Influence of ultrasound pretreatment on the subsequent glycation of dietary proteins

Journal Pre-proofs Influence of ultrasound pretreatment on the subsequent glycation of dietary proteins Dan Xu, Lin Li, Yi Wu, Xia Zhang, Ming Wu, Yuting Li, Zuoqi Gai, Bing Li, Di Zhao, Chunbao Li PII: DOI: Reference:

S1350-4177(19)31029-6 https://doi.org/10.1016/j.ultsonch.2019.104910 ULTSON 104910

To appear in:

Ultrasonics Sonochemistry

Received Date: Revised Date: Accepted Date:

3 July 2019 7 November 2019 26 November 2019

Please cite this article as: D. Xu, L. Li, Y. Wu, X. Zhang, M. Wu, Y. Li, Z. Gai, B. Li, D. Zhao, C. Li, Influence of ultrasound pretreatment on the subsequent glycation of dietary proteins, Ultrasonics Sonochemistry (2019), doi: https://doi.org/10.1016/j.ultsonch.2019.104910

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier B.V.

Influence of ultrasound pretreatment on the subsequent glycation of dietary proteins Dan Xu a, Lin Li a, b, Yi Wu a, Xia Zhang a, Ming Wu a, Yuting Li b, Zuoqi Gaid, Bing Li a, *,

a

Di Zhao c, * Chunbao Li c

College of Food Science and Engineering, Guangdong Province Key Laboratory for

Green Processing of Natural Products and Product Safety, South China University of Technology, 381 Wushan Road, Tianhe District, Guangzhou, 510640, China b

School of Chemical Engineering and Energy Technology, Dongguan University of

Technology, College Road 1, Dongguan, 523808, China c

Key Laboratory of Meat Processing and Quality Control, MOE; Jiangsu Collaborative

Innovation Center of Meat Production and Processing, Quality and Safety Control; Key Laboratory of Meat Products Processing, MOA; Nanjing Agricultural University; Nanjing, 210095, P.R. China d College

of Life Science and Engineering, Foshan University, Foshan, 528231, China

* Corresponding author: Bing Li, E-mail: [email protected], Tel number: +86 2087113252 Mail address: College of Food Science and Engineering, South China University of Technology, 381 Wushan Road, Guangzhou, 510640, China Di Zhao, E-mail: [email protected], Tel number: +86 15951911680 Mail address: College of Food Science and Technology, Nanjing Agricultural University, Nanjing, 210095, China

Abstract: The influence of ultrasound treatment on the subsequent glycation process of proteins is controversial. Glycation behaviors of bovine serum albumin (BSA), β-lactoglobulin (β-Lg) and β-casein (β-CN) after ultrasound pretreatment (UP) were compared by both evaluating glycation kinetics and analyzing structural changes of proteins. UP resulted in both unfolding and aggregation behavior in protein samples, which altered the accessibility of the Lys and Arg. Five cycles of UP up-regulated the glycation degree of BSA and β-Lg, possibly due to the unfolding behavior induced by UP, which exposed additional glycation sites. In contrast, 30 cycles of UP induced a dramatic increase (by 97.9 nm) in particle size of BSA, thus burying portions of glycation sites and suppressing the glycation process. Notably, UP had minimal influence on glycation kinetics of β-CN, due to its intrinsic disordered structure. Based on proteomics analysis, the preference of Lys and Arg during glycation was found to be changed by UP in BSA and β-Lg. Four, 3 and 3 unique carboxyethylated lysine residues were identified in glycated BSA after 0, 5 and 30 cycles of UP, respectively. This study suggests that the protein glycation can be affected by UP, depending on the ultrasonication duration and native structure of the protein. KEYWORDS: ultrasound pretreatment; protein glycation; unfolding; aggregation

1. Introduction The Maillard reaction or glycation is a non-enzymatic modification of amino groups from amino acids, peptides, or proteins by reducing sugars or other reactive carbonyls [1]. Glycation largely contributes to attractive colors and flavors of foods [1, 2]. In addition, glycation endows the dietary proteins with new functionalities and bioactivities [3]. The solubility, thermal stability, antioxidant properties, and antiproliferative ability of dietary proteins can be largely improved by glycation [4, 5]. Moreover, glycated proteins have been revealed to be excellent units to build functional macrostructures in foods. [7, 8]. Therefore, regulation of the protein glycation has aroused lots of interests among food researchers. Protein glycation can be affected by many factors, including temperature, pressure, pH, reducing sugar, metal ions and physical field application (e.g. microwave and ultrasound). The application of high-pressure treatment was shown to reduce the degree of glycation of proteins [9]. Fe2+ was reported to promote the glycation in a concentration range of 20 to 100 mg/L, whereas Cu2+ and Zn

2+

were

revealed to promote glycation in the concentration range of 1-5 mg/mL [10]. Additionally, phosphate groups have been found to accelerate the glycation process by stabilizing the Schiff base [11]. It is worth noting that proteins are the macromolecules with high-order structures. Therefore, the glycation process of a protein can also be affected by altering its structures, which is essentially determined by the location and accessibility of glycation sites (Lys and Arg residues). Compared with conventional thermal processing, microwave treatment was proved to slow down the glycation kinetics of ovalbumin [12]. Breaking the disulfide bonds increased the glycation degree of ovalbumin by exposing more Lys residues [13].

