LWT - Food Science and Technology 152 (2021) 112306

Contents lists available at ScienceDirect

LWT
journal homepage: www.elsevier.com/locate/lwt

Changes of antioxidative activities and peptidomic patterns of

Auxenochlorella pyrenoidosa protein hydrolysates: Effects of enzymatic
hydrolysis and decoloration processes
Wenhan Zhang a, Nan Jia a, Zihao Zhu b, Yanchao Wang a, *, Jingfeng Wang a, Changhu Xue a, c
a
College of Food Science and Engineering, Ocean University of China, Qingdao, 266003, China
b
College of Food Science and Engineering, Shandong Agricultural University, Tai’an, 271018, China
c
Laboratory for Marine Drugs and Bioproducts, Qingdao National Laboratory for Marine Science and Technology, Qingdao, 266237, China

A R T I C L E I N F O A B S T R A C T

Keywords: Enzymatic hydrolysis and decoloration are indispensable processes for the production of biological peptides from
Microalgae microalgae materials. Herein, this study aimed to investigate the effects of enzymatic hydrolysis and activated
Antioxidative activity carbon decoloration on antioxidative activities and peptidomic patterns of microalgae protein hydrolysates based
Peptidomic pattern
on a peptidomic approach. Protein hydrolysates with different hydrolysis time (2–16 h) showed significantly
Enzymatic hydrolysis
Decoloration
different antioxidative activities based on DPPH and ABTS assays. Regarding the peptidomic pattern, principal
component analysis (PCA) models revealed that hydrolysates were classified into discriminable clusters
depending on the hydrolysis and decoloration processes. Protein hydrolysates contained many aromatic amino
acid containing peptides, such as GDSLYPG, SADPETF, FPALN, FPGDY, and LTDWV, contributing to the clas­
sification of PCA models. Activated carbon decoloration at the concentration of 5 mg/mL significantly increased
the color lightness by 38.20%, reduced antioxidative activities by 29.78% and 33.59% based on DPPH and ABTS
methods, and reduced relative abundances of peptides, especially aromatic amino acid containing peptides.
Fifteen selected peptides showed strong ABTS radical scavenging abilities and ten of these peptides were
correlated with the antioxidative properties of hydrolysates. The monitoring of dynamic production processes at
a molecular scale would contribute to the production and quality control of bioactive peptides from Auxeno­
chlorella pyrenoidosa.

1. Introduction et al., 2018, 2021). Therefore, researchers are working to screen anti­
oxidative peptides from novel potential protein materials, and investi­
Excessive free radicals in the human body or food system are gate the relationship between biological activities and preparation
accountable to progress numerous harmful effects on human health or processes.
food quality (Alamed, Chaiyasit, Mcclements, & Decker, 2009; Stohs, Microalgae is recognized as one type of promising and novel bio­
1995). Many protein hydrolysates could help delaying or inhibiting the resources for food industry due to its rapid accumulation of valuable
oxidation process by scavenging radicals and chelating transition metals components, such as protein, carbohydrate, lipid and pigment (Van­
due to the presence of antioxidative peptides (Chai et al., 2021; Chen, thoor-Koopmans, Wijffels, Barbosa, & Eppink, 2013; Williams & Lau­
Ning, Jiao, Xu, & Cheng, 2021; Zhang et al., 2021). Antioxidative pep­ rens, 2010). The utilization of microalgae as high-value added products
tides have been prepared and identified from pecan meal (Hu et al., in food industry has attracted increasing attention in recent years.
2018), tropical jackfruit seed (Chai et al., 2021), rice dreg (Chen et al., Auxenochlorella pyrenoidosa could be cultivated through heterotrophic
2021), and algae (Sheih, Wu, & Fang, 2009) protein hydrolysates. The and autotrophic systems, and is generally comprised of approximately
antioxidative activities of peptides have been reported to attribute to the 50% protein on a dry weight basis (Song et al., 2018; Zhao, Tan, Yang,
presence and position of specific amino acids, such as sulfur-containing Liao, & Li, 2019). Moreover, Auxenochlorella pyrenoidosa, being recom­
amino acids (cysteine and methionine) and aromatic amino acids (Wu mended as the Generally Regarded as Safe material (FDA, 2014), has

* Corresponding author.
E-mail address: wangyanchao@ouc.edu.cn (Y. Wang).

https://doi.org/10.1016/j.lwt.2021.112306
Received 24 May 2021; Received in revised form 22 July 2021; Accepted 13 August 2021
Available online 13 August 2021
0023-6438/© 2021 Elsevier Ltd. All rights reserved.
W. Zhang et al. LWT 152 (2021) 112306

