Degradation of edible mushroom waste by Hermetia illucens L. and consequent adaptation of its gut microbiota

The edible fungus industry is one of the pillar industries in the Yunnan–Guizhou Plateau, China. The expansion of the planting scale has led to the release of various mushroom residues, such as mushroom feet, and other wastes, which are not treated adequately, resulting in environmental pollution. This study investigated the ability of black soldier fly (Hermetia illucens L.) larvae (BSFL) to degrade mushroom waste. Moreover, this study analyzed changes in the intestinal bacterial community and gene expression of BSFL after feeding on mushroom waste. Under identical feeding conditions, the remaining amount of mushroom waste in Pleurotus ostreatus treatment group was reduced by 18.66%, whereas that in Flammulina velutipes treatment group was increased by 31.08%. Regarding gut microbial diversity, compared with wheat bran-treated control group, Dysgonomonas, Providencia, Enterococcus, Pseudochrobactrum, Actinomyces, Morganella, Ochrobactrum, Raoultella, and Ignatzschineria were the most abundant bacteria in the midgut of BSFL in F. velutipes treatment group. Furthermore, Dysgonomonas, Campylobacter, Providencia, Ignatzschineria, Actinomyces, Enterococcus, Morganella, Raoultella, and Pseudochrobactrum were the most abundant bacteria in the midgut of BSFL in P. ostreatus treatment group. Compared with wheat bran-treated control group, 501 upregulated and 285 downregulated genes were identified in F. velutipes treatment group, whereas 211 upregulated and 43 downregulated genes were identified in P. ostreatus treatment group. Using Kyoto Encyclopedia of Genes and Genomes and Gene Ontology enrichment analyses, we identified 14 differentially expressed genes (DEGs) related to amino sugar and nucleotide sugar metabolism in F. velutipes treatment group, followed by 12 DEGs related to protein digestion and absorption. Moreover, in P. ostreatus treatment group, two DEGs were detected for fructose and mannose metabolism, and two were noted for fatty acid metabolism. These results indicate that feeding on edible mushroom waste can alter the intestinal microbial community structure of BSFL; moreover, the larval intestine can generate a corresponding feedback. These changes contribute to the degradation of edible mushroom waste by BSFL and provide a reference for treating edible mushroom waste using BSFL.


Source of the tested insect and mushrooms
The BSFL used in this study were obtained from a colony reared at the Industrial Research Group, College of Biology and Engineering.The BSFL were reared under controlled environmental conditions (T = 28 °C ± 1 °C, relative humidity [RH] = 70% ± 10%) until the 5th-instar larval stage and were used as the insect source in this study.The mushroom waste used in this study were obtained from the canteen of Guizhou Medical University and were selected as the experimental material.

Treatment of mushroom waste with 5th-instar BSFL
The experimental group was divided into nine parallel groups, as follows: three groups each for treatment with F. velutipes, P. ostreatus, and wheat bran (which was used as the control).Each parallel group was weighed to ensure that it contained 105 g of BSFL with a similar body size.The BSFL used weighs about 0.105 g each and has a body length of 12-19 mm.Subsequently, the three parallel groups were incubated with 750 g of F. velutipes mushroom waste, three parallel groups were incubated with 750 g of P. ostreatus mushroom waste, and the remaining three parallel groups were incubated with 750 g of wheat bran.Nine additional parallel groups were set up that contained only equal amounts of mushroom waste and wheat bran, without BSFL, as a control group, and were incubated at a feeding temperature of 25-28 °C and humidity of 70-75%.After 5 days of treatment, the amount of remaining edible mushroom waste and quality of wheat bran were recorded.BSFL was also raised in the same way from the 4th-instar stage, and after 12 days of treatment, the intestine and intestinal contents were separated for total DNA extraction, intestinal bacterial library construction, and sequencing.
We collected 5th-instar BSFL with consistent size and health status, rinsed them with sterile water for 30 s, disinfected the body surface with 75% alcohol, and finally washed away the residual alcohol from the body surface using sterile water.Under aseptic conditions, we sliced open the body surface of the larvae from the side, carefully removed the intact intestinal tract, and cut off the midgut.We removed the intestinal contents and collected them in 2-mL sterile centrifuge tubes.The intestinal contents of 30 BSFL were placed in each collection tube, and 3 tubes were prepared for each treatment group.We used PowerSoil® DNA Isolation kit to extract DNA.After quality control, we used Solexa PCR library preparation kit to construct a gene library and sequenced the gene library using the PacBio Sequel sequencing platform.

