Multidimensional Analysis of Silkworm Cocoons Produced with Different Feeding Diets

Simple Summary

The sericulture industry traditionally relies on mulberry leaves to produce high-quality silk, but the increasing adoption of artificial diets to improve rearing efficiency can alter the intrinsic quality of cocoons. Differentiating between mulberry leaf-fed and artificial diet-fed cocoons remains a practical challenge in the sericultural market. In this study, we developed an authentication method that accurately discriminates between these two cocoon types based on their microstructural and chemical fingerprints. Mulberry leaf-fed cocoons retained greater structural integrity and contained diet-derived flavonoids, while artificial diet-fed cocoons exhibited distinct biomarkers associated with artificial feed formulations. Moreover, mulberry leaf-fed cocoons extract demonstrated superior bioactive properties, including enhanced antioxidant and anti-inflammatory activities. This work provides a reliable tool for quality verification and fraud prevention in the silk industry, ensures product authenticity for consumers, and supports the potential application of quality-assured silk in biomedical and nutraceutical applications.

Abstract

Artificial diets are increasingly utilized in sericulture, yet they often yield cocoons with properties that differ from those produced by mulberry leaf-fed silkworms. A reliable method to distinguish between these two types of cocoons remains lacking, compromising quality control in silk-related industries. In this study, we report a multidimensional analysis method integrating chemical, structural, and biomarker analyses to distinguish cocoons produced by mulberry leaf-fed (Mul-fed) silkworms from those produced by artificial diet-fed (Art-fed) silkworms. The SEM images showed that after the process of biomarker extraction, Mul-fed cocoons had a more complete morphological structure than Art-fed cocoons, and the sericin layer of Mul-fed cocoons was less damaged. Thermogravimetric and amino acid analyses revealed no significant differences between the two types of cocoons. Biomarker analyses via ultra-performance liquid chromatography–mass spectrometry (UPLC-MS) revealed that quercitrin and quercetin were enriched in Mul-fed cocoons, while daidzein and genistein were enriched in cocoons produced by artificial diet-fed silkworms. Furthermore, materials extracted from Mul-fed cocoons demonstrated significantly superior bioactivity than those from Art-fed cocoons in in vitro assays. This study provides a reliable and accurate method for assessing cocoon quality and distinguishing cocoons from different feeding methods, laying a robust basis for quality evaluation and silk product development.

1. Introduction

The silkworm (Bombyx mori) is an oligophagous insect that primarily feeds on mulberry leaves and produces cocoons via its natural spinning process [1]. Proteins extracted from these cocoons are currently widely exploited in the development of biomedical, biomaterials, and functional foods [2,3,4,5,6,7,8]. However, the adoption of artificial diets in modern sericulture has resulted in significant variations in cocoon quality [9,10]. As a result, distinguishing cocoons produced by mulberry leaf-fed (Mul-fed) silkworms from those produced by artificial diet-fed (Art-fed) silkworms has become critical for the advancement and practical application of silk fibers. Recent studies have demonstrated that the chemical composition of silk cocoons, particularly secondary metabolites such as flavonoids and protein profiles, can be significantly influenced by the silkworms’ diet and rearing environment [11,12]. Furthermore, Art-fed silkworms typically display slower growth and reduced silk protein synthesis efficiency. Metabolomic analysis of larval frass has revealed a marked decrease in amino acids, carbohydrates, and lipids, alongside elevated levels of urea and citric acid, suggesting that artificial diets disrupt the tricarboxylic acid (TCA) cycle in silkworms [13].

Traditional methods for evaluating cocoon quality primarily rely on morphological indices, physical parameters (e.g., cocoon layer thickness and compactness), and subjective manual sensory assessment [14,15]. However, these approaches are limited by operator-dependent variability, high subjectivity, and inadequate accuracy. More importantly, they fail to capture the intrinsic chemical composition and structural differences among cocoons produced under different feeding modes, rendering them insufficient to meet the increasingly stringent requirements of modern traceability and quality control. Therefore, there is an urgent need for a more reliable and precise analytical method capable of qualitatively and quantitatively characterizing cocoons at the molecular level, thereby facilitating accurate assessment of cocoon quality.

Current consensus indicates that Mul-fed silkworms produce cocoons of superior quality compared to those reared on artificial diets. A comparative study of three rearing methods (exclusive artificial diet, exclusive mulberry leaves, and diet-to-mulberry transition) confirmed that mulberry feeding preserves key silk properties, particularly tensile strength and fiber fineness [16]. Furthermore, studies have suggested that silk from Art-fed silkworms may exhibit weaker sericin–fibroin adhesion, resulting in more facile sericin detachment from the core fibroin fibers [11,17]. Statistical analysis of cocoon economic traits revealed that while Art-fed silkworms produced significantly higher whole cocoon yield than Mul-fed ones, their cocoon shell weight and cocoon shell rate were markedly lower. Further metabolome profiling has identified distinct amino acid metabolism pathways between feeding methods, with biomarker screening identifying six compounds associated with cocoon traits: homocitrulline, glycitein, valyl-threonine, propyl gallate, 3-amino-2,3-dihydrobenzoic acid, and 3-dimethylallyl-4-hydroxyphenylpyruvate [18]. Despite these differences, reliably distinguishing cocoons in the market, where sources and quality are often mixed, remains challenging. Assessment of silk mechanical properties and cocoon economic traits often fluctuates due to genetic and regional variations and requires large sample sizes. Additionally, no significant structural differences have been detected in silk proteins between Art-fed and Mul-fed cocoons [19]. Variations in silkworm diets substantially impact the nutritional profile of pupae and frass. Specifically, Art-fed silkworms produce pupae and frass with higher protein content and lower cellulose levels compared to Mul-fed silkworms [13]. Incorporating mulberry and dandelion into the diet notably enhances total amino acid and linolenic acid content in pupae [20]. Biomarker monitoring has clarified the metabolic changes driven by dietary conditions, raising the prospect of using biomarker profiles to distinguish Mul-fed from Art-fed cocoons.

