Ultrasound-assisted construction of pea protein isolate-folic acid covalent complex as self-assembled nanocarrier: Enhancing the stability, precise release property, and bioaccessibility of curcumin

Graphical abstract

Highlight

• •Applying an efficient and green ultrasound-assisted grafting method to prepare pea protein isolate folic acid conjugates.
• •The covalent complexation of folic acid improved the binding affinity of pea protein isolate for curcumin.
• •Encapsulation within pea protein isolate-folic acid realized the gastrointestinal controlled release for curcumin.
• •Pea protein isolate-folic acid-curcumin complex showed excellent stability, cellular uptake efficiency, and bioaccessibility.

Abstract

Curcumin (Cur) is a hydrophobic phenolic compound with superior biological activity, but the limited water solubility, chemical instability, and poor bioaccessibility of curcumin restrict its application. This study developed pea protein isolate (PPI)-folic acid (FA) covalent complexes as nanocarriers for the encapsulation, protection, and precise delivery of curcumin. The results of SDS-PAGE, XPS, and FTIR demonstrated that covalent complexation between PPI and FA was formed through an ultrasound-assisted free radical grafting method. Notably, ultrasonication for 15 min (PPI-FA-US15) achieved a higher FA loading capacity (44.78 ± 0.37 µg/mg) than the traditional free radical grafting for 24 h (36.45 ± 0.43 µg/mg). The ultrasonic treatment and covalent complexation of FA greatly improved the solubility of PPI (from 61.34 ± 0.67% to 97.02 ± 0.88%). Curcumin was efficiently encapsulated within the core of PPI-FA-US15 conjugates to form PPI-FA15-Cur nanocomplexes by hydrogen bonding and hydrophobic interaction. The complexation of FA improved the binding affinity of PPI for curcumin. PPI-FA15-Cur demonstrated a 1.97-fold superior loading capacity for curcumin compared to PPI-Cur. Meanwhile, PPI-FA15-Cur displayed a spherical morphology, and maintained a stable particle size distribution during 77-d storage. The heating stability, photochemical stability and antioxidant property of curcumin were significantly improved. Furthermore, in vitro digestion studies demonstrated that PPI-FA15-Cur showed a sustained release of curcumin. The bioaccessibility of curcumin within PPI-FA15-Cur was about 4.75-fold and 1.11-fold higher than that of free curcumin and PPI-Cur, respectively. This study provides a promising approach for developing plant protein-based carriers that enable precise delivery of curcumin in functional foods.

1. Introduction

In the field of functional food, active substances from plants have garnered significant attention due to their environmental friendliness, health benefits, and sustainability [1], [2]. Curcumin (Cur) is a typical hydrophobic flavonoid extracted from turmeric, exhibiting potent antibacterial, antioxidant, and antitumor properties [3]. The application of curcumin, particularly in fat-free functional food systems, is severely limited by its significant hydrophobicity and inherent instability. These disadvantages particularly affect the absorption and uptake of curcumin. In fact, curcumin is primarily absorbed in the small intestine, while most ingested curcumin is already degraded in the gastric environment before reaching the intestines. Therefore, developing advanced delivery strategies is essential to improve the precise release property and absorption efficiency of curcumin, thereby enhancing its bioavailability and biological activity [4].

Due to the favorable biocompatibility, protein nanoparticles are one of the desirable vehicles for delivering curcumin. The small size of protein-based nanocarriers often enables faster penetration through the mucus layer and endothelial cells, which could greatly improve the bioaccessibility of curcumin [5]. For instance, walnut protein isolate [6], pea protein isolate [7], and casein [8] have been used to form nanocomplexes with curcumin via hydrophobic interactions, enhancing the antioxidant property and bioaccessibility of curcumin. In comparison to animal proteins, plant proteins are characterized by abundant resources, cost-effectiveness, and reduced allergenic potential [9], [10]. They also exhibit excellent encapsulation properties, enabling effective delivery of bioactive components [9]. Pea protein isolate (PPI), a high-quality plant protein, has garnered widespread attention due to its high yield, low-cost, exceptional nutritional value, and excellent biocompatibility [10], [11]. As a delivery vehicle, native PPI typically presents limited loading capacity and undesirable solubility, as well as a lack of precise release property.

Recently, growing evidence suggests that the properties, like surface charge, loading capacity, and targeted release, of protein-based nanocarriers can be precisely modulated by the complexation with some small-molecule compounds, such as folic acid (FA) [12], apigenin [13], and tannic acid [14]. FA, composed of glutamic acid, p-amino-benzoic acid, and pterin, is an essential water-soluble vitamin for humans [15]. The carboxyl groups (–COOH) of FA were conjugated with the amino groups (–NH2) of zein, thereby enhancing the dispersibility and stability of zein [12]. Meanwhile, FA is a key component in the construction of precise delivery systems, due to the fact that the pterin of FA could specially bind with the folate receptors on the (cancer or inflammatory) cells. The cellular uptake of citrus essential oil was greatly enhanced after encapsulation within the zein-FA particles [12].

The complexation of PPI with FA can occur in two fundamental forms: stable covalent bonding and reversible, transient non-covalent interactions. The latter is typically driven by transient forces like hydrophobic effects and electrostatic interactions, leading to spontaneous and reversible associations. In contrast, covalent interactions exhibit more persistent conjugation capabilities and superior stability. As delivery vehicles, the covalent complexes generally demonstrate superior oxidation stability and better protection for encapsulated polyphenols than non-covalent complexes [13]. The traditional method for preparing protein-FA covalent complexes is the carbodiimide catalysis [16]. Nevertheless, carbodiimide shows poor biosafety, and must be avoided in food applications. Another method commonly used for fabricating protein-based conjugates is the free radical grafting method, which utilizes H₂O₂ and ascorbic acid as the reaction medium [17]. Whey protein-epigallocatechin gallate conjugates have been successfully fabricated by the free radical grafting method. The covalent conjugates-entrapped bioactives showed better stability than those encapsulated within native protein [18]. Similarly, the covalent conjugation between the –NH₂ group in chitosan and the –COOH in ferulic acid has been achieved via the free radical grafting method [19]. Despite the applicability of the radical grafting method in the food industry, it suffers from drawbacks such as prolonged reaction times and low grafting degrees.

Ultrasonication is a non-thermal and eco-friendly physical modification technology. It regulates the conformation changes and aggregation states of native proteins via the synergistic effects of microfluidic effects, shear stress, and cavitation [20], [21]. Ultrasonic treatment can improve the solubility of PPI and enhance the stability of the nanocomplexes. Additionally, ultrasonic technology can mechanically disrupt the hydrogen bonds and hydrophobic interactions, inducing protein structural rearrangement to expose active sites, thereby enhancing the ability of protein to form covalent and non-covalent bonds with other molecules [22], [23]. Simultaneously, ultrasonic treatment can enhance the frequency of collisions between proteins and small molecules, accelerating the free radical grafting reaction [24]. The duration of ultrasonic treatment has great effects on the properties, yield, and structure of the conjugates [24]. Excessively long ultrasonic treatment leads to protein aggregation and decomposition of small molecules [22]. Therefore, the conjugation efficiency of PPI and FA via radical grafting can be significantly enhanced under appropriate ultrasonication duration. The resulting PPI-FA conjugate can be utilized to construct a highly efficient and precise delivery system for curcumin.

This study aims to develop an ultrasonic-assisted free radical grafting method to fabricate PPI-FA conjugates to effectively deliver curcumin. The effect of ultrasonic duration on the covalent conjugation between PPI and FA was analyzed using SDS-PAGE, FT-IR, XPS, and far-UV circular dichroism spectroscopy. The binding interaction between curcumin and PPI-FA conjugates, as well as the microscopic morphology of the nanocomplexes, was analyzed using fluorescence quenching, three-dimensional fluorescence, AFM, and TEM. Subsequently, the loading capacity, thermal stability and photostability of curcumin were determined. Finally, the improvements in the precise release characteristics, bioaccessibility, and in vitro uptake properties of encapsulated curcumin were investigated. Overall, this study provides important insights for developing efficient delivery systems and opening new avenues for precision nutrition.

2. Materials and methods

2.1. Materials

White pea grains, sourced from a local supermarket in Hunan, China, were subjected to dehulling and milling to obtain pea flour. The resulting flour was soaked in hexane, and defatted via Soxhlet extraction to yield defatted pea flour. Folic acid (FA, purity > 97%), curcumin (Cur, purity > 98%), 2,2-diphenyl-1-picrylhydrazyl (DPPH), 1-anilino-8-naphthalenesulfonate (ANS), KBr (spectrum pure), 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS), and o-phthalaldehyde (OPA, purity ≥ 98%) were purchased from Solarbio Science & Technology Ltd. (Beijing, China). The CCK-8 Cell Counting Kit was supplied from Vazyme Biotech Co., Ltd. (Nanjing, China). All other analytical-grade chemical reagents were supplied by Aladdin (Shanghai, China).