Isolation and identification of extracellular enzyme-producing bacteria in the midgut of BSFL
The midgut of BSFL was collected and rinsed with sterile water for 30 s, and phosphate-buffered saline was added for sample grinding.The grinding liquid suspensions was spread on a plate, and the purification process was repeated via streaking until a single colony appeared on the plate.Next, a single colony from LB liquid medium was picked up and incubated at 37 °C until the bacterial solution became cloudy.Subsequently, 10 μL of the bacterial solution was steadily dropped onto a pre-prepared enzyme activity screening plate (the center of the cellulase enzyme screening plate; protease-, urease-, and asparaginase-detection plate), which was placed in a 37 °C incubator for incubation and observation.After sequencing the isolated strains, NCBI sequence alignment was performed, and a phylogenetic tree was constructed for strain identification.

Transcriptome RNA extraction, library construction, and sequencing
Total RNA was extracted from the midgut of each treatment group of BSFL using R6834 Total RNA Kit I. Three replicates were used for each treatment group, with each replicate comprising 15 midgut samples of BSFL.After quality control, VAHTS Universal V6 RNA-seq Library Prep Kit (Illumina®, NR604-01/02) was used to select different index tags for library construction.The qualified libraries were sequenced using the HiSeq sequencing platform.

Data quality control and comparison with the reference genome
Quality control of the filtered data was performed using fastac 26 .The filtered transcriptome sequences were aligned with reference genes according to Star 27 .Finally, statistical analysis was performed.

Differentially expressed genes (DEGs) and their functional enrichment analysis
The expression level of genes was estimated by counting the sequences located in the genomic region or gene exon region, and functional annotation of transcripts was performed using seven databases, including Nr 28 , Pfam 29 , Uniprot, KEGG 30 , Gene Ontology (GO) 31 , KOG/COG 32 , and PATHWAY 30 .The number of reads that were mapped to each transcript in all samples was obtained using RSEM 33 , and FPKM conversion was performed.Double-ended sequences from the same fragment were counted as one fragment, and the expression levels of genes and transcripts were determined.The quantitative expression levels of the genes in each sample were used for differential expression analysis using DESeq2 software 34 .The screening threshold was Padj < 0.05 and |log2FoldChange|> 1, and the main metabolic pathways or signaling pathways associated with the DEGs were determined.

Ethics approval
This article does not contain any studies with human participants performed by any of the authors.

Degradation of edible mushroom waste and wheat bran by 5th-instar BSFL
After treatment with BSFL for 5 days, differences were observed in the consumption of P. ostreatus waste, F. velutipes waste, and wheat bran (Fig. 1).The consumption was compared between BSFL-treated (red) and non-BSFL-treated (blue) groups to determine whether BSFL treatment resulted in statistically significant differences in feed surplus.The remaining amount of BSFL-treated F. velutipes waste was 328.3 g, whereas the consumption in the non-BSFL-treated group was 321.733 g, with no statistically significant differences between the two groups.The remaining amount of P. ostreatus waste in the BSFL-treated group was 203.733 g, whereas that in the non-BSFL-treated group was 366.533 g, with a statistically significant difference between the two groups.As a control, the amount of wheat bran that remained after BSFL treatment was 250.467 g, whereas that in the non-BSFL-treated group was 263.367 g, with no statistically significant differences between the two groups.BSFL exhibited varying consumption rates for different edible mushroom waste and wheat bran, with the consumption of P. ostreatus waste being greater than that of wheat bran and the consumption of P. ostreatus waste after BSFL treatment being significantly increased compared with the untreated group.At the same time, the weight of BSFL in three different treatment groups was statistically analyzed.The average weight of BSFL in the Flammulina velutipes treatment group decreased by 36.4 g, the average weight of BSFL in the wheat bran control group increased by 56.5 g, and the average weight of BSFL in Pleurotus ostreatus treatment group increased by 79.1 g.