Liquid chromatography–mass spectrometry (LC-MS) offers high separation efficiency and sensitive qualitative capabilities, enabling precise determination of molecular weights and detailed structural information for analytes [21,22,23]. High-performance liquid chromatography (HPLC) has been previously employed to detect biomarkers in silkworm larval powders [20]. In this study, ultra-performance LC-MS (UPLC-MS) was utilized for refined biomarker detection in cocoons produced by Art-fed and Mul-fed silkworms, with the aim of identifying molecular biomarkers and establishing distinctive chemical fingerprints for each feeding method. This approach provides a scientific and objective basis for rapid discrimination and standardized molecular-level quality evaluation of cocoons.

To accurately distinguish between Mul-fed and Art-fed cocoons, characteristic marker compounds derived from their respective dietary sources were investigated. Mulberry leaves contain various bioactive constituents [24,25]. Among them, flavonoids such as quercitrin (C₂₁H₂₀O₁₁), isoquercitrin (C₂₁H₂₀O₁₂), and quercetin (C₁₅H₁₀O₇) are typical components of mulberry leaves [12], whereas their levels in artificial diets are extremely low or absent, providing compelling chemical evidence for mulberry leaf feeding. In contrast, artificial diets predominantly composed of soybeans and corn introduce distinct metabolic markers, including isoflavones derived from soybeans, such as genistein (C₁₅H₁₀O₅) and daidzein (C₁₅H₁₀O₄) [26,27], alongside carotenoids from corn, such as lutein (C₄₀H₅₆O₂) and zeaxanthin (C₄₀H₅₆O₂) [28]. In this study, a multidimensional analysis method integrating chemical, structural, and biomarker analyses was developed to distinguish Mul-fed and Art-fed cocoons.

2. Materials and Methods

2.1. Materials

Both the Art-fed and Mul-fed cocoons were provided by Cathaya (Hangzhou, Zhejiang, China). Artificial diet rearing typically requires specialized silkworm strains with high adaptability to non-mulberry feed, whereas traditional sericulture uses strains optimized for mulberry leaves. The Art-fed cocoons were produced by Zhong 2016 × Ri 2016 line, while the Mul-fed cocoons were produced by 781 line. These 781 lines were reared exclusively on fresh mulberry leaves (Nongsang-12) at 25 to 28 ° C and 70–85% humidity, and cocoons were harvested in October 2024 from a sericulture base in Banqiao Town, Baoshan City, Yunnan Province, China. The Zhong 2016 × Ri 2016 lines were reared on a commercial artificial diet provided by Shengzhou Mulsun Biotech Co., Ltd. (Shaoxing, Zhejiang, China), with a 12 h light and 12 h dark cycle at 25 to 28 ° C. The diet is primarily composed of soybean meal, corn flour, mulberry leaf powder, vitamin complex, and inorganic salts [29]. All samples were dried and stored at room temperature prior to analysis. Standards of quercetin, quercitrin, and isoquercitrin were purchased from Beyotime Biotechnology Co., Ltd. (Shanghai, China). Daidzein and genistein were purchased from Yuanye Biotechnology Co., Ltd. (Shanghai, China). Ethanol and acetonitrile (ACN, HPLC-grade) were purchased from Hushi Reagent Co., Ltd. (Shanghai, China). Enzyme-linked immunosorbent assay (ELISA) kits, superoxide dismutase (SOD) assay kits, and reactive oxygen species (ROS) detection kits were all purchased from Beyotime Biotechnology Co., Ltd.

2.2. Assessment of Degumming Kinetics

Silkworm cocoons were degummed as described previously [30]. Briefly, the weight of the cocoons before degumming was accurately weighed and recorded. The cocoons were then degummed in a boiling 0.02 M sodium carbonate (Na₂CO₃) solution at a 1:400 (w/v) ratio. After 1 min of degumming, the resulting degummed silk was rinsed with purified water at least three times, subsequently dried in a 60 °C oven, and its weight was recorded. Following this, the degummed silk was subjected to an additional 9 min of degumming, rinsed at least three times, dried, and the final weight was recorded, at which point it was assumed that all sericin had been completely removed. Based on the weights recorded after the two degumming steps, the degumming rate of the cocoon during the initial 1 min was calculated to compare the different cocoons.

2.3. Extraction of Biomarkers from Cocoons

Biomarkers in the cocoons were extracted according to previous reports with some modifications [20,31]. Briefly, precisely 0.5 g of dried silkworm cocoon sample was weighed, cut into small pieces, and subjected to ultrasonic-assisted extraction for 30 min using 75% ethanol aqueous solution at a ratio of 1:10 (w/v), followed by heating extraction for 4 h. Then incubate overnight, the supernatant was collected and filtered through a 0.22 μm membrane filter to obtain extracts for subsequent analysis.