been used as food supplements in the form of tablets and capsules. More 2.2. Microalgae protein hydrolysate preparation
recent researches have attempted to improve the functional properties
of Chlorella through the production of bioactive peptides. Ko, Kim, and 2.2.1. Microalgae protein extraction
Jeon (2012) prepared antioxidative peptides from Chlorella ellipsoidea Microalgae powder was suspended in distilled water (1/6, w/w), and
through peptic enzymatic hydrolysis, and identified one novel anti­ subjected to high-pressure homogenization at a pressure of 1000 bar for
oxidative hexapeptide LNGDVW with the IC50 values of 0.92 mM and three cycles. After homogenization, the mixture was centrifuged at
1.42 mM based on DPPH and hydroxyl radical scavenging abilities. 5000 g for 10 min, and the supernatant was collected, freeze-dried, and
Sheih et al. (2009) investigated the production of antioxidative peptides then defatted with ethanol. The obtained microalgae protein was used
from Chlorella vulgaris, and a novel antioxidative peptide for subsequent enzymatic hydrolysis.
VECYGPNRPQF was identified. This peptide could efficiently quench a
variety of radicals, such as peroxyl, DPPH and ABTS radicals, and 2.2.2. Enzymatic hydrolysis of microalgae protein
perform more efficiently than BHT and Trolox. Accordingly, Auxeno­ Microalgae protein was suspended in distilled water (1/9, w/w), and
chlorella pyrenoidosa might be employed as a sustainable protein mate­ the mixture was adjusted to appropriate pH. Different types of enzymes
rial for antioxidative peptide production and identification in food were added at an enzyme to substrate concentration of 2.0% (w/w) at
industry. appropriate pH values (alcalase - pH8.5, neutrase - pH7.5, flavorzyme -
Generally, antioxidative peptides have been commonly produced pH7.5). Then, the mixture was incubated at 55 ◦ C with shaking at 150
through enzymatic hydrolysis using different commercial proteases rpm. Samples were collected with different hydrolysis time (2 h, 4 h, 8 h,
from plants, animals or microorganisms, and subsequent purification or 16 h), and incubated in a water boiling bath for 10 min. After centri­
refinement processes (Nikoo & Benjakul, 2015; Sarteshnizi et al., 2021). fugation, the supernatant was freeze-dried, and stored at − 20 ◦ C for
During the enzymatic hydrolysis process, polypeptides, oligopeptides or further analysis. Hydrolysates by alcalase with different hydrolysis time
amino acids are released from respective parent proteins, accompanied were coded as Alc2, Alc4, Alc8 and Alc16 respectively. The yield of
with the release of pigments and other components. Many consumers microalgae protein hydrolysates was determined according to the
showed good acceptance for nutritional ingredients from microalgae, as weight ratio of protein hydrolysates to microalgae protein.
being sustainable and natural, however, the presence of pigments such
as chlorophyll and their degradation products might prevent their 2.3. Activated carbon adsorption
application in food products (Heaton & Marangoni, 1996). Considering
the presence of pigments in microalgae protein hydrolysates, a mass of Microalgae protein hydrolysates by alcalase with the hydrolysis time
consumers might reject them due to their poor sensory quality, and of 4 h were decolorized using powdered activated carbon (200 mesh).
therefore decoloration steps, such as using activated carbon, anion ex­ Briefly, the sample was dissolved in aqueous solution at a concentration
change macroporous resin or H2O2 aqueous, might be involved in the of 40 mg/mL, and then powdered activated carbon was added into the
production of bioactive peptides (Shi et al., 2019; Xie, Shen, Nie, Li, & solution at the concentration of 5 mg/mL and 20 mg/mL respectively.
Xie, 2011). Of these approaches, activated carbon is highly efficient at The mixture was incubated at 25 ◦ C for 1 h with shaking at 150 rpm.
removing pigments from protein products (Sessa & Palmquist, 2008), After centrifugation, the supernatant was collected, filtrated through
and has been widely adopted in the decoloration of proteins or peptides 0.45 μm microfiltration membrane, freeze-dried, and stored at − 20 ◦ C
in an industrial scale. However, a number of studies have reported a for further analysis. Hydrolysates with activated carbon adsorption at
successful production of high Fischer ratio peptides from protein hy­ the concentration of 5 mg/mL and 20 mg/mL were coded as Alc4-0.5AC
drolysates through activated carbon treatment, which could lead to a and Alc4-2.0AC respectively.
marked increase in the ratio of total branched-chain amino acids to total
aromatic amino acids (Udenigwe & Aluko, 2010; Wang, Song, Feng, & 2.4. Color measurement
Cui, 2019). This suggested that aromatic amino acid containing peptides
and free aromatic amino acids might be selectively adsorbed during the The color property was determined by using a portable chroma meter
decoloration process. At present, changes of peptide patterns and anti­ (ATAGO, Guangzhou, China), and reported as L* (lightness), a* (+for
oxidative capacities of protein hydrolysates during the enzymatic hy­ redness and – for greenness) and b* (yellowness). The color difference
drolysis and decoloration processes remained little studied. index (ΔE) was calculated as (ΔL2 + Δa2 + Δb2)1/2. Three measurements
Herein, the aim of this study was to investigate changes of anti­ were performed at different locations on each sample.
oxidative activities and peptidomic patterns of Auxenochlorella pyr­
enoidosa protein hydrolysates depending on the enzymatic hydrolysis 2.5. Antioxidative activity determination
and decoloration processes based on a peptidomics approach, including
effects of hydrolysis time and activated carbon decoloration on the 2.5.1. ABTS assay
antioxidative activities and peptidomic patterns, and the relationship Each microalgae protein hydrolysate sample was suspended in
between antioxidative activities and peptidomic patterns. phosphate buffer (50 mM, pH7.4) at the concentration of 0.1 mg/mL.
ABTS radical scavenging activity was determined according to the
2. Materials and methods method described by Zheng, Zhao, Xiao, Zhao, and Su (2016) with slight
modifications. Briefly, the stock ABTS•+ solution was produced by
2.1. Materials reacting ABTS solution with potassium persulfate solution for 12–16 h at
room temperature in the dark. The stock solution was diluted with
Auxenochlorella pyrenoidosa was cultivated according to the method phosphate buffer (50 mM, pH7.4) to obtain the ABTS•+ reaction solu­
of Song et al. (2018). Alcalase 2.4L, Flavourzyme 1000L and Neutrase tion with an absorbance of 0.7 ± 0.05 at 734 nm. Ten microlitres of
0.8L were purchased from Novozymes (Copenhagen, Danmark). sample solution were added into two hundred microlitres of ABTS•+
MS-grade acetonitrile and formic acid were purchased from Thermo reaction solution, and then the mixture was incubated for 60 min at
Fisher Scientific (MA, USA). 2,2-Diphenyl-1-picrylhydrazyl (DPPH) was room temperature in the dark. An equivalent volume of phosphate
purchased from Solarbio Science & Technology (Beijing, China). Total buffer (50 mM, pH7.4) was used as the control sample. Acontrol and
antioxidant capacity assay kit with 2,2′ -azino-bis(3-ethyl­ Asample represented the absorbance of control and samples at 734 nm
benzthiazoline-6-sulfonic acid) (ABTS) method was purchased from respectively. The scavenging percentage was calculated as follows: In­
Beyotime Biotechnology (Shanghai, China). All other chemicals were of hibition (%) = (Acontrol – Asample)/Acontrol × 100%. Trolox was used as
analytical grade. the standard, and the antioxidative capacities of samples were expressed

