Fitness effects of synthetic and natural diet preservatives on the edible insect *Bombyx mori*

The study addresses the challenge of microbial spoilage in high-moisture, nutrient-rich artificial diets used to rear Bombyx mori year-round for silk and nutritionally valuable pupae. Limited mulberry leaves and labor shortages drive the need for artificial diets, but non-sterile rearing conditions lead to spoilage and adverse impacts on silkworm rearing. Preservatives such as ethylparaben (EP), a widely used synthetic antimicrobial, and medium-chain fatty acids (MCFA), natural antimicrobials abundant in milk fat and certain plant oils, are candidate additives to suppress microbial growth. The research question is whether EP and/or MCFA can preserve artificial diet by inhibiting pathogens without compromising silkworm fitness or gut microbiota, and which option is safer for silkworm health. The study evaluates antibacterial efficacy in vitro and in diet, effects on larval growth and pupation, gut microbiota composition, and host transcriptomic responses, with the hypothesis that MCFA will be an effective and safer preservative than EP.

Background highlights: Silkworm pupae are protein- and micronutrient-rich and used as food and feed across Asia; replacing conventional feed with pupae supports comparable animal growth. Artificial diets enable continuous rearing but are susceptible to microbial contamination (bacteria, molds). Preservatives extend shelf-life by inhibiting microbial growth. EP (a paraben) acts by disrupting mitochondrial function and oxidative phosphorylation and is widely used in foods, cosmetics, and drugs; concerns exist about endocrine modulation and potential health risks of parabens. Natural alternatives with lower toxicity, including bacteriocins and organic acids, show antimicrobial activity. MCFA (C6–C12), present in milk fat and oils (coconut, palm, cuphea), exhibit antimicrobial properties via membrane destabilization, inhibition of bacterial lipases, and activation of autolytic enzymes. Prior in vitro work demonstrates MCFA activity against bacteria, viruses, and parasites. This context motivates comparing EP and MCFA as preservatives for silkworm artificial diets.

Antibacterial assays: Three bacteria were tested: Serratia marcescens (Sm; Gram-negative entomopathogen), Bacillus thuringiensis (Bt; Gram-positive entomopathogen), and Lactobacillus plantarum (Lp; food spoilage bacterium in silkworm artificial diet). Bacteria were cultured on LB agar 12–24 h at 37 °C; single colonies identified by 16S rRNA gene, then grown in LB broth. EP and MCFA were sourced from commercial suppliers. Preservatives were incorporated into LB agar at 0.05%, 0.1%, and 0.15%; plates without preservative served as controls (CK). Inoculated plates were incubated at 37 °C for 24 h to assess antibacterial efficacy. A concentration of 0.1% exhibited the most effective activity and was used for subsequent experiments.

Antimicrobial efficacy in artificial diet: Artificial diet composition: 40% mulberry leaf powder, 30% soybean meal, 25% corn starch, 2.5% compound vitamins, 2% inorganic salts; hydrated 1:1.6 (diet:water). Diets were assigned to three groups: CK (sterile water), 0.1% EP, or 0.1% MCFA, then sterilized at 105 °C for 40 min. Diets were inoculated with Sm, Bt, or Lp. At 0, 12, and 48 h post-inoculation, equal volumes were plated on LB agar and incubated at 37 °C for 24 h; colony-forming units (CFU) were counted.

Silkworm rearing and trait measurements: Fifth-instar silkworms (strain Jingsong Haoyue) were randomized into cohorts of 30 individuals (n=30 per group) and fed ad libitum on CK, EP, or MCFA diets at 25 °C and 60% RH; diet refreshed every 3 days. Body weight was recorded daily; pupal weight recorded.

Gut bacterial load and microbiota profiling: For CFU load, four silkworms (n=4) per group were dissected on days 1, 3, and 5 of the fifth instar; gut contents were plated on LB agar, incubated 24 h at 37 °C, and CFU counted. For microbiota composition, gut contents from individual insects (n=6 per group) on day 5 were processed. DNA was extracted from 20 mg homogenate (MasterPure kit). 16S rRNA gene amplicons targeted the V4 region using primers 515F/806R (after initial quality check with 27F/1492R). Sequencing was performed on Illumina MiSeq. Denoising used QIIME2 q2-dada2; taxonomy assigned with a Naïve Bayes classifier against SILVA v138. Singletons and chloroplast/mitochondrial ASVs were removed. Alpha/beta diversity metrics were computed in R 4.3.1.

