. Introduction 

Protein scarcity is one of the key challenges in achieving food security for a growing global population. According to recent projections, 
approximately 600 million people are expected to suffer from chronic undernourishment by 2030 (FAO, IFAD, UNICEF, WFP, & WHO, 2023). 
Expanding the sources of food, particularly protein, is essential for addressing the population growth and food security crisis. Insects and 
related products offer a feasible and necessary solution because of their high efficiency in protein conversion and their minimal environmental 
footprint. This attribute makes them an attractive option for sustainable protein production (van Huis & Gasco, 2023). Silkworms have played a significant role in the textile industry for centuries, producing silk, a natural protein product with diverse applications. Silk consists of two main components: sericin and fibroin. Sericin makes up approximately 25 %–35 % of silk, whereas fibroin 
comprises the remaining 65 %–75 %. In the silk reeling industry, the manufacturer removes the sericin covering the outer layer of silk by 
boiling the cocoons to obtain textile-usable fibroin fibers and the removed sericin is discharged into the reeling wastewater. The global 
production of cocoons exceeds 100,000 tons per year, which means that more than 30,000 tons of sericin is wasted without proper application. 
This wastage presents an opportunity for resource optimization and sustainable development within the silk industry (Seo, Das, Shin, & Patra, 2023). Previous studies have demonstrated the remarkable properties of sericin and its potential application in the food industry. Sericin exhibits 
excellent biocompatibility, with virtually no toxicity or allergenicity. In addition, it possesses a range of functional properties, such as antioxidant and bacteriostatic activities (Saad, El-Samad, Gomaa, Augustyniak, & Hassan, 2023; Zhu, Zhang, Lu, Chen, Chen, Han, et al., 2020). The 
addition of sericin as a food ingredient has shown promising results in improving the compatibility and texture of various food products. For 
instance, studies have observed positive effects when sericin was added to salad dressings, bread, and potato dough, enhancing their sensory 
characteristics (Ghosh et al., 2019; Gong et al., 2019; Sabu Mathew, Maria, Allardyce, Rajkhowa, & Thomas, 2024). These studies provide opportunities for innovative food formulations, improved food quality,

The Maillard reaction rate was indicated by the brown pigment level 

produced by the Maillard reaction. The sample (250 mg) was dissolved in 10 mL of phosphate buffer (pH = 8.0) and stirred at room temperature 
for 60 min, followed by the addition of 12 μL of alcalase (P4860, ≥2.4 U/ g, Sigma-Aldrich, Inc., USA) and incubation in a water bath at 55 ?C for 
15 min. The enzyme reaction was stopped by the addition of 1 mL of trichloroacetic acid (TCA) (80 %, ω/v), and filtered. The absorbance of 
the filtrate was measured using FlexA-200 (Hangzhou Allsheng Instruments Co., Ltd., China) at 420 nm to indicate the Maillard reaction 
rate in the sample. Nε-carboxymethyllysine (CML) and Nε-carboxyethyllysine (CEL) are two typical glycation end products (AGEs) formed at the advanced stage of the Maillard reaction (Zhu, Huang, Cheng, Khan, & Huang, 2020). CML ELISA and CEL ELISA kits (Shanghai Lianzu Biotechnology Co., Ltd.) were used to measure the CML and CEL levels in the samples, respectively (G′omez-Ojeda et al., 2018). 
 Carbonyl content analysis 
The protein carbonyl content is widely used to assess the degree of 
protein oxidation. Protein carbonyl reacts with 2,4-dinitrophenylhydrazine (DNPH) to form a red precipitate with a characteristic absorption 
peak at 370 nm. The protein carbonyl content can be quantified by measuring the change in absorbance value (Poojary & Lund, 2022). For 
this, 5 mL of the sample solution was added with 5 mL of 10 mmol/L DNPH solution containing 2 mol/L HCl, was allowed to react at room 
temperature away from light, and was shaken every 15 min. After 1 h, 5 mL of 20 % TCA was added to the mixture, followed by centrifugation at 10,000 g for 5 min. The obtained precipitate was mixed with 5 mL of ethyl acetate–ethanol mixture (1:1) and 10 mL of 6 mol/L guanidine 
hydrochloride solution, and the absorbance at 370 nm was measured after a water bath at 37 ?C for 15 min and centrifugation at 10,000 g for 
3 min before the calculation of the carbonyl content of myosin. Carbonyl Content (nmol/mg) = 10^6 × n × A370/(C × D × 22,000), 
where A370-absorbance value of the sample at a wavelength of 370 nm, C-concentration of the protein solution (mg/mL), D-colorimetric 
diameter, 22,000-molar absorbance coefficient [L/(mol × cm)], and ndilution factor. The results obtained were normalized and analyzed 
using data from day 0. 
Determination of the free sulfhydryl content 
5,5′-Dithiobis(2-nitrobenzoic acid), i.e., DTNB, is commonly known as Ellmann’s reagent. In the presence of sulfhydryl compounds, the 
colorless DTNB was converted to yellow 5-mercapto-2-nitrobenzoic acid with a maximum absorption peak at 412 nm. The sample solution (1.0 
mL) before and after heat treatment was mixed with 4.0 mL of Tris-Gly buffer (0.086 M Tris, 0.09 M Gly, 5 mM EDTA, pH 8.0), and then, 50 μL of Ellmann’s reagent at 4 mg/mL (4.0 mg of DTNB dissolved in 1.0 mL of Tris-Gly buffer) was added. Subsequently, the reaction was performed at 37 ?C for 15 min and the absorbance value was measured at 412 nm. Free Sulfhydryl Content (nmol/mg) = (73.53 × A412 × n)/C, 
where A412-absorbance of the sample at a wavelength of 412 nm, Cconcentration of the protein solution (mg/mL), and n-number of dilutions. The results obtained were normalized and analyzed using data from day 0. 

 

Analysis of the molecular weight distribution 

MALDI-TOF-MS was used to measure the molecular weights of glycated-HPNBs according to the methods of Liu and Kahlif. Sinapic acid 
(5 mg/mL) and trifluoroacetic acid (0.1 %, ω/v) were added to acetonitrile (50 %, ω/v). Trifluoroacetic acid (0.1 %, ω/v) was used the 
substrate. Proteins were diluted with MillQ-H2O (1:100) and mixed with the previous matrix (1:1). After the crystallization of both samples by 
blowing, 1.5 μL of each sample was placed in a MALDI target cup, dried, and analyzed in positive ion mode using an Ultraflex mass spectrometer 
(Bruker Daltousics, Germany). The results obtained were further analyzed using mMass (Niedermeyer & Strohalm, 2012). 
Statistical analysis 
Three replicate samples (n = 3) were used to eliminate experimental errors. At a significance level of p < 0.05, data were expressed as mean 
± standard deviation. SPSS Statistics 20.0 (IBM, Inc., USA) was used for one-way ANOVA, followed by Duncan’s multiple range test. Excel 2019 
(Microsoft, USA) was used to count and calculate the data. GraphPad Prism (version 8.4.2.679, GraphPad Software, USA) and Origin Lab 
2018C (OriginLab, USA) were used to prepare graphs. Adobe Illustrator 2021 (Adobe, USA) was used to trim and manipulate the graphs.