Splicing and assembly of the 16srDNA sequence of the microorganisms present in the midgut of BSFL
A total of 66,971 reads were obtained from the midgut samples of the 5th-instar larvae of the black soldier flies.After quality control, the available sequence reads were clustered at a similarity level of 97.0% (Table 1).The annotation results based on OTUs showed that 117 OTUs were obtained from the combination of 9 samples, with 7 phyla, 9 classes, 22 orders, 42 families, 78 genera, and 101 species being annotated.

Sequencing depth and α-diversity analysis of the midgut microbial community of BSFL
The horizontal species accumulation curve (Fig. 2a), rarefaction curve (Fig. 2b), and Shannon diversity index dilution curve (Fig. 2c) tended to be smooth at the end, indicating a sufficient sequencing sample size.The rank abundance curve (Fig. 2d) was wide and flat, indicating the presence of abundant sample species and high uniformity.The α-diversity of BSFL intestinal microorganisms in different treatment groups was analyzed using the Chao index, Shannon index, Ace index, and Simpson index.Figure 2e presents the mean α-diversity index of different treatment groups.As shown in Table 2, the mean Shannon index, Chao index, and Simpson index of BSFL intestinal microorganisms in mushroom treatment group were 4.32, 95.57, and 0.92, respectively.Furthermore, the mean Shannon index, Chao index, and Simpson index of BSFL intestinal microorganisms in wheat bran-treated control group were 3.81, 72.21, and 0.88, respectively.Moreover, in mushroom treatment group, the corresponding mean values were 3.4, 78.33, and 0.83.The species abundance and diversity of BSFL intestinal microorganisms in mushroom treatment group were higher than those in wheat bran-treated control group.www.nature.com/scientificreports/

Analysis of bacterial β-diversity in the midgut of BSFL
The results of PCoA (Fig. 3a) showed that there were nine samples from F. velutipes treatment group, wheat bran-treated control group, and P. ostreatus treatment group, and different treatment samples were clustered in different quadrants without any overlap.UPGMA hierarchical clustering (Fig. 3b) of the samples revealed that the samples within the same treatment group were close to each other and had short branches, whereas those   in different treatment groups were far away from each other and had long branches.Anosim analysis (Fig. 3c) revealed that the R value was 0.942, which was close to 1, and the difference between the groups was greater than the difference within the groups.The P-value was 0.002, which was far below 0.05, with high reliability, indicating significant differences in β-diversity between different treatment groups.Differences between the samples in distinct treatment groups were greater than those observed within the same treatment group.

Comparison of the species and abundance of the microorganisms identified in the midgut of BSFL in the different groups
Figure 4a shows that the top 3 bacterial species in the BSFL of F. velutipes treatment group at the phylum level were Proteobacteria (58.22%),Bacteroidota (22.57%), and Firmicutes (9.68%).The top 3 bacteria in the BSFL of wheat bran-treated control group at the phylum level were Proteobacteria (56.58%),Bacteroidota (22.12%), and Firmicutes (12.43%).The top 3 bacteria in the BSFL of P. ostreatus treatment group at the phylum level were Proteobacteria (32.82%),Bacteroidota (27.89%), and Campylobacterota (20.70%).The top 10 bacteria at the genus level were analyzed (Fig. 4b).The top