Transcriptomics and functional analyses: Midgut transcriptomes were sequenced; after filtering, 40–54 million clean reads per sample (90.39–97.42%) were obtained across CK, EP, MCFA. Unigenes were annotated using GO, KEGG, COG, NR, Swiss-Prot, and Pfam; >14,000 functional genes identified. Differentially expressed genes (DEGs) were identified between treatments and CK. GO and KEGG enrichment analyses were conducted to interpret DEGs. Gene set enrichment analysis (GSEA) evaluated coordinated pathway shifts beyond DEGs. Protein–protein interaction (PPI) networks were built with STRING; hub genes identified via CytoHubba.

qRT-PCR validation: Nine genes spanning antioxidant activity (Bmcat, Bmsod, GST01), digestion (Bmlipase, Alpha amylase, Trypsin-like protease), and antimicrobial peptides (Cecropin B, Lysozyme, Attacin) were validated by qRT-PCR. Correlations between RNA-seq and qRT-PCR log2FC were assessed via linear fitting (R²) and Pearson correlation.

Statistics: For growth and CFU comparisons, one-way ANOVA with Tukey’s post hoc test was applied; Wilcoxon rank-sum for alpha diversity; NMDS (Bray-Curtis) with ANOSIM for beta diversity. Significance thresholds reported as p<0.05, p<0.01, p<0.001 where applicable.

• Antimicrobial efficacy: Both 0.1% EP and 0.1% MCFA significantly inhibited Sm, Bt, and Lp growth on agar and within the artificial diet compared to CK. In diet, significant inhibition was evident at 12 and 48 h; for Sm, CK CFU were at least 10-fold higher than EP or MCFA groups. Similar trends were observed for Bt and Lp.
• Silkworm performance: Larval body weight increased similarly across groups; EP-fed larvae were significantly heavier on day 5 versus CK (p<0.05), while MCFA matched CK. Pupal weights did not differ among CK, EP, and MCFA, indicating no adverse effects on economic traits.
• Gut bacterial load and microbiota: CFU loads did not differ on day 1; on day 3, EP and MCFA groups showed higher CFU than CK, suggesting transient microbiota perturbation; by day 5, differences were not significant, indicating restored homeostasis. 16S profiling (52 ASVs) showed Enterococcus dominated in all groups (relative abundance: CK 95.96%, EP 97.74%, MCFA 95.84%). Alpha diversity (Shannon, Chao1, Pielou-e, Simpson) showed no significant differences (p>0.05). NMDS (Bray-Curtis) showed overlapping communities (stress≈0.01; ANOSIM p>0.05).
• Transcriptomic responses: EP elicited more DEGs (up 127, down 113) than MCFA (up 12, down 24). 8815 genes (91.34%) were shared across CK and treatments; uniquely expressed genes: CK 149, EP 109, MCFA 108. Five DEGs were common to EP and MCFA; one (nose resistance to fluoxetine protein 6) linked to homeostasis was downregulated.
• GO/KEGG enrichment under EP: Significant enrichment of metabolism of xenobiotics by cytochrome P450, drug metabolism-cytochrome P450, Toll and IMD signaling, steroid hormone biosynthesis, porphyrin metabolism, ascorbate/aldarate metabolism, pentose and glucuronate interconversions, chemical carcinogenesis-DNA adducts, retinol metabolism, fat digestion/absorption. Detoxification genes, notably Ugt2 (UGTs pathway), were upregulated. Immune effector gene Cecropin B (Cec B1) was upregulated in Toll/IMD pathways, indicating enhanced defense responses. GO terms with upregulation included glycosyltransferase activity, UDP-glycosyltransferase activity, serine-type peptidase activity, toxin activity, and response to other organisms.
• KEGG enrichment under MCFA: Enriched pathways included protein digestion and absorption, pancreatic secretion, cholesterol metabolism, valine/leucine/isoleucine biosynthesis, and lysosome. Digestive enzyme genes chymotrypsins (CTRB1), pancreatic elastase (CELA), and carboxypeptidase B (CPB) were upregulated, suggesting improved protein digestion and absorption.
• GSEA: EP showed significant upregulation of ribosome (NES 3.83; FDR 0.00; p=0.00), ribosome biogenesis in eukaryotes (NES 1.39; FDR 0.11; p=0.00), oxidative phosphorylation (NES 1.50; FDR 0.17; p=0.00), and retinol metabolism (NES 1.47; FDR 0.13; p=0.00). MCFA showed upregulation of ECM-receptor interaction (NES 1.69; FDR 0.09; p=0.01) and proteasome (NES 1.60; FDR 0.10; p=0.02).
• PPI network: In EP-fed silkworms, 18 DEGs formed a network (17 edges) mainly in apoptosis and UDP-glucuronosyl/glucosyl transferase metabolism; hub genes included EcR, FIBH, FIBL, P25, and Ppbp2. No interacting proteins were found in MCFA-fed silkworms.
• qRT-PCR validation: Expression patterns for nine genes agreed with RNA-seq. Correlations: EP R²=0.448, Pearson r=0.669; MCFA R²=0.616, Pearson r=0.785 (p<0.05). Antioxidant-related genes were generally downregulated, suggesting reduced antioxidant capacity; digestive enzyme genes were upregulated in MCFA, supporting improved digestion.

