Structural, color, and stability properties of hydrogen peroxide-treated phycoerythrin-based colorant proteins from dried laver (Pyropia yezoensis)

Abstract

This study aimed to develop a strategy for use of the water-soluble protein fraction extracted from laver biomass to establish a seaweed biorefinery system. By leveraging the oxidative reactivity differences of hydrogen peroxide (H₂O₂), a phycoerythrin (PE)-based colorant protein (PBCP) was produced from the water-soluble proteins of dried laver (Pyropia yezoensis). The highest a* value and PE/PC ratio (13.85) for PBCP were achieved after a 1.5 h treatment with 4.5 mol/L H₂O₂. The protein content (3.48 mg/mL) of the PBCP was approximately 1.5-fold to 2-fold greater than that of PE obtained through ammonium sulfate precipitation and ion-exchange chromatography. H₂O₂ treatment induced interactions between proteins and saccharides in the water-soluble protein, causing aggregation through β-structures primarily stabilized by disulfide bonds. These structural changes led to enhanced emulsifying properties and color stability under various temperature, light, and pH conditions. These findings establish PBCP as a stable, seaweed-derived protein pigment for food applications.

Graphical abstract

Highlights

• •Oxidative process for phycoerythrin (PE)-based colorant protein (PBCP) examined.
• •H₂O₂ treatment degraded phycocyanin, yielding red PBCP from Pyropia yezoensis.
• •PBCP had higher protein than PE from (NH₄)₂SO₄/ion exchange process.
• •H₂O₂-induced aggregation and structural rearrangement enhanced PBCP stability.
• •PBCP offers a natural pigment for food applications from seaweed biorefineries.

1. Introduction

The seaweed-based biorefinery industry has attracted considerable attention because of its high sustainability and the consistent availability of carbohydrate-rich biomass (Baghel, 2023; Yun et al., 2015). Red seaweeds, including laver (Pyropia spp.), are rich in galactose-based polysaccharides, making them excellent candidates for use as biorefinery feedstocks. Studies on laver-based biorefineries have primarily focused on polysaccharide utilization to produce biofuels, biodegradable plastics, and similar products (Greetham et al., 2020; Qiang et al., 2025). However, the establishment of a comprehensive laver biorefinery requires not only polysaccharide utilization but also strategies for the valorization of water-soluble protein fractions generated before and after polysaccharide extraction.

Water-soluble proteins derived from laver are rich in pigment proteins, including phycobiliproteins (Basri et al., 2025; Lee et al., 2024). Phycoerythrin (PE), a major phycobiliprotein in red algae, has considerable potential as a natural dye, fluorescent material, and bioactive ingredient because of its high molar extinction coefficient, broad excitation spectrum, and superior fluorescence properties. However, PE is sensitive to physicochemical factors such as heat and light. Therefore, it is necessary to establish extraction and fractionation conditions that maintain its color and structural integrity. (Mittal et al., 2017).

Ammonium sulfate (AS) precipitation, ion-exchange (IE) chromatography, membrane filtration, and aqueous two-phase systems have been extensively used to extract and purify PE from seaweeds (Devi et al., 2025; Ji et al., 2022; Liu et al., 2005; Sun et al., 2009). However, these methods have limitations, including poor selectivity, pigment degradation, lengthy processing durations, and significant equipment and operational costs, which impede their industrial-scale use. Therefore, developing a selective fractionation strategy that minimizes processing stages while maintaining the color and structural integrity of PE is essential.

Hydrogen peroxide (H₂O₂) treatment leads to oxidative and structural changes in proteins, and when appropriately controlled, the varying oxidative susceptibilities of proteins can be exploited for selective denaturation and fractionation (Meissner et al., 2026). Moreover, oxidation-induced structural modifications can promote protein aggregation and alter the reactivity of functional groups, which may enhance functional properties such as emulsifying capacity, foaming properties, and processing stability (Aider et al., 2023; Akbarbaglu et al., 2025). However, residual H₂O₂ may cause continued oxidation, resulting in excessive structural changes and decreased storage stability. Catalase effectively terminates oxidative reactions by decomposing residual H₂O₂ into water and oxygen. The combined use of H₂O₂ and catalase has been demonstrated to be applicable and safe in various food-related systems, including cheese, eggs, and fish roe (EFSA Panel on Food Contact Materials et al., 2023). This H₂O₂ treatment, combined with catalase-mediated reaction termination, offers a simplified technique for the selective fractionation of specific proteins from complex water-soluble protein extracts of laver.

This study aimed to develop an H₂O₂–catalase-based selective fractionation process to produce a PE-containing colorant fraction with enhanced physicochemical stability from laver extracts. The color and structural characteristics of the resulting PE, its stability under different processing conditions, and its functional properties relevant to food applications were evaluated. The findings of this study provide a process-based foundation for the high-value utilization of water-soluble protein fractions in laver-based biorefineries and may contribute to expanding the potential use of PE-containing pigments as functional ingredients in foods.

2. Materials and methods

2.1. Sample preparation

Dried laver (P. yezoensis) was purchased from a local market in Seocheon, South Korea. The samples were pulverized using an electric blender (SHMF-3370SS, Hanil Co., Ltd., Seoul, Korea), sieved through a 400-mesh sieve, and stored at −20 °C until analysis.

2.2. Methods

2.2.1. Extraction of water-soluble protein

To extract water-soluble proteins, powdered dried laver was mixed with distilled water in a 1:20 (w/w) ratio and stirred for 24 h at 4 °C. The mixture was then filtered through a 400-mesh sieve and centrifuged at 10,000 ×g and 4 °C for 15 min to remove insoluble particles. The resulting supernatant, referred to as Pyropia extract (control), was collected for further analysis.