2
W. Zhang et al. LWT 152 (2021) 112306

in Trolox equivalents, as mmol TE/g sample. expressed as mean ± SD. Statistical analysis was performed using one-
way analysis of variance (ANOVA). Tukey’s post hoc test was made
2.5.2. DPPH assay for multiple comparisons if one-way ANOVA test was found to be sta­
Each microalgae protein hydrolysate sample was suspended in so­ tistically significant. Statistical significance was set at p < 0.05. The
dium acetate buffer (50 mM, pH5.5) at different concentrations. DPPH correlation between antioxidative activities and selected peptides were
radical scavenging activity was determined according to the method determined using Pearson’s linear correlation coefficient. Statistical
described by Zheng, Lin, Su, Zhao, and Zhao (2015) with slight modi­ analysis was performed using SPSS 25.0 software (IBM, New York, USA).
fications. Briefly, the DPPH• solution was prepared by dissolving DPPH
in ethanol at a final concentration of 150 μM. Two milliliters of sample 3. Results and discussion
solution were mixed with 2 mL of DPPH• solution, and then the mixture
was incubated for 60 min at room temperature in the dark. An equiva­ 3.1. Effects of enzymatic hydrolysis on the antioxidative activity and
lent volume of sodium acetate buffer (50 mM, pH5.5) was used as the peptidomic pattern
control sample. Acontrol and Asample represented the absorbance of
control and samples at 525 nm respectively. The scavenging percentage Microalgae protein hydrolysates were prepared by using different
was calculated as follows: Inhibition (%) = (Acontrol – Asample)/Acontrol types of proteases, including alcalase, neutrase, and flavorzyme.
× 100%. IC50 value was defined as the concentration of peptides that Microalgae protein hydrolysates by alcalase with the hydrolysis time of
could scavenge 50% radicals. 4 h showed significantly higher antioxidative activity (7.27 ± 0.27
mmol TE/g) than those by neutrase and flavorzyme (5.35 ± 0.33 and
2.6. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) 5.54 ± 0.52 mmol TE/g) depending on the ABTS radical scavenging
analysis ability (p < 0.05). Moreover, the yield of protein hydrolysates by alca­
lase (56.62 ± 2.68%) was significantly higher than those by neutrase
Each sample was pretreated by deproteinization according to the and flavorzyme (46.31 ± 0.62% and 41.82 ± 1.85%) (p < 0.05). Find­
method of Khaldi et al. (2014), and then desalted by using a ZipTip C18 ings of the current study were consistent with those of Khantaphant,
column. LC-MS/MS analysis was performed by using an Ultimate 3000 Benjakul, and Kishimura (2011) who reported that protein hydrolysates
UPLC system (Thermo, MA, USA) coupled to a Q Exactive mass spec­ from brownstripe red snapper by alcalase showed higher DPPH and
trometer (Thermo, MA, USA). The system was equipped with an ACQ­ ABTS radical scavenging activities while those by flavorzyme displayed
UITY UPLC Peptide CSH C18 column (1 mm × 150 mm, 1.7 μm, Waters, higher ferrous chelating activity. Considersing the yield and ABTS
MA, USA). Desalted sample was dissolved in buffer A (H2O, 0.1% formic radical scavenging ability of microalgae protein hydrolysates, alcalase
acid), and separated at a flow rate of 0.15 mL/min. A binary gradient was selected as the appropriate protease and used for subsequent
elution system, comprising buffer A (H2O, 0.1% formic acid) and buffer analysis.
B (acetonitrile, 0.1% formic acid), was applied as follows: 5% to 20%B Effects of hydrolysis time on the yields and antioxidative activities of
from 0 to 10 min, 20% to 35%B from 10 to 25 min, 35% to 90%B from 25 protein hydrolysates by alcalase were shown in Table 1. Protein hy­
to 45 min, 90%B from 45 to 55 min. Data were acquired with the drolysates by alcalase with the hydrolysis time of 2 h showed signifi­
following MS conditions: ESI source, positive mode; acquisition mode, cantly lower yield compared with those with longer hydrolysis time (p <
data dependent top 3 mode; MS scan range, 100–1500 m/z; MS scan 0.05), and no significant difference was observed among those with
resolution, 70000; MS/MS scan resolution, 17500; stepped collision hydrolysis time longer than 4 h (Table 1). For DPPH radical scavenging
energy, 20%, 30%, 40%. ability, Alc2 and Alc16 showed significantly lower IC50 values in
Raw data files were searched using Proteome Discoverer 2.4 against comparison with Alc4 and Alc8 (p < 0.05) (Table 1). For ABTS assay,
Auxenochlorella pyrenoidosa database. The search parameters were set as Alc16 displayed significantly higher ABTS radical scavenging ability
follows: enzyme specificity was set to unspecific; dynamic modifications than other three samples (p < 0.05) (Table 1). Zheng, Si, Ahmad, Li, and
were oxidation (M); precursor mass tolerance was set to 10 ppm; frag­ Zhang (2018) found that the reducing power of chicken blood protein
ment mass tolerance was set to 0.02 Da. Peptide identification results hydrolysates declined as a function of hydrolysis time until around 5 h,
were filtered with a false discovery rate (FDR) ≤ 1%. Each sample was after which the reducing power increased with further increasing of
measured in two biological duplicates, and only peptides that were hydrolysis time. Regarding the changes of radical scavenging abilities,
detected in two duplicate samples were used for quantification. The this indicated that enzymatic hydrolysis by alcalase generated protein
peptide abundance was determined according to the peak areas of pre­ hydrolysates with different antioxidative activities depending on hy­
cursor ions, normalized with the internal standard, transformed into the drolysis time.
logarithmic format (log2Abundance), and then normalized among Previous studies have revealed that the antioxidative activities of
different samples within the same variable. Euclidean distance was used peptides were related with several properties, such as peptide length,
for hierarchical clustering analysis. Peptide quantification results were molecular weight, amino acid composition and sequence (Wong, Xiao,
used for principal components analysis (PCA) only if there was a fold Wang, Ee, & Chai, 2020). Alcalase belongs to one type of serine endo­
change ≥2.0 or ≤ 0.5. Heatmap graphs and PCA analysis were per­ peptidases with broad specificity and a preference to cleave after hy­
formed using R software (version 3.6.3, R Foundation for Statistical drophobic amino acids. As shown in Fig. 1, dynamic changes in the
Computing, Vienna, Austria).
Table 1
2.7. Peptide synthesis and validation Yields and antioxidative activities of microalgae protein hydrolysates by
alcalase.
Fifteen antioxidative peptides were screened from the identified
Alc2 Alc4 Alc8 Alc16
peptides of microalgae protein hydrolysates using our previously
Yield (%) 45.22 ± 56.62 ± 57.01 ± 58.83 ±
established quantitative structure-activity relationship (QSAR) model.
2.10a 2.68b 0.85b 1.18b
These fifteen peptides were synthesized and validated for the anti­ DPPH-IC50 (mg/ 1.15 ± 0.01a 2.54 ± 0.02c 1.73 ± 1.31 ± 0.23a
oxidative activities using ABTS assay. mL) 0.04b
ABTS (mmol TE/ 4.56 ± 0.23a 7.27 ± 3.74 ± 0.59a 8.78 ± 0.74c
2.8. Statistical analysis g) 0.27b

