFU/g on PCA. In the same study, the members of *Enterobacteriaceae* family were 2.6 log CFU/g for mealworm and below the detection level in house cricket powder, while bacterial endospores were detected only in *A. domesticus* powder and counted at 1.5 log CFU/g. LAB were also the object of investigation by Klunder et al. [11], but their levels (ranging from 7.9 to 8.9 log CFU/g) were evaluated only after fermentation of the mixture mealworm powder/water at 30 ◦C. Information about the microbiology of *T. molitor* and *A. domesticus* are also available for the entire insects after

freeze-drying [15]. Furthermore, Osimani et al. [46] also investigated on the hygiene of these insects reared under controlled conditions.

Only the study of Klunder et al. [11] considered the microbiological changes occurring during the refrigeration (4 ◦C) and ambient temperature (25 ◦C) storage of *A. domesticus* powder, but the monitoring period lasted 16 days. Thus, due to the different preparation and storage duration/conditions and the samples analyzed, a real comparison of data with those available in the literature is difficult. In general, the levels of TMM and enterobacteria registered for *A. domesticus* and *T. molitor* powders after 12 months of storage were comparable to those reported for the insects raised in the open field [11].

#### *3.2. Culture-Independent Microbiological Analysis*

After processing of the demultiplexed FASTQ files using DADA2 package, we obtained 69,818 and 133,305 reads for house cricket and mealworm powder, respectively. Only taxonomic groups with, at least, two representative sequences per taxonomic unit were retained and the relative abundances were reported in Figure 2.

**Figure 2.** Relative abundances (%) of bacterial groups identified by MySeq Illumina in insect powders. Only taxonomic groups with at least two representative sequences per taxonomic unit were retained.

The diversity evaluation done using MiSeq Illumina identified members of the *Citrobacter* genus as the major components of the mealworm powder, which are commonly associated with the mid-gut of Lepidoptera insects [47], followed by *Enterobacteriaceae*. Members of *Porphyromonadaceae* family constituted the major bacterial group of house cricket powder. Many species of the family *Porphyromonadaceae* are part of the indigenous microbiota of the human and animal gastrointestinal tract and oral cavity [48]. To retrieve information at the species level, the most representative sequences of the two samples were manually blasted against the NCBI database. All *Enterobacteriaceae* OTUs were identified as belonging to *Salmonella enterica*/*Pseudocitrobacter faecalis*/*Cronobacter sakazakii*, while *Porphyromonadaceae* OTUs belonged to uncultured bacteria. No significant differences were found between observed and predicted (Chao1estimator) OTUs. Therefore, it is possible to capture the majority of OTUs present in each sample. Statistical analysis using the Kruskal-Wallis test revealed significant differences among *A. domesticus* and *T. molitor* (*p* < 0.05). This analysis showed a highest biodiversity in terms of bacterial species for *A. domesticus* powder. Regarding the lactic acid bacteria group, Illumina identified *Enterococcus* and *Lactobacillus* genera, generally found in entire mealworms [46].

Even though NGS analysis, as performed in this study, was based on DNA and this approach does not provide any indication on the viability of the detected species some safety issues arose from the composition of the microbiotas of the two insect powders analysed. In particular, the major groups belonged to enteric bacteria that are commonly found in several raw materials, such as *Enterobacteriaceae* family members in raw milk [49], meat [50] and vegetables [51]. Thus, considering insect powders as raw materials to be added as ingredient in food matrix formulations rather than as foods themselves, no particular food safety alerts concerning the major foodborne pathogens were shown by this study, even though the presence of *Salmonella* spp. deserves deepen investigations. However, several enterococci were isolated in viable form. Due to the antibiotic resistance gene transfer that could occur in some *Enterococcus* strains [52], a more comprehensive investigation on the isolates of this study should be done once insects flours will be used in food and feed formulations.

#### *3.3. Characterization of LAB*

The colonies grown on the media (MRS, M17 and KAA) specifically used for mealworm and house cricket powder LAB enumeration were collected and characterized. Barely 40 cultures were still considered presumptive LAB after Gram determination and catalase test. The microscopic investigation showed a coccus shape for all bacteria even though MRS allows the growth of rod LAB. The genetic typing showed nine different RAPD profiles, which corresponded to nine distinct strains (Figure 3). The sequencing of rRNA genes allotted all strains into *Enterococcus* genus.