2.4. Ultra-High-Performance Liquid Chromatography–Mass Spectrometry

Biomarker analysis of the above extracts was performed using a previously described method with slight modifications [20]. Briefly, the chemical components of the cocoon extracts were systematically analyzed using an ultra-high-performance liquid chromatography–mass spectrometry system (UPLC-MS; Waters H-Class/SQD2, Milford, MA, USA). The detection was performed using a UPLC coupled with a Photo-Diode Array (PDA) detector (Milford, MA, USA) and a Single Quadrupole Detector (SQD2) (Milford, MA, USA). Chromatographic separation was achieved on a C18 column (1.7 μm, 2.1 × 50 mm³). The mobile phase system consisted of Phase A (Water with 0.1% Formic Acid (FA)) and Phase D (Acetonitrile, ACN). A gradient elution program was employed at a flow rate of 0.400 mL/min: initial conditions were 80.0% A; held at 80.0% A until 1.00 min; then Phase A was decreased to 0.0% and Phase D increased to 100% and held until 10.00 min; at 10.10 min, Phase A was returned to 80.0% and Phase D decreased to 20.0%, and this composition was held until the end of the run at 12.00 min. The UPLC detection wavelength was set at 254 nm, with a scanning range of 200–600 nm. Mass Spectrometry utilized an Electrospray Ionization (ESI) source, operated in negative ion mode (ES⁻), with a scanning range of 50–1000 Da.

2.5. Principal Component Analysis

Principal Component Analysis (PCA) was performed for dimensionality reduction and differentiation analysis using the quantitative content data of the five specific biomarkers across Mul-fed and Art-fed cocoon samples [5,32,33,34]. First, the contents of the five biomarkers were set as raw variables, and the data were subjected to centering and standardization prior to PCA to eliminate the influence of dimensional differences [18,35]. Second, eigenvalues and eigenvectors of the covariance matrix were calculated to extract the principal components (PCs) that explained most of the data variance. Subsequently, a two-dimensional PC score plot was constructed with the extracted PCs (PC1 and PC2) as axes, with the two sample groups clearly distinguished, and 95% confidence ellipses or convex hulls used to assess the significance of group separation. Finally, the contribution of each original biomarker to the principal components was analyzed via Loading Vectors to identify the differential compounds critical for distinguishing between Mul- and Art-fed cocoons.

2.6. Attenuated Total Reflectance Fourier Transform Infrared Spectroscopy Analysis

Samples of cocoon shells from Mul-fed and Art-fed silkworms, both before and after biomarker extraction, were collected separately for attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopy analysis. The silkworm cocoon was placed directly on the ATR crystal surface with appropriate pressure applied to ensure good optical contact. The scanning range was set from 400 to 4000 cm⁻¹, with a resolution of 4.0 cm⁻¹ and 32 accumulated scans, to characterize the secondary structure of silk proteins in the two cocoon groups. The peaks at 1620 and 1698 cm⁻¹ were assigned to the β-sheet structure, while the Gaussian peaks at 1645 and 1685 cm⁻¹ were assigned to the random coil/helix and β-turn structures, respectively [2].

2.7. Scanning Electron Microscopy Analysis

Small pieces of silk fiber samples from Mul-fed and Art-fed cocoons were cut and fixed onto an aluminum stub. To enhance conductivity and obtain clear images, the samples were coated with a thin layer of gold film under vacuum. A Scanning Electron Microscope (SEM, Zeiss Gemini 450, Carl Zeiss Microscopy GmbH, Jena, Germany) was subsequently used to observe and photograph the surface morphology, diameter, and microstructure of the silk filament under an accelerating voltage of 5 kV.

2.8. Thermal Analysis

Thermal property analysis was conducted using a simultaneous thermogravimetric Analysis/differential scanning calorimetry analyzer (TGA/DSC 3+/1600 HT, Mettler Toledo, Greifensee, Switzerland). Approximately 5 mg of the cocoon sample was accurately weighed and placed into an alumina crucible. The test was performed under a nitrogen atmosphere with a gas flow rate of 50 mL/min. The heating rate was 10 °C/min, and the temperature range was scanned from room temperature to 400 °C to determine the thermal decomposition temperatures and assess the thermal stability differences among the protein components of the two cocoon groups.

2.9. Amino Acid Analysis

10 mg of the cocoon sample was weighed into a hydrolysis tube, and 6 M hydrochloric acid (HCl) solution was added at a 1:10 (w/v) ratio for complete hydrolysis under heating at 110 °C for 24 h. Following hydrolysis, 1 mL of the hydrolysate was dried under a nitrogen stream, dissolved in the sample dissolution solution, diluted 100 times, and filtered through a 0.22 μm membrane filter. The filtrate was analyzed using an automated amino acid analyzer (MembraPure A388, Bodenheim, Germany) to quantitatively determine the composition and content of amino acids in the Mul-fed and Art-fed cocoons.

2.10. Cytotoxicity Assessment

Raw cocoons from both Mul-fed and Art-fed silkworms were collected separately and subjected to extraction using a saturated lithium bromide (LiBr) solution at a ratio of 1:5 (w/v). Subsequently, the mixture was heated in an oven at 60 °C for 4 h. The resulting solution was then centrifuged at 12,000 g for 30 min, followed by dialysis for 4 days. Finally, the solution was lyophilized to obtain silkworm cocoon extracts. The RAW 264.7 cell line was used for culture. Cells were seeded into a 96-well plate at an appropriate density and cultured for 24 h to allow for adherence. After removing the old medium, fresh medium containing a gradient of cocoon extract concentrations (0.1–1000 μg/mL) was added, and the cells were incubated for 24 h. Subsequently, 10 μL of CCK-8 reagent was added to each well and incubated in the dark at 37 °C for 2 h. Finally, the absorbance was measured at 450 nm using a microplate reader, and the relative cell viability was calculated to assess the cytotoxicity of the cocoon extracts.