Data were shown as mean ± SD. Different superscript letters in the same raw
Data were analyzed from three independent experiments and represented the significant difference (p < 0.05).

3
W. Zhang et al. LWT 152 (2021) 112306

(2021) who reported a great differentiation among the peptidomic

patterns of porcine liver protein hydrolysates regarding the effect of
hydrolysis time.
The loading plot showed a strong correlation between the relative
abundances of certain peptides and the discrimination of protein hy­
drolysates with different hydrolysis time (Fig. 2B). Considering the score
and loading plots, a mass of peptides, such as NEDGLNYL, GFFDPL,
GLADDPDTF, SADPETF, and GDSLYPG, with a high negative loading
score on PC1 scale accounted for the discrimination of Alc2 (Fig. 2C).
Peptides with a positive loading score and a negative loading score on
PC1 and PC2 scales respectively contributed to the classification of Alc4,
while peptides with both positive loading scores on PC1 and PC2 scales
accounted for the score plot of Alc8 and Alc16 (Fig. 2C). Results of the
loading plot were in accordance with the relative abundance distribu­
tion profiles of peptides in protein hydrolysates in Fig. 1.
Based on the peptidomics analysis, the significantly different anti­
oxidative activities of protein hydrolysates with different hydrolysis
time might result from their differing peptide compositions. Interest­
ingly, Alc16 displayed significantly strong antioxidative activities
compared to Alc8 depending on DPPH and ABTS radical scavenging
abilities (Table 1). However, according to the PCA analysis, Alc8 and
Alc16 shared much similarity for the quantification results of identified
peptides with amino acid residues above five. Based on peptide sequence
searching in the BIOPEP database, many bioactive peptides found in the
database were small fractions with sizes smaller than five amino acids
(Minkiewicz, Iwaniak, & Darewicz, 2019). As shown in Fig. 2D, several
peptides sharing homology of sequences with known antioxidative
peptides in the BIOPEP database were detected in microalgae protein
hydrolysates, suggesting that prolonged hydrolysis time by alcalase
might lead to extensive hydrolysis, and then result in the production of
short peptides (Minkiewicz et al., 2019). Chi, Wang, Wang, Zhang, and
Deng (2015) found a novel antioxidative tripeptide GPP from the pro­
tein hydrolysates of bluefin leatherjacket head, and this peptide showed
strong scavenging abilities on DPPH and ABTS radicals with IC50 values
Fig. 1. Peptide abundance distribution profiles in protein hydrolysates with of 1.93 mg/mL and 2.47 mg/mL respectively. Four short peptides (GAA,
different hydrolysis time. GFVG, ELLI, and KFPE) with strong antioxidative properties were pu­
rified from proteins hydrolysate of spotless smoothhound muscle, and
relative abundances of 318 identified peptide sequences in microalgae displayed IC50 values ranging from 0.33 mg/mL to 1.75 mg/mL (Wang
protein hydrolysates by alcalase with different hydrolysis time were et al., 2014). Thus, the outstanding antioxidative activities of Alc16
observed based on a peptidomics approach. Several peptide sequences, might attribute to the generation of short peptides with extensive hy­
such as NEDGLNYL, GFFDPL, GLADDPDTF, SADPETF, and GDSLYPG, drolysis of protein materials.
showed relatively high abundances in protein hydrolysates with the
hydrolysis time of 2 h compared with those with longer hydrolysis time, 3.2. Effects of decoloration on the antioxidative activity and peptidomic
and most of these sequences contained aromatic amino acids (Fig. 1). pattern
Prolonged hydrolysis time resulted in dynamic changes of identified
peptide abundances in the protein hydrolysates. However, several pep­ Due to the fact that crude microalgae protein hydrolysates contained
tide sequences, such as LGLPY, FADPL, FGDPL, GEFPGDY, and QFDPL, many impurities, especially pigments, decoloration was recognized as
showed relatively high abundances in protein hydrolysates with an important process in the preparation of antioxidative peptides from
different hydrolysis time (Fig. 1). microalgae. Adsorption treatment is typically a highly effective decol­
According to the score plot of the PCA model, the first two principal oration method which produces few harmful byproducts, and activated
components (PC1 and PC2) explained 74.88% of the total variance carbon has been extensively used for the adsorption of pigments (Kon­
(Fig. 2A). The PCA model revealed that Alc2 distributed at the PC1 sowa, Ossman, Chen, & Crittenden, 2010). As shown in Fig. 3A, acti­
negative coordinate, while the other three samples appeared at the PC1 vated carbon treatment significantly increased the color lightness value
positive coordinate (Fig. 2A). Moreover, Alc4 and the other two samples (L) (p < 0.05), indicating that activated carbon decoloration could
could be discriminated along the second principal component PC2. Alc4 effectively remove pigments and lead to the production of microalgae
distributed along PC2 negative coordinate, while Alc8 and Alc16 were protein hydrolysates with light color. The color difference indexes (ΔE)
projected along with PC2 positive coordinate (Fig. 2A). This indicated of Alc4-0.5AC and Alc4-2.0AC were significantly increased by activated
that the relative abundances of identified peptide sequences changed carbon treatment compared to that of Alc4. No significant difference was
significantly within the first 8 h of enzymatic hydrolysis. The PCA model observed between the color lightness value (L) of Alc4-0.5AC and
was unable to provide a good cluster between Alc8 and Alc16 (Fig. 2A), Alc4-2.0AC, suggesting that increased activated carbon concentration
which might result from the similar relative abundances of identified showed no significant improvement of decoloration.
peptides in Alc8 and Alc16. The sample distribution profile attributed to Activated carbon treatment resulted in a significant decrease in the
the dynamic changes of peptidomic patterns along with the increasing of antioxidative activities depending on DPPH and ABTS radical scav­
hydrolysis time. Findings of the current study were in good agreement enging abilities (p < 0.05) (Fig. 3B and C). Activated carbon decolor­
with those of Lopez-Pedrouso, Borrajo, Amarowicz, Lorenzo, and Franco ation at the concentration of 5 mg/mL significantly increased the IC50
value of DPPH radical scavenging by 29.78% (Fig. 3B) and reduced the