2.2. Fabrication of pea protein isolate-folic acid (PPI-FA) conjugates

Following the previous method, PPI was extracted from the defatted flour [25], [26]. The obtained PPI was dissolved in deionized water (12 mg/mL) and stirred for 2 h. The solution was then allowed to hydrate fully for 6 h, followed by centrifugation at 6500 g for 20 min. The resulting supernatant was collected. Then, the PPI solution was conditioned to a pH of 9.0. H₂O₂ and ascorbic acid were added to the solution to achieve concentrations of 50 mM and 2.5 mg/mL, respectively. After stirring at 500 rpm for 2 h, FA (1 mg/mL) was incorporated. The solution was subjected to different ultrasonication times by an ultrasound device (Scientz Sonifier II D, Xin Zhi Ultrasonics Co., Ltd., Ningbo, China). The ultrasonication time was performed for 5, 15, 30, 45 and 60 min, utilizing 450 W ultrasonic intensity with temperature maintained at 40 °C. After ultrasonication, the solution was dialyzed with a dialysis bag (5000 Da) in a 4 °C refrigerator. After 48 h, the solution was lyophilized to obtain PPI-FA complexes. Based on ultrasonic time, the prepared PPI-FA complexes were named as PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60, respectively. Additionally, to compare the differences between ultrasound-assisted and traditional free radical grafting reaction, a control group was prepared by stirring for 24 h after adding FA (without ultrasound treatment), with all other steps consistent with the PPI-FA-US15, named as PPI-FA-9C.

2.3. Characterization of PPI-FA conjugates

2.3.1. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE)

PPI, PPI-FA-9C, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60 (3 mg/mL protein concentration) were combined with 5 × sample buffer [27]. After incubating the samples in boiling water for 10 min, they were loaded into a gel composed of a 5% concentrated gel and 15% separator gel. After electrophoresis, Coomassie Brilliant Blue G-250 was applied to color the gel for 5 min.

2.3.2. Free amino group (–NH₂) and free sulfhydryl (−SH) contents

The content of free –NH₂ in PPI, PPI-FA-9C, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60 was determined by the OPA assay. 200 μL of samples was blended with 4 mL of OPA solution. After incubation at 35 °C for 2 min, the absorbance of the mixture at 340 nm was determined [12].

To determine the amount of free sulfhydryl groups (−SH), 100 μL of PPI/PPI-FA complexes was mixed with 5 mL of Tris-glycine buffer. Then, the solution was mixed with dithio-bis-nitrobenzoic acid (4 mg/mL) for 60 min [12]. The −SH content was measured by the following formula:

$$-\mathrm{S}\mathrm{H}(\mu \mathrm{m}\mathrm{o}\mathrm{l}/\mathrm{g})=\frac{A_{412}\times D\times 73.53}{C}$$ (1)

Herein, C is protein concentration (mg/mL), A₄₁₂ is absorbance values, and D is dilution factors.

2.3.3. Solubility

PPI, PPI-FA-9C, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60 (15 mg/mL) were solubilized in deionized water for 2 h, and then centrifuged (6500 g, 20 min). The supernatant was mixed Coomassie Brilliant Blue solution. After 5 min, the absorbance at 595 nm was determined to calculate the soluble protein concentration (mg/mL) in the solution. The percentage of protein in the supernatant relative to the total protein was the soluble protein content (%).

2.3.4. Loading capacity of FA

A Multiskan Go 1510 microplate reader (Thermo Fisher Scientific, Massachusetts, USA) was employed to collect the absorbance of PPI, PPI-FA-9C, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60 at 360 nm to calculated the FA content [28]. The FA load capacity (LC, μg/mg) was obtained by calculating the ratio of FA content to the total mass of the complex.

2.3.5. UV–vis absorbance spectra

A UV–visible spectrometer (UV-7500, Shanghai, China) was used to scan 200 μL of PPI, PPI-FA-9C, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60 placed in a microcuvette in the range of 200–600 nm.

2.3.6. Fourier transform infrared spectroscopy (FT-IR)

A Nicolet 6700 spectrometer (Thermo Electron Co., Madison, USA) was employed to measure the FT-IR spectra of free curcumin, free FA, PPI, PPI-FA-9C, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60 [29]. KBr was mixed with curcumin, FA, PPI or PPI-FA complexes (100:1, w/w), and pressed into transparent pellets. The spectral range acquired was 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹.

2.3.7. X-ray photoelectron spectroscopy (XPS)

The interaction between PPI and FA were analyzed by XPS spectroscopy following previous reported method [29]. The spectra of samples were obtained using a Kratos Axis spectrometer (Ultra Kratos Analytical, Manchester, UK) equipped with a 180° hemispherical electron spectrometer and a monochromatic Al K_α source.

2.3.8. UV light stability of FA

To investigate the UV stability of free FA and encapsulated FA, free FA and PPI-FA-US15 were diluted to a FA concentration of 10 μM and then exposed to UV radiation at 365 nm for 240 min. The fluorescence intensity of the samples was determined at excitation wavelength 360 nm and emission wavelength 445 nm. After 240 min of UV irradiation, FA background values were subtracted from all samples. Subsequently, normalization was performed using the fluorescence intensity of 10 μM free FA under identical conditions as the ref. [30]. Additionally, the FA content in samples after 240 min of UV irradiation was determined by a HPLC system (Waters, Milford, MA, USA). FA retention rate was calculated as the ratio of FA content after 240 min of UV irradiation to the original FA content.

2.3.9. Surface hydrophobicity (H₀)

The ANS was employed as a fluorescent probe to measure the H₀ of PPI and PPI-FA-US15. The samples were mixed with ANS (8 mmol/L) at a 40:1 vol ratio [31]. After 15 min, the fluorescence intensity of the mixture was determined at an excitation wavelength of 390 nm and an emission wavelength of 484 nm. A curve was plotted for each sample showing the relationship between fluorescence intensity and corresponding concentration. H₀ was the slope of the linear regression equation.

2.3.10. Intrinsic fluorescence spectroscopy analysis

The intrinsic fluorescence spectra of PPI and PPI-FA complexes were measured by a fluorescence spectrophotometer (F-4500, Hitachi, Japan) [32]. The emission spectra were recorded in the range of 300 to 460 nm under excitation conditions of 280 nm wavelength and 5 nm slit width.

2.3.11. Far-UV circular dichroism (CD) spectroscopy analysis

A Chirascan spectrophotometer (JASCO, Tokyo, Japan) was applied to analyze the far-UV CD spectra of PPI and PPI-FA complexes at 25 ℃ following previous reported method [33]. The range of scanning was set between 190 nm and 260 nm.

2.3.12. Particle size, polydispersity index (PDI), and zeta-potential

A Malvern Zetasizer 2000 (Malvern Instruments, Southborough, MA, US) was applied to measure the particle size, PDI, and zeta-potential of PPI and PPI-FA complexes by the dynamic laser scattering (DLS) [26].

2.4. Fabrication of pea protein isolate-folic acid-curcumin (PPI-FA-Cur) nanocomplexes

The native PPI and PPI-FA-US15 complexes (20 mg/mL) were hydrated in deionized water. After 6 h, the solutions were conditioned to pH 12.0 and agitated for 1 h. Curcumin was dissolved at various concentrations of 0.5, 1,1.5, 2, 2.5, and 3 mg/mL in pH 12.0 deionized water, and then mixed with PPI/PPI-FA-US15 solution at a volume of 1:1. The samples contained curcumin at final concentrations of 0.25, 0.5, 0.75, 1, 1.25, and 1.5 mg/mL. Then, the mixtures were acidified to pH 7.0. After stirred for 2 h, the solutions were centrifuged (6500 g, 20 min), and the supernatant was collected as PPI-FA-Cur nanocomplexes. The obtained nanoparticles by PPI and curcumin were named as PPI-Cur_0.25, PPI-Cur_0.5, PPI-Cur_0.75, PPI-Cur₁, PPI-Cur_1.25, PPI-Cur_1.5, respectively. The obtained nanoparticles by PPI-FA-US15 and curcumin were named as PPI-FA15-Cur_0.25, PPI-FA15-Cur_0.5, PPI-FA15-Cur_0.75, PPI-FA15-Cur₁, PPI-FA15-Cur_1.25, PPI-FA15-Cur_1.5, respectively.

2.5. Characterization of PPI-Cur/PPI-FA15-Cur nanocomplexes

2.5.1. Loading capacity of curcumin

The new solutions of PPI-Cur and PPI-FA15-Cur nanocomplexes were diluted with absolute ethanol [8]. After centrifugation (6500 g, 20 min), the absorbance at 422 nm of the solution was measured to calculate the encapsulated curcumin content. The loading capacity of curcumin (μg/mg) was the relative ratio of encapsulated curcumin to total protein.

2.5.2. XRD analysis

The PPI, PPI-FA-US15, free curcumin, free FA, PPI-Cur, and PPI-FA15-Cur were analyzed using XRD with an X-ray diffractometer (Ultima IV, Rigaku, Japan) [24]. The measurement range for the 2θ angle was established as 5°–60°.

2.5.3. FT-IR and intrinsic fluorescence spectroscopy analysis

The FT-IR and intrinsic fluorescence spectra of free FA, free curcumin, PPI, PPI-FA-US15, PPI-Cur, and PPI-FA15-Cur was measured as 2.3.6, 2.3.10.