2.2.2. Preparation of phycoerythrin (PE)-based colorant protein (PBCP) and PE samples

To prepare the PE-based colorant protein (PBCP), Pyropia extract was treated with H₂O₂ (Sakai et al., 2022). Briefly, 30% (w/w) H₂O₂ was added to the extract to achieve concentrations ranging from 0.5 to 5.0 mol/L, and the mixture was incubated for 8 h at 25 °C. During incubation, samples were collected at 30 min intervals for up to 3 h and subsequently at 1-h intervals for up to 8 h. The residual H₂O₂ was decomposed by adding catalase to achieve a concentration of 60 U/mL. The samples were kept in an ice bucket to minimize temperature fluctuations. The chilled samples were centrifuged at 10,000 ×g and 4 °C for 15 min to remove denatured proteins. Samples obtained after H₂O₂- and catalase treatment were defined as PBCP and compared with PE samples prepared using AS precipitation and IE chromatography.

PE was recovered from the Pyropia extract through a sequential AS precipitation process (Kumar et al., 2025). Initially, the extract was adjusted to 15% concentration with saturated AS and centrifuged at 10,000 ×g for 15 min to remove the precipitated non-colorant proteins. The supernatant was then adjusted to 30% saturation and centrifuged under identical conditions to remove PC. Finally, the supernatant was increased to 40% saturation, and the resulting precipitate (PE) was collected by centrifugation at 10,000 ×g for 15 min. The precipitate was then suspended and dialyzed against distilled water for 48 h at 4 °C, with water being replaced every 8 h. The PE obtained after 40% saturation and completing dialysis was collected for further analysis (AS).

The dialyzed AS sample was further purified using DEAE-Sepharose at pH 6.2 (Simovic et al., 2022). A SciLab® DURAN glass chromatography column (30Ф, 400 mm, 280 mL) was packed with the resin, equilibrated with a 20 mmol/L sodium phosphate buffer (pH 6.2), and PE was eluted using the same buffer supplemented with 150 mmol/L sodium chloride. The purified PE was dialyzed in distilled water for 48 h at 4 °C under identical conditions and collected for further analysis (IE).

2.2.3. Characterization of PBCP and PE samples

2.2.3.1. PC, PE, and allophycocyanin (APC) contents

The concentrations of PC, PE, and allophycocyanin (APC) were determined using a colorimetric method (Vieira et al., 2025). The PBCP and PE samples were scanned across a wavelength range from 450 to 700 nm. using a UV–visible spectrophotometer (μ2; Microdigital Co., Ltd., Seongnam, Korea). PC, PE, and APC contents were calculated using Eqs. (1), (2), (3):

$$\mathit{PC}\left(\mathit{mg}/\mathit{mL}\right)=\frac{{\mathit{Abs}}_{\mathbf{620}}-\mathbf{0.7}{\mathit{Abs}}_{\mathbf{650}}}{\mathbf{7.38}}$$ (1)

$$\mathit{APC}\left(\mathit{mg}/\mathit{mL}\right)=\frac{{\mathit{Abs}}_{\mathbf{650}}-\mathbf{0.19}{\mathit{Abs}}_{\mathbf{620}}}{\mathbf{5.56}}$$ (2)

$$\mathit{PE}\left(\mathit{mg}/\mathit{mL}\right)=\frac{{\mathit{Abs}}_{\mathbf{565}}-\mathbf{2.8}\mathit{PC}\left(\mathit{mg}/\mathit{mL}\right)-\mathbf{1.34}\mathit{APC}\left(\mathit{mg}/\mathit{mL}\right)}{\mathbf{12.7}}$$ (3)

2.2.3.2. Protein yield

Protein concentration was determined according to the Biuret method (Lee et al., 2025) using a UV–visible spectrophotometer (Microdigital Co., Ltd.) set at 540 nm, with bovine serum albumin as the standard.

2.2.3.3. H₂O₂ content

The concentration of H₂O₂ was determined using a previously published method with minor modifications (Rajamma et al., 2021). PBCP was mixed with Reagent A (0.4 M KI, 0.05 M NaOH, and 1.6 × 10⁻⁴ M [NH₄]₆Mo₇O₂₄·4H₂O and Reagent B (0.1 M KHC₈H₄O₄) in a 1:1:1 (v/v/v) ratio and vortexed for 30 s. The absorbance was measured at 350 nm using a UV–visible spectrophotometer (μ2; Microdigital), and H₂O₂ concentration was calculated using the Beer–Lambert law (4), as follows:

$$\mathbf{Concentration\; of}\mathbf{H}₂\mathbf{O}₂\left(\mathit{mol}\right)=\frac{{\mathit{Abs}}_{\mathbf{350}}}{\mathit{\epsilon }\mathit{l}}\times \mathit{D}$$ (4)

where Abs₃₅₀ is the absorbance at 350 nm, ε is the molar absorptivity of the triiodide complex (26,400 mol⁻¹ cm⁻¹), l is the optical path length (1 cm), and D is the dilution factor of the sample.

2.2.3.4. Color

The color of the extracted protein was measured using a Lovibond RT Series colorimeter (Tintometer Ltd., Amesbury, UK), as described by Lee et al. (2025). The results were represented in three-dimensional CIE Lab* coordinates: L* for lightness, a* for redness, and b* for yellowness. The standard plate values were L* = 94.91, a* = 0.17, and b* = 4.77. The average values were calculated from three independent measurements for each sample.