4
W. Zhang et al. LWT 152 (2021) 112306

Fig. 2. Effects of hydrolysis time on peptidomic patterns of protein hydrolysates. A, score plot of the PCA model. B, loading plot of the PCA model. C, peptides
contributing to the classification of the PCA model. D, homologous short peptides with antioxidative activities (shaded by different colors) in the peptides of protein
hydrolysates. . (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

ABTS radical scavenging capacities by 33.59% (p < 0.05) (Fig. 3C). demonstrated that activated carbon adsorption preferred aromatic
Protein hydrolysates by activated carbon decoloration at different con­ amino acid containing peptides rather than branched-chain amino acid
centrations (5 mg/mL and 20 mg/mL) showed similar antioxidative containing peptides (Wang et al., 2019). The antioxidative activities of
activities based on DPPH radical scavenging ability (Fig. 3B). Increasing peptides have been reported to be related with their amino acid com­
the activated carbon concentration from 5 mg/mL to 20 mg/mL resulted positions and sequences, especially aromatic amino acids. Tagliazucchi,
in a significant decrease in the ABTS radical scavenging ability of protein Helal, Verzelloni, and Conte (2016) found that the presence of aromatic
hydrolysates by 66.43% (p < 0.05) (Fig. 3C). Previous studies have amino acids played an important role in determining the antioxidative

5
W. Zhang et al. LWT 152 (2021) 112306

Fig. 3. Effects of activated carbon decoloration on the color and antioxidative properties, and peptidomic patterns of protein hydrolysates. A, color property. B,
DPPH radical scavenging ability. C, ABTS radical scavenging ability. D, peptide abundance distribution profiles in protein hydrolysates. E, score plot of the PCA
model. F, loading plot of the PCA model. Different superscript letters for the same variable represented the significant difference (p < 0.05). . (For interpretation of
the references to color in this figure legend, the reader is referred to the Web version of this article.)

6
W. Zhang et al. LWT 152 (2021) 112306

activities of peptides due to their strong capabilities to donate a proton. hydrolysates might result from the adsorption of aromatic amino acid
Based on peptidomics analysis, microalgae protein hydrolysates con­ containing peptides and free aromatic amino acids.
tained a mass of peptides containing aromatic amino acids, which might Activated carbon treatment resulted in the decrease of relative
contribute to their strong antioxidative activities. After activated carbon abundances of most peptides based on peptidomics analysis (Fig. 3D). A
treatment, the reduction of antioxidative activities of microalgae protein list of 205 identified peptides and their abundances were included in the

Fig. 4. The antioxidative activities and abundance distribution profiles of selected peptides. A, ABTS radical scavenging ability. B, peptide abundance distribution
profiles in protein hydrolysates with different hydrolysis time. C, peptide abundance distribution profiles in protein hydrolysates before and after activated carbon
decoloration.