2.5.4. DSC analysis

3 mg of free curcumin, PPI-Cur_0.75, and PPI-FA15-Cur_0.75 nanocomplexes were put in the sealed aluminum crucible, with a blank aluminum crucible serving as the control. The crucibles were warmed from 20 °C to 200 °C under a nitrogen atmosphere (flow rate: 20 mL/min).

2.5.5. Measurement on hydrophilicity and hydrophobicity of nanocomplexes

A dynamic contact angle meter (SL200KS, KINO, USA) was employed to assess the hydrophilicity and hydrophobicity of curcumin (using the Beijing HARKE-SPCA model apparatus). Powders of were distributed on a glass slide. After that, a single drop of deionized water was added to the sample surface. Images of the droplets were captured using a high-speed camera after 10 s, and modeled via T. Young’s fitting method to acquire the three-phase contact angle [34].

2.5.6. Fluorescence quenching and fluorescence lifetime measurement

A fluorescence spectrometer (F-4500, Hitachi, Japan) employed to conduct the fluorescence quenching experiments. The sample solution was blended with curcumin solution (with final curcumin concentrations of 10, 20, 40, 50, 60, 70, and 80 µM). The emission wavelength was 300–450 nm, and the excitation wavelength was 280 nm.

Measurement of fluorescence lifetime for PPI-FA-US15 with or without curcumin was performed by an Edinburgh FLS1000 fluorescence spectrometer (Edin burgh instruments ltd, Edinburgh, UK). A 280 nm EPLED light source was used for excitation, with the emission wavelength monitored at 330 nm. Data were performed by the software provided with the instrument [35].

2.5.7. Three-dimensional fluorescence spectroscopy

Three-dimensional fluorescence spectroscopy measurements were performed using excitation wavelength ranges from 200 nm to 400 nm and emission wavelength ranges from 200 nm to 500 nm [36].

2.5.8. TEM and AFM

The particle shape of PPI, PPI-FA-US15, PPI-Cur_0.75, and PPI-FA15-Cur_0.75 was observed using a TEM operating at 80 kV (JEM-1400 Flash, Japan) [37]. The samples were diluted to 20 μg/mL. The diluted solution applied to a copper grid, and excess solution was removed. Simultaneously, an AFM (model Multimode 8, Bruker Corp., Santa Barbara, CA, USA) was applied to observed the morphology of nanoparticles [8]. 10 μL of diluted solution were deposited onto a mica surface. After overnight drying, each diluted sample was scanned at the intermittent contact mode with a nominal spring constant of 2.8 N/m and a scanning speed of 75 Hz.

2.6. Stability of nanocomplexes

2.6.1. Storge stability

The samples were kept at 4 °C for 77 d. Particle size changes of PPI-Cur_0.75 and PPI-FA15-Cur_0.75 were determined every 7 d to assess their storage stability. The particle size of the PPI-Cur_0.75 and PPI-FA15-Cur_0.75 measured according to section 2.3.12.

2.6.2. UV light stability and thermal stability

Free curcumin, PPI-Cur_0.75 and PPI-FA15-Cur_0.75 were exposed to UV radiation at 365 nm for 240 min to analyze the UV light stability [36]. Free curcumin, PPI-Cur_0.75 and PPI-FA15-Cur_0.75 were heated at different temperature (65 °C, 75 °C, and 85 °C) for 30 min and 60 min to analyze the thermal stability. The curcumin content was measured as section 2.5.1.

2.7. Antioxidant activity

In antioxidant experiments, free curcumin, PPI-Cur, and PPI-FA15-Cur were diluted to the same curcumin concentration, with Vc at the same concentration serving as the control. The curcumin concentrations used in the ABTS and DPPH radical scavenging assays were 10 μg/mL and 20 μg/mL, respectively. PPI and PPI-FA-US15 were maintained at the same protein concentration as PPI-Cur_0.75, PPI-FA15-Cur_0.75, respectively. The ABTS scavenging capacity and DPPH scavenging capacity were characterized as previously reported [25].

2.8. Bioavailability and in vitro release property

The bioavailability of curcumin was evaluated using an in vitro simulated gastrointestinal tract (GIT) model [22]. A 10 mL aliquot of sample was combined with 10 mL of 0.1 M HCl (pH 1.5) and pre-incubated for approximately 10 min at 37 °C with shaking at 100 rpm. Subsequently, gastric digestion was initiated by adding 20 mg of pepsin powder. After 60 min, the pepsin digestates were immediately neutralized to pH 7.0, followed by the addition of 400 mg of bile extract powder under continuous shaking for 10 min. Intestinal digestion was then started by introducing 20 mg of pancreatin powder. Following 120 min of incubation, the resulting mixtures were centrifuged at 10,000 g for 30 min. The curcumin bioaccessibility was calculated as the percentage of curcumin detected in the supernatant relative to the initial amount present prior to digestion.

The release characteristics of free curcumin and encapsulated curcumin were also investigated [31]. Simulated gastric digestion involves adjusting the sample solution (10 mL) to pH = 2.0 followed by mixing with 10 mL simulated gastrointestinal fluid for 2 h. After 2 h, the mixture was adjusted to pH 6.8, and further mixed with simulated intestinal fluid (20 mL) for 4 h to simulate small intestinal digestion. During digestion, the sample was put into a dialysis bag (molecular weight: 5000 Da) and sequentially immersed in 30 mL of PBS solution (pH = 2.0 and 6.8) containing 1 wt% Tween 80, then shaken at 100 rpm and 37 °C. At the designated time point, 2 mL of solution was withdrawn from the exterior of the dialysis bag, and replenished with an equal volume of PBS. The concentrations of curcumin in the extracted solution were measured according to section 2.5.1, respectively, to calculate their cumulative release from the nanoparticles.

2.9. Cell viability assessment

The CCK-8 assay was applied to examine the impact of samples on the cell viability of HCT-116 cells [36]. Cell viability was performed by the following formula:

$$\mathrm{C}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{s}\mathrm{v}\mathrm{i}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}(\%)=\frac{{\mathit{OD}}_{\mathrm{S}}-{\mathit{OD}}_{\mathrm{B}}}{{\mathit{OD}}_{\mathrm{C}}-{\mathit{OD}}_{\mathrm{B}}}\times 100$$ (2)

OD_S, OD_C, and OD_B were defined as follows: the optical density (OD) value of the cells co-cultured with the sample, the OD value of the cells co-cultured with an equal volume of CCK-8 solution, and the OD value of the medium mixed with CCK-8 solution, respectively.

2.10. In vitro blood compatibility

The hemolytic activity of PPI-FA-US15 was determined. The red blood cells (RBCs) were separated from fresh sheep blood via defibrination and centrifugation, followed by washing with PBS. The RBCs were resuspended in PBS (2%, V/V). 200 μL sample solution was combined with 800 μL of the RBC suspension, and placed at 37 °C for 12 h. After centrifugation (8000 g, 5 min), the supernatant was measured at 570 nm. Deionized water was positive and PBS acted as and negative controls, respectively. The hemolysis rate was performed by the following formula:

$$\mathrm{H}\mathrm{e}\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{s}\mathrm{i}\mathrm{s}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}(\%)=\frac{{\mathit{OD}}_{\mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}}-{\mathit{OD}}_{\mathrm{n}\mathrm{e}\mathrm{g}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}}}{{\mathit{OD}}_{\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}}-{\mathit{OD}}_{\mathrm{n}\mathrm{e}\mathrm{g}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}}}\times 100$$ (3)

2.11. In vitro cellular uptake measurement

To facilitate observation of cellular uptake of nanocomplexes, C6 was used to label the particles for subsequent analysis [38]. After co-incubation with the sample for 2 h, the cells underwent three PBS washes followed by fixation with formaldehyde solution for 20 min. Subsequently, the fluorescence imaging was performed using a CLSM to visualize the cellular uptake process. Untreated cells were used as the control group.

2.12. Statistical methods

All experiments were performed in triplicate, and obtained results were expressed as mean ± SD values. Data were analyzed by SPSS 24.0 software (SPSS Incorporated, Chicago) for one-way ANOVA analysis with Duncan’s test (p < 0.05). The graphs were created using Origin 2019 software.

3. Results and discussion

3.1. Formation mechanism of ultrasound-assisted PPI-FA conjugates

3.1.1. SDS-PAGE

The covalent interaction between PPI and FA by ultrasound-assisted free radical grafting was verified by SDS-PAGE [17]. The lanes 1–8 of SDS-PAGE image were PPI, Marker, PPI-FA-9C, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60, respectively (Fig. 1A). Compared with PPI (lane 1), new bands were observed at the top of lanes 3–8, and the corresponding bands shifted upward, demonstrating the formation of PPI–FA conjugates with higher molecular weight via the free radical grafting method. Compared with PPI-FA-9C, the new bands at the top of PPI-FA-US5, PPI-FA-US15, and PPI-FA-US30 lanes were darker, with greater upward shifts observed near the 75 kDa and 65 kDa bands. This indicated that FA was more readily grafted to PPI by the ultrasonication treatment, which could be due to the fact that the sonication generated high mechanical energy and shear stress, thereby inducing protein unfolding and exposing more grafting sites [12]. Among all ultrasound-assisted covalent grafting complexes, the bands at the top of PPI-FA-US15 and PPI-FA-US30 lanes were denser than those of PPI-FA-US45, and PPI-FA-US60. This suggested that more PPI-FA conjugates were produced with 15 and 30 min of sonication, as excessive energy from prolonged ultrasonication could cause protein aggregation and reduce the availability of active sites for FA reaction [24].