2.2.4. Quantum chemical calculation

Quantum-chemical computations were performed to evaluate the effect of H₂O₂ treatment on the chromophore in PBCP, as previously described by Rajan et al. (2018). Phycoerythrobilin (PEB) and phycocyanobilin (PCB), the chromophores of PE and PC, respectively, were modeled using the Avogadro software (v1.2.0). Geometry optimization and quantum-chemical calculations were executed with the ORCA program (v6.0.1) at the B3LYP/6–31 + G(d,p) level of density functional theory (DFT) for both the ground (0) and cationic (+1) states. Aqueous conditions were modeled using the CPCM (water) solvation model, and all optimizations adhered to the TightSCF convergence criteria. The adiabatic ionization potential (AIP), vertical ionization potential (VIP), and reorganization energy (RE⁺) were calculated using Eqs. (5), (6), (7):

$$\mathit{AIP}={\mathit{E}}_{\mathit{C}}-{\mathit{E}}_{\mathit{N}}$$ (5)

$$\mathit{VIP}={\mathit{E}}_{\mathit{CN}}-{\mathit{E}}_{\mathit{N}}$$ (6)

$${\mathit{RE}}^{+}=\mathit{VIP}-\mathit{AIP}$$ (7)

where E_N and E_C denote the energies of the optimized neutral and cationic species in their most stable geometries, respectively, and E_CN represents the energy of the charged state at the optimized neutral geometry. The optimized geometries, highest occupied molecular orbital (HOMO), and singly occupied molecular orbital (SOMO) states of PEB and PCB were visualized using the IboView software (v20211019-RevA).

2.2.5. Surface morphology of PBCP and PE

The surface morphologies of the PBCP and PE proteins were examined using field-emission scanning electron microscopy (FE-SEM; SIGMA 560, Carl Zeiss AG, Land Baden-Württemberg, Germany), as described previously, with some modifications (Minuti et al., 2023). The samples were diluted with distilled water to a concentration of 1 mg/10 mL, and 5 μL aliquots were carefully added dropwise onto siliconized glass slides (HR3–231, Hampton Research Co., CA, USA). The samples were dried in a desiccator for 48 h at 25 °C, sputter-coated with platinum, and observed under FE-SEM at an accelerating voltage of 2 kV.

2.2.6. Molecular characteristics of PBCP and PE samples

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

The molecular weights of PBCP and PE samples were determined using sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) (Lee et al., 2025). A stacking gel with 5% acrylamide and a separating gel with 15% acrylamide were prepared. Electrophoresis was conducted using a Mini-PROTEAN Tetra Cell (Bio-Rad, Hercules, CA, USA). Each well was loaded with a 10 μL sample containing 5 μg of protein; the voltage was initially set to 60 V for 30 min and then increased to 110 V for 2 h to facilitate protein migration. The molecular weights were assessed by comparing them with standard protein markers (10–250 kDa; Thermo Fisher Scientific Inc., Waltham, MA, USA).

2.2.6.2. Peptide mass fingerprinting (MALDI-TOF and TOF/TOF)

Protein spots excised from SDS-PAGE gels were enzymatically digested in-gel using modified porcine trypsin (Promega), as described previously (Shevchenko et al., 1996). Gel pieces were washed with 50% acetonitrile to remove SDS, salts, and stains, dehydrated, vacuum-dried, rehydrated with trypsin (8–10 ng/μL) in 50 mM ammonium bicarbonate (pH 8.7) and then digested for 8–10 h at 37 °C. The samples were analyzed using MALDI-TOF mass spectrometry with LIFT™ ion optics (Bruker autoflex max, Bruker Daltonics GmbH, Bremen, Germany). MS and MS/MS data were acquired using a smart beam laser (2 kHz repetition rate) with up to 4000 shots per spectrum. The MS/MS mode was operated at 2 keV collision energy with air as the collision gas under nominal single-collision conditions. The resolution for precursor selection was set to 100 for single-component analytes. MS and MS/MS data were acquired using the instrument's default calibration without any internal or external calibration. MS/MS ion searches were performed using Mascot (Matrix Science, London, UK) against the NCBInr database, with taxonomy limited to Porphyra (16,826 sequences; 6,819,571 residues).

2.2.6.3. Fourier transform-infrared (FT-IR) spectroscopy

Freeze-dried PBCP and PE samples were analyzed using a Spectrum 3 Fourier transform-infrared (FT-IR) spectrometer (PerkinElmer, Inc., Waltham, MA, USA), as described previously (Sun et al., 2023). FT-IR spectra were captured at 25 °C across a wavelength range of 4000–650 cm⁻¹ with a resolution of 1 cm⁻¹ and involved four repeated scans. Each sample was measured in triplicate. Spectral data were processed and corrected using Origin 2025 (OriginLab Corporation, Northampton, MA, USA) to determine the protein secondary structures.

2.2.6.4. Free sulfhydryl content

The free sulfhydryl content of PBCP was determined using the 5,5′-dithiobis (2-nitrobenzoic acid) (DTNB) method (Ellman's method), with some modifications (Xiong et al., 2025). The absorbance was recorded at 412 nm using a UV–visible spectrophotometer (Microdigital Co., Ltd.). The free sulfhydryl content was calculated using Eq. (8):

$$\mathbf{Free\; sulfhydryl\; content}\left(\mathit{nmol}/\mathit{mg}\right)=\frac{{\mathit{A}}_{\mathbf{412}}\times \mathbf{73.53}\times \mathit{D}}{\mathit{C}}$$ (8)

where A₄₁₂ is the absorbance at 412 nm, C is the protein concentration (mg/mL), 73.53 is the molar absorptivity, and D is the dilution factor of the sample.

2.2.6.5. Carbonyl content

The carbonyl content of PBCP was determined using the 2,4-Dinitrophenylhydrazine (DNPH) method, with some modifications (Xiong et al., 2025). The absorbance was recorded at 370 nm using a UV–visible spectrophotometer (Microdigital Co., Ltd.). The carbonyl content was calculated using Eq. (9):

$$\mathbf{Carbonyl\; content}\left(\mathit{nmol}/\mathit{mg}\right)=\frac{{\mathit{A}}_{\mathbf{370}}\times \mathbf{45.45}\times \mathit{D}}{\mathit{C}}$$ (9)

where A₃₇₀ is the absorbance at 370 nm, C is the protein concentration (mg/mL), 45.45 is the molar absorptivity, and D is the dilution factor of the sample.