7
W. Zhang et al. LWT 152 (2021) 112306

multivariate statistical analysis of PCA model. The PCA model exhibited exhibited the strongest ABTS radical scavenging abilities among these
a good discriminant ability with the first two principal components (PC1 four samples (Table 1 and Fig. 4B). It could be speculated that lots of
and PC2) explaining 93.88% of the total variance (Fig. 3E). The score short peptides (2–4 amino acid residues) with strong antioxidative ac­
plot illustrated that protein hydrolysates could be clearly classified into tivities were generated during prolonged enzymatic hydrolysis process,
three clusters regarding the effects of activated carbon decoloration. contributing to the excellent antioxidative activities of Alc16. These
Considering the score and loading plots (Fig. 3E and F), a mass of pep­ fifteen peptides could be identified in protein hydrolysates both before
tides contributed to the discrimination of protein hydrolysates after and after activated carbon treatment (Fig. 4C). Nevertheless, activated
decoloration. Peptide sequences containing aromatic amino acids, such carbon decoloration resulted in a significant decrease in the relative
as FGDPL, GEFPGDY, YPGGF, NEDGLNYL, and PGGLW, showed abundances of these fifteen peptides, which then might impair the
distinctly reduced abundances in protein hydrolysates after activated antioxidative activities of protein hydrolysates. Based on amino acid
carbon adsorption, which might be consistent with the significantly composition, all these fifteen peptides contained aromatic amino acids,
decreased antioxidative activities. which could be preferentially adsorbed by activated carbon. Accord­
ingly, activated carbon at the concentration of 5 mg/mL was selected for
the decoloration of microalgae protein hydrolysates.
3.3. Relationship between the antioxidative activity and peptidomic Enzymatic hydrolysis and decoloration processes play critical roles
pattern in the preparation and purification of bioactive peptides, which are
extensively used in the industrial production of peptides. Activated
Fifteen antioxidative peptides were screened from microalgae pro­ carbon is one of the commonly used material in the decoloration process
tein hydrolysates by using our previously established quantitative of protein hydrolysates. Previous researches mainly focused on the ef­
structure-activity relationship model. These fifteen peptides were syn­ fects of proteases on the composition and biological activities of protein
thesized and subjected to antioxidative activity validation based on hydrolysates. However, effects of hydrolysis time and activated carbon
ABTS assay. Thirteen of fifteen exhibited Trolox equivalent ABTS radical decoloration on the changes of peptide sequences and biological activ­
scavenging activities above 6 mmol TE/g (Fig. 4A). Yang et al. (2020) ities, and their relationship remained unclear. Our previous studies have
reported that GSH showed a Trolox equivalent antioxidative capacity of demonstrated that changes of hydrolysis time could significantly influ­
0.17 mmol TE/g, which was lower than those of these synthesized ence the biological activities of protein hydrolysates. Herein, this study
peptides. Two antioxidative peptides SKGFTSPLF and LDDPVFRPL were confirmed the effects of hydrolysis time on the changes of antioxidative
identified from cauliflower by-products and showed an IC50 value of activities and peptidomic patterns of protein hydrolysates based on a
10.35 μmol/L and 8.29 μmol/L based on the ABTS radical scavenging peptidomics approach. Moreover, this study revealed that peptides,
activity (Montone et al., 2018). Cai et al. (2015) demonstrated the especially aromatic amino acid containing peptides, could be adsorbed
presence of antioxidative peptides PYSFK and GFGPEL in the protein by activated carbon during the decoloration process, which would then
hydrolysates of grass carp skin, which exhibited high scavenging ac­ result in the decrease of antioxidative activities.
tivities on ABTS radicals with IC50 values of 0.28 mM and 0.53 mM
respectively. These reports were consistent with the results of this study 4. Conclusion
that peptides containing aromatic amino acids were more likely to
exhibit strong antioxidative properties. The present study revealed that the antioxidative activities and
For these fifteen selected peptides, relative abundances of ten pep­ peptidomic patterns of Auxenochlorella pyrenoidosa protein hydrolysates
tides were correlated with the antioxidative activities of protein hy­ dynamically changed regarding the production processes, including
drolysates based on DPPH and ABTS radical scavenging abilities enzymatic hydrolysis and decoloration processes. Increasing of hydro­
(Table 2). These fifteen antioxidative peptides were identified from lysis time within 8 h resulted in significant changes of antioxidative
microalgae protein hydrolysates with different hydrolysis time and the properties of protein hydrolysates, which was consistent with the dy­
relative abundances of these peptides dynamically changed along with namic changes of peptidomic patterns in protein hydrolysates, espe­
the increasing of hydrolysis time (Fig. 4B). Alc4 displayed relatively cially aromatic amino acid containing peptides. Prolonged hydrolysis
high abundances of these fifteen peptides and significantly higher ABTS time led to a significant increase in the antioxidative capacities of pro­
radical scavenging ability compared to Alc2 and Alc8 (Table 1 and tein hydrolysates, which might attribute to the generation of short
Fig. 4B). Interestingly, the relative abundances of peptides in Alc16 were peptides. Activated carbon decoloration resulted in a significant
relatively low in comparison to those in other samples, but Alc16 decrease in the antioxidative activities of protein hydrolysates and
observably reduced relative abundances of aromatic amino acid con­
Table 2 taining peptides. Several identified aromatic containing peptides, such
Pearson’s linear correlation coefficient between peptide quantification and as DALYPGE, GEFPGDY, and LGDLPY, were correlated with the DPPH
antioxidative activities of protein hydrolysates based on DPPH and ABTS assays. and ABTS radical scavenging capacities of protein hydrolysates. This
Sequence Length DPPH-IC50 ABTS study suggested that Auxenochlorella pyrenoidosa protein hydrolysates
AGAPVYL 7 ¡0.724 0.402
could be used as an appealing antioxidative ingredient in the food in­
DALYPGE 7 ¡0.865* 0.021 dustry, and both enzymatic hydrolysis and decoloration processes
EGGLDYL 7 0.494 − 0.425 should be taken into account as important determinant of antioxidative
FPGDY 5 0.558 0.650 properties.
GDSLYPG 7 ¡0.554 − 0.324
GEFPGDY 7 ¡0.781 0.448
GPAGAVY 7 − 0.266 0.760 CRediT authorship contribution statement
LGDLPY 6 ¡0.843* 0.783
LGGDYL 6 − 0.212 0.026 Wenhan Zhang: Investigation, Methodology, Data curation, Writing
LGLPY 5 − 0.046 − 0.273 – original draft. Nan Jia: Investigation, Methodology, Data curation.
PGGSGLW 7 ¡0.611 0.713
QPVYI 5 ¡0.531 0.492
Zihao Zhu: Investigation, Data curation. Yanchao Wang: Conceptual­
WVTPI 5 ¡0.653 0.564 ization, Methodology, Writing – original draft, Writing – review &
YAIGDL 6 0.189 − 0.197 editing, Supervision, Funding acquisition, Project administration.
YPGGF 5 0.202 0.322 Jingfeng Wang: Writing – review & editing. Changhu Xue: Writing –
Asterisk represented the significant pearson correlation with p < 0.05, and review & editing.
correlation coefficients above 0.5 were labeled in bold type.