Ultrasound treatment has also been found to affect the glycation of proteins in a few studies, whereas their conclusions were largely inconsistent. More glycated Lys residues in bovine serum albumin (BSA) was identified after ultrasound pretreatment (UP) [14]. Ultrasonication was found to enhance the formation of Maillard reaction products (MRPs) in BSA and β-conglycinin [15, 16]. In contrast, UP was found to reduce the formation of MRPs in the subsequent glycation of casein [17]. UP decreased the glycation process of β-Lg by accelerating its aggregation behavior in the initial heating which buried portions of glycation sites [18]. It can be deduced that the glycation of proteins under ultrasonic treatment would be closely associated with their structures and the ultrasonic parameters. However, the underlying mechanism have not been revealed. To verify these hypotheses, the influences of UP on the glycation behaviors of bovine serum albumin (BSA) which has heart-shaped and highly helical structure [19], β-lactoglobulin (β-Lg) which is a rigid globular protein with a high level of β-sheet structure [20], and β-casein (β-CN) which is an intrinsically disordered protein [21] were compared in this study. Changes in glycation kinetics, protein structures and glycation sites were analyzed to uncover the mechanism. 2. Materials and methods 2.1. Materials BSA (≥98%), β-Lg (≥90%), β-CN (≥98%), glucose (Glu), o-Phthaldialdehyde (OPA), 1-anilinonaphthalene-8-sulfonate (ANS) and trypsin (≥10,000 BAEE units/mg protein) were from Sigma-Aldrich. (Shanghai, China). 2.2. Ultrasound pretreatment (UP)

In this experiment, the volume of the solution in reaction vessel was 100 ml. The protein solution was treated using a 13 mm probe sonicator (20 kHz, VCX 500, SONIC & MATERIALS, USA), the length of the flash was 26 mm in the solution, and the probe was fixed in the center of each sample solution during ultrasonic pretreatment. The protein samples dissolved in phosphate buffer (12 mg/mL, 25 mM, pH 7) were ultrasonicated for 0 to 60 cycles (one cycle include 30 s ultrasonication and 30 s pause). An ice-water bath was used to minimize the thermal effect induced by ultrasonication. New ice was added after every 5 cycles of UP during the 30 s of pause to ensure the constant temperature. 2.3. Glycation The protein solution (12 mg/mL) after UP was mixed with glucose solution (75 mM, in phosphate buffer, pH 7) in a ratio of 2:1 so as to obtain a final protein concentration of 8 mg/mL and a final glucose concentration of 25 mM in the mixture. These mixtures were heated at 90 °C for a set period. Control samples were prepared by mixing protein solutions after UP with aqueous solution in a ratio of 1:2. At the set points of 0, 1, 2, 4, 8, 16 and 24 h, the pH of each sample was measured. 2.4. Surface hydrophobicity The changes of the surface hydrophobic sites of a protein was detected by using the ANS assay [22]. ANS (2 mM) 10 μL was added to the diluted protein samples (290 μL, 20-100 μg/mL). The microplate reader (Tecan, Grödig, Austria) was used for detecting the fluorescence intensity of the mixture. The λex and λem was 390 nm and 488 nm, respectively. The fluorescent values of ANS was subtracted as the blank. The surface hydrophobicity value was determined as the initial slope when the fluorescence intensity (arbitrary unit) was plotted against the protein concentration

(μg/mL). 2.5. Particle size and zeta-potential Protein solutions were diluted to 1 mg/mL by using phosphate buffer (25 mM, pH 7) before measurement, the particle size of each sample were measured via a Nano Zsp MPT 2 dynamic light scattering (Malvern Instruments, Worcestershire, UK), the zeta-potential of each glycated protein sample was obtained by the same instrument. 2.6. o-Phthaldialdehyde (OPA) analysis The concentration of free amino groups during glycation was analyzed with OPA assay [23]. The working solution was prepared by mixing 200 mg OPA and 125 mL sodium tetraborate (10 mM) with 20% (w/w) sodium dodecyl sulfate (SDS) (12.5 mL), then 500 μL β-mercaptoethanol was added to the mixture. Afterward, a 250 mL mixture was obtained by adding 250 mL with deionized water. In each measurement, 0.6 mL OPA solution was mixed with each sample solution (30 μL), then incubated using water bath at 40 °C for 5 min. The UV absorbances at 340 nm of samples (200 μL) were measured using a microplate reader (Tecan, Grödig). The absorbances at 412 nm of samples with or without OPA were deducted. The relative percentage of free amino groups content was measured, and the absorbance of protein without glycation was considered to be 100%. The decrease kinetics of free amino groups of glycated samples was evaluated using the following equation; y = q ― Ae ―𝑘𝑡 [24] where y represents the estimated loss of free amino groups (%), q represents the

decrease in free amino groups at equilibrium (%), A is the total decrease in free amino groups (%), t is the glycation time (h), and k is the apparent rate constant (h-1). 2.7. UV absorbance analysis The UV absorbance at 294 nm represents the MRPs that are generated at intermediate state of the Maillard reaction, and absorbance at 420 nm indicates the brown polymers that are produced during advanced stages [25, 26]. The UV absorbances at 294 nm and 420 nm of each sample were detected by a microplate reader (Tecan, Grödig). Each sample was diluted with phosphate buffer (25 mM, pH 7) to appropriate concentration before measurement. 2.8. Transmission electron microscopy (TEM) Protein samples (5 μL, 1 mg/mL) were transferred to carbon-stabilized grids and stayed for dry naturally. Afterwards, the superfluous sample was absorbed by filter papers, and then 0.5% phosphotungstic acid (PTA) was used to stain the protein for 30 s. The remaining staining solution was removed, and the samples were air-dry before being observed using a TEM (2100F, JEOL, Japan). 2.9. Identification of glycated Lys and Arg residues Glycation sites of proteins were analyzed using the proteomics method [13]. Samples were digested as follows: 10 μL of each sample was blended with 60 μL ammonium bicarbonate solution (50 mM) and 15 μL DTT solution (100 mM). After that, each sample was incubated at 90 °C for 5 min, and then cooled using an ice bath. Then, 12 μL of iodoacetamide (100 mM) was added at room temperature and stood in dark for 20 min. Subsequently, 8 μL trypsin solution (0.1 mg/mL) was added and incubated at 37 °C for 18 h before analysis.