2.11. Assessment of Cellular Anti-Inflammatory Activity

Immune cell lines, specifically RAW 264.7 macrophages, were selected to establish the inflammation model. Cells were seeded into a 48-well plate and cultured for 24 h. Subsequently, the medium was removed, and fresh medium containing various concentrations of cocoon extracts was added for a two-hour pre-incubation. Following pre-incubation, the cells were stimulated with 100–1000 ng/mL Lipopolysaccharide (LPS) for 24 h to construct the inflammation model. The cell supernatant was collected, and ELISA kits were used to quantitatively detect the secretion levels of pro-inflammatory cytokines, including tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), interleukin-1β (IL-1β), and transforming growth factor-β (TGF-β).

2.12. Assessment of Intracellular Antioxidant Activity

RAW 264.7 macrophages were used and seeded into a 96-well plate. The medium was removed, and fresh medium containing various concentrations of cocoon extracts was added for a 24 h pre-incubation. After pre-incubation, the cells were induced to undergo oxidative stress by the addition of hydrogen peroxide (H₂O₂). The total cellular SOD activity was measured. The decrease in fluorescence intensity reflects the extract’s capacity to scavenge intracellular reactive oxygen species (ROS).

2.13. Statistical Analysis

To account for individual biological variability, samples for the UPLC-MS biomarker analysis were obtained from a bulk batch of >1 kg of cocoons for each group. We analyzed five independent biological replicates per group (n = 5). Each replicate consisted of pooled material from multiple randomly selected cocoons to ensure a representative chemical fingerprint and minimize individual bias. For the in vitro bioactivity assays (cytotoxicity, ROS scavenging, and anti-inflammatory tests) and physical characterizations (degumming kinetics, TGA, and FTIR), we conducted three independent experiments (n = 3), with each experiment performed in technical triplicate where applicable. Statistical analysis was performed using Prism 8.0.1. Statistical analysis was performed using one-way ANOVA or two-way ANOVA, followed by Tukey’s post hoc test for multiple comparisons. All data are represented as mean ± SD.

3. Results

3.1. Calculation of Sericin Removal Efficiency

Sericin serves as a globular adhesive protein that coats and binds the two fibroin fibrils, ensuring the structural integrity of natural silk fibers. Given that silkworm diets differ substantially between natural mulberry leaves and artificial feed, we hypothesized that changes in nutrient composition could modify the molecular structure or interfacial interactions of silk proteins, thereby influencing the adhesion between sericin and fibroin. To verify this hypothesis, we conducted a degumming test to assess sericin detachment behavior in cocoons from the two feeding types. As shown in Figure 1B, the initial degumming rate of Art-fed cocoons within the first minute showed an increasing trend compared to the Mul-fed cocoons. But statistical analysis revealed no significant difference between the two groups (Figure 1B), likely due to individual biological variability. This finding suggests that while physical parameters like degumming kinetics may hint at structural differences, they lack the sensitivity required for definitive authentication, highlighting the need for the molecular biomarker analysis developed in this study. These results align with previous reports [11,17]. Furthermore, it was found that Mul-fed cocoons typically exhibited an initial 1 min degumming rate of below 65.0%, which might provide a simple indicator for the identification of feeding methods.

Figure 1.

Degumming of Mul-fed cocoons and Art-fed cocoons. (A) The cocoon weight of the two types of cocoons before degumming (0 min), after 1 min of degumming (1 min), and after 10 min of degumming (10 min). (B) The degumming rate of the two types of cocoons in the first minute. (C) The photo of the cocoon shell. The left is Mul-fed cocoons, and the one on the right is Art-fed cocoons. The comparison among groups is performed using one-way ANOVA (ns p > 0.05).

3.2. Physicochemical Properties of Cocoons Before and After Biomarker Extraction

Biomarkers were extracted from cocoon shells using sequential ultrasonic and thermal treatments in 75% ethanol. The physicochemical characteristics of Mul-fed and Art-fed cocoons after extraction were investigated, with particular focus on fiber morphology, thermal stability, amino acid compositions, and protein secondary structures.

To assess the initial silk fiber morphology, scanning electron microscopy (SEM) was performed on cocoon samples prior to biomarker extraction, confirming that the silk fibers from both cocoon types remained largely intact (Figure 2A,B). Following biomarker extraction, the Art-fed cocoon showed a disorganized and loose silk fiber network, presenting a rough and irregular surface. Moreover, the sericin layer of Art-fed cocoon displayed pronounced damage, exposing the underlying fibroin, which suggests weaker sericin binding and increased susceptibility to hydrolysis for Art-fed cocoons during biomarker extraction (Figure 2A). In contrast, Mul-fed cocoons retained a smooth, compact silk fiber network and were encased by a continuous and integral sericin layer (Figure 2B). This coherent morphology is highly consistent with the observed lower initial degumming rate, indicating that mulberry feeding may enhance the adhesion of sericin to fibroin and improve overall fiber integrity.

Figure 2.