The study demonstrates that both EP and MCFA at 0.1% effectively suppress growth of common silkworm pathogens in vitro and within artificial diets, maintaining feed hygiene without reducing larval growth or pupal weight. Despite transient increases in gut bacterial load on day 3 for preservative groups, gut microbiome diversity and composition remained stable by day 5, with Enterococcus dominance across treatments, indicating resilience and restoration of homeostasis. Transcriptomic analyses reveal divergent host responses: EP triggers broad xenobiotic and drug metabolism pathways (notably cytochrome P450 and UGTs), with upregulation of detoxification gene Ugt2 and immune effector Cecropin B via Toll/IMD pathways. GSEA indicates increased ribosomal and oxidative phosphorylation activity under EP, consistent with elevated metabolic and stress responses that can be deleterious for longevity and may induce oxidative stress. These molecular signatures align with concerns about paraben-associated endocrine and cellular effects, suggesting EP imposes a measurable physiological burden despite maintaining performance metrics. By contrast, MCFA elicits fewer DEGs and enriches pathways linked to digestion (protein digestion/absorption, pancreatic secretion) and cellular maintenance (ECM-receptor interaction, proteasome). Upregulation of CTRB1, CELA, and CPB implies improved protein digestion, aligning with the nutritional role of fatty acids and the biocompatibility of MCFA as natural feed components. The absence of an observable PPI network under MCFA underscores a milder host response compared to EP. Collectively, these findings address the central question: both preservatives function effectively, but MCFA provides comparable antimicrobial benefits with fewer adverse host transcriptomic effects, making it a safer and more suitable preservative for silkworm artificial diets. This has practical relevance for scalable, safe production of edible insect pupae and for designing preservatives that balance antimicrobial efficacy with host health.

Adding EP or MCFA to silkworm artificial diets effectively inhibits spoilage/pathogenic bacteria without detrimental effects on larval growth or pupal weight and preserves gut microbiota homeostasis. However, EP induces pronounced detoxification and immune responses (e.g., Ugt2 and Cecropin B upregulation) and broader metabolic pathway activation, indicating potential physiological stress. MCFA, a natural preservative, achieves antimicrobial control with minimal host perturbation and may enhance digestion via upregulation of digestive enzymes. From a safety perspective, MCFA are recommended as preservatives for silkworm artificial diets. Future work should optimize MCFA formulations and dosages, assess long-term and developmental outcomes across silkworm strains and life stages, evaluate residue profiles in pupae, and test generalizability to other edible/resource insects.

• The study primarily used a single effective preservative concentration (0.1%) for most in-diet experiments; dose–response effects on host physiology and microbiota were not extensively profiled.
• Only three bacterial species were tested; broader microbial communities (including fungi/molds) relevant to diet spoilage were not comprehensively evaluated.
• Assessments focused on short-term larval performance and midgut transcriptomics; long-term health, reproductive fitness, lifespan, and potential systemic toxicological endpoints were not measured.
• One silkworm strain and controlled laboratory conditions were used; results may not fully generalize to other strains or production environments.
• Chemical analyses of preservative residues in diet or pupal tissues and potential impacts on consumer safety were not reported.