**Figure 3.** Dendrogram obtained with combined random amplification of polymorphic DNA (RAPD)-PCR patterns of *Enterococcus* strains.

In particular, *Enterococcus faecium* were identified from *A. domesticus* powder, while *E. faecium* and *Enterococcus lactis* were observed in mealworm powder as confirmed by species-specific multiplex PCR.

Although some information is available on the presence of LAB in edible insects [46], this is the first study aimed at increasinging the characterization of insect powder LAB. Mealworm larvae and their frass analyzed for the presence of LAB by culture-independent tools were found to host *Lactobacillus, Pediococcus*, and *Leuconostocaceae* when the investigation was done through Illumina, while *Lactococcus* spp., *Enterococcus* spp. and *Lactobacillus* spp. when denaturing gradient gel electrophoresis was used [17], but no information on their viability was reported. As anticipated above, Klunder et al. [11] investigated LAB during the fermentation of mealworm powder, starting from a batch prepared mixing the insect powder with water at a ratio of 40:60, which was subjected to 5 fermentation cycles using 10% inoculums from the previous cycle. LAB increased in time, but the acidifying species were not identified.

It was seen that the dominant LAB of *T. molitor* and *A. domesticus* powders were all members of *Enterococcus* genus. Enterococci are bacteria of intestinal origin often associated with food matrices, but they rarely represent starter cultures for the fermentation processes [53]. The presence of *Enterococcus* in food products is a direct consequence of faecal contaminations [53]. In the case of insects, they are transferred from their intestinal tracts. In general, the enterococci from insect powder did not show dangerous risk factors since all of them were not virulent and were sensitive to the 12 antibiotics commonly used for the treatment of human and animal infections.

#### *3.4. Proximate Composition, Energy and Fatty Acid Profile of Insect Powders*

The proximate composition of insect powders is summarized in Table 1.

**Table 1.** Proximate composition (%), chitin (%), energy content (kJ/100g) and fatty acid composition (%) of the insect powders. Values are mean (three replications) ± standard deviation.


n.d. = not detected.

The lipid value is 26.17% (Table 1) in *T. molitor* and 21.66% in *A. domesticus*. As expected, both insect species showed a high proportion of protein, 52.95% for *T. molitor* and 63.62% for *A. domesticus*. Insects contain chitin, a primary component of the exoskeleton of arthropods [54]; our analyses showed a chitin percentage of 14.42 for *T. molitor* and 5.50 for *A. domesticus*. Regarding moisture and ash, the values were between 3 and 5% for *T. molitor* and *A. domesticus*, respectively. The energy content

(Table 1) varied slightly between the species, ranging from 1868.59 ± 2.26 kJ/100g for *T. molitor* to 1882.88 ± 3.65) kJ/100 g for *A. domesticus*. These results are comparable to the energy supplied by beef (1735 kJ/100 g) or fish (1662 kJ/100 g) [55]. The results of the proximate composition were comparable to those reported in literature [18,54–56] and put in evidence the high potential of insects as alternative sources of new and renewable animal proteins and fat [54,57].

Table 1 shows the fatty acid composition of edible insect's powder. The amount of saturated fatty acids (SFA) ranged from 23.65% for *T. molitor* to 39.68% for *A. domesticus*. The two main components of the SFA were palmitic acid (C16:0) and stearic acid (C18:0). The highest amount of these fatty acids was observed in *A. domesticus*. Values detected in *T. molitor* powder were in agreement to those reported from Zieli ´nska et al. [55] that showed 18% of palmitic acid and 3.8% of stearic acid in the larvae of *T. molitor*. The content of monounsaturated fatty acids (MUFA) varied from 24.66% in *A. domesticus* to 48.06% in *T. molitor*. The major MUFA of edible insect's powder is oleic acid (C18:1 *n*-9). The highest content of oleic acid was observed in *T. molitor*. The fraction of polyunsaturated acids (PUFA) ranged from 28.28% in *T. molitor* to 35.66% (Table 1) in *A. domesticus*. Yang et al. [58] obtained similar values for crickets (33.8%). In particular, in this study, in *T. molitor* powder, the omega-6 (*n*-6) PUFA were higher than omega-3 (*n*-3) (2.3%). For the n-6 class, the most abundant FA was linoleic acid. Regarding *A. domesticus* powder, n-6 PUFA were higher than omega-3 (*n*-3). Linoleic acid was the most abundant fatty acid of the n-6 class similar to values observed by Yang et al. [58] in ground crickets (32.2%) and by Osimani et al. [18] in the same species. Furthermore, various data regarding the body composition of insects showed the variable composition of FA between species, origin and developmental stages [56].