Morphological and structural characterizations of cocoons. (A) SEM images of Art-fed cocoon before biomarker extraction (Art-BE) and after biomarker extraction (Art-AE). (B) SEM images of Mul-fed cocoon before biomarker extraction (Mul-BE) and after biomarker extraction (Mul-AE). (C) The FTIR spectra of different cocoons before and after biomarker extraction. (D) Comparison of β-sheet, β-turn, and random coil/helix content of silk proteins in different cocoons before and after biomarker extraction. The comparison among groups is performed using one-way ANOVA (ns p > 0.05, *** p < 0.001).

To further investigate the underlying mechanism responsible for these morphological differences, ATR-FTIR spectroscopy was employed to analyze the secondary structures of silk proteins in both types of silkworm cocoons, before and after biomarker extraction (Figure 2C). Before biomarker extraction, the β-sheet content of silk fibers in both Mul-fed and Art-fed cocoons was approximately 20% (Figure 2D). After biomarker extraction, the β-sheet content of silk fibers in Mul-fed cocoons was around 20%, suggesting that the sericin layer remained relatively intact and was not significantly compromised by the extraction process. In contrast, the β-sheet content of silk fibers in Art-fed cocoons increased significantly from 20% to 38%, accompanied by a reduction in random coil/helix content (Figure 2D). This shift is likely due to the significant disruption of the sericin layer in Art-fed cocoons, exposing the underlying fibroin, consistent with the SEM results described above.

The thermal stability of Mul-fed and Art-fed cocoons was investigated using TGA and DSC analysis. Both types of cocoons show primary thermal decomposition in the range of 200 to 400 °C (Figure 3A). There were no significant differences observed between the two types of cocoons in either the decomposition peak temperature or the total weight loss percentage (Figure 3B,D). Specifically, the average main decomposition temperature remained at ~319 °C, regardless of whether the samples were analyzed before or after extraction (Figure 3C). It can be concluded that while different feeding methods lead to variations in chemical composition and structural stability of cocoons, the core thermal stability of their major protein components remains broadly similar.

Figure 3.

Thermal analysis and amino acid analysis results of Mul-fed and Art-fed cocoons before and after extraction. (A) Thermogravimetric analysis curves before and after extraction. (B) Quantitative comparison of the main decomposition peak temperature. (C) Derivative thermogravimetric curves before and after extraction. (D) Quantitative comparison of total weight loss percentage. (E) Amino acid analysis of Mul-fed and Art-fed cocoons before and after biomarker extraction. The comparison among groups is performed using one-way ANOVA (ns p > 0.05, ** p < 0.01, *** p < 0.001).

The amino acid compositions of Mul-fed and Art-fed cocoons before and after biomarker extraction were subsequently analyzed. Silk proteins from both types of cocoons were predominantly composed of glycine, alanine, and serine, consistent with previous reports [17,19]. And no significant differences were observed in the amino acid compositions of silk proteins between Mul-fed and Art-fed cocoons (Figure 3E). After biomarker extraction, the molar fractions of amino acids, such as serine and aspartic acid/asparagine, significantly decreased. Concurrently, the relative contents of glycine and alanine, which constitute the crystalline core of silk fibroin, significantly increased (Figure 3E). These combined findings from SEM, TGA/DSC, and amino acid analysis indicated that partial hydrolysis of sericin occurred during the biomarker extraction process, and sericin from Art-fed cocoons was more susceptible to hydrolysis than that from Mul-fed cocoons.

3.3. Analysis of Biomarkers in Cocoons

Biomarkers extracted from cocoons with 75% ethanol were tested to distinguish between Mul-fed and Art-fed cocoons at the molecular level. Quercitrin, isoquercitrin, and quercetin, typical components of mulberry leaves, along with daidzein and genistein in artificial diets, were selected as candidate biomarkers for comparative analysis using UPLC-MS. Firstly, the five biomarkers were identified, and their mass-to-charge ratios (m/z) were analyzed by UPLC-MS (Figure 4A,B; Table 1). Subsequently, Art-fed and Mul-fed cocoon extracts used for biomarker analysis were also subjected to UPLC-MS identification. Preliminary examination of the total ion chromatograms (TICs) from both groups revealed distinct differences in the 3–6 min retention time range (Figure 4C,D and Figure S1A). Detailed analysis of the mass spectrometry (MS) data, particularly the [M-H]⁻ mass-to-charge ratio (m/z), indicated the presence of key biomarkers, including quercitrin, isoquercitrin, quercetin, daidzein, and genistein. Direct comparison with the standards confirmed the identities of these biomarkers in the cocoon extracts, with matching retention times and mass-to-charge ratios (m/z). The identification of these five biomarkers in Mul-fed and Art-fed cocoons provides a clear molecular basis for subsequent quantitative analysis.

Figure 4.

Analysis of biomarkers in silkworm cocoons. (A) Total ion chromatograms of the five standards. (B) Total ion chromatograms within 3–6 min of five standards. (C) Total ion chromatograms of extracts from Art-fed and Mul-fed cocoons. (D) Total ion chromatograms (3–6 min) of extracts from Art-fed and Mul-fed cocoons. The characteristic peaks of five identified biomarkers are highlighted by colored vertical lines: rose red for isoquercitrin, yellow for daidzein, green for genistein, purplish red for quercitrin, and blue for quercetin.

Table 1.

The LC-MS results of five standards.