Detection of extracellular enzyme activity and identification of bacterial species in the gut of BSFL
Three strains (preliminarily termed 47, 66, and 71) were isolated and purified from the gut of BSFL.After enzyme activity detection on the plate, strain 71 exhibited a clear circle around the cell body on the protease-detection plate (Fig. 5a), indicating the presence of protease activity.Furthermore, strain 47 showed a clear circle around the cell body on the cellulase-detection plate after Congo red staining (Fig. 5b), indicating the presence of cellulase activity.Strains 47, 66, and 71 showed blue areas around the cell body on the urease-detection plate (Fig. 5c-e), indicating that all three strains had urease activity.Strains 47, 66, and 71 showed blue areas around the cell body on the asparaginase-detection plate (Fig. 5f-h), indicating that all three strains had asparaginase activity.Finally, based on sequencing, strains 47, 66, and 71 were identified as Providencia sp.(strain 47) and Proteus mirabilis (strains 66 and 71).www.nature.com/scientificreports/

Statistical analysis and quality assessment of the gut transcriptome sequencing data of BSFL in each treatment group
After the administration of wheat bran to F. velutipes and P. ostreatus waste, the total amount of data obtained from the transcriptome sequencing of the BSFL intestinal tract was 630 303 022 sequences, and the sample data amount of each treatment group was 70 004 922-70 050402 sequences.After data filtering and quality control, 100% of the data was retained as clean reads, and 630 303 022 sequences were obtained for downstream analysis, with 70 004 922-70 050402 sequences identified for each treatment group, among which the F. velutipes treatment group exhibited the highest number of sequences and the wheat bran-treated control group had the lowest number of sequences.The Q20% of each treatment group was above 97.387%, and the Q30 was above 91.990%.The GC content was 39.733-43.939%.The transcriptome sequencing quality was high, and the data met the requirements for bioinformatics analysis.

Analysis of DEGs in the gut of BSFL in each treatment group
Compared with wheat bran-treated control group, 786 DEGs were identified in F. velutipes treatment group, among which 501 genes were upregulated and 285 genes were downregulated (Fig. 6a).Furthermore, 254 DEGs were identified in P. ostreatus treatment group, among which 211 genes were upregulated and 43 genes were downregulated (Fig. 6b).The main DEGs related to nutrient composition and energy metabolism are listed in  Table 3.As shown in the table, the FPKM value of genes related to the intestinal amino sugar and nucleotide sugar metabolism of BSFL in F. velutipes treatment group was 1.57, which was 102.87 times higher than that in wheat bran-treated control group; moreover, the FPKM value of these genes in P. ostreatus treatment group was 64.53, which was 41.10 times higher than that in wheat bran-treated control group.In addition, the FPKM value of genes related to the intestinal methionine and cysteine metabolism of BSFL in wheat bran-treated control group was 0.34, which was 44.24 times higher than that in F. velutipes treatment group.The FPKM value of these genes in P. ostreatus treatment group was 13.88, which was 40.82 times higher than that in wheat bran-treated control group.

GO and KEGG enrichment analyses of DEGs
Compared with wheat bran-treated control group, GO enrichment analysis of F. velutipes treatment group showed that 11 DEGs were enriched in serine-type endopeptidase activity (Fig. 7a).KEGG enrichment analysis revealed that 204 DEGs were enriched in 20 KEGG pathways (Fig. 7b), among which significant enrichment (q < 0.05) was observed in antigen processing and presentation, legionellosis, amino sugar and nucleotide sugar metabolism, and protein digestion and absorption.The category exhibiting the greatest number of enriched DEGs was amino sugar and nucleotide sugar metabolism (14 DEGs), followed by protein digestion and absorption (12 DEGs).GO enrichment analysis of P. ostreatus treatment group showed that one DEG was associated with ATPasecoupled transmembrane transporter activity (Fig. 8a).KEGG enrichment analysis revealed that 41 DEGs were enriched in 20 KEGG pathways (Fig. 8b), such as fructose and mannose metabolism, fatty acid metabolism, galactose metabolism, and glycolysis/gluconeogenesis; however, these differences were not significant.