Ultimate 3000 RSLC nano system (Thermo Scientific, Fremont, CA) was connected to a RP-C18 sampling column (Acclaim PepMap, Thermo Scientific) and a RP-C18 separation column (Acclaim PepMap RSLC, Thermo Scientific) for the separation of peptide mixtures. Eluent A refers to formic acid (FA, 0.1%) and eluent B refers to acetonitrile containing 0.1% FA. Gradient elution was applied for each sample at the flow rate of 30 nL/min: 4% eluent B from 0-6min; 6-50 min, 4% to 25% eluent B; 50-58 min, from 25% to 40% eluent B; 58-61 min, from 40% to 85% eluent B; 61-66 min, 85% eluent B; and 66-73 min, from 85% to 4% elution B. Orbitrap Fusion (Thermo Scientific) was used to detect the MS/MS of peptide at the mode of positive ion. The MS and MS/MS mass were scanned from 350 to 1800 m/z and from 150 to 1800 m/z, respectively. The obtained data were performed by Proteome Discover 1.4 as follows: False discovery rate was set as 1%; carbamidomethyl (Cys) was

set

as

the

fixed

modification;

glucosylation

(Lys,

+162.14

Da),

carboxymethylation (Lys and Arg, +58.04 Da), carboxyethylation (Lys and Arg, +72.06 Da), pyrraline (Lys, +108.10 Da), glyoxal derived hydroimidazolone (G-H) (Arg, +40.02 Da) and methylglyoxal derived hydroimidazolone (MG-H) (Arg, +54.05) were set as variable modification [26, 27]. The tolerance of peptide and MS/MS were 20 ppm and 0.1 Da, respectively. 2.10. Statistical analysis The data were made and analyzed by one-way ANOVA (Dunca’s multiple-range test) in SPSS 20 software. Significant difference was defined when P<0.05. 3. Results and discussion 3.1. Structural changes in proteins during UP

Changes in surface hydrophobicity and particle sizes of each protein during 60 cycles of UP are shown in the Fig 1. Surface hydrophobicity of BSA (Fig. 1, A1) and β-CN (Fig. 1, C1) increases in the initial 25 cycles of ultrasonication and decreases after 25 and 50 cycles of UP, respectively. In contrast, surface hydrophobicity in β-Lg demonstrates an increasing trend throughout the UP, possible due to its more rigid structure which slows down its structural changes during UP [20]. These results correspond to the study of Miriani et al. [28] who reported an elevated surface hydrophobicity in the myofibrillar protein during ultrasonication. The elevated surface hydrophobicity indicates that unfolding of BSA/β-Lg and dissociation of β-CN micelles had occurred during UP, resulting in the exposure of the hydrophobic residues [29]. Analogously, glycation sites inside the folded BSA/β-Lg and micelles of β-CN could also be exposed along with their unfolding or dissociation behavior. Particle sizes in three protein samples are found to increase along with the increase in UP time (Fig. 1 A2, B2 and C2). For example, the size peak of β-CN is shown to increase from 13.8 nm to 413.7 nm after 60 cycles of UP. Ultrasonication have been reported to denature milk protein, resulting in aggregation of whey protein and β2-microglobulin [30, 31]. Therefore, the elevation in particle sizes of all proteins suggests their aggregation behavior during UP, particularly in β-CN samples. β-CN was identified as a typical intrinsic disordered protein that possesses a flexible structure and high thermal stability [21, 32]. Therefore, the cavitation effect of UP can easily change its native structure and result in severer aggregation of β-CN. Aggregation of a protein may bury partial hydrophobic residues inside the aggregates, which corresponds to the decreased surface hydrophobicity in BSA and β-CN samples after 25 and 50 cycles of UP, respectively. Correspondingly, portions of glycation sites could also be assembled into the aggregates during aggregation, which may reduce

their accessibility to reducing sugar in the subsequent glycation. The cavitation zone generated by UP in a liquid medium can lead to extremely high temperatures and pressures, and thereby producing free radicals. The thermal, mechanical and chemical effects of ultrasound have been attributed to the rapid formation and collapse of cavitation bubbles [33, 34]. These phenomena are then responsible for the structural changes of tested proteins. In this section, unfolding and aggregation of a protein were proved to have occurred spontaneously during UP. Unfolding of proteins may have exposed additional Lys and Arg, whereas aggregation behavior may had buried a portion of glycation sites inside the aggregates. Their influences may compete in the subsequent glycation. In addition, cavitation-induced hydrogen peroxide may modify the cysteine, tyrosine and phenylalanine, which may change the progress of glycation that also consist of complex radical reactions [34, 35]. These speculations were checked in the following analysis of glycation process. 3.2. UP-regulated glycation process of BSA and β-Lg The BSA, β-Lg and β-CN samples after 0, 5 or 30 cycles of UP were then reacted with glucose at 90 C for 24 h. The main glycation sites of a protein are the Lys (ε-amino group) and Arg (guanidyl group) residues [36-39]. Therefore, the content of free amino group was analyzed to detect the glycation process, and the results are shown in Fig. 2. The fitting curves of all samples properly fit the single-exponential equation (R2 > 0.947), and parameters of the fitting curves are shown in Table 1. UP obviously induced discrepancies in the decreasing trend of free amino groups in BSA and β-Lg, whereas no significant difference was found among β-CN samples (P>0.05). BSA and β-Lg samples after 5 cycles of UP are shown to lose more free amino groups in the subsequent glycation, compared with those without UP. For example, the free