2.2.6.6. Schiff base content

The Schiff base content of PBCP was determined using fluorescence spectroscopy (Fluorolog-QM, Horiba Co. Ltd., Kyoto, Japan) (Utrera & Estévez, 2013). An aliquot (1 mL) of PBCP was diluted in 9 mL of 20 mM sodium phosphate buffer (pH 7.4). The excitation wavelength was set at 350 nm, and the emission spectrum was scanned from 400 to 500 nm using a UV–visible spectrophotometer (Microdigital Co., Ltd.). The scanning speed was set to 60 nm/min, and the excitation and emission slit widths were set to 1 nm.

2.2.7. Physicochemical characteristics of PBCP and PE samples

2.2.7.1. Determination of emulsifying activity index (EAI) and emulsion stability index (ESI)

The emulsifying activity index (EAI) and emulsion stability index (ESI) were determined using a turbidimetric method (Zhang et al., 2023). The pH of PBCP and PE samples was adjusted to 6.2, and the samples were diluted to a concentration of 0.5 mg/mL in distilled water. The diluted samples were mixed with soybean oil (4:1, w/w) and homogenized at 10,000 rpm for 1 min. A portion of the homogenate (100 μL) was added to 5 mL of a 0.1% SDS solution and gently agitated, and its turbidity was measured at 500 nm (A₁) using a UV–visible spectrophotometer (Microdigital Co., Ltd.). After allowing the mixture to stand for 10 min, the homogenates were mixed with SDS solution, and the turbidity was measured (A₂). The EAI and ESI values were calculated using Eqs. (10), (11), respectively:

$$\mathit{EAI}\left({\mathit{m}}^{\mathbf{2}}/\mathit{g}\right)=\frac{\mathbf{2}\times \mathbf{2.303}\times {\mathit{A}}_{\mathbf{1}}\times \mathit{D}}{\mathit{C}\times \mathit{l}\times \left(\mathbf{1}-\mathit{\theta }\right)\times \mathbf{10},\mathbf{000}}$$ (10)

$$\mathit{ESI}\left(\%\right)=\frac{{\mathit{A}}_{\mathbf{1}}}{{\mathit{A}}_{\mathbf{2}}}\times \mathbf{100}$$ (11)

where D is the dilution factor, C is the protein concentration (g/mL), l is the optical path length (1 cm), and θ is the ratio of the oil phase in the emulsion (θ = 0.25).

2.2.7.2. Zeta potential

PBCP and PE samples were adjusted to pH 6.2 and diluted with distilled water to achieve a protein concentration of 0.25 mg/mL. Zeta potential was determined at 25 °C using dynamic light scattering (Litesizer500, Anton Paar GmbH, Graz, Austria) (Yu et al., 2023). The solvent refractive index was maintained at 1.3303, with an equilibration time of 30 s.

2.2.7.3. Surface hydrophobicity (SH) determination

Surface hydrophobicity (SH) was measured as previously described, with minor modifications (Zhou & Yang, 2020). Absorbance was measured at 595 nm using a UV spectrometer (Microdigital Co., Ltd.). SH was calculated using Eq. (12):

$$\mathit{SH}\left(\mathit{\mu g}\right)=\frac{\mathbf{200}\mathit{\mu g}\times \left({\mathit{A}}_{\mathbf{0}}-{\mathit{A}}_{\mathbf{1}}\right)}{{\mathit{A}}_{\mathbf{0}}}$$ (12)

where A₀ and A₁ are the absorbance values of the control and sample groups, respectively. 200 μg is the total amount of bromophenol blue initially added.

2.2.7.4. Stability of PBCP and PE samples

PBCP and PE samples were adjusted to pH 6.2 and diluted to a concentration of 0.3 mg/mL with distilled water to assess their stability to heat, irradiation, and pH (Zhang et al., 2023).

The thermal stability was assessed by monitoring the thermal degradation of the samples by measuring the absorbance at 565 nm using a UV–visible spectrophotometer (Microdigital Co., Ltd.). The samples were incubated in a water bath at 25, 40, 60, and 80 °C for 30 min. The thermal stability was calculated using Eq. (13):

$${\mathit{Thermal\; stability}}_{\mathit{t}}=\frac{{\mathit{A}}_{\mathbf{565},\mathit{t}}}{{\mathit{A}}_{\mathbf{565},\mathbf{0}}}$$ (13)

where ${\mathit{A}}_{\mathbf{565,0}}$is the absorbance at 565 nm before thermal treatment at 25 °C, and ${\mathit{A}}_{\mathbf{565}\mathit{,}\mathit{t}}$is the absorbance at 565 nm after thermal treatment.

Photostability was assessed by exposing the samples to natural light or UV irradiation (254 nm) at 25 °C for 12 h. The absorbance at 565 nm was recorded every 3 h. Photostability was calculated using Eq. (14):

$${\mathit{Photostability}}_{\mathit{t}}=\frac{{\mathit{A}}_{\mathbf{565},\mathit{t}}}{{\mathit{A}}_{\mathbf{565},\mathbf{0}}}$$ (14)

where ${\mathit{A}}_{\mathbf{565,0}}$is the absorbance at 565 nm before light exposure, and ${\mathit{A}}_{\mathbf{565}\mathit{,}\mathit{t}}$is the absorbance at 565 nm after $\mathit{t}$h of natural light or UV exposure.

To assess pH stability, the samples were adjusted to pH 3, 5, 7, 9, and 11 using 1 mol/L HCl or NaOH, incubated at 4 °C for 30 min, and the absorbance was measured at 565 nm. pH stability was calculated using Eq. (15):

$$\mathit{pH}{\mathit{stability}}_{\mathit{p}}=\frac{{\mathit{A}}_{\mathbf{565},\mathit{p}}}{{\mathit{A}}_{\mathbf{565},\mathbf{0}}}$$ (15)

where ${\mathit{A}}_{\mathbf{565,0}}$is the absorbance at 565 nm before pH adjustment, and ${\mathit{A}}_{\mathbf{565}\mathit{,}\mathit{p}}$is the absorbance at 565 nm after pH adjustment.