8
W. Zhang et al. LWT 152 (2021) 112306

Declaration of competing interest α-amylase, and DPP-IV enzymes. Lebensmittel-Wissenschaft und -Technologie- Food
Science and Technology, 142, 111019.
Sessa, D. J., & Palmquist, D. E. (2008). Effect of heat on the adsorption capacity of an
The authors declare that they have no known competing financial activated carbon for decolorizing/deodorizing yellow zein. Bioresource Technology,
interests or personal relationships that could have appeared to influence 99, 6360–6364.
the work reported in this paper. Sheih, I. C., Wu, T. K., & Fang, T. J. (2009). Antioxidant properties of a new antioxidative
peptide from algae protein waste hydrolysate in different oxidation systems.
Bioresource Technology, 100, 3419–3425.
Acknowledgement Shi, S., Zhang, W. T., Ren, X. Y., Li, M., Sun, J., Li, G. L., et al. (2019). An advanced and
universal method to high-efficiently deproteinize plant polysaccharides by dual-
functional tannic acid-fe(III) complex. Carbohydrate Polymers, 226, 115283.
This work was supported by National Natural Science Foundation of Song, X. J., Wang, J., Wang, Y. C., Feng, Y. G., Cui, Q., & Lu, Y. D. (2018). Artificial
China (No. 31801479). creation of Chlorella pyrenoidosa mutants for economic sustainable food production.
Bioresource Technology, 268, 340–345.
Stohs, S. J. (1995). The role of free radicals in toxicity and disease. Journal of Basic and
References Clinical Physiology and Pharmacology, 6, 205–228.
Tagliazucchi, D., Helal, A., Verzelloni, E., & Conte, A. (2016). Bovine milk antioxidant
Alamed, J., Chaiyasit, W., Mcclements, D. J., & Decker, E. A. (2009). Relationships properties: Effect of in vitro digestion and identification of antioxidant compounds.
between free radical scavenging and antioxidant activity in foods. Journal of Dairy Science & Technology, 96, 657–676.
Agricultural and Food Chemistry, 57, 2969–2976. Udenigwe, C. C., & Aluko, R. E. (2010). Antioxidant and angiotensin converting enzyme-
Cai, L., Wu, X., Zhang, Y., Li, X., Ma, S., & Li, J. (2015). Purification and characterization inhibitory properties of a flaxseed protein-derived high Fischer ratio peptide
of three antioxidant peptides from protein hydrolysate of grass carp mixture. Journal of Agricultural and Food Chemistry, 58, 4762–4768.
(Ctenopharyngodon idella) skin. Journal of Functional Foods, 16, 234–242. Vanthoor-Koopmans, M., Wijffels, R. H., Barbosa, M. J., & Eppink, M. H. (2013).
Chai, T. T., Xiao, J., Mohana Dass, S., Teoh, J. Y., Ee, K. Y., Ng, W. J., et al. (2021). Biorefinery of microalgae for food and fuel. Bioresource Technology, 135, 142–149.
Identification of antioxidant peptides derived from tropical jackfruit seed and Wang, B., Gong, Y. D., Li, Z. R., Yu, D., Chi, C. F., & Ma, J. Y. (2014). Isolation and
investigation of the stability profiles. Food Chemistry, 340, 127876. characterisation of five novel antioxidant peptides from ethanol-soluble proteins
Chen, M. L., Ning, P., Jiao, Y., Xu, Z., & Cheng, Y. H. (2021). Extraction of antioxidant hydrolysate of spotless smoothhound (Mustelus griseus) muscle. Journal of Functional
peptides from rice dreg protein hydrolysate via an angling method. Food Chemistry, Foods, 6, 176–185.
337, 128069. Wang, Y. C., Song, X. J., Feng, Y. G., & Cui, Q. (2019). Changes in peptidomes and
Chi, C. F., Wang, B., Wang, Y. M., Zhang, B., & Deng, S. G. (2015). Isolation and Fischer ratios of corn-derived oligopeptides depending on enzyme hydrolysis
characterization of three antioxidant peptides from protein hydrolysate of bluefin approaches. Food Chemistry, 297, 124931.
leatherjacket (Navodon septentrionalis) heads. Journal of Functional Foods, 12, 1–10. Williams, P. J. B., & Laurens, L. M. L. (2010). Microalgae as biodiesel & biomass
FDA. (2014). Agency response letter GRAS notice No. GRN 0005192014. feedstocks: Review & analysis of the biochemistry, energetics & economics. Energy &
Heaton, J. W., & Marangoni, A. G. (1996). Chlorophyll degradation in processed foods Environmental Science, 3, 554–590.
and senescent plant tissues. Trends in Food Science & Technology, 7, 8–15. Wong, F. C., Xiao, J., Wang, S., Ee, K. Y., & Chai, T. T. (2020). Advances on the
Hu, F., Ci, A. T., Wang, H., Zhang, Y. Y., Zhang, J. G., Thakur, K., et al. (2018). antioxidant peptides from edible plant sources. Trends in Food Science & Technology,