amino group decreased by 25.98 % (A=24.47) and 29.29 % (A=28.69) in glycated BSA and β-Lg without UP after 24 h of incubation, which were significantly lower than that of 31.69 % (A=28.58) and 36.05% (A=33.33) in glycated BSA and β-Lg with 5 cycles of UP, respectively (Table 1). Corresponding with the glycation kinetics, the increased UV absorption of glycated BSA (0.78 ± 0.05 at 294 nm, 0.36 ± 0.02 at 420 nm) and β-Lg (0.56 ± 0.01 at 294 nm, 0.28 ± 0.01 at 420 nm) with 5 cycles of UP are significantly higher than those of the samples without UP (Table 2). Additionally, Table 2 shows lower pH values in glycated BSA and β-Lg that were subjected to 5 cycles of UP, compared with those without UP. Decrease in pH is a well-known indicator of the progress of the Maillard reaction. Therefore, these results distinctly suggested an elevation in glycation in BSA and β-Lg which experienced 5 cycles of UP. The cavitation effect during 5 cycles of UP would benefit the unfolding of BSA and β-Lg. The UP-induced unfolding behavior could expose more glycation sites, thus accounting for this elevation in glycation process [14]. In addition, ultrasonication has been reported to be capable of breaking the disulfide bond of soybean proteins [41, 42], thus could also promote the unfolding behavior of the protein samples. Interestingly, in contrast to 5 cycles of UP, 30 cycles of UP was shown to suppress the glycation of BSA and β-Lg according to the results in Fig. 2 and Table 1. In BSA samples, 30 cycles of UP significantly decreased the loss of amino group from 25.98% (A=24.47) to 21.77% (A=19.56), and significantly reduced the UV absorption by 0.17 and 0.09 at 294 and 420 nm, respectively. UV absorption values at 294 nm and 420 nm for glycated β-Lg were also found to be decreased significantly by 0.08 and 0.06 after 30 cycles of UP. According to these results, different effects of UP on the subsequent glycation of BSA and β-Lg are clearly illustrated. A mild UP (5 cycles) had unfolded

BSA and β-Lg and did not induce obvious aggregation, which exposed more glycation sites and thus promoted protein glycation. In contrast, prolonged UP (30 cycles of UP) had not only unfolded the proteins but also induced drastic aggregation, as shown in Fig. 1 A2 and B2. The aggregation behavior of protein could have buried some of the glycation sites, thus suppressing glycation of BSA and β-Lg. With the increase of UP cycles, the predominant influential factor of UP on the glycation of proteins changed from unfolding that enhanced glycation to aggregation that decreased glycation. That is why different results were obtained in samples after 5 and 30 cycles of UP. Particles sizes and morphologies of BSA and β-Lg samples are displayed in Fig. 3 (A, B) and Fig. 4 (A, B), which help to further elucidate the influence of UP on their subsequent glycation process. Particle sizes of unglycated BSA and β-Lg continue to increase during heat treatment at 90 °C. In contrast, the particle sizes of glycated BSA and β-Lg without UP increase rapidly in the initial 6 h and remain constant in the subsequent incubation. Compared with unglycated BSA and β-Lg aggregates (Fig. 4, A1 and B1), smaller globular aggregates (Fig. 4, A2) and shorter fibrils (Fig. 4, B2) are observed in glycated BSA and β-Lg, respectively, in TEM images. These results demonstrate distinct suppression of glycation on the aggregation of BSA and β-Lg at 90 °C, which can be attributed to glycation-induced change in the surface hydrophobicity, electrostatic interactions and steric hindrance [43-45]. Notably, particle sizes of ultrasonicated BSA and β-Lg samples increased into larger values than their counterparts without UP. For example, particle size of glycated β-Lg without UP increases by 23.6 nm after 1 h of glycation, which is less than the increase of 34.4 and 45.8 nm in glycated β-Lg with 5 and 30 cycles of UP, respectively. In line with this result, UP increased the diameter of glycated BSA aggregates (Fig. 4, A3 and A4), and resulted in longer fibril in glycated β-Lg aggregates (Fig. 4, B3 and B4). UP denatured

the BSA and β-Lg, rendering them less stable, and therefore aggregated faster than proteins without UP in the subsequent glycation [18, 46]. The faster aggregation process of ultrasonicated BSA and β-Lg in glycation could also had buried some glycation sites and decreased their glycation process. The changes in zeta potentials during heat treatment were shown in Fig. 5. After UP, zeta-potentials of three tested proteins increased significantly. For example, the zeta potential of control BSA (without UP) was -14.1 mV, which was significantly lower than sample with 5 cycles (19.5 mV) of UP or 30 cycles (21.3 mV) of UP. This phenomenon should be related with the unfolding behavior of tested proteins during UP, which exposed more charged residues. The negative charges in all samples increased rapidly in the first 4 h of incubation and then tended to be stable afterward. Notably, glycation process generally increased the negative charge of tested proteins, as reported in several previous studies [47, 48]. This result directly explained the suppression of glycation on the aggregation of BSA and β-Lg by increasing the electrostatic repulsion. Additionally, UP was shown to further elevate the zeta potential in BSA and β-Lg. This result reflects the influence of UP on the following aggregation of BSA/β-Lg and could be account for the different morphologies in Figure 4. 3.3. Glycation of β-CN was affected minimally by UP Notably, no significant changes in the loss of free amino groups and increase in UV absorption at 420 nm are shown in Table 1. These results suggest that glycation of β-CN was changed minimally by UP. Fig. 3C and Fig. 4C illustrate the changes in particles sizes and representative TEM images of β-CN. Particle sizes of β-CN evlolved into a size of 53.5 ± 3 nm after 2