2.2.8. Statistical analysis

All experiments were performed in triplicate, and the results are expressed as mean ± standard deviation (SD). Statistical analysis was performed using the SPSS software (v27; IBM Corp., Armonk, NY, USA). A one-way analysis of variance (ANOVA) followed by Tukey's multiple-comparison test was applied, with differences considered significant at P < 0.05.

3. Results and discussion

3.1. Effects of H₂O₂ on the color and phycobiliprotein contents of PBCP

The rates of change in color intensity and phycobiliprotein content of PBCP increased with increasing H₂O₂ concentrations. The L* values of PBCP increased consistently with reaction duration, particularly at ≥3.0 mol/L H₂O₂ (Fig. 1A). Initially, the a* value increased to ∼35 before declining, with the peak occurring earlier at higher H₂O₂ concentrations (Fig. 1B). Conversely, the b* values remained relatively stable regardless of H₂O₂ concentration and reaction duration (Fig. 1C).

Fig. 1.

Changes in color characteristics, residual hydrogen peroxide (H₂O₂) content, and phycobiliprotein levels in PE-based colorant protein (PBCP) treated with different H₂O₂ concentrations (0.5–5.0 mol/L) and reaction times: (A) L* value, (B) a* value, (C) b* value, (D) residual H₂O₂ concentration after catalase addition, (E) phycoerythrin (PE) concentration, and (F) phycocyanin (PC) concentration.

The phycobiliprotein content steadily declined during the H₂O₂ reaction (Fig. 1E and F), with PC oxidizing more rapidly than PE. At 5.0 mol/L H₂O₂, approximately 90% of the PC was removed within 1 h, whereas approximately 60% of the PE remained. At 4.5 mol/L H₂O₂, approximately 94% of the PC was removed after 1.5 h, whereas approximately 45% of the PE remained. The increase in L* values likely reflects protein oxidation and decolorization, along with increased light scattering because of changes in particle size. The initial increase in a* values can be attributed to the selective oxidation of PC, which accentuated the red color of the remaining PE. Subsequent oxidation of PE led to a decrease in a* values.

Residual H₂O₂ was removed by catalase treatment. In samples treated with ≤4.5 mol/L H₂O₂, no residual H₂O₂ was detected post-treatment (Fig. 1D). However, under the 5.0 mol/L treatment, approximately104.3 mg/kg of H₂O₂ persisted, and the levels were approximately 109.6 mg/kg after 30 min of catalase treatment. Catalase decomposes H₂O₂ into H₂O and O₂; however, its activity can be reduced or nullified in highly concentrated H₂O₂ solutions (Sakai et al., 2022; Sooch et al., 2014). Consequently, 4.5 mol/L was chosen as the optimal concentration for preparing PBCP to ensure consistent reaction termination and product safety.

3.2. Comparison of quantum chemical stability of PE and PC

The distinct red and blue colors of PE and PC are determined by their respective chromophores, PEB and PCB. The reactivity of these chromophores depends on their chemical structure and conjugation length (Mroginski et al., 2021). Quantum chemical calculations based on localized chromophore modeling can provide insights into protein changes in response to physicochemical treatments (Huang et al., 2026). The optimized geometries and orbital distributions of PCB and PEB in the HOMO and SOMO states are shown in Fig. 2.

Fig. 2.

Optimized structures were obtained using the B3LYP/6–31 + G(d,p) level of theory and frontier molecular orbitals—highest occupied molecular orbital (HOMO) and singly occupied molecular orbitals (SOMO-α and SOMO-β). Orbital isosurfaces are colored to indicate opposite phases (HOMO: blue and purple; SOMO: orange and pink). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

PCB has nine conjugated double bonds extending to the D-pyrrole ring, whereas PEB has six conjugated double bonds. Typically, an increase in conjugation results in a shift in the absorption wavelength of the red peak (Dagnino-Leone et al., 2022) and a greater number of double bonds enhances π-electron density and electron delocalization. Thus, longer conjugated system of PCB suggests that it is structurally more susceptible to oxidation than PEB. The quantum chemical energies associated with PCB and PEB oxidation are summarized in Table 1. Lower ΔE and RE⁺ values signify higher reactivity, whereas lower VIP indicates more favorable oxidation (Bakheit et al., 2024; Safi et al., 2022). For PEB and PCB, the ΔE values were 2.838 and 2.364 eV, the VIP values were 4.955 and 4.929 eV, and the RE⁺ values were 0.187 and 0.150 eV, respectively. These findings suggest that PCB can be ionized with slightly less energy than PEB, establishing PCB as the more chemically vulnerable chromophore.

Table 1.

Quantum chemical characteristics of the phycobilin chromophore in neutral and cationic states.

	HOMO1)	LUMO2)	ΔE	VIP3)	AIP4)	RE+
Phycoerythrobilin	−5.158	−2.320	2.838	4.955	4.768	0.187
Phycocyanobilin	−5.133	−2.769	2.364	4.929	4.779	0.150

Note: reorganization energy, All Values are in eV.

¹Highest occupied molecular orbital.

²Singly occupied molecular.

³Vertical ionization potential.

⁴Adiabatic ionization potential.

3.3. Comparison of PBCP and PE samples

To evaluate the characteristics of PBCP relative to those of conventionally purified PE samples, PBCP produced after treatment with 4.5 mol/L H₂O₂ was collected at different reaction intervals. The samples collected at 0, 0.5, 1.5, 3.0, and 5.0 h were designated as Control, PBCP0.5, PBCP1.5, PBCP3.0, and PBCP5.0, and compared to PE produced via AS precipitation and IE chromatography.