Discussion
In this study, we selected the two most common edible mushrooms in the market as the research subjects for treatment of edible mushroom waste with BSFL.P. ostreatus, which has a short growth cycle, is an edible mushroom that is relatively easy to digest and absorb 35 .The cultivation of F. velutipes produces a large amount of waste; moreover, it is difficult for mammals to completely digest and absorb the waste because it is rich in dietary fiber.Several studies have attempted to use the waste as an additive in poultry feed 36,37 .In the current study, from the perspective of the remaining amount of feed in different treatment groups, P. ostreatus mushroom waste were the most suitable for BSFL feeding, yielding the smallest remaining amount, which was 0.81 times higher than that in wheat bran-treated control group.Second, the largest remaining amount was observed for F. velutipes mushroom waste, which was 1.31 times higher than that in wheat bran-treated control group.These results demonstrated that the amount of F. velutipes waste remaining after BSFL treatment was greater than the consumption of F. velutipes in the natural environment.Based on the difficulty of digesting and absorbing chitin, which is observed in F. velutipes, it is speculated that the F. velutipes waste are less suitable for BSFL feeding and more difficult to digest among the three types of feed used in this study.This result validates the feasibility of treating edible mushroom foot waste with BSFL.In particular, the amount of P. ostreatus waste remaining after BSFL treatment was considerably lower than that remaining after wheat bran control treatment, demonstrating that BSFL can degrade edible mushroom waste.www.nature.com/scientificreports/ The relative abundance of intestinal bacterial species in BSFL from different treatment groups varied.Compared with wheat bran-treated control group, the abundance of the cellulolytic bacterial genus Providencia 38 in F. velutipes and P. ostreatus treatment groups increased by 16.68% and 12.36%, respectively, which were 6.05 times and 4.75 times higher than that of wheat bran-treated control group, with significant differences (P < 0.05).F. velutipes has a high dietary fiber content and is difficult to digest and absorb.F. velutipes and P. ostreatus both contain adequate amounts of high-quality protein and are good sources of dietary fiber, vitamin C, B vitamins and minerals 39 .The wheat bran as the control group contained about 70% carbohydrate, 33% pentosan, 8% protein, 2% fiber, and 8.3% fat 40 .The highest crude fat content of BSFL was 28.4%, and the lowest crude protein content was 38% 41 .To meet their growth requirements, BSFL change the abundance of Providencia individuals in their gut, to enhance the decomposition and use of dietary fiber.Compared with wheat bran-treated control group (1.42%), the abundance of Pseudochrobactrum 42 in F. velutipes treatment group (3.07%) and P. ostreatus  treatment group (4.33%) increased by 1.65% and 2.91%, respectively, which were 2.16 and 3.05 times higher than that of the wheat bran-treated control group, with significant differences (P < 0.05).Morganella is a conditional pathogenic bacterium of the gut, and F. velutipes can exert immune regulatory effects by stimulating immune responses, producing cytokines, and exerting antibacterial, antiviral, and antifungal effects 43,44 , resulting in lower levels of this bacterium in the gut of the BSFL-treated groups.Studies have also shown that changes in insect diet can affect their immune system while affecting their growth 45,46 .After feeding BSFL with agricultural waste and liquid waste, it was observed that their antibacterial peptides were significantly increased 47 .Subsequently, strains with cellulase activity, protease activity, urease activity, and asparaginase activity were isolated from the gut of BSFL, among which bacteria in the Providencia genus were identified as strains with significantly increased abundance after feeding the larvae with edible mushroom waste, thus further verifying the degradation ability of BSFL on edible mushroom waste.In a subsequent, unpublished study, we successfully screened proteaseproducing strains from the BSFL intestine and demonstrated that they can enhance BSFL intestinal protease activity.In summary, the gut flora of BSFL can regulate the abundance of different bacterial species based on the nutrient composition of the edible mushroom waste.With the assistance of the gut flora, BSFL can better degrade edible mushroom waste and convert them into their own biomass, thus reducing environmental waste while becoming a more nutritious feed source.
The FPKM value of genes related to the intestinal amino sugar and nucleotide sugar metabolism of BSFL in F. velutipes treatment group was 161.50, which was 102.87 times higher than that in wheat bran-treated control group (1.57).Furthermore, the FPKM value of these genes in P. ostreatus treatment group was 64.53, which was 41.10 times higher than that in wheat bran-treated control group.In addition, the FPKM value of genes related to the intestinal methionine and cysteine metabolism of BSFL in F. velutipes treatment group was 15.04, which was 44.24 times higher than that in wheat bran-treated control group (0.34).The FPKM value of these genes in P. ostreatus treatment group was 13.88, which was 40.82 times higher than that in wheat bran-treated control group.Feeding with mushroom waste upregulated the expression of intestinal energy metabolism-related genes of BSFL, thus enhancing the absorption and metabolism of energy substances.From the perspective of metabolic pathways, KEGG analysis revealed that compared with wheat bran-treated control group, the intestinal DEGs of BSFL in F. velutipes treatment group exhibited significant enrichment in energy-related metabolic pathways, such as amino sugar and nucleotide sugar metabolism and protein digestion and absorption (P < 0.02).Taken together with the changes in the intestinal microbial community of BSFL, these findings indicated that because of the difficulty in digestion and utilization of dietary fiber and polysaccharides present in mushroom waste by the black soldier fly, BSFL are forced to upregulate the expression of energy metabolism-related genes.This increases the abundance of bacteria with intestinal digestive enzyme activity, thereby enhancing the absorption and utilization of nutrients.
In summary, this study used BSFL as a treatment medium to organically degrade mushroom waste, which are a waste product of the edible mushroom industry.Mushroom waste showed a certain impact on the abundance of different types of bacteria in the gut of BSFL and expression of gut genes, ultimately increasing the consumption of mushroom waste by BSFL.In conclusion, BSFL can be used to treat organic mushroom waste, thus providing a new alternative for treating edible mushroom waste and protecting the highland environment.