h of heat treatment and remained constant in the following glycation. In accordance with this result, all glycated β-CN samples are found to have the similar size and morphology in Fig. 4 (C2, C3 and C4). Notably, the size of β-CN aggregates after 30 cycles of UP decreased sharply from 215.4 nm to 60.6 nm after 1 h glycation at 90 °C, and almost remained constant after that. Comparatively, the size of BSA and β-Lg aggregation after 30 cycles of UP continued to grow in the subsequent heat treatment. The instability of UP-induced β-CN aggregates could minimize the effect of UP-induced aggregation on glycation behavior. The structure of β-CN is of high flexibility and prone to aggregate as incompact micelles [21, 49]. In contrast, BSA assembles into compact amorphous aggregates and β-Lg turns to elongate into ordered fibrils [20, 50]. Therefore, the glycation sites inside the β-CN aggregates could be more accessible for glucose than BSA and β-Lg aggregates that have more well-organized structures. These structural and aggregation characteristics of β-CN rendered its glycation process to be affected minimally by UP. 3.4. UP changed the preference of Lys and Arg to be glycated Proteomics that integrates application of proteases, mass-spectrometric technique and protein database has been widely used to identify the post-transcriptional modifications of protein, including glycation, phosphorylation, acetylation, oxidation and deamination [36]. The glucosylation, carboxymethylation, carboxyethylation, pyrraline, G-H and MG-H modifications were investigated, and the results are shown in Table 3. Several representative MS/MS spectra of trypsin-digested glycated peptides were shown in Fig. 6. BSA with 5 cycles of UP is shown to have the largest number of glycation sites in the modifications of glucosylation, carboxymethylation, pyrraline and MG-H. These results are consistent with the glycation parameters

(Table1 and Table 2) and demonstrate a promotion of glycation in BSA and β-Lg by 5 cycles of UP. In addition, BSA with 30 cycles of UP were shown to have the least glycation sites in the modifications of carboxymethylation, carboxyethylation and G-H, which corresponded to their least degree of glycation. Changes in glycation sites were shown in the Fig. 7. Some Lys residues of BSA and β-Lg were suggested to be preferentially glycated [36-39]. These Lys residues were prone to be glycated than other residues, possibly due to their more exposed locations in the protein. Therefore, the changes of the glycation sites in Fig. 7 directly demonstrate changes in the preference of the Lys or Arg residues in BSA or β-Lg. The difference in the number of glycation structures, including carboxymethylation, carboxyethylation and pyrraline modifications in BSA, are shown in Fig. 7. Notably, samples without UP shared more glycation sites with the samples with 5 cycles of UP, compared with the samples with 30 cycles of UP. For example, BSA without UP shared 46 glucosylated, 24 carboxymethylated and 5 carboxyethylated Lys in common with BSA with 5 cycles of UP, while it shared 45 glucosylated, 21 carboxymethylated and 4 carboxyethylated Lys in common with BSA with 30 cycles of UP. This result shows that longer UP leading to greater changes in the preference of Arg or Lys or residues in BSA. Additionally, 9 and 2 unique carboxymethylated Lys in BSA with 5 and 30 cycles of UP are shown, respectively; 2, 4 and 3 unique Lys are found to be glycated into pyrraline in BSA with 0, 5 and 30 cycles of UP. These results further demonstrate the influence of UP on the preference of Lys residues during the following glycation. The instantaneous extreme high temperature and pressure being generated during UP lead to conformational changes in selected proteins, as shown in Figure 1. These structural changes should have largely changed the locations of Lys and Arg, thus subsequently alter their preference to be glycated.

The location of these residues in BSA could be more exposed and then be carboxymethylated or glycated into pyrraline. 4. Conclusion In this work, UP induced unfolding and aggregation behavior of BSA and β-Lg samples, which change the accessibility of glycation sites. With the increase of UP cycles, the predominant influential factor of UP on the glycation of proteins changed from unfolding that enhanced glycation to aggregation that decreased glycation. That’s why different results were obtained in samples after 5 and 30 cycles of UP. Therefore, 5 cycles of UP enhanced the glycation of BSA and β-Lg, and 30 cycles of UP suppressed their glycation process. UP has the minimal influence on glycation process of β-CN, due to its disordered structure. The preference of Lys and Arg to be glycated in BSA and β-Lg was also largely regulated based on proteomics analysis. Accordingly, UP is a promising method to be used to up-regulate or down-regulate the glycation behavior of protein depending on both the UP duration and different structures of the proteins. Compared with conventional methods, including adjusting salt concentration, pH and metal ions, UP avoids the use of chemicals and could attract more interests in the future. Acknowledgments National Key R&D Program of China (No. 2016YFD0400203), the National Natural Science Foundation of China (No. 31671961 & 31701727), Key Projects of Guangdong Natural Science Foundation (No. 2017A030311021), Key Project of Guangzhou S&T Program (No. 201904020005), Start-up Fund of Natural Sciences Foundation of Guangdong Province (No. 2015A030310189), and the CIUC of Zhongshan (No. 2016C1013).

REFERENCES [1] M. Hellwig, T. Henle, Baking, Ageing, Diabetes: A short history of the Maillard reaction, Angew. Chem. Int. Ed. 53 (2014) 10316-10329. [2] M.W. Poulsen, R.V. Hedegaard, J.M. Andersen, B. de Courten, S. Bugel, J. Nielsen, L.H. Skibsted, L.O. Dragsted, Advanced glycation endproducts in food and their effects on health, Food Chem. Toxicol. 60 (2013) 10-37. [3] C.M. Oliver, L.D. Melton, R.A. Stanley, Creating proteins with novel functionality via the Maillard reaction: A review, Crit. Rev. Food Sci. Nutr. 46 (2006) 337-350. [4] N. Yamabe, Y.-J. Kim, S. Lee, E.-J. Cho, S.-H. Park, J. Ham, H.Y. Kim, K.S. Kang, Increase in antioxidant and anticancer effects of ginsenoside Re-lysine mixture by Maillard reaction, Food Chem. 138 (2013) 876-883. [5] Q. Wang, B. Ismail, Effect of Maillard-induced glycosylation on the nutritional quality, solubility, thermal stability and molecular configuration of whey protein, Int. Dairy J. 25 (2012) 112-122. [6] S. Drusch, S. Berg, M. Scampicchio, Y. Serfert, V. Somoza, S. Mannino, K. Schwarz, Role of glycated caseinate in stabilisation of microencapsulated lipophilic functional ingredients, Food Hydrocolloid. 23 (2009) 942-948. [7] Y. Zhang, C. Tan, S. Abbas, K. Eric, S. Xia, X. Zhang, Modified SPI improves the emulsion properties and oxidative stability of fish oil microcapsules, Food Hydrocolloid. 51 (2015) 108-117. [8] J. Qiu, Q. Zheng, L. Fang, Y. Wang, M. Min, C. Shen, Z. Tong, C. Xiong, Preparation and characterization of casein-carrageenan conjugates and self-assembled microcapsules for encapsulation of red pigment from paprika, Carbohydr. Polym. 196 (2018) 322-331. [9] M. Kijewska, K. Radziszewska, M. Kielmas, P. Stefanowicz, Z, Nonenzymatic modification of ubiquitin under high-pressure and-temperature treatment: mass spectrometric studies, J. Agric.