3.3.1. Changes in the color of PBCP and PE samples

The color of PBCP and PE samples varied distinctly depending on the H₂O₂ reaction duration and the type of sample (Fig. 3A). PBCP0.5 appeared reddish-purple, PBCP1.5 exhibited a vivid red hue, and extended H₂O₂ treatment gradually decolorized the samples to a yellowish-white color. In contrast, the AS and IE samples appeared dark purple and pink, respectively These color differences likely reflect the combined contributions of native phycobiliproteins, particularly PE and PC, as well as proteins and polysaccharides in the Pyropia extract. Color analysis showed that the L* value increased progressively with H₂O₂ reaction time, whereas the a* value was reduced after PBCP1.5 (Fig. 3B). The L* and a* values of the AS and IE samples were comparable to those of PBCP0.5 and PBCP1.5, respectively, but their b* values were markedly lower. The higher b* values in PBCP can be attributed to yellowing and changes in color caused by protein oxidation during H₂O₂ treatment (Sakai et al., 2022).

Fig. 3.

Appearance, color parameters, phycobiliprotein and total protein contents, and UV–Vis absorption spectra of Pyropia extracts (control), PE-based colorant protein (PBCP) treated with 4.5 mol/L H₂O₂ for different reaction times (0.5–5.0 h), and PE samples obtained through ammonium sulfate (AS) precipitation, and ion-exchange (IE) chromatography: (A) visual appearance, (B) color parameters, (C) PE, PC, and total protein contents, and (D) UV–Vis absorption spectra. Different letters indicate significant differences at P < 0.05.

3.3.2. Concentration of phycobiliproteins and UV–vis spectra of PBCP and PE samples

Changes in the PE, PC, and total protein contents of the PBCP and purified PE samples were evaluated to determine the optimal H₂O₂ treatment condition by balancing the selective removal of PC against the retention of PE and total protein (Fig. 3C). During the first 1.5 h of treatment, PC was preferentially oxidized. In PBCP1.5, the PC and PE contents decreased by 96% and 70%, respectively, compared with those of the control, indicating that most of the PC was removed After 1.5 h, however, PE degradation accelerated markedly. PBCP3.0 retained only 6% of the initial PE content.

Notably, PBCP1.5 exhibited the highest PE/PC ratio of 13.85 and the highest a* value (Fig. 3B). In addition, PBCP1.5 retained a total protein concentration of 3.48 mg/mL, which was higher than those of the AS and IE-purified PE samples (<2.0 mg/mL). These results indicate that 1.5 h of H₂O₂ treatment was the most promising condition for producing a PE-rich red colorant protein.

The UV–Vis spectra showed distinct absorption peaks at approximately 495, 565, and 620 nm in the untreated Pyropia extract, corresponding to PE and PC (Vieira et al., 2025) (Fig. 3D). In PBCP, these peaks diminished in intensity and broadened with H₂O₂ treatment, with the 620-nm peak disappearing after 1.5 h of treatment. In contrast, AS and IE maintained absorption peaks at 495, 565, and 620 nm, suggesting that PC in PBCP was largely degraded and decolorized after 1.5 h of H₂O₂ oxidation.

3.3.3. Protein identification

Protein components in the Pyropia extract, PBCP, and PE samples were identified using SDS-PAGE, in-gel digestion, and peptide mass fingerprinting (PMF). In the Pyropia extract, bands at approximately 50–60, 50, and 15–20 kDa were identified as RubisCO, PE, and PC subunits, respectively (Fig. 4A and C). In AS and IE, bands were detected primarily in the 10–25 kDa range, with no bands near 50 kDa (Fig. 4D and E). PMF analysis identified the 15–20 kDa bands as PE and PC subunits.

Fig. 4.

SDS-PAGE patterns and protein identification using MALDI-TOF-PMF and MALDI-TOF/TOF-PMF: (A) SDS-PAGE profiles of PE-based colorant protein (PBCP) and PE obtained using ammonium sulfate precipitation (AS), and ion-exchange (IE) chromatography, (B) changes in band intensity of PBCP, (C) MALDI-TOF protein identification of Pyropia extracts (control), (D) MALDI-TOF protein identification of PE obtained using AS, (E) MALDI-TOF protein identification of PE obtained using IE, and (F) MALDI-TOF protein identification of PBCP (4.5 mol/L, 1.5 h).

The SDS-PAGE profile of PBCP was comparable to that of Pyropia extract, showing bands corresponding to RubisCO, PE, and PC subunits, along with an additional strong band at approximately 80–90 kDa (Fig. 4F). CP1 and CP3 could not be reliably identified by MALDI-TOF-based PMF alone, probably because oxidized, aggregated, or fused proteins produced complex peptide patterns after trypsin digestion. Further MALDI-TOF/TOF and PMF searches against fungal and Pyropia databases yielded peptide matches for these lineages. The intensity of the aggregate band at approximately 80–90 kDa increased with the duration of H₂O₂ reaction, whereas the PE subunit band (15–20 kDa) gradually decreased (Fig. 4B). This suggests that although the PBCP production process enables the stable recovery of PE, prolonged H₂O₂ treatment may progressively induce the formation of high-molecular-weight aggregates.

3.3.4. Surface morphology of PBCP and PE samples

The microstructures of PBCP and PE samples are shown in Fig. 5. The Pyropia extract featured oval-shaped particles averaging ∼0.5 μm in size, whereas IE-purified PE was composed of smaller spherical particles that aggregated into clusters (Fig. 5A and B). In PBCP, both particle size and aggregation increased as the H₂O₂ reaction time extended. PBCP0.5 contained small irregular particles, PBCP1.5 exhibited rod-like aggregates, and PBCP3.0 and PBCP5.0 were primarily composed of larger rod-like aggregates (Fig. 5C–F). The evolution of PBCP microstructure aligned with the changes in protein patterns observed through SDS-PAGE. The aggregation of particles was accompanied by the formation of three-dimensional structures through increased cross-linking, which can physically shield chemically vulnerable regions, such as chromophores, thereby enhancing physicochemical stability (Bharmoria et al., 2020).

Fig. 5.

