Double-spotted pufferfish (Takifugu bimaculatus) skin collagen: Preparation, structure, cytocompatibility, rheological, and functional properties

2.3.1 Extraction of tetrodotoxin

Tetrodotoxin extraction was performed using an enzyme-linked immunosorbent assay (ELISA) kit (RNA9701, REAGEN, USA) according to the manufacturer’s instructions. Double-spotted pufferfish skins were homogenized and then boiled in 0.1 M acetic acid at a solid/solvent ratio of 1:5 (w/v) for 30 min. The boiled supernatant samples were centrifuged at 4500 rpm for 10 min, and the precipitate was treated as described above. Then the samples were placed into a separating funnel with an equal volume of ether for multiple degreasings. The pH of each sample was adjusted to 6.5–7.4 with 1 M NaOH. PBST was added to a final volume of 50 mL, and the extracted samples were stored at 4 °C.

2.3.2 ELISA for tetrodotoxin determination

Tetrodotoxin was determined according to the indirect ELISA kit directions provided by the manufacturer. The decreased signal, as indicated by the ratio of absorbance (Bi/B₀, where Bi is the absorbance of the sample, and B₀ is the absorbance of the blank with no tetrodotoxin added), had a linear relationship with the logarithmic concentration of tetrodotoxin in the samples, within a certain concentration range. The tetrodotoxin concentration of the sample was calculated according to Eq. (3):

TTX concentration (µg/kg) = C × K, (3).

where C is the tetrodotoxin concentration (ng/mL) in the diluted sample extract, and K is the sample dilution multiple (in this extraction method, K was 10).

2.4.1 Protein type analysis

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) was used to identify DPC. SDS-PAGE was performed by the method described in Chen et al. (2019) with a slight modification. Electrophoretic patterns of collagens were determined using an 8 % running gel and 3 % stacking gel and electrophoresed in a mini-protein vertical slab electrophoresis system (PowerPacTM Basic, Bio-Rad Laboratories, Hercules, CA, USA). The gel was stained with a mixture of 0.125 % Coomassie Brilliant Blue R-250 in 50 % (v/v) ethanol and 10 % (v/v) acetic acid. Finally, the gel was destained with methanol/acetic acid/distilled water (5:1:4, v/v/v). Molecular weights were calculated using Quantity-One 4.6.0 software (Bio-Rad Laboratories, USA) and band intensities were quantified by ImageJ analysis software version 1.8.0.

2.4.2 Primary structure analysis

DPC primary structure was determined by using a combination of SDS-PAGE, liquid chromatography–mass spectrometry (LC-MS)/MS methods, and a UniProt database search. Protein identification was performed as described in Chen et al., 2016 with modifications. The gels obtained by SDS-PAGE were cut to obtain the collagen subunits. The gels were washed and then pounded to particles using microtube pestles. The gel pieces were reduced using a reducing solution (10 mM dithiothreitol and 100 mM NH₄HCO₃) at 56 °C for 45 h and then alkylated with a solution containing 50 mM iodoacetamide and 100 mM NH₄HCO₃. The gel pieces were dried and reconstituted with a hydrolysate solution (10 ng/µL trypsin in a 50 mM NH₄HCO₃ solution with 10 % acetonitrile) at room temperature in the dark for 45 min. LC-MS/MS (QEactive Easy-nLC1200, Thermo Fisher Scientific, USA) analysis using an Eksigent nanoLC-Ultra ™ 2D LC system for gradient elution was performed to determine the collagen protein sequences. The UniProt database was used to analyze all MS/MS data (PLGS v2.3, Matrix Science, London, UK).

2.4.3 Secondary structure analysis

2.4.4 Tertiary structure analysis

DPC tertiary structure was identified by X-ray diffraction (XRD). XRD was performed according to the method described in Chen et al. (2016) with a slight modification. Collagen was scanned by an X-ray diffraction instrument (DX-1000, Dan-dong Fangyuan Instrument Co., ltd.). Wavelengths were obtained within a 2θ range of 5°–90° at an angular speed of 2°min⁻¹, and the distance between molecular chains (d) was determined using Eq. (4):

where λ is the X-ray wavelength (11.5406 Å) and θ is the Bragg diffraction angle.

2.5.1 Determination of denaturation temperature (Td)

The determination of the Td of the DPC was performed by the method of Li et al., 2022 with a slight modification. Collagen was dispersed in 0.5 M acetic acid at a concentration of 20 mg/mL. The viscosity was measured at different temperatures (10–50 °C) by using a rheometer (MCR 302, Anton Paar, Austria), using a CP25-2 measuring cone (2° cone angle, 25 mm cone diameter). Td was defined as the temperature at which the fractional viscosity was 0.5.