Food. Chem. 63 (2015), 614-619. [10] D.T. Ramonaityte, M. Kersience, A. Adams, K.A. Tehrani, N. De Kimpe, The interaction of metal ions with Maillard reaction products in a lactose-glycine model system, Food Res. Int. 42 (2009) 331-336. [11] L.N. Bell, Maillard reaction as influenced by buffer type and concentration, Food Chem. 59 (1997) 143-147. [12] H. Wang, Z.-C. Tu, G.-X. Liu, C.-M. Liu, X.-Q. Huang, H. Xiao, Comparison of glycation in conventionally and microwave-heated ovalbumin by high resolution mass spectrometry, Food Chem. 141 (2013) 985-991. [13] X. Huang, Z. Tu, H. Wang, Q. Zhang, Y. Shi, H. Xiao, Increase of ovalbumin glycation by the Maillard reaction after disruption of the disulfide bridge evaluated by liquid chromatography and high resolution mass spectrometry, J. Agric. Food. Chem. 61 (2013) 2253-2262. [14] Q. Zhang, Z. Tu, H. Wang, X. Huang, Y. Shi, X. Sha, H. Xiao, Improved glycation after ultrasonic pretreatment revealed by high-performance liquid chromatography-linear ion trap/orbitrap high-resolution mass spectrometry, J. Agric. Food. Chem. 62 (2014) 2522-2530. [15] W.-H. Shi, W.-W. Sun, S.-J. Yu, M.-M. Zhao, Study on the characteristic of bovine serum albumin-glucose model system, treated by ultrasonic, Food Res. Int. 43 (2010) 2115-2118. [16] B. Zhang, Y.J. Chi, B. Li, Effect of ultrasound treatment on the wet heating Maillard reaction between beta-conglycinin and maltodextrin and on the emulsifying properties of conjugates, Eur. Food Res. Technol. 238 (2014) 129-138. [17] W. Bi, W. Ge, X. Li, L. Du, G. Zhao, H. Wang, & X. Qu, Effects of ultrasonic pretreatment and glycosylation on functional properties of casein grafted with glucose. J. Food Process Preserv. 41 (2016) e13177 [18] D. Zhao, L. Li, D. Xu, B. Sheng, D. Qin, J. Chen, B. Li, X. Zhang, Application of ultrasound pretreatment and glycation in regulating the heat-induced amyloid-like aggregation of

beta-lactoglobulin, Food Hydrocolloids 80 (2018) 122-129. [19] M. Bhattacharya, N. Jain, S. Mukhopadhyay, Insights into the mechanism of aggregation and fibril formation from bovine serum albumin, J. Phys. Chem. B 115 (2011) 4195-4205. [20] B.Y. Qin, M.C. Bewley, L.K. Creamer, H.M. Baker, E.N. Baker, G.B. Jameson, Structural basis of the tanford transition of bovine beta-lactoglobulin, Biochemistry 37 (1998) 14014-14023. [21] S. Perticaroli, J.D. Nickels, G. Ehlers, E. Mamontov, A.P. Sokoov, Dynamics and rigidity in an intrinsically disordered protein, beta-casein, J. Phys. Chem. B 118 (2014) 7317-7326. [22] C.A. Haskard, E.C.Y. Li-Chan, Hydrophobicity of bovine serum albumin and ovalbumin determined using uncharged (PRODAN) and anionic (ANS(-)) fluorescent probes, J. Agric. Food. Chem. 46 (1998) 2671-2677. [23] P.M. Nielsen, D. Petersen, C. Dambmann, Improved method for determining food protein degree of hydrolysis, J. Food Sci. 66 (2001) 642-646. [24] Y. Li, S. Jongberg, M.L. Andersen, M.J. Davies, M.N. Lund, Quinone-induced protein modifications: Kinetic preference for reaction of 1, 2-benzoquinones with thiol groups in proteins, Free Radical Bio. Med. 97 (2016) 148-157. [25] F. Liu, C. Sun, D. Wang, F. Yuan, Y. Gao, Glycosylation improves the functional characteristics of chlorogenic acid-lactoferrin conjugate, Rsc Advances 5 (2015) 78215-78228. [26] D. Zhao, L. Li, T.T. Le, L.B. Larsen, G. Su, Y. Liang, B. Li, Digestibility of glyoxal-glycated beta-casein and beta-lactoglobulin and distribution of peptide-bound advanced glycation end products in gastrointestinal Digests, J. Agric. Food. Chem. 65 (2017) 5778-5788. [27] D. Zhao, T.T. Le, L.B. Larsen, L. Li, D. Qin, G. Su, B. Li, Effect of glycation derived from alpha-dicarbonyl compounds on the in vitro digestibility of beta-casein and beta-lactoglobulin: A model study with glyoxal, methylglyoxal and butanedione, Food Res. Int. 102 (2017) 313-322. [28] M. Miriani, M. Keerati-u-rai, M. Corredig, S. Iametti, F. Bonomi, Denaturation of soy proteins in solution and at the oil-water interface: A fluorescence study, Food Hydrocolloids 25