Relevant legislations, permitting and consent
All experimental procedures were performed in accordance with the standards of the Animal Management Committee of Guizhou Medical University, China.

Figure 2 .
Figure 2. (a) Species relative abundance accumulation curve of BSFL gut microbes in each treatment group.(b) BSFL gut microbial dilution curve of each treatment group.(c) BSFL gut microbial Shannon index curve of each treatment group.(d) Grade abundance curves of BSFL intestinal microorganisms in different treatment groups.(e) BSFL gut microbial α-diversity of each treatment group, as estimated using Chao1, Shannon, ACE, and Simpson indexes.

Figure 3 .
Figure 3. (a) PCoA of BSFL gut microbes in each treatment group based on the weighted UniFrac distance matrix.(b) UPGMA clustering tree of BSFL gut microbes in each treatment group.(c) Anosim analysis box of BSFL gut microbes in each treatment group.As the R value became closer to 1, the difference between the groups became greater than the difference within groups, and P-values of < 0.05 indicated high reliability of the test.The ordinate represents the β distance.The blue box diagram represents the β distance data of all intergroup samples, the orange box diagram represents the β distance data of all intragroup samples, and the subsequent box diagram represents the β distance data of intragroup samples of different groups.

Figure 4 .
Figure 4. Relative abundance of the dominant phyla (a) and genera (b) of gut bacteria in the 5th-instar larvae of Hermetia illucens L.

Figure 5 .
Figure 5. Detection of the extracellular enzyme activity of intestinal bacteria present in the 5th-intar larvae of Hermetia illucens L.

Figure 6 .
Figure 6.Volcano plot of the gut in the 5th-instar larvae of Hermetia illucens L. based on different Flammulina velutipes (a) and Pleurotus ostreatus (b) treatment groups.

Figure 7 .
Figure 7. GO (a) and KEGG (b) enrichment analyses of differently expressed genes between Flammulina velutipes treatment group and control group.

Figure 8 .
Figure 8. GO (a) and KEGG (b) enrichment analyses of differently expressed genes between Pleurotus ostreatus treatment group and control group.

Table 2 .
α-diversity indexes of bacteria in the midgut of the 5th-instar larvae of Hermetia illucens L.

Table 3 .
Main differentially expressed genes in various treatment groups.