Name	Retention Time/min	m/z [M-H]−
Isoquercitrin	4.281	463.4
Quercitrin	4.466	447.2
Quercetin	4.991	301.2
Genistein	4.850	253.2
Daidzein	5.263	269.2

To further validate the candidate biomarkers, we performed a targeted analysis using Extracted Ion Chromatograms (EICs). The EIC at m/z = 463, corresponding to the [M-H]⁻ ion of isoquercitrin, confirmed the presence of this flavonol glycoside in every cocoon sample analyzed. However, its content exhibited considerable individual-to-individual variation, suggesting that while its presence is universal, its accumulation may be influenced by additional physiological factors (Figure S1B, Table S1). In contrast, a distinct pattern emerged for quercitrin (m/z = 447), which was consistently detected in all Mul-fed cocoons but was significantly reduced in the Art-fed cocoons (Figure S1C). Quantitative analysis showed the average concentration of quercitrin was approximately five times higher in the Mul-fed cocoons (Table S2), underscoring its strong association with the natural mulberry diet. A more pronounced disparity was observed for the aglycone quercetin (m/z = 301). It was detected in all Mul-fed cocoons but was entirely absent in a substantial subset of Art-fed cocoons, leading to an average abundance nearly forty-fold greater in the Mul-fed group (Figure S1D, Table S3). This stark contrast suggests that the flavonoids quercetin and quercitrin detected in cocoons predominantly originate from mulberry leaves, while their content in the artificial diets is negligible. Conversely, the profile for isoflavones was inverted. Daidzein (m/z = 253) and genistein (m/z = 269) were reliably identified in all Art-fed cocoons, whereas their levels were negligible or below the detection limit in virtually all Mul-fed samples (Figure S1E,F; Tables S4 and S5). This clear inverse relationship not only confirms their status as negative markers for a mulberry diet but also implies that the artificial feed is the primary source of these isoflavone precursors.

Biomarker analysis revealed distinct compositional differences between Mul-fed and Art-fed cocoons. Mul-fed cocoons were enriched in flavonoids such as quercetin and quercitrin, while Art-fed cocoons uniquely contained isoflavones such as genistein and daidzein (Figure 5A,B; Table 2). Mass spectra of Art-fed cocoons also showed more and higher-intensity impurity peaks, indicating greater hydrolytic degradation of sericin during extraction, leading to the formation of small peptides and amino acid fragments.

Figure 5.

Quantitative analysis of biomarkers. (A) Content and differential statistical analysis of five biomarkers in Mul-fed and Art-fed cocoons. (B) Enrichment analysis of five biomarkers. The comparison among groups is performed using one-way ANOVA (ns p > 0.05, * p < 0.05, *** p < 0.001).

Table 2.

The average contents of flavonoids and isoflavonoids in cocoons. The comparison among groups is performed using one-way ANOVA (^ns p > 0.05, * p < 0.05, *** p < 0.001).

Name	Mul-Fed Cocoon (%)	Art-Fed Cocoon (%)
Isoquercitrin	0.1140 ± 0.09730 ns	0.06850 ± 0.03830
Quercitrin	0.1570 ± 0.09680 *	0.03110 ± 0.01900
Quercetin	0.1116 ± 0.05430 ***	0.002800 ± 0.001900
Genistein	0.001400 ± 0.001000	0.1615 ± 0.08410 ***
Daidzein	0.004000 ± 0.002800	1.589 ± 0.9472 ***

To validate the accuracy of biomarkers for cocoon differentiation, principal component analysis (PCA) was employed. The score plot revealed a distinct separation between the Mul-fed and Art-fed groups along the first two principal components (PC1 and PC2), which accounted for 57.0% and 22.5% of the total variance, respectively (Figure 6A). This pronounced clustering pattern indicates that feeding type is a predominant factor influencing the chemical composition of the cocoon. The distribution plot, incorporating 95% confidence ellipses, further corroborated the intergroup differences, with minimal overlap between the ellipses (Figure 6B). Notably, the ellipse for the Art-fed group appeared both broader and more elongated compared to the Mul-fed group, suggesting increased chemical heterogeneity (e.g., batch-to-batch or individual variation) within the Art-fed cocoons.

Figure 6.

Principal component analysis (PCA) of biomarkers extracted from Art-fed and Mul-fed cocoons. (A) PCA score plot (convex hull). (B) PCA score plot (confidence ellipse). (C) PCA loading biplot. The arrows represent the loading vectors of the identified metabolites, indicating their respective contributions to PC1 and PC2. The direction and length of the arrows reflect the correlation and importance of each biomarker in differentiating the Art-fed and Mul-fed groups. (D) 3D PCA score plot.

The loading vectors (depicted by red arrows) superimposed on the PCA score plot identified the key metabolites driving the observed group separation (Figure 6C). Vectors oriented towards the Art-fed group (on the negative side of PC1) were primarily associated with isoflavones, including daidzein and genistein, which are highly enriched in Art-fed cocoons. Conversely, vectors directed towards the Mul-fed group (on the positive side of PC1) corresponded to flavonoids, such as quercitrin and quercetin, that are predominantly found in Mul-fed cocoons.

The three-dimensional PCA model, which integrated PC1, PC2, and PC3, explained 92.2% of the total variance (57.0% + 22.5% + 12.7%) (Figure 6D). In this 3D space, the separation between the Mul-fed and Art-fed groups remained apparent. Importantly, the Art-fed samples exhibited greater dispersion along PC2 and PC3, reflecting a higher degree of chemical variability. These findings provide compelling evidence for the use of UPLC-MS-based biomarker profiling, combined with PCA modeling, as a robust and reliable approach for distinguishing Mul-fed from Art-fed cocoons at the molecular level.