Fig. 1.

Formation of PPI-FA conjugates: (A) SDS-PAGE profiles (Lanes 1–8: PPI, Maker, PPI-FA-9C, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60), (B) free amino content, (C) free sulfhydryl content, (D) FT-IR, (E) UV–vis spectra, (F) loading capacity of FA, (G) soluble protein content, (H) particle size, and (I) PDI values. Means followed by different letters were significantly different (p < 0.05).

3.1.2. Reactive group content

Free –NH₂ and −SH groups in proteins exhibit high reactivity, and the variation in their content serves as an indicator for evaluating the protein's ability to bind with FA. The free radical grafting method produced hydroxyl radicals by utilizing H₂O₂ and ascorbic acid. These free radicals could attack the free –NH₂ and −SH groups on PPI to generate protein radicals, which could react with FA to form PPI-FA covalent conjugates [17]. As presented in Fig.1B and C, compared to the native PPI, the –NH₂ and −SH groups contents showed a significant reduction in that of the PPI-FA conjugates, indicating the successful covalent conjugation between PPI and FA by C-N and C-S bonds. Additionally, PPI-FA conjugates treated with ultrasound presented more significant decrease in –NH₂ groups contents than PPI-FA-9C (without ultrasound), which suggested that ultrasonication promoted the covalent binding between FA and PPI. This promotion resulted from ultrasonic processing inducing the unfolding of the PPI structure, thereby providing additional binding sites for FA. Meanwhile, the mechanical effects produced by ultrasound increased the reaction efficiency between H₂O₂ and ascorbic acid in producing hydroxyl radicals, thereby increasing the yield of protein radicals with activity to react with FA. As the ultrasonication time increased from 5 min to 15 min, both the free –NH₂ and −SH group contents exhibited a decreasing trend. Moreover, the –NH₂ and −SH groups contents of PPI-FA-US15 were 84.80 nmol/mg and 25.98 μmol/g, respectively, significantly lower than those of PPI-FA-9C (105.91 nmol/mg and 27.21 μmol/g), indicating that the grafting efficiency of the conjugate obtained via ultrasonic-assisted free radical grafting method for 15 min was significantly higher than that of the conjugate obtained via the common reaction for 24 h. However, as the ultrasonication time was extended (from 15 min to 60 min), the NH₂ and −SH contents were increased and those of PPI-FA-US60 were 88.82 nmol/mg and 26.47 μmol/g, respectively. The reason for this result was that prolonged ultrasound generated excessive energy, causing protein aggregation and reducing the number of reactive sites, resulting in a decrease in the grafting efficiency [24].

3.1.3. FT-IR and UV spectral analysis

In Fig. 1D, the typical protein characteristic peaks, including amide I (1653 cm⁻¹) related to C=O stretching, amide II related to N–H bending vibrations at 1514 cm⁻¹, amide III related to C–N stretching and N–H bending vibrations at 1240 cm⁻¹, of PPI could be observed. Compared with the PPI, the absorption of bands at 1640 cm⁻¹ (C–O), 1514 cm⁻¹ (N–H) and 1240 cm⁻¹ (C–N, N–H) in PPI-FA-9C, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60 were decreased, indicating that the –NH₂ in PPI was consumed and reacted with FA to form covalent conjugates by the ultrasound-assisted free radical grafting method [29].

As illustrated in the ultraviolet spectra (Fig. 1E), native PPI exhibited an absorption peak at 280 nm which was related to the tryptophan and tyrosine residues in protein [18]. Compared to native PPI, PPI-FA-9C showed a new absorption peak at 360 nm, further suggesting the successful conjugation of FA with PPI. It is noteworthy that PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60 conjugates showed higher absorbance, indicating that ultrasound-assisted conjugates complexed more FA. This could be explained by the fact that ultrasound treatment promoted the unfolding of proteins, thus exposing more binding sites, proving that ultrasound significantly improved the grafting efficiency of PPI and FA.

3.1.4. Loading capacity of FA

As shown in Fig. 1F, the FA loading capacity of the ultrasound-assisted covalent grafted conjugates was higher than that of PPI-FA-9C, indicating that the conjugates obtained from a short-term ultrasound-assisted free radical grafting reaction (5 min) loaded more FA than those obtained from the conventional free radical grafting reaction (24 h). Among these samples, PPI-FA-US15 showed the highest FA loading capacity (44.78 µg/mg), which was about 1.2-fold higher than that of the PPI-FA-9C conjugate (36.45 µg/mg). These results also demonstrated that the ultrasound-assisted free radical grafting method could enhance the binding efficiency between PPI and FA, enabling PPI to bind more FA in a short time, consistent with the results in section 3.1.2. Besides, the FA loading capacity of PPI-FA-US15 conjugates in this study was about 1.78-fold higher than that (25 µg/mg) of lysozyme-FA complexes prepared by Ma et al. [39], further demonstrating the high efficiency of the ultrasound-assisted free radical grafting reaction.

3.1.5. Soluble protein content

A favorable solubility is essential for the delivery system, as it can improve the aqueous solubility of the delivered nutrients and reduce protein aggregation during digestion, thereby stabilizing the entire system until reaching the intestines [40]. As demonstrated in Fig. 1G, all the PPI-FA conjugates showed higher soluble protein content than that of native PPI (61.34%), indicating that conjugation with FA effectively enhanced the solubility of PPI. Meanwhile, the PPI-FA-US15 conjugates showed higher soluble protein content than that of PPI-FA-9C and PPI-FA-US5, which could be attributed to the fact that ultrasonic treatment disrupted the aggregate structures and improved dispersion of PPI, thereby increasing the solubility [41]. However, PPI-FA-US60 (80.64%) displayed lower soluble protein content than PPI-FA-US15 (97.53%), because prolonged ultrasonication (60 min) generated excessive energy, leading to protein aggregation [24].

3.1.6. Particle size and PDI

Characterization of particle size and PDI is crucial for evaluating the performance of curcumin-loaded nanocomplexes, as these properties directly affect the stability, bioavailability, and functional characteristics of the complex in food systems [42]. The particle size of PPI (Fig. 1H) was 100.92 nm smaller than that of the PPI-FA-9C conjugate (129.45 nm), which could be ascribed to the unfolding of the protein structure induced by the alkaline pH (pH 9) [43]. Compared with PPI and PPI-FA-9C, the particle size of the ultrasonically treated samples was reduced. This decrease could be due to the powerful shear forces produced by ultrasonic cavitation, which disrupted protein aggregates and consequently diminished the particle size of covalent complexes. Among all ultrasound-assisted free radical grafting conjugates, the particle size values reduced with the increase in ultrasonication time, achieving reaching a minimum at 15 min of ultrasound (PPI-FA-US15, 77.28 nm), and then increased significantly (p < 0.05). The reason was that appropriate ultrasonic treatment could promote a more uniform particle distribution, while excessive ultrasonic treatment would cause particle aggregation [22]. Meanwhile, the PDI value of PPI-FA-US15 was 0.2978, which was less than 0.3, indicating that PPI-FA-US15 was a stable nanoparticle with uniform particle size distribution.

3.1.7. XPS analysis

To further investigate the covalent interactions between PPI and FA, XPS technology was used to analyze the surface elemental composition of the conjugate [44]. As presented in Fig. 2A, three peaks corresponding to O1s, N1s, and C1s were observed at ∼ 529.7 eV, ∼397.9 eV, and ∼ 284.5 eV, respectively, indicating that the surfaces of PPI and PPI-FA complexes mainly consisted of oxygen, nitrogen, and carbon (Table 1). In addition, the C1s peak in the spectra of PPI and PPI-FA conjugates could be fitted into four component peaks, located at 287.9 eV (C=O), 286.3 eV (C–O), 284.8 eV (C–C), and 285.8 eV (C–N), respectively. In Fig. 2C, the intensities of the C–C peak in PPI-FA-9C was lower than that in PPI. Meanwhile, Table 2 also showed that the content of C–N in PPI-FA-9C was 20.89%, higher than that (15.00%) in native PPI, suggesting that the –NH₂ in PPI reacted with the –COOH of FA to form amide (–CONH-) bonds by the free radical grafting method to produce PPI-FA conjugates. Meanwhile, the contents of C–N in the ultrasound-assisted conjugates were higher than that in PPI-FA-9C, also suggesting that the ultrasonication could promote the covalent grafting reaction of FA with PPI. Among all the samples, PPI-FA-US15 exhibited the highest C–N content (25.11%), indicating that the grafting rate of PPI with FA was the highest at 15 min of ultrasonication.

Fig. 2.