Surface morphology of PE-based colorant protein (PBCP) treated with 4.5 mol/L H₂O₂ for different reaction times (0.5–5.0 h) and PE obtained using IE chromatography. Images were taken at 5000× (left, scale bar = 10 μm), 15,000× (middle, scale bar = 2 μm), and 50,000× (right, scale bar = 1 μm): (A) Pyropia extracts, (B) PE obtained using IE, (C) PBCP for 0.5 h, (D) PBCP for 1.5 h, (E) PBCP for 3.0 h, and (F) PBCP for 5.0 h.

3.3.5. Changes in the structural properties of PBCP and PE samples

The alterations in the FT-IR spectra of PBCP following H₂O₂ treatment reflect the structural modifications and chemical interactions during the oxidation process (Fig. 6A). The distinct peaks observed at 1211, 851, and 1037 cm⁻¹ in PBCP were attributed to the symmetric and asymmetric stretching vibrations of S O and C–O–S, respectively, whereas the peak at 919 cm⁻¹ corresponded to the C–O–C vibration of galactose derivatives, such as 3,6-anhydrogalactose (Hong et al., 2021). The peak at 1352 cm⁻¹ was identified as amide III, which is related to the interactions between polysaccharides and proteins (Feng et al., 2025). The increased intensities of the peaks at 851, 919, and 1352 cm⁻¹ compared with those of the IE and AS samples suggest that PBCP aggregation was accompanied by enhanced protein–polysaccharide interactions. In contrast to PBCP, the IE and AS samples exhibited distinct peaks corresponding to amide I, amide II, and 3290 cm⁻¹. The amide I and II peaks are associated with peptide bond vibrations, whereas the band at 3290 cm⁻¹ is related to hydrogen bonding (Chu et al., 2025). The relatively high intensities of the amide-related bands in the IE and AS samples suggest that proteins were enriched during purification.

Fig. 6.

Structural characteristics of Pyropia extracts (control), PE-based colorant protein (PBCP) treated with 4.5 mol/L H₂O₂ for different reaction times (0.5–5.0 h), and PE samples obtained using AS precipitation and IE chromatography: (A) FT-IR spectra (650–4000 cm⁻¹), and (B) secondary structure composition, (C) changes in carbonyl content and free sulfhydryl (SH) content of PBCP, (D) changes in fluorescence intensity of PBCP. Different letters indicate significant differences at P < 0.05.

Deconvolution analysis of the amide I band was performed to evaluate protein secondary structures. In PBCP0.5, there was an increase in the random coil content, and the β-structure content markedly increased with prolonged H₂O₂ treatment. These findings indicate that proteins initially undergo structural unfolding during H₂O₂ treatment, followed by reorganization into β-structure-rich aggregates through oxidative modification (Fig. 6B). The formation of protein aggregates rich in β-structures has been associated with disulfide bond formation and hydrophobic interactions (Lee et al., 2025; Wang et al., 2022) Conversely, AS exhibited a relatively high content of random coils, likely resulting from partial unfolding of the peptide structure caused by ammonium sulfate treatment. Edward et al. (2023) noted that ammonium sulfate can weaken intermolecular hydrogen bonds and disrupt protein secondary structures.

The free sulfhydryl and carbonyl groups and fluorescence intensity (FI) of PBCP indicate the chemical and structural transformations that occur during PBCP production. The free sulfhydryl content of PBCP decreased with increasing H₂O₂ treatment duration, suggesting that disulfide bond formation through thiol oxidation led to greater intermolecular crosslinking (Fig. 6C) (Niu et al., 2025). Before catalase treatment, the carbonyl content of PBCP increased up to 0.5 h and then declined after 1.5 h. Although carbonyl content is commonly used as an indicator of protein oxidation (Xiong et al., 2025), H₂O₂-induced aggregation may reduce detectable carbonyl levels by limiting DNPH access to carbonyl groups because of steric hindrance (Estévez et al., 2022). Furthermore, the formation of Schiff bases between carbonyl and amino groups during protein aggregation may further reduce the number of detectable carbonyl groups. After catalase treatment, the carbonyl content of PBCP decreased more prominently.

FI, indicative of Schiff base formation and oxidation-induced cross-links, increased with prolonged H₂O₂ treatment and exhibited a marked increase after catalase treatment. (Fig. 6D). This implies that oxidation, followed by catalase treatment, facilitated protein aggregation and enhanced intermolecular cross-linking, aligning with the noted reduction in carbonyl levels. The oxidative aggregation of PBCP was likely driven by a combination of disulfide bond formation through sulfhydryl oxidation, intermolecular β-sheet-rich associations, and non-covalent interactions. These structural rearrangements and cross-linking could lead to an increase in PBCP particle size and reinforce the internal network structure.

3.4. Physicochemical properties of colorants

3.4.1. Emulsion and aggregation properties of PBCP and PE samples

The EAI, ESI, zeta potential, and surface hydrophobicity of the PBCP and PE samples are shown in Fig. 7. Compared with the PE samples, PBCP exhibited a lower EAI and absolute zeta potential, suggesting reduced electrostatic repulsion between PBCP particles, increased tendency toward aggregation, and lower adsorption efficiency at the oil–water interface (Fig. 7A and B). The reduction in zeta potential, which signifies particle dispersion stability, with an increase in H₂O₂ treatment duration corroborates the progressive aggregation of PBCP. These changes may be attributed to the formation of larger aggregates and oxidation-induced changes in surface charge (Bielska et al., 2022).

Fig. 7.

Physicochemical properties and stability of Pyropia extracts (control), PE-based colorant protein (PBCP) treated with 4.5 mol/L H₂O₂ for different reaction times (0.5 and 1.5 h), and PE samples obtained using AS precipitation and IE chromatography: (A) emulsifying activity index (EAI) and emulsion stability index (ESI), (B) zeta potential and surface hydrophobicity, (C) changes in absorbance ratio (A_t/A₀, λ = 565 nm) following natural light irradiation, (D) changes in absorbance ratio (A_t/A₀, λ = 565 nm) following UV irradiation, (E) changes in absorbance ratio (A_p/A₀, λ = 565 nm) after different pH treatments, and (F) changes in absorbance ratio (A_t/A_25°C, λ = 565 nm) following thermal treatment. Different letters indicate significant differences at P < 0.05.