Identification and hydrolysis kinetic of a novel antioxidant peptide from pecan meal 99, 44–57.
using Alcalase. Food Chemistry, 261, 301–310. Wu, R. B., Huang, J. F., Huan, R., Chen, L. L., Yi, C. P., Liu, D., et al. (2021). New insights
Khaldi, N., Vijayakumar, V., Dallas, D. C., Guerrero, A., Wickramasinghe, S., into the structure-activity relationships of antioxidative peptide PMRGGGGYHY.
Smilowitz, J. T., et al. (2014). Predicting the important enzymes in human breast Food Chemistry, 337, 127678.
milk digestion. Journal of Agricultural and Food Chemistry, 62, 7225–7232. Wu, R. B., Wu, C. L., Liu, D., Yang, X. H., Huang, J. F., Zhang, J., et al. (2018).
Khantaphant, S., Benjakul, S., & Kishimura, H. (2011). Antioxidative and ACE inhibitory Antioxidant and anti-freezing peptides from salmon collagen hydrolysate prepared
activities of protein hydrolysates from the muscle of brownstripe red snapper by bacterial extracellular protease. Food Chemistry, 248, 346–352.
prepared using pyloric caeca and commercial proteases. Process Biochemistry, 46, Xie, J. H., Shen, M. Y., Nie, S. P., Li, C., & Xie, M. Y. (2011). Decolorization of
318–327. polysaccharides solution from Cyclocarya paliurus (Batal.) Iljinskaja using
Ko, S. C., Kim, D., & Jeon, Y. J. (2012). Protective effect of a novel antioxidative peptide ultrasound/H2O2 process. Carbohydrate Polymers, 84, 255–261.
purified from a marine Chlorella ellipsoidea protein against free radical-induced Yang, J., Huang, J. C., Dong, X. L., Zhang, Y. L., Zhou, X. H., Huang, M., et al. (2020).
oxidative stress. Food and Chemical Toxicology, 50, 2294–2302. Purification and identification of antioxidant peptides from duck plasma proteins.
Konsowa, A. H., Ossman, M. E., Chen, Y., & Crittenden, J. C. (2010). Decolorization of Food Chemistry, 319, 126534.
industrial wastewater by ozonation followed by adsorption on activated carbon. Zhang, X. X., Zuo, Z. Y., Yu, P. B., Li, T., Guang, M., Chen, Z. X., et al. (2021). Rice
Journal of Hazardous Materials, 176, 181–185. peptide nanoparticle as a bifunctional food-grade Pickering stabilizer prepared by
Lopez-Pedrouso, M., Borrajo, P., Amarowicz, R., Lorenzo, J. M., & Franco, D. (2021). ultrasonication: Structural characteristics, antioxidant activity, and emulsifying
Peptidomic analysis of antioxidant peptides from porcine liver hydrolysates using properties. Food Chemistry, 343, 128545.
SWATH-MS. Journal of Proteomics, 232, 104037. Zhao, X. C., Tan, X. B., Yang, L. B., Liao, J. Y., & Li, X. Y. (2019). Cultivation of Chlorella
Minkiewicz, P., Iwaniak, A., & Darewicz, M. (2019). BIOPEP-UWM database of bioactive pyrenoidosa in anaerobic wastewater: The coupled effects of ammonium,
peptides: Current opportunities. International Journal of Molecular Sciences, 20. temperature and pH conditions on lipids compositions. Bioresource Technology, 284,
Montone, C. M., Capriotti, A. L., Cavaliere, C., La Barbera, G., Piovesana, S., Zenezini 90–97.
Chiozzi, R., et al. (2018). Characterization of antioxidant and angiotensin-converting Zheng, L., Lin, L. Z., Su, G. W., Zhao, Q. Z., & Zhao, M. M. (2015). Pitfalls of using 1,1-
enzyme inhibitory peptides derived from cauliflower by-products by diphenyl-2-picrylhydrazyl (DPPH) assay to assess the radical scavenging activity of
multidimensional liquid chromatography and bioinformatics. Journal of Functional peptides: Its susceptibility to interference and low reactivity towards peptides. Food
Foods, 44, 40–47. Research International, 76, 359–365.
Nikoo, M., & Benjakul, S. (2015). Potential application of seafood-derived peptides as Zheng, Z. J., Si, D. Y., Ahmad, B., Li, Z. X., & Zhang, R. J. (2018). A novel antioxidative
bifunctional ingredients, antioxidant–cryoprotectant: A review. Journal of Functional peptide derived from chicken blood corpuscle hydrolysate. Food Research
Foods, 19, 753–764. International, 106, 410–419.
Sarteshnizi, R. A., Sahari, M. A., Gavlighi, H. A., Regenstein, J. M., Nikoo, M., & Zheng, L., Zhao, M. M., Xiao, C. Q., Zhao, Q. Z., & Su, G. W. (2016). Practical problems
Udenigwe, C. C. (2021). Influence of fish protein hydrolysate-pistachio green hull when using ABTS assay to assess the radical-scavenging activity of peptides:
extract interactions on antioxidant activity and inhibition of α-glucosidase, Importance of controlling reaction pH and time. Food Chemistry, 192, 288–294.