3.4. In Vitro Bioactivity Analysis

We dissolved the cocoon shell in LiBr solution, then dialyzed and lyophilized to obtain the cocoon extraction materials. To evaluate the potential of silk cocoons for future food applications, the extraction material was assessed for cytotoxicity and biosafety. The results demonstrated that both Mul-fed and Art-fed cocoon extracts exhibited no apparent cytotoxicity on RAW 264.7 cells across the entire experimental concentration range 0.1 to 1000 μg/mL, with cell viability consistently maintained above 95%, thus confirming the excellent biocompatibility of both extracts (Figure 7A). Similarly, there was no statistically significant change in the ROS scavenging activity of the two groups (Figure 7B). However, the antioxidant activity of both groups showed a dose-dependent manner, that is, higher concentrations resulted in higher ROS scavenging activity (Figure 7B). In terms of anti-inflammatory activity, two types of cocoons both demonstrated a dose-dependent anti-inflammatory effect, and the Mul-fed cocoon extracts exhibited a significantly higher effect than the Art-fed cocoon extracts in high concentration (1000 μg/mL), both in inhibiting the secretion of pro-inflammatory cytokines (IL-6, IL-1β, and TNF-α) and in promoting the secretion of the anti-inflammatory factor (TGF-β) (Figure 7C–F). We propose that the superior antioxidant and anti-inflammatory activities observed in Mul-fed cocoon extracts indicate considerable potential for future food industry applications.

Figure 7.

In vitro biological evaluation of Mul-fed and Art-fed cocoon extracts. (A) Cytocompatibility assessment of RAW 264.7 cells treated with different concentrations of extracts 0.1–1000 μg/mL. (B) Intracellular reactive oxygen species scavenging rate of the extracts. (C–F) Inhibitory effects of the extracts on the secretion levels of pro-inflammatory cytokines IL-6 (C), IL-1β (D), TNF-α (E), and TGF-β (F) in the LPS-induced inflammatory model. Data present mean values ± SD from three independent experiments. The comparison among groups is performed using one-way ANOVA (ns p > 0.05, * p < 0.05, ** p < 0.01).

4. Discussion

This study established a multidimensional analytical framework to distinguish cocoons produced by Mul-fed and Art-fed silkworms. While artificial diets offer economic advantages and enable year-round rearing, the resulting decline in cocoon quality has been a persistent concern for the sericulture industry. The findings provide a scientific basis for differentiating these cocoons not just by physical appearance, but by intrinsic molecular and structural fingerprints.

Morphological and thermal analyses revealed that the primary difference between the two cocoon types lies in the stability of the sericin layer rather than the thermal properties of the core fibroin. The SEM and degumming results demonstrated that Art-fed cocoons possess a more loosely organized fiber network and a sericin layer that is significantly more susceptible to hydrolysis. This finding aligns with previous studies suggesting that artificial diets may disrupt the synthesis or secretion balance of sericin components, thereby weakening adhesion between sericin and fibroin [11,17]. The increase in β-sheet content in Art-fed cocoons after biomarker extraction likely reflects the rapid loss of the amorphous sericin layer, exposing the crystalline fibroin core. In contrast, the robust sericin layer in Mul-fed cocoons suggests that natural dietary components contribute to a more cohesive silkworm cocoon structure, which is critical for downstream textile and biomaterial processing.

The most defining distinction identified in this study is the specific accumulation of dietary metabolites in the cocoon shell. It was successfully validated that secondary metabolites derived from the diet are transferred to the silk gland and deposited in the cocoon. The detection of flavonoids (quercitrin and quercetin) exclusively or predominantly in Mul-fed cocoons serves as a positive biomarker for mulberry leaf feeding. These compounds are abundant in mulberry leaves and are known to contribute to the unique properties of silk [12,24]. Conversely, the presence of genistein and daidzein in Art-fed cocoons acts as a definitive negative biomarker. These isoflavones are characteristic of soybean meal, a primary protein source in artificial diets [26]. This clear “metabolic mirroring” confirms that LC-MS-based biomarker profiling is a far more specific and objective tool for traceability than traditional sensory evaluation. Furthermore, the PCA model further demonstrates that these chemical signatures can statistically separate the two groups with high confidence, overcoming the limitations associated with batch-to-batch variability.

Beyond structural and compositional characteristics, this study highlights the functional superiority of Mul-fed cocoons. The in vitro assays demonstrated that extracts from Mul-fed cocoons possess significantly higher intracellular antioxidant and anti-inflammatory activities compared to Art-fed counterparts. This enhanced bioactivity is likely correlated with the enrichment of the identified flavonoids (quercetin and quercitrin), which are potent antioxidants and anti-inflammatory agents [7,12]. This finding is particularly relevant for the application of silk sericin in the food and biomedical sectors. It suggests that Mul-fed cocoons are not only structurally superior for textiles but also biologically superior for developing functional foods or wound-healing materials.