(2011) 620-626. [29] I. Gulseren, D. Guezey, B.D. Bruce, J. Weiss, Structural and functional changes in ultrasonicated bovine serum albumin solutions, Ultrason. Sonochem. 14 (2007) 173-183. [30] A. Shanmugam, J. Chandrapala, M. Ashokkumar, The effect of ultrasound on the physical and functional properties of skim milk, Innov. Food Sci. Emerg. Technol. 16 (2012) 251-258. [31] M. So, H. Yagi, K. Sakurai, H. Ogi, H. Naiki, Y. Goto, Ultrasonication-dependent acceleration of amyloid fibril formation, J. Mol. Biol. 412 (2011) 568-577. [32] A.M. Sulewska, K. Olsen, J.C. Sorensen, L.H. Ogendal, Chaperone-like activity of beta-casein and its effect on residual in vitro activity of horseradish peroxidase, Int. J. Food Sci. Technol. 49 (2014) 2538-2545. [33] D.C. Kang, Y.H. Zou, Y.P. Cheng, L.J. Xing, G.H. Zhou, W.G. Zhang, Effects of power ultrasound on oxidation and structure of beef proteins during curing processing. Ultrason. Sonochem. 33 (2016), 47-53. [34] İ. Gülseren, D. Güzey, B. D. Bruce, J. Weiss, Structural and functional changes in ultrasonicated bovine serum albumin solutions. Ultrason. Sonochem. 14 (2007), 173-183. [35] F.U. Shanlin, R. Stocker, M.J. Davies, Biochemistry and pathology of radical-mediated protein oxidation. Biochem. J. 324(1997), 1-18. [36] S. Milkovska-Stamenova, R. Hoffmann, Identification and quantification of bovine protein lactosylation sites in different milk products, J. Proteomics 134 (2016) 112-126. [37] G. Renzone, S. Arena, A. Scaloni, Proteomic characterization of intermediate and advanced glycation end-products in commercial milk samples, J. Proteomics 117 (2015) 12-23. [38] S. Arena, G. Renzone, G. Novi, & A. Scaloni, Redox proteomics of fat globules unveils broad protein lactosylation and compositional changes in milk samples subjected to various technological procedures. Journal of Proteomics 74 (2011) 2453-2475.

[39] S. Arena, G. Renzone, G. Novi, A. Paffetti, G. Bernardini, A. Santucci, & A. Scaloni, Modern proteomic methodologies for the characterization of lactosylation protein targets in milk. Proteomics 10 (2010) 3414-3434. [40] K.S. Suslick, Y. Didenko, M.M. Fang, T. Hyeon, K.J. Kolbeck, W.B. McNamara III, M.M. Mdleleni, M. Wong, Acoustic cavitation and its chemical consequences, Philosophical Transactions of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences 357 (1999) 335-353. [41] H. Hu, J. Wu, E.C.Y. Li-Chan, L. Zhu, F. Zhang, X. Xu, G. Fan, L. Wang, X. Huang, S. Pan, Effects of ultrasound on structural and physical properties of soy protein isolate (SPI) dispersions, Food Hydrocolloids 30 (2013) 647-655. [42] H. Hu, X. Fan, Z. Zhou, X. Xu, G. Fan, L. Wang, X. Huang, S. Pan, L. Zhu, Acid-induced gelation behavior of soybean protein isolate with high intensity ultrasonic pre-treatments, Ultrason. Sonochem. 20 (2013) 187-195. [43] C. Wang, X. Ren, Y. Su, Y. Yang, Application of glycation in regulating the heat-Induced nanoparticles of egg white protein. Nanomaterials, 8 (2018), 943. [44] G. Liu, Q. Zhong, Glycation of whey protein to provide steric hindrance against thermal aggregation, J. Agric. Food. Chem. 60 (2012) 9754-9762. [45] M.D.S. Pinto, S. Bouhallab, A.F. De Carvalho, G. Henry, J.-L. Putaux, J. Leonil, Glucose slows down the heat-induced aggregation of beta-lactoglobulin at neutral pH, J. Agric. Food. Chem. 60 (2012) 214-219. [46] Z. Zhang, J.M. Regenstein, P. Zhou, Y. Yang, Effects of high intensity ultrasound modification on physicochemical property and water in myofibrillar protein gel, Ultrason. Sonochem. 34 (2017) 960-967. [47] A. Achouri, J.I. Boye, V.A. Yaylayan, F.K. Yeboah, Functional properties of glycated soy 11S glycinin, J. of Food Sci. 70 (2005), C269-C274.

[48] G. Liu, Q. Zhong, Thermal aggregation properties of whey protein glycated with various saccharides. Food Hydrocolloid. 32 (2013), 87-96. [49] C.G. de Kruif, R. Tuinier, C. Holt, P.A. Timmins, H.S. Rollema, Physicochemical study of kappa- and beta-casein dispersions and the effect of cross-linking by transglutaminase, Langmuir 18 (2002) 4885-4891. [50] A. Kroes-Nijboer, P. Venema, J. Bouman, E. van der Linden, Influence of protein hydrolysis on the growth kinetics of beta-Ig fibrils, Langmuir 27 (2011) 5753-5761.