(A) XPS spectra, (B-H) C1s peak fitting spectra of PPI, PPI-FA-9C,PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, and PPI-FA-US60, (I) fluorescence spectra of native PPI and PPI-FA conjugates. (J) Schematic diagram of producing PPI-FA conjugates by traditional and ultrasound-assisted free radical method.

Table 1.

Atomic contents of PPI and PPI-FA covalent complexes.

Sample	C (%)	N (%)	O (%)
PPI	68.03	9.91	22.06
PPI-FA-9C	66.69	10.62	22.70
PPI-FA-US5	69.55	10.35	20.10
PPI-FA-US15	66.15	11.36	22.49
PPI-FA-US30	68.46	11.38	20.16
PPI-FA-US45	67.83	10.14	22.03
PPI-FA-US60	69.26	10.14	20.60

Table 2.

Different C contents of PPI and PPI-FA covalent complexes determined by XPS.

Sample	C–C (%)	C-N (%)	C-O (%)	C=O (%)
PPI	59.59	15.00	11.51	14.14
PPI-FA-9C	54.66	20.89	10.90	13.55
PPI-FA-US5	54.98	21.11	9.39	14.53
PPI-FA-US15	50.25	25.11	13.21	14.94
PPI-FA-US30	54.31	21.23	9.35	15.11
PPI-FA-US45	52.19	19.37	12.79	15.65
PPI-FA-US60	58.51	20.35	8.88	12.27

The mechanism of ultrasound-assisted PPI grafting with FA was shown in Fig. 2J. The reaction between ascorbic acid and H₂O₂ generated hydroxyl radicals, which subsequently attacked the –NH₂ and −SH groups in PPI, leading to the formation of reactive protein radicals. The protein radicals could conjugate with the –COOH of FA, forming the covalent cross-link between PPI and FA. During this process, ultrasound treatment induced PPI to expose more active sites (–NH₂ and −SH), and increased the collision frequency between PPI and FA, thereby significantly accelerating the covalent conjugation with FA.

3.1.8. Intrinsic fluorescence spectroscopy

The interactions between proteins and ligands and the alterations in the tertiary structure of proteins could lead to changes in the intrinsic fluorescence spectra [33]. As illustrated in Fig. 2I, the maximum emission intensity of native PPI was observed at 334 nm. After conjugation with FA, the fluorescence intensity was greatly decreased, and the maximum emission wavelength underwent a red shift from 334 to 335 nm. The redshift could be explained by the fact that the alkaline pH and ultrasonic treatment unfolded the protein structure, leading to the chromophore being exposed to a more hydrophilic environment. The cause of fluorescence quenching could be attributed to the fact that the tryptophan residues participated in covalent binding reactions between PPI and FA [6].

3.1.9. Far-UV CD spectroscopy

The effect of ultrasonic treatment on the secondary structure composition of PPI was determined using far-UV CD spectroscopy. In the far-UV CD spectrum of PPI (Fig. 3A), the negative band at 208 nm corresponded to the α-helical structure, while the positive band at 195 nm was associated with the β-sheet structure [45]. The secondary structure of native PPI had 35.30% α-helix, 16.90% β-sheet, 16.87% β-turn, and 30.90% random coil forms. Following ultrasonic treatment, a shift in peak positions and altered ellipticity of the negative band were observed, indicating the significant changes in the secondary structure of PPI. Compared to PPI, the proportion of α-helix (31.03%) in PPI-FA-US5 (Fig. 3B) was decreased, while the proportion of β-turn (19.30%) and random coil (31.10%) increased. This indicated that the ultrasonic treatment could promote the transformation of highly ordered α-helix into more disordered β-turn, leading to the unfolding of protein structure. However, as the ultrasonication time was extended to 15 min, the ratio of α-helix increased to 33.43% (PPI-FA-US15), but the ratio of random coil (29.23%) decreased, indicating that excessive ultrasonication caused the protein molecules to reaggregate [24]. Therefore, considering the combined factors of FA loading capacity, sample characteristics, and the cost, PPI-FA-US15 was selected for further investigation.

Fig. 3.

(A) Far-UV CD spectroscopy and (B) secondary structure content of PPI, PPI-FA-US5, PPI-FA-US15, PPI-FA-US30, PPI-FA-US45, PPI-FA-US60. (C) Surface hydrophobicity of PPI and PPI-FA-US15. The visual appearance of (D) PPI and (E) PPI-FA-US15 before and after centrifugation with the addition of curcumin (0.25 mg/mL to 1.5 mg/mL). (F) Loading capacity of PPI-Cur and PPI-FA15-Cur nanocomplexes. (G) Particle size and PDI of PPI-Cur nanocomplexes. (H) Particle size and PDI of PPI-FA15-Cur nanocomplexes. (I) Zeta-potential of PPI-Cur and PPI-FA15-Cur nanocomplexes. Means followed by different letters were significantly different (p < 0.05).

3.1.10. Surface hydrophobicity (H₀)

The H₀, serving as an indicator of the hydrophobic group number on the surface of protein molecules, can be used to evaluate the variations in the physicochemical properties of nanocomplexes. In Fig. 3C, the surface hydrophobicity of PPI-FA-US15 conjugates was remarkably higher than that of PPI, which was likely attributed to the ultrasonic cavitation effect exposing the hydrophobic groups originally buried within the PPI molecules, thereby increasing the H₀ value. Typically, higher surface hydrophobicity indicates more hydrophobic sites on the protein surface, facilitating binding to more hydrophobic active substances [46].

3.2. Characteristics of PPI-Cur and PPI-FA15-Cur nanocomplexes

3.2.1. Curcumin loading capacity

The encapsulation properties of PPI-FA-US15 for curcumin were evaluated, and the influence of curcumin concentration on the loading capacity of nanocomplexes was examined (Fig. 3D–F). Fig. 3D and Fig. 3E showed the visual appearance of PPI and PPI-FA-US15 before and after centrifugation with the addition of curcumin. After centrifugation, the mixed solutions of PPI/PPI-FA-US15 and curcumin exhibited clear and transparent states at all curcumin concentrations, indicating the successful construction of PPI-Cur and PPI-FA15-Cur nanocomposites. The solution of free curcumin presented a turbid state, demonstrating that free curcumin was poorly soluble in deionized water. However, after complexation with the protein-based nanocarrier, curcumin was uniformly dispersed in water, demonstrating that complexation with the PPI-FA-US15 effectively improved the dispersibility of curcumin in aqueous solutions. As the curcumin concentration increased from 0.25 mg/mL to 1.5 mg/mL, curcumin loading capacity (Fig. 3F) of the PPI-FA15-Cur nanocomplexes showed an upward trend, which also indicated the successful encapsulation of curcumin within PPI-FA-US15 nanoparticles.

The curcumin loading capacity of PPI-FA15-Cur nanocomplexes steadily enhanced with the increasing curcumin concentration, reaching a plateau at 0.75 mg/mL, with a value of 64.31 μg/mg (Fig. 3F) which surpassed the loading capacity (40.30 μg/mg) of oat protein isolate-β-glucan-curcumin nanocomplexes prepared by Zhong et al. [47]. The PPI-FA15-Cur_1.5 exhibited the highest loading capacity of 69.67 μg/mg, significantly higher than that of PPI-Cur_1.5 (35.29 μg/mg) (p＜0.05), which could be attributed to the fact that PPI-FA-US15 showed higher surface hydrophobicity, thereby enhancing the number of hydrophobic sites available for curcumin binding, resulting in a higher curcumin loading capacity.

3.2.2. Particle size, PDI, and zeta-potential

Fig. 3G and Fig. 3H showed the particle size and PDI value of PPI-Cur and PPI-FA15-Cur nanocomplexes. Compared to PPI and PPI-FA-US15, PPI-Cur and PPI-FA15-Cur exhibited smaller particle sizes, likely due to curcumin binding within the PPI structure, which resulted in a more compact nanoparticle structure and consequently reduced the particle size [37]. Nevertheless, the particle size of PPI-Cur nanocomplexes was increased when the concentration of curcumin exceeded 0.75 mg/mL. Meanwhile, in Fig. 3I, the absolute value of zeta-potential of PPI-Cur nanocomplexes decreased significantly as the curcumin concentration increased from 0.75 mg/mL to 1.5 mg/mL, which could be due to the electrostatic shielding effect produced by the binding of curcumin. The reduction in the absolute value of zeta potential resulted in weakened electrostatic repulsion between particles, which in turn caused the aggregation of the particles. In comparison to PPI-Cur, the zeta-potential value of the PPI-FA15-Cur nanocomposite was less affected by curcumin binding, which could be due to the fact that the negative charge in PPI-FA15-Cur was provided not only by the carboxyl groups of the protein but also by the carboxyl groups of FA. Therefore, the conjugation of FA enhanced the stability and curcumin loading capacity of PPI. Increasing the curcumin concentration from 0 to 1.5 mg/mL, the absolute potential values of PPI-FA15-Cur nanocomplexes were all maintained above 25 mV, which represented that the stability of all PPI-FA15-Cur nanocomplexes was favorable [22].