#### *3.5. Enzymatic Hydrolysis*

The enzyme selection is the most important factor in protein hydrolysis affecting the yield and physico-chemical properties of the final product. From the comparison of the three types of commercial proteases used, Alcalase from *Bacillus licheniformis* had provided the best results in hydrolyzing the total proteins of the *T. molitor* powder (Figure 4a) in terms of DH %, with increasing trend, followed by Protamex and Flavourzyme, which instead showed rather stable values over time (*p* < 0.05). The maximum DH% (between 20.1% of the enzyme Protamex and 25.8% of the enzyme Alcalase) were reached after 180 min of reaction.

The enzymatic reaction of hydrolysis to the powder of *A. domesticus* (Figure 4b) was better in terms of DH%, also in this case, with the enzyme Alcalase (*p* < 0.05). The plateau phase was reached at 195 min, where the maximum value of 24.6% (DH) (Figure 4d) was observed, which remains unchanged in subsequent measurements. Several authors had reported that, compared to other proteolytic enzymes, Alcalase allows superior protein recovery and provide hydrolysates with good functional properties [59–63]. Generally, alkaline proteases, including Alcalase, exhibit greater proteolytic activity than acid or neutral proteases such as Flavourzyme [62]. Yang et al. [64] reported that, among the proteases used, Alcalase had higher DH during the hydrolysis period, which suggested that Alcalase is more efficient than the other enzymes for preparing protein hydrolysates from edible insects. Tang et al. [23] obtained better results in terms of DH%, in *T. molitor* larvae, with a combination of Alcalase and Flavourzyme followed by slightly lower values obtained only with the enzyme Alcalase. Mizani et al. [60] reported that Alcalase, when used in combination with sodium sulphite and triton x-100, increased the yield of protein hydrolysates from *Penaeus semisulcatus* shrimp waste (heads) from 45.1% to 62–65%. This is related to a reduction of disulphide bonds and increased the solubility of proteins, as has been previously demonstrated in the case of soy products [65]. Recently, Alcalase has been used extensively in the hydrolysis of plant and animal proteins [20,21,66].

**Figure 4.** Physico-chemical properties: (**a**) Degree of hydrolysis of *T. molitor* powder; (**b**) Degree of hydrolysis of *A. domesticus* powder, different lowercase letters indicate significant differences between different enzymes (a, b, c...: *p* < 0.05), different uppercase letters indicate significant differences between different time points for each enzyme (A; B; C ... *p* < 0.05); (**c**) SDS-PAGE of the total proteins (TM) and proteins hydrolisates, obtained with the enzymes alcalase AL, protamex Pr and flavourzyme Fl, from *T. molitor* powder at the end of the enzymatic process; (**d**) SDS-PAGE of the total proteins (TM) and proteins hydrolisates, obtained with the enzymes alcalase AL, protamex Pr and flavourzyme Fl, from *A. domesticus* powder at the end of the enzymatic process; (**e**) DPPH radical scavenging activity of protein hydrolysates of *T. molitor* powder; (**f**) DPPH radical scavenging activity of protein hydrolysates of *A. domesticus* powder. Different lowercase letters indicate significant differences between different enzymes at the same concentration (a, b, c...: *p* < 0.05), different uppercase letters indicate significant differences between different concentrations tested for each enzyme (A; B; C ... *p* < 0.05); values are mean (three replications) ± standard deviations.

#### *3.6. SDS PAGE*

The SDS-PAGE electrophoresis (Figure 4c,d) showed that the electrophoretic pattern of the protein hydrolysates with respect to the total proteins extracted from the whole insect meals (T. m. and A. d. in Figure 4c,d), had a progressive loss of the bands at higher molecular mass, attesting to the efficiency of the reaction. According to the low DH obtained, the electrophoretic profile of Flavourzyme hydrolysates showed protein bands with a higher relative molecular mass than the other two enzymes, comparable to the profile of the total proteins of the whole samples (T. m. and A. d. in Figure 4c,d). In the study done by Kristinsson and Rasco [66], peptides with a lower molecular weight were observed with a higher DH. The enzyme Alcalase gave peptides at 65 kDa, while with Flavourzyme most of the peptides present a molecular mass of 70 kDa. At a higher degree of hydrolysis the bands with greater mass begin to disappear; this confirmed the fact that the molecular weight of the peptides formed by hydrolysis is associated with the degree of hydrolysis.