Figure captions Fig. 1. Change in surface hydrophobicity (A1-C1) and particle size (A2-C2) of BSA (A), β-Lg (B), and β-CN (C) in 60 cycles of UP. The error bars indicate the standard deviation of triplicates. The black, red, green, blue, purple and orange lines refer to size distribution of each protein after 0, 5, 10, 15, 30 and 60 cycles of UP. Fig. 2. UP-induced changes of free amino groups during 24 h of glucose-derived glycation of BSA (A), β-Lg (B), and β-CN (C). The symbol colored black indicates glycated sample without UP; the symbol colored red indicates glycated sample with 5 cycles of UP and symbol colored blue indicates glycated sample with 30 cycles of UP. The error bars indicate the standard deviation of triplicates. Solid lines represent the fitting curves by single exponential equation in Originpro 8.0. Fig. 3. UP-induced changes in particles size of the BSA (A), β-Lg (B), and β-CN (C) with and without UP during 24 h of heat treatment. The symbol colored black indicates control sample; symbol colored red indicates glycated sample without UP; symbol colored blue indicates glycated sample with 5 cycles of UP and symbol colored cyan indicates glycated sample with 30 cycles of UP. The error bars indicate the standard deviation of triplicates. Fig. 4. Representative TEM image of unglycated and glycated sample of BSA (A), β-Lg (B), and β-CN (C) after 24 h of incubation: A1, B1 and C1 indicate samples without UP and glycation; A2, B2 and C2 indicates glycated samples without UP; A3, B3 and C3 indicates glycated samples with 5 cycles of UP; and A4, B4 and C4 indicates glycated samples with 30 cycles of UP. Fig. 5. Changes in zeta potential of the BSA (A), β-Lg (B), and β-CN (C) with and

without UP during heat treatment. The symbol colored black indicates control sample; symbol colored red indicates glycated sample without UP; symbol colored blue indicates glycated sample with 5 cycles of UP and symbol colored cyan indicates glycated sample with 30 cycles of UP. The error bars indicate the standard deviation of triplicates. Fig. 6. MS/MS spectrum of trypsin-digested glycated peptides. A refers to glucosylated (K) peptide; B refers to carboxymethylated (K) peptide; C refers to carboxyethylated peptide; D and E refer to peptides with pyrraline (K) and MG-H1 (R) structure, respectively. Fig. 7. UP-induced changes in glycation sites of BSA (A) and β-Lg (B), as shown by Venn diagrams.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Table 1. Kinetics parameters for the fitting curves in Fig. 2, as evaluated by the single exponential equation. Sample ID

A

k (h -1×10-3)

adj R2

BSA+ Glu

24.47

0.1571

0.9518

BSA (5) + Glu

28.58

0.1551

0.9508

BSA (30) + Glu

19.56

0.1441

0.9736

β-Lg + Glu

28.69

0.1383

0.9887

β-Lg (5) + Glu

33.33

0.2887

0.9773

β-Lg (30) + Glu

29.59

0.1668

0.9967

β-CN + Glu

35.95

0.1018

0.9563

β-CN (5) + Glu

35.26

0.0888

0.9639

β-CN (30) + Glu

34.81

0.1458

0.9474

les

Table 2. The pH, decreases in free amino groups, and increase in UV absorption at 294 nm and 420 nm in glycated (for 24 h) samples with 0, 5 and 30 cycles of UP. The error bars indicate the standard deviation of triplicates. Sample

Loss of (–NH2) (%)

UV (294 nm)

UV (420 nm)

pH

BSA + Glu

25.98 ± 1.52 b1

0.64 ± 0.03 b1

0.29 ± 0.02 b1

6.28 ± 0.07 a1b1

BSA (5) + Glu

31.69 ± 2.41 a1

0.78 ± 0.05 a1

0.36 ± 0.02 a1

6.12 ± 0.10 b1

BSA (30) + Glu

21.77 ± 1.18 c1

0.47 ± 0.03 c1

0.20 ± 0.01 c1

6.35 ± 0.08 a1

β-Lg + Glu

29.29 ± 2.22 b2

0.53 ± 0.02 a2

0.25 ± 0.01 a2

6.17 ± 0.05 a2b2

β-Lg (5) + Glu

36.05 ± 2.35 a2

0.56 ± 0.01 a2

0.28 ± 0.01 a2

6.10 ± 0.09 b2

β-Lg (30) + Glu

28.86 ± 1.30 b2

0.45 ± 0.02 b2

0.19 ± 0.01b2

6.24 ± 0.05 a2

β-CN+ Glu

35.10 ± 0.83 a3

0.42 ± 0.02 b3

0.21 ± 0.01 a3

6.37 ± 0.06 a3

β-CN (5) + Glu

35.84 ± 0.88 a3

0.44 ± 0.03 a3, b3

0.23 ± 0.02 a3

6.40 ± 0.12 a3

β-CN (30) + Glu

33.78 ± 1.85 a3

0.49 ± 0.01 a3

0.23 ± 0.01 a3

6.34 ± 0.11 a3

Table 3. UP-induced changes in the number of identified glycation sites of BSA and β-Lg after 24 h of glycation. Number of Glycation sites Glucosylation

Carboxymethylation

Carboxyethylation

Pyrraline

G-H1

MG-H1

(K, +C6H10O5)

(K, +C2O2H2)

(K, +C3O3H4)

(K, +C6O2H4)

(R, +C2O)

5

—

—

—

—

—

+ Glu

48

28

10

6

2

4

(5) + Glu

52

38

10

9

2

7

(30) + Glu

50

25

9

10

1

5

3

—

—

—

—

+ Glu

11

7

2

6

0

1

(5) + Glu

13

10

3

7

1

1

(30) + Glu

12

10

3

7

1

0

(R, + C3OH

Graphical Abstract

Highlights 1. Ultrasound pretreatment (UP) induced unfolding and aggregation behavior of bovine serum albumin (BSA) and β-lactoglobulin (β-Lg). 2. The structure change of BSA and β-Lg induced by UP could change the preference of Lys residues to be glycated. 3. UP has the minimal influence on glycation process of β-casein, due to its disordered structure. 4. UP is a promising method to be used to up-regulate or down-regulate the glycation behavior of protein .