3.2.3. UV light stability of FA

Under the UV radiation, FA readily degraded to 6-formylpterin (FPT) and p-aminobenzoylglutamate (PGA). The fluorescence quantum yields of the degradation products of FA ranged from 0.07 to 0.33, while that of FA was extremely low (< 0.005). Therefore, the degradation of FA could be analyzed by detecting changes in the fluorescence intensity of the solution in the process of UV irradiation [30]. In Fig. 4A, the fluorescence intensity of free FA was increased quickly, suggesting that FA was decomposed to PGA and FPT, and free FA had a poor photostability. However, the FA in PPI-FA-US15 exhibited a very slow degradation rate, suggesting that the aromatic groups and double bonds in PPI effectively absorbed UV light, thus decreasing the photodegradation of FA. After UV irradiation (240 min), the retention rate of free FA (Fig. 4B) was 53.68%, but the FA retention rate of the PPI-FA-US15 reached 95.14%, confirming that the complexation with PPI could significantly enhance the photostability of FA. In addition, the photostability of FA in PPI-FA15-Cur nanocomplexes was evaluated. The FA retention rate of PPI-FA15-Cur reached 99.10% after 240 min of UV irradiation, indicating that co-encapsulation with curcumin further enhanced the protective effect for FA.

Fig. 4.

(A) Fluorescence intensity during 240 min of ultraviolet irradiation and (B) folic acid retention after 120 min of ultraviolet irradiation of free FA, PPI-FA-US15, and PPI-FA15-Cur_0.75. (C) XRD spectrum and (D) XRD strip heat map of free Cur and PPI-Cur nanocomplexes. (E) XRD spectrum and (F) XRD strip heat map of free Cur and PPI-FA-Cur nanocomplexes. FT-IR spectra of (G) PPI-Cur nanocomplexes and (H) PPI-FA15-Cur nanocomplexes. Typical intrinsic fluorescence spectra of (I) PPI-Cur nanocomplexes and (J) PPI-FA15-Cur nanocomplexes. (K) DSC curves of PPI, PPI-Cur_0.75, and PPI-FA15-Cur_0.75. The water contact angle of (L) free Cur, (M) PPI-Cur_0.75, and (N) PPI-FA15-Cur_0.75. Means followed by different letters were significantly different (p < 0.05).

3.2.4. XRD analysis

XRD analysis was used to study the crystal diffraction of curcumin. Typically, an amorphous curcumin is more favorable for absorption and uptake than its crystalline state [48]. As displayed in Fig. 4C–F, free curcumin showed sharp peaks at 12.16°, 14.52°, 17.16°, 19.32°, 21.16°, 23.24°, and 24.64° in the 2θ range, indicating that free curcumin was highly crystalline. Notably, the sharp peaks were absent in PPI-Cur and PPI-FA15-Cur, indicating that the complexation with protein-based carriers inhibited the crystallization of curcumin, and the curcumin within nanocomplexes was in an amorphous form. The two peaks appearing at 31° and 45° in the spectrum correlate with NaCl crystals produced upon pH adjustment. In addition, free FA showed sharp peaks at 5.3°, 10.6°, 12.8°, 21.5°, 26.6°, and 29.2° (Fig. 4E and F), confirming its high crystallinity. In the XRD spectrum of PPI-FA15-Cur, the sharp peak of FA was vanished, indicating that FA was amorphous within the PPI-FA15-Cur nanocomplexes.

3.2.5. FT-IR and intrinsic fluorescence spectra analysis

The FT-IR spectra of PPI-Cur nanocomplexes and PPI-FA15-Cur were displayed in Fig. 4G and H, respectively. The absorbance peaks of free curcumin were observed at 961 cm⁻¹ (benzoate trans-CH vibrations), 1027 cm⁻¹ (C–O–C), 1274 cm⁻¹ (enol C–O), and 3512 cm⁻¹ (phenolic − OH), while these peaks vanished in the spectra of PPI-Cur and PPI-FA15-Cur [8]. This result suggested that hydrogen bonding and hydrophobic interactions between PPI/PPI-FA-UA15 and curcumin existed, which restricted the stretching and flexural vibrations of curcumin [25]. Meanwhile, the sharp absorbance peaks of free curcumin were not visible in the PPI-FA15-Cur nanocomplexes, indicating that curcumin was fully wrapped within the core of PPI-FA-US15 [48].

Fig. 4I and J showed the fluorescence spectra of curcumin interacting with PPI and PPI-FA-US15, respectively [31]. In general, the fluorescence intensity of the protein gradually reduced with curcumin concentration increasing from 0 to 1.5 mg/mL, indicating that curcumin could bind with the proteins via hydrophobic interactions, causing a regular quenching of the endogenous fluorescence. Considering the curcumin load capacity, particle size distribution, and zeta potential value, PPI-Cur_0.75 and PPI-FA15-Cur_0.75 were selected for further analysis.

3.2.6. DSC analysis

The thermal properties of the free curcumin, PPI-Cur_0.75 and PPI-FA15-Cur_0.75 were characterized by DSC, and the thermograms were depicted in Fig. 4K [8]. Free curcumin exhibited a sharp endothermic peak at 178.9 °C, demonstrating the occurrence of curcumin crystal melting. However, this peak disappeared in the nanocomposites. A broad peak was found in the curves of PPI-Cur and PPI-FA15-Cur likely related to the protein denaturation. The smooth curves presented in the DSC thermograms of PPI-Cur and PPI-FA15-Cur, along with the disappearance of the characteristic melting peak of curcumin, confirmed the absence of crystalline curcumin in the nanocomposite. These results indicated that the curcumin was encapsulated within the nanocomplexes in the non-crystalline state, consistent with XRD results. Furthermore, the result of optical contact angle suggested that, upon encapsulation within PPI-FA-US15, the contact angle of free curcumin was decreased from 146.16° (Fig. 4L) to 51.78° (Fig. 4N), indicating that loading in PPI-FA-US15 greatly improved the hydrophilicity of the curcumin, which was contributed to changes in the water-solubility (Section 3.2.1).

3.2.7. Quenching of PPI and PPI-FA-US15 fluorescence by curcumin

In order to elucidate the binding mechanism between curcumin and PPI/PPI-FA-US15, the effects of varying curcumin concentrations on the fluorescence spectra of PPI/PPI-FA-US15 were investigated. As shown in Fig. 5A and Fig. 5B, curcumin induced a concentration-dependent quenching of fluorescence intensity, further confirming that curcumin formed nanocomplexes with PPI/PPI-FA-US15 via hydrophobic interactions, thereby triggering the fluorescence quenching. Fluorescence quenching is categorized into dynamic quenching and static quenching, with static quenching representing the formation of fluorophore-quencher complexes. The Stern-Volmer equation can be employed to elucidate the mechanism of fluorescence quenching:

$$\frac{F}{F_0}=1+K_{\mathit{SV}}\left[Q\right]=1+K_q{\tau }_0[Q]$$ (4)

Fig. 5.

(A) Fluorescence emission spectra of PPI and (D) PPI-FA-US15 at various concentrations of curcumin (0–80 μM). The Stern-Volmer plots for the quenching of (B) PPI and (E) PPI-FA-US15 by curcumin. The logarithmic plots for the quenching of (C) PPI and (F) PPI-FA-US15 fluorescence by curcumin. (G) Fluorescence lifetime decay plots of PPI-FA-US15 in the absence and presence of curcumin. Three-dimensional fluorescence spectra of (H) PPI, (I) PPI-FA-US15, (J) PPI-Cur_0.75, and (K) PPI-FA15-Cur_0.75. Means followed by different letters were significantly different (p < 0.05).

Here, F₀ and F represent the fluorescence intensity without or with the addition of curcumin, respectively; K_SV and K_q are the Stern–Volmer quenching constant and quenching constant, respectively; [Q] represents the curcumin concentration; τ₀ (10⁻⁸ s) is the average fluorescence lifetime of the protein without quencher. K_q is applied to estimate the quenching mechanism. The maximum value of the collisional quenching constant is 2 × 10¹⁰ s⁻¹M⁻¹. The Stern–Volmer plots of PPI and PPI-FA-US15 quenched by curcumin were depicted in Fig. 5B and Fig. 5E, respectively. K_SV and K_q were calculated and displayed in Table 3. The K_q of PPI and PPI-FA-US15 were 6.43 × 10¹¹ s⁻¹M⁻¹ and 4.81 × 10¹¹ s⁻¹M⁻¹, which were higher than 2 × 10¹⁰ s⁻¹M⁻¹, indicating that all of the quenching reactions between curcumin and PPI/PPI-FA-US15 were static quenching [31].

Table 3.

Stern-Volmer modeling, binding parameters, and n of curcumin binding to PPI and PPI-FA-US15.