#### *3.7. DPPH Radical Scavenging Activity*

The DPPH assay showed that the protein hydrolysate of *T. molitor* powder (Figure 4e) had antioxidant properties, attested by inhibition of the DPPH radical up to 14.0% (Figure 4e). In particular, the hydrolysate obtained from the enzyme Protamex showed a higher antioxidant power in respect to the hydrolysate obtained with Alcalase and Flavourzyme only at the highest concentrations tested (5 and 1 mg/mL) (Figure 4e) (*p* < 0.05).

The DPPH assay on *A. domesticus* powder hydrolysate (Figure 4f) showed that the best antioxidant properties, resulting in an inhibition of DPPH radical up to 26.5%, were observed in the hydrolysates obtained using the enzyme Alcalase (*p* < 0.05). In Figure 4f it is evident a linear and dose-dependent relationship between the scavenging properties of the DPPH radical and the various concentrations tested, the linearity has been tested through the linear regression test obtaining the following values of R2: *A. domesticus* Alcalase 0.982 (*p* > 0.001); *A. domesticus* Protamex 0.976 (*p* > 0.001); *A. domesticus* Flavourzyme 0.994 (*p* > 0.001). These results were comparable to previous reports [23,64].

Similar results have already been reported in the literature, confirming that the degree of hydrolysis can significantly influence the antioxidant activity of the resulting hydrolysates, probably due to the presence of a high amount of low molecular weight active peptides [67–69]. Ahn et al., [70] stated that the molecular weight of peptides was related to their functional properties, with greater efficacy in bioactive peptides at a molecular weight of about 1.0–3.0 kDa. Taheri et al., [71] showed that protein hydrolysate fractions between 1.0 and 10 kDa had higher antioxidant power than higher molecular weight fractions. Therefore, since high DH means that more peptide bonds were cleaved, the protein would release lower molecular weight peptides to the hydrolysates, endowing the hydrolysates with high antioxidant activity, indicating that a certain degree of DH is necessary to the physico-chemical activity of hydrolysates. Our future analyses will be directed towards the specific amino-acid analysis of the hydrolysates obtained.

#### **4. Conclusions**

Insect powders have been subjected to a deep microbiological characterization in view of their application in wheat powder fermentations to obtain fortified products. The fortification of traditional food products represents a successful strategy to provide the necessary nutrients without substantial modification of the alimentary habits [72]. Generally, insects were found to be highly nutritious and to represent good sources of protein, fat, minerals, vitamins, and energy [54]. They are traditionally used as a food source in different countries, but nowadays, they are becoming globally increasingly attractive as a protein and fat source for humans and many types of pet and farm animals [73]. Insects are useful not only for their nutritional composition [74] but also for the transfer of other indispensable nutrients and micronutrients to the recipients [75]. Insect protein is readily available with protein quality values similar to, or slightly higher than, fish meat or soybean powder [74]. Moreover, proteins analyzed showed to be a suitable source of biologically active peptides to generate multifunctional hydrolysates that could be incorporated into functional foods or used as nutraceuticals or as natural alternatives to synthetic antioxidants. Further study is also needed in the characterization of edible insect peptides, optimization of functional properties, sensory evaluation, and establishing applications of these hydrolysates in food formulations [20,21].

**Author Contributions:** Conceptualization, C.M.M., N.F. and L.S.; methodology, R.G., M.M. and R.A.; software, R.G. and M.T.; formal analysis, R.G., M.M. and R.A.; investigation, R.G., M.M. and R.A.; data curation, R.G., M.M., M.T. and N.F.; writing—original draft preparation, C.M.M. and L.S.; writing—review and editing, C.M.M., R.G., M.M., M.T. and L.S.; supervision, G.M. and A.S.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors are grateful to Emilio Calligaris and Riccardo Zamponi for their precious collaboration during samples collection.

**Conflicts of Interest:** All authors certify that there is no conflict of interests in this study.