Conversely, ESI increased, and surface hydrophobicity decreased with extended H₂O₂ treatment duration (Fig. 7A and B). The reduction in surface hydrophobicity could stem from oxidation-induced structural rearrangements that limit the exposure of hydrophobic regions while enhancing the accessibility of hydrophilic regions (Wang et al., 2014). These changes may facilitate the formation of a denser and more stable interfacial emulsion layer (Zhang et al., 2023). Proteins and peptides can form three-dimensional interfacial networks with high elasticity and viscoelasticity (Schröder et al., 2017), while polysaccharides can interact with protein networks to enhance interfacial stability (Evans et al., 2013).

Overall, although polysaccharide-associated PBCP may initially exhibit a lower emulsifying capacity due to its larger particle size, the interactions between polysaccharides and the protein network can stabilize the interfacial network, thereby enhancing emulsion stability. The differences in the physical properties of PBCP and PE were consistent with the FT-IR spectral characteristics, indicating enhanced protein–polysaccharide interactions.

3.4.2. Thermal, light, and pH stability of colorant

The stability of PE in PBCP and PE samples when exposed to thermal, light, and pH conditions is shown in Fig. 7. Under both natural light and UV irradiation, the absorbance of PE decreased over time in all samples (Fig. 7C and D). Among the samples, PBCP0.5 exhibited the least photostability, with its absorbance falling below 50% after 12 h, whereas PBCP1.5 retained over 70% of its initial absorbance in both lighting conditions.

The photodegradation of PE under natural and UV light is attributed to photooxidation by reactive oxygen species and thermal denaturation during irradiation (Rinalducci et al., 2008; Ruzza et al., 2021). Aromatic amino acids, such as tryptophan and tyrosine, which have strong UV absorbance, can undergo photochemical reactions with excessive exposure (Pattison et al., 2012). Thus, the reduced photostability of PBCP0.5 may result from increased exposure of functional groups after partial oxidation, whereas aggregation in PBCP1.5 could reduce photosensitivity by concealing aromatic residues within the aggregates. The photostability of chromophores in colorant matrices can be enhanced by carbohydrate- or protein-mediated physical aggregation, which shields photosensitive moieties from light exposure (Liu et al., 2022; Liu et al., 2026). The higher photostability of PBCP1.5 compared with the control and PBCP0.5 coincided with the H₂O₂ treatment time-dependent increase in particle aggregation, suggesting that aggregate formation contributed to chromophore shielding.

The pH stability study revealed that the absorption of all samples peaked at pH 5 and declined under strongly acidic and alkaline conditions (Fig. 7E), consistent with previous findings on PE (Munier et al., 2014). PE is relatively stable within the pH range of 5–8; however, its structure and characteristics are altered at pH values below 5 and above 9 (Orta-Ramirez et al., 2000). PBCP exhibited greater absorbance retention than AS between pH 3 and 7 and surpassed IE in absorbance retention at pH 9, indicating a broader pH stability range than AS and IE. This enhanced stability likely reflects the oxidized and aggregated structure of PBCP, which partially mitigates the pH-induced structural changes.

During the thermal stability assessment, all samples remained stable at 20 and 40 °C but exhibited a significant decrease in stability at ≥60 °C (Fig. 7F). At 80 °C, AS and IE retained less than 40% of their initial absorbance, whereas PBCP0.5 and PBCP1.5 retained more than 70% absorbance, indicating superior thermal stability for PBCP. Protein aggregation can suppress excessive unfolding and safeguard chromophores within the aggregates, thereby enhancing resistance to external denaturing agents (Tounsi et al., 2023). Additionally, aggregation can elevate the glass transition temperature and particle compactness, contributing to thermal and photostability (Davies, 2016; Tounsi et al., 2023). Moreover, proteins with high molecular weights that are crosslinked may exhibit partially neutralized surface charges and reduced surface hydrophobicity, which enhances pigment stability against pH changes (Li et al., 2021). Therefore, the enhanced stability of PBCP is likely associated with oxidation-induced aggregation, increased structural compactness, and decreased exposure of sensitive chromophores.

4. Conclusions

In this study, a PE-based protein colorant (PBCP) was produced from an aqueous extract of dried laver (P. yezoensis) using H₂O₂ treatment. The H₂O₂ treatment selectively oxidized the PC chromophore, which was more susceptible to oxidation than PE, thus transforming the extract into a red protein pigment with a high a* value. The H₂O₂-induced protein oxidation and aggregation facilitated interactions between proteins and polysaccharides in the laver extract, resulting in the formation of β-structure-rich aggregates. The resulting PBCP exhibited enhanced physical properties, such as improved stability against heat, light, pH, and emulsions. Compared with conventional AS precipitation and IE chromatography, the H₂O₂-based process achieved approximately 1.5–2-fold higher protein recovery and demonstrated greater efficiency in producing PE-based pigments However, the mechanism underlying the enhanced stability of H₂O₂-produced PBCP, particularly concerning polysaccharide composition and binding structures, needs further investigation. These findings demonstrate the potential of PBCP as a stable seaweed-derived natural protein pigment for food applications and offer a promising strategy for the valorization of protein fractions in seaweed biorefinery systems.

CRediT authorship contribution statement

Jong Bong Lee: Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation. Min Jae Kim: Writing – original draft, Validation, Software, Methodology, Investigation, Formal analysis. Si Hwan Kim: Formal analysis. Jun Hee An: Formal analysis. Suk Kyung Sohn: Formal analysis. Eun Ji Park: Formal analysis. Ha Eun Park: Formal analysis. Kil Bo Shim: Writing – review & editing, Project administration, 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.

Data availability

Data will be made available on request.