	$\frac{F}{F_0}=1+K_{\mathit{SV}}\left[Q\right]=1+K_q{\tau }_0[Q]$			$\mathrm{l}\mathrm{o}\mathrm{g}[\frac{{\mathrm{F}}_0-\mathrm{F}}{\mathrm{F}}]=\mathrm{l}\mathrm{o}\mathrm{g}\mathrm{K}+\mathrm{n}\mathrm{l}\mathrm{o}\mathrm{g}\left[\mathrm{Q}\right]$
	R2	Ksv (103 M− 1)	Kq(1011 s−1M−1)	R2	K (106 M−1)	n
PPI	0.9900	6.43	6.43	0.9893	3.2107	1.4112
PPI-FA-US15	0.9849	4.81	4.81	0.9957	8.9269	1.5095

Based on the result of static quenching, the K and n of the interaction of between curcumin and PPI/PPI-FA-US15 were determined by Eq. (5):

$$\mathrm{l}\mathrm{o}\mathrm{g}[\frac{{\mathrm{F}}_0-\mathrm{F}}{\mathrm{F}}]=\mathrm{l}\mathrm{o}\mathrm{g}\mathrm{K}+\mathrm{n}\mathrm{l}\mathrm{o}\mathrm{g}\left[\mathrm{Q}\right]$$ (5)

Herein, K is the binding constant; n is the binding sites number. The corresponding curves were shown in Fig. 5C (PPI) and Fig. 5F (PPI-FA-US15). The values of K and n were also calculated, and the results were displayed in Table 3. The binding constant and binding sites of PPI-FA-US15 to curcumin were 8.93 M⁻¹ and 1.51, respectively, both higher than those of PPI (3.21 M⁻¹ and 1.42), indicating that PPI-FA-US15 showed superior binding affinity for curcumin compared to PPI. This phenomenon could be due to the free radical grafting reaction and ultrasonic treatment, which induced the exposure of hydrophobic groups within the protein structure, increasing the binding sites for curcumin [49].

In order to reconfirm the fluorescence quenching mechanism of PPI-FA-US15 by curcumin, the fluorescence lifetimes of PPI-FA-US15 and PPI-FA15-Cur were determined. Fig. 5G showed that the decay data of PPI-FA-US15 and PPI-FA15-Cur were overlapped, indicating that the presence of curcumin had no influence on the decay curves of PPI-FA-US15. In Table 4, the average fluorescence lifetime of PPI-FA-US15 was 2.66 ns and that of PPI-FA15-Cur was 2.62 ns, which could be regarded as the fluorescence lifetimes of PPI-FA-US15 before and after the addition of curcumin were unchanged. These results verified that the fluorescence quenching mechanism of curcumin on PPI-FA-US15 was static quenching.

Table 4.

Lifetime decay parameters of PPI-FA-US15 in the absence and presence of curcumin.

Sample	τ1 (ns)	τ2 (ns)	f1	f2	τ (n)	χ2
PPI-FA-US15	1.0321	3.4386	0.32	0.68	2.6584 ± 0.058	0.8378
PPI-Cur	0.9300	3.2820	0.28	0.72	2.6199 ± 0.053	0.9053

Three-dimensional fluorescence spectroscopy was applied to further investigate the curcumin-induced changes in the microenvironment and conformation of PPI. Fig. 5H–K showed the three-dimensional fluorescence spectra of PPI, PPI-FA-US15, PPI-Cur_0.75, and PPI-FA15-Cur_0.75, respectively. The three peaks are marked as peak a (Rayleigh scattering peak), peak b (associated with Tyr and Trp residues) and peak c (related to the polypeptide backbone) [36]. The peak b fluorescence intensity of PPI-FA-US15 was significantly reduced compared to that of PPI, and the peak b fluorescence intensity of PPI-FA15-Cur_0.75 was further reduced compared with that of PPI-FA-US15, which suggested that the binding of both FA and curcumin altered the microenvironment around the fluorescent chromophores of PPI, resulting in the fluorescence quenching [50].

3.2.8. Microstructure of PPI-Cur_0.75/PPI-FA15-Cur_0.75

As shown in Fig. 6, TEM and AFM were employed to observe the morphology of PPI, PPI-FA-US15, PPI-Cur_0.75 and PPI-FA15-Cur_0.75. Compared to PPI, PPI-FA-US15 exhibited smaller particle sizes, likely due to the ultrasonic cavitation effect dispersing the protein aggregates, thereby reducing the particle size [24]. Furthermore, the particle size of the curcumin-loaded nanocomplexes (PPI-Cur_0.75/PPI-FA15-Cur_0.75) was smaller than that of their corresponding protein-based carrier (PPI/PPI-FA-US15). This reduction was likely because the encapsulated curcumin induced a more compact particle structure [37]. The PPI-FA15-Cur_0.75 nanocomplexes had a spherical shape and uniform dispersion, suggesting that the resulting nanocomplexes possessed a homogeneous and stable structure, which was beneficial for enhancing the absorption rates and facilitating the application in food products.

Fig. 6.

TEM images of (A) PPI, (B) PPI-FA-US15, (C) PPI-Cur_0.75, and (D) PPI-FA15-Cur_0.75. The curves in the TEM images represented the particle size distribution. AFM images of (E) PPI, (F) PPI-FA-US15, (G) PPI-Cur_0.75, and (H) PPI-FA15-Cur_0.75.

3.3. Stability evaluation

3.3.1. Storage stability

Storage stability is a critical property of colloidal delivery systems, and provides valuable insight into the potential shelf life of commercial products [8]. PPI-Cur_0.75 and PPI-FA15-Cur_0.75 were stored at 4 °C for 77 d, and their particle size changes were analyzed to evaluate the storage stability of the nanocomplexes. In Fig. 7A, PPI-Cur_0.75 nanocomposites exhibited turbidity at the bottom of the solution after 63 d of storage. Furthermore, the particle size distribution of PPI-Cur_0.75 (Fig. 7B) revealed the emergence of macromolecular aggregates after storage for 63 d. In contrast, the PPI-FA15-Cur_0.75 nanocomplexes formed macromolecular aggregates only after storage for 77 d, and maintained a relatively uniform particle size distribution throughout the storage period (0–70 d). These results indicated that the PPI-FA15-Cur_0.75 showed better storage stability than PPI-Cur_0.75, which was attributed to the binding of FA increasing the surface charge of nanoparticles (Fig. 3I), thereby enhancing electrostatic repulsion between PPI-FA15-Cur particles and inhibiting the aggregation of particles.

Fig. 7.

(A) Appearance of the nanocomplexes during 77 d of storage. Particle size distribution of (B) PPI-Cur_0.75 and (C) PPI-FA15-Cur_0.75 during storage. Curcumin retention after heating at 65 ℃, 75 ℃, and 85 ℃ for (D) 30 min and (E) 60 min. (F) Curcumin retention after 240 min of ultraviolet irradiation. (G) ABTS scavenging activity and (H) DPPH scavenging activity of Vc, free curcumin, PPI, PPI-FA-US15, PPI-Cur_0.75, and PPI-FA15-Cur_0.75. Means followed by different letters were significantly different (p < 0.05).

3.3.2. Thermal stability of curcumin

Samples were heated at 65 °C, 75 °C, and 85 °C for 30 min (Fig. 7D) or 60 min (Fig. 7E), followed by determination of curcumin retention rates. As shown in Fig. 7D and E, the retention rate of free curcumin was significantly lower than that of the encapsulated curcumin under all heating conditions (p < 0.05). Notably, after heating at 85 °C for 60 min (Fig. 7E), only 64.81% of free curcumin remained, further validating the necessity of encapsulation to enhance its thermal stability. The curcumin retention rate of the PPI-FA15-Cur_0.75 nanocomplexes reached 86.33%, indicating that encapsulation within PPI-FA-US15 significantly enhanced protection against curcumin thermal degradation. This could be explained by the electrostatic interactions provided by the protein-based shell, which suppressed the decomposition and aggregation of particles during heating process, and delayed the degradation of actives, enhancing the thermal stability of curcumin [51].

3.3.3. UV light stability of curcumin

Curcumin underwent photochemical degradation upon exposure to ultraviolet or visible light by an oxidative reaction [46]. Herein, the stability of free and encapsulated curcumin was assessed after exposing to ultraviolet irradiation for 240 min at 25 °C (Fig. 7F). The retention rate of free curcumin rapidly reduced with the prolonged exposure time. After 240 min of irradiation, only 23.42% of free curcumin remained. However, the degradation rate of curcumin in PPI-Cur_0.75 or PPI-FA15-Cur_0.75 was significantly slower than that of free curcumin. After 240 min of irradiation, the curcumin retention rates of PPI-Cur_0.75 and PPI-FA15-Cur_0.75 reached 62.79% and 63.40%, respectively. According to the overall tendency, the curcumin retention rate of PPI-FA15-Cur was higher than that of PPI-Cur which could be explained by the fact that FA provided antioxidant activity and UV filtering effects, offering further photoprotection for curcumin [46]. These findings indicated that the photostability of curcumin was effectively enhanced by encapsulation within PPI-FA-US15.