It is important to acknowledge that this study employed a specific commercial artificial diet and a single cultivar of mulberry leaves. Although the absolute concentrations of the identified biomarkers may exhibit variation depending on the feed manufacturer, mulberry cultivar, or geographical region, the fundamental metabolic divergence—characterized by the presence of soy-derived isoflavones in artificial diet-fed cocoons and the enrichment of specific flavonoids in mulberry-fed cocoons—is expected to serve as a consistent distinguishing feature. Future investigations encompassing a diverse range of diet formulations and geographical origins will further substantiate the universality and robustness of these markers. Similarly, the flavonoid profile of mulberry leaves can fluctuate based on geographic location, season, and cultivar, potentially affecting the quantitative baseline of positive markers. Secondly, while the study confirmed the presence of these metabolites in the cocoon shell, the specific metabolic pathways and transport mechanisms governing their transfer from the hemolymph to the silk gland remain to be fully elucidated. Finally, the bioactivity assays were conducted in vitro; in vivo studies are necessary to confirm the clinical relevance of the observed anti-inflammatory and antioxidant effects.

Future research should focus on expanding the library of biomarker library through the analysis of a broader range of silkworm strains, diet formulations, and rearing environments to establish a more robust and comprehensive molecular database. Building on the “metabolic mirroring” concept validated in this study, significant potential exists for optimizing artificial diets via supplementation with specific mulberry-derived flavonoids, which may help narrow the quality gap between the two rearing methods. From an application perspective, translating these sophisticated LC-MS findings into rapid, cost-effective detection tools—such as portable spectral sensors or colorimetric test kits targeting these specific biomarkers—would represent a major advancement for industrial quality control in sericulture. Furthermore, given the superior bioactivity of Mul-fed cocoons, future work should explore their specific development into high-value functional biomaterials, including wound healing dressings or bioactive scaffolds for tissue engineering.

5. Conclusions

This study successfully employed a multi-dimensional analytical approach integrating chemical profiling, structural characterization, and bioactivity evaluation to comprehensively differentiate and compare the quality of Mul-fed and Art-fed cocoons. UPLC-MS analysis identified five compounds, daidzein, genistein, isoquercitrin, quercitrin, and quercetin, as characteristic biomarkers. Quercitrin and quercetin were significantly enriched in Mul-fed cocoons, while daidzein and genistein were uniquely present in Art-fed cocoons. This distinction enabled the establishment of a reliable and quantifiable molecular fingerprinting method for accurate cocoon source authentication. In addition, this study confirmed the superior structural stability of Mul-fed cocoons, as evidenced by a more uniform surface morphology and a more regular, compact fiber network after biomarker extraction. Furthermore, in vitro assays showed that Mul-fed cocoon extracts exhibited significantly greater biological activity than Art-fed cocoon extracts, including enhanced cell proliferation, excellent anti-inflammatory effects, and higher intracellular antioxidant activity. This study provides a scientific basis for cocoon identification and value assessment, as well as insights into future biomaterial development based on cocoons.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/insects17020188/s1, Figure S1: Quantitative analysis of biomarkers; Figure S2: The mass spectrum of the quercetin (2-(3,4-Dihydroxyphenyl)-3,5,7-trihydroxy-4H-chromen-4-one) standard; Figure S3: The mass spectrum of the quercitrin (2-(3,4-Dihydroxyphenyl)-5,7-dihydroxy-3-[(2S,3R,4R,5R,6S)-3,4,5-trihydroxy-6-methyloxan-2-yl]oxy-4H-chromen-4-one) standard; Figure S4: The mass spectrum of the isoquercitrin (3-[[(2S,3R,4S,5S,6R)-3,4,5-Trihydroxy-6-(hydroxymethyl)oxan-2-yl]oxy]-2-(3,4-dihydroxyphenyl)-5,7-dihydroxy-4H-chromen-4-one) standard; Figure S5: The mass spectrum of the daidzein (7-Hydroxy-3-(4-hydroxyphenyl)-4H-chromen-4-one) standard; Figure S6: The mass spectrum of the genistein (7-Hydroxy-3-(4-hydroxyphenyl)-4H-chromen-4-one) standard; Figure S7: The extracted ion chromatogram (EIC) results for Art-fed cocoon-1 at the same mass-to-charge ratios as the five aforementioned standards; Figure S8: The EIC results for Art-fed cocoon-2 at the same mass-to-charge ratios as the five aforementioned standards; Figure S9: The EIC results for Art-fed cocoon-3 at the same mass-to-charge ratios as the five aforementioned standards; Figure S10: The EIC results for Mul-fed cocoon-1 at the same mass-to-charge ratios as the five aforementioned standards; Figure S11: The EIC results for Mul-fed cocoon-2 at the same mass-to-charge ratios as the five aforementioned standards; Figure S12: The EIC results for Mul-fed silkworm cocoon-3 at the same mass-to-charge ratios as the five aforementioned standards; Table S1: Content of isoquercitrin in different samples, Table S2: Content of quercitrin in different samples; Table S3: Content of quercetin in different samples; Table S4: Content of daidzein in different samples; Table S5: Content of genistein in different samples.

Author Contributions

Conceptualization, C.G., S.Y. and X.M.; methodology, C.G., S.Y., X.M. and J.M.; software, X.M.; validation, C.G., S.Y., R.H. and Y.S.; formal analysis, C.G. and X.M.; investigation, C.G., S.Y. and X.M.; resources, C.G. and R.H.; data curation, X.M.; writing—original draft preparation, S.Y. and X.M.; writing—review and editing, C.G., S.Y., R.H. and Y.S.; visualization, X.M. and J.M.; supervision, C.G.; project administration, S.Y. and J.M.; funding acquisition, Y.S. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

All relevant data are included in the paper.

Conflicts of Interest

Authors Ran Huang and Yuwei Song were employed by the company Zhejiang Cathaya International Co., Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding Statement

This work was supported by the Foundation of Westlake University.