3.4. Antioxidant activity

The antioxidant properties of PPI, PPI-FA-US15, free curcumin, PPI-Cur_0.75, and PPI-FA15-Cur_0.75 samples were analyzed by ABTS and DPPH radical scavenging assays (Fig. 6H and I). In the ABTS radical scavenging assay, PPI-FA-US15 exhibited higher scavenging capacity than PPI, indicating that the binding of FA endowed PPI-FA-US15 with stronger antioxidant activity than PPI. Additionally, at the same curcumin concentration, the ABTS radical scavenging rates of free curcumin, PPI-Cur_0.75, and PPI-FA15-Cur_0.75 were 9.64%, 79.48%, and 83.06%, respectively. Free curcumin exhibited the lowest antioxidant capacity due to the poor water solubility, which limited its reaction with ABTS radicals. Encapsulation within PPI/PPI-FA-US15 suppressed the crystalline precipitation and improved the water dispersibility of curcumin, thereby promoting the reaction of curcumin with the ABTS radicals and enhancing the antioxidant ability of curcumin [46]. Additionally, since the antioxidant capacity of FA and the conjugation with FA further enhanced the solubility of particles, PPI-FA15-Cur_0.75 exhibited the highest free radical scavenging activity among all the samples. Similar results were obtained in the DPPH radical scavenging experiment, indicating that the PPI-FA15-Cur_0.75 nanocomplexes possessed a superior antioxidant capacity.

3.5. In vitro release behavior

Sustained release of curcumin could increase its bioactivity and stability, as most curcumin is absorbed in the small intestine. Therefore, the in vitro digestion process of PPI-FA15-Cur_0.75 was simulated to analyze the release and bioaccessibility of curcumin under relevant physiological conditions. The free curcumin and PPI-FA15-Cur_0.75 nanocomplexes were inserted in SGF and SIF for 120 and 240 min, respectively. Fig. 8A showed that free curcumin rapidly diffused in SGF (with a release rate of 68.26%), leading to the degradation and low bioavailability (16.82%) of curcumin (Fig. 8C). As shown in Fig. 8D, the PPI-FA15-Cur_0.75 nanocomplexes exhibited a slow-release rate for curcumin during simulated gastric digestion. At the end of SGF digestion, the cumulative release of curcumin was only 6.50% (Fig. 8D), indicating that the majority of it could be released in the intestinal tract, thereby achieving precise intestinal targeting.

Fig. 8.

Release of (A) free and (D) encapsulated curcumin under simulated gastric fluid (SGF) and simulated intestinal fluid (SIF). Fitting curves of released curcumin with different models in (B) SGF and (E) SIF. (C) Bioavailability of free curcumin, curcumin within PPI-Cur_0.75 and PPI-FA15-Cur_0.75. (F) Hemolysis ratios of PPI-FA-US15. Cell viability of (G) PPI-FA-US15, (H) free curcumin and PPI-FA-Cur_0.75. (I) Mean fluorescence intensity of CLSM images of free C6 (K₃), PPI-C6 (L₃) and PPI-FA15-C6 (M₃). CLSM images of cells after 2 h incubation with (J) PBS, (K) free C6, (L) PPI-C6, and (M) PPI-FA15-C6. Scale bar = 100 μm. Means followed by different letters were significantly different (p < 0.05).

Furthermore, the PPI-FA15-Cur_0.75 nanocomposite showed a sustained release of curcumin during simulated intestinal fluid digestion (Fig. 8D). These results indicated that the PPI-FA15-Cur_0.75 nanocomplexes could slowly release curcumin during digestion, protecting it from digestive enzymes and pH effects, thereby effectively enhancing the stability and bioavailability of curcumin. This could be attributed to that PPI-FA15-Cur_0.75 encapsulated curcumin within the core of protein structure by hydrogen bonds and hydrophobic interactions to effectively control the release of curcumin. Based on these properties, the bioavailability of curcumin (Fig. 8C) in PPI-FA15-Cur_0.75 nanocomplexes was 76.55%, significantly higher than that of free curcumin (approximately 17%) [46]. Notably, the bioavailability of curcumin encapsulated in PPI-FA15-Cur_0.75 (Fig. 8C) was higher than that in PPI-Cur_0.75 (68.81%), owing to the superior solubility of PPI-FA-US15 compared to native PPI, resulting in more favorable bioavailability.

For the further evaluation of the curcumin release mechanism as a function of time from PPI-FA15-Cur nanocomplexes in SGF and SIF, the experimental data were fitted by the different mathematical kinetic models, including first order, zero order, Higuchi and Ritger-Peppas kinetic models. The details were shown in Table 5. For curcumin release in SGF digestion, zero-order and first-order models were the best fitting models, indicating that the curcumin release resulted from the combined effects of carrier degradation by gastric enzymes (zero-order characteristic) and curcumin molecule diffusion (first-order characteristic). During SIF digestion, the first-order model and Higuchi model best simulated the release of curcumin, indicating that the release mechanisms in the intestinal digestion were based on Fick’s law and dissolution, resulting in a steady release over the digestive process [53]. In the Ritger-Peppas kinetic model, the release indices (n) for the curcumin encapsulated within the PPI-FA15-Cur_0.75 was 0.7170, further confirming that the controlled-release mechanism of curcumin in SIF was a combination of Fick’s diffusion and dissolution [53].

Table 5.

Release constants and adjusted correlation coefficients of different mathematical equations for curcumin release from PPI-FA15-Cur_0.75 nanocomplexes.

	Kinetic model
SGF	Zero-Order	First-Order	Higuchi	Ritger-Peppas
	R22	R22	R22	R22	n2
	0.9906	0.9927	0.8849	0.9938	1.100
SIF	Zero-Order	First-Order	Higuchi	Ritger-Peppas
	R22	R22	R22	R22	n2
	0.8946	0.9344	0.9307	0.91502	0.7170

R² was the correlation coefficients obtained from the release curves according to the different mathematical models. The parameter n is the release index applied to the Ritger-peppas model.

3.6. In vitro cytotoxicity and cellular uptake

To investigate the biosafety of PPI-FA-US15 as a food ingredient, its hemolytic activity was evaluated [52]. As shown in Fig. 8F, no significant hemolysis was observed across all sample groups (concentration range 0–1000 μg/mL), and the hemolysis rates were consistently below 1%. These results indicated that PPI-FA-US15 exhibited a high level of biocompatibility. Furthermore, Fig. 8G showed the effect of PPI-FA-US15 on cell viability as detected by the CCK-8 assay. After incubation with PPI-FA-US15 conjugates for 24 h, the cell viability remained above 95%. These findings indicated that PPI-FA-US15 possessed a favorable biosafety, making it suitable for food applications.

To further evaluate the biological activity of curcumin after encapsulation, the in vitro antiproliferative activity of free curcumin, and PPI-FA15-Cur_0.75 against tumor cells (HCT-116) was investigated. In Fig. 8H, as curcumin concentration increased, HCT-116 cells showed a dose-dependent antiproliferative effect. Within the studied concentration range, PPI-FA15-Cur_0.75 demonstrated stronger antiproliferative activity than free curcumin. The results demonstrated that encapsulation within PPI-FA-US15 significantly enhanced the inhibition of HCT-116 cell proliferation of curcumin, which could be attributed to the fact that the FA could bind with the folate receptor on the HCT-116 cells, and greatly promote the cellular uptake of curcumin.

To validate the enhancement of active ingredient cellular uptake efficiency by nanoencapsulation, PPI-FA-US15 was labeled with C6 (a hydrophobic fluorophore). Free C6, PPI-C6, and PPI-FA15-C6 were incubated with cells for 4 h. The fluorescence intensity of PPI-FA15-C6 (Fig. 8M3) in cells was significantly higher than that of free C6 (Fig. 8K3) and PPI-C6 (Fig. 8L3), and exhibited a broader distribution range. As shown in Fig. 8I, the mean fluorescence intensity of PPI-FA15-C6 was about 3.07-fold and 1.93-fold higher than that of free C6 and PPI-C6, respectively. The results showed that PPI-FA-US15 could effectively improve the affinity with cells and absorption efficiency of active substances.

4. Conclusion

This study successfully prepared PPI-FA conjugates through an ultrasonic-assisted free radical grafting method, and achieved the efficient encapsulation of curcumin. Ultrasonic treatment greatly enhanced the efficiency of covalent bonding between FA and PPI. The binding of FA enhanced the solubility of PPI, and increased the loading capacity of PPI for curcumin. The PPI-FA complexes showed a higher binding affinity to curcumin than native PPI, which was supported by the results of fluorescence quenching. Meanwhile, the hydrogen bond and hydrophobic interaction played crucial roles in the formation of PPI-FA15-Cur nanocomplexes. The photostability, thermal stability and antioxidant properties of curcumin were remarkably enhanced after encapsulation within PPI-FA conjugates. Moreover, the PPI-FA15-Cur nanocomplexes showed a sustained and precise release of curcumin during the in vitro digestion process. Overall, these findings offer novel insights into protein-based nanocarrier systems for the precise and efficient delivery of hydrophobic nutrients.

5. Data availability

Data will be made available on request.

CRediT authorship contribution statement

Zijun Wang: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Data curation, Conceptualization. Huan Li: Methodology, Investigation, Conceptualization. Hanlu Yu: Formal analysis, Data curation. Xinyao Wang: Methodology, Investigation. Jia Guo: Formal analysis, Data curation. Jia Qing: Methodology, Investigation. Haiying Yang: Software, Methodology, Conceptualization. Xiaoqing Xiao: Methodology, Investigation. Rongrong Wang: Methodology, Investigation. Yang Shan: Writing – review & editing, Project administration, Funding acquisition, Conceptualization. Shenghua Ding: Writing – review & editing, Resources, Project administration, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.