2.5.2 Zeta potential

Zeta potential was determined by the method described in Yousefi et al. (2017) with a slight modification. Collagen samples were dissolved in 0.1 M acetic acid to a final concentration of 0.2 mg/mL of DPC solution. The pH of the solution was adjusted to 2, 3, 4, 5, 6, 7, 8, 9, or 10 with 1 M HNO₃ or 1 M KOH. The samples were measured using a Zeta autotitrator (Zetasizer Nano ZS90, Malvern Instruments, UK).

2.6.1 Dynamic sweep tests

Dynamic frequency sweep tests and dynamic temperature sweep measurements were performed by the method described in Zheng et al. (2021) with a slight modification. Dynamic frequency sweeps for DPC sample solutions with different concentrations (5, 10, 15, 20, and 25 mg/mL) were performed from 0.01 to 10 Hz at 20 °C at a constant strain of 30 %. To evaluate the thermal stability of the DPC sample solutions under a shear, dynamic temperature sweeps were measured. DPC solutions (10 mg/mL) at different temperatures (20 °C, 25 °C, 30 °C, 35 °C, and 40 °C) were tested from 0.01 to 10 Hz at a constant strain of 30 %.

2.6.2 Steady shear tests

Steady shear tests were performed by the method described in Zhu et al. (2017) with a slight modification. Steady-state viscosity curves were determined for the DPC solution at different concentrations (5, 10, 15, and 20 mg/mL) at 20 °C, and for the DPC solution (10 mg/mL) at different temperatures (20 °C, 25 °C, 30 °C, 35 °C, and 40 °C).

2.7.1 Water holding capacity (WHC) and oil absorption capacity (OAC)

WHC and OAC were determined by the method described in Li et al. (2022) with modifications. DPC (200 mg) was added to distilled water (20 mL) or peanut oil (10 mL) and mixed using a vortex mixer (MX-E, SCILOGEX, USA) for 2 min at room temperature. The mixture was held at room temperature for 1 h, then centrifuged at 6000 rpm for 10 min, and the supernatant was carefully removed. The tube was weighed before and after centrifugation and supernatant removal. WAC and OHC were calculated according to Eq. (5):

where W₀ is the DPC dry weight (g), W₁ is the weight of the tube and dry sample (g), and W₂ is the total weight of the precipitate and tube after centrifugation (g).

2.7.2 Foaming capacity (FC) and foaming stability (FS)

FC and FS of DPC solutions were determined by the method described in Ogunwolu et al. (2009) with some modifications. A DPC solution (5 mg/mL) was homogenized (JS25, JUNRUI, Yangzhou, China) at 10,000 rpm for 2 min at room temperature. The sample was then rapidly transferred into a 50-mL graduated cylinder and the volume of foam at different pH values (2, 4, 6, 6.8, 8, or 10) was recorded. After 1 h, the foam stability was determined by measuring the decrease in the volume of foam at each pH. FC and FS were calculated according to Eq. (6) and Eq. (7), respectively:

where V₀ represents the volume after whipping at 0 min, and V₃₀ represents the total volume remaining after 30 min.

2.7.3 Emulsifying activity index (EAI) and emulsifying stability index (ESI)

The EAI and ESI of the DPC sample were determined by the method described in Çelik et al., (2019) with modifications. EAI and ESI were tested by using a UV–visible spectrophotometer (UV-2550, Shimadzu, Suzhou, China). The sample solution (5 mg/mL) and corn oil at a 1:3 ratio (v/v) were homogenized at 16000 rpm for 60 s in a homogenizer. The emulsion from the bottom of the sample was removed (50 μL), mixed with 5 mL of SDS (0.1 %, v/v), and the absorbance at 500 nm was measured. The 500 nm absorbances measured at 0 min (A₀) and 10 min (A₁₀) were used to calculate EAI and ESI according to Eq. (8) and Eq. (9), respectively:

where, N is the dilution coefficient (1 0 0), c is the DPC concentration (0.005 g/mL), A₀ is the absorbance determined at 0 min, A₁₀ is the emulsion absorbance measured after 10 min, and φ is the volume fraction of the oil in the emulsion (6.6/26.6 = 0.248).

2.8.1 Cell viability and proliferation

Cell viability and proliferation were assessed using a CCK-8 assay according to Ge et al. (2020) with modifications. MC3T3-E1 cells were cultured in a 96-well plate at a density of 1x10⁴ cells/well at 37 °C with 5 % CO₂ for 1, 3, or 5 d. The optical density (OD) was measured using a multifunctional microplate reader (SpectralMax M5, Molecular Devices, USA) at 490 nm, and cell viability was calculated using an equation for the relative growth rate (RGR):

where A₀, A_t, and A are the absorbances at 490 nm of the control cell group, the experimental group, and the group without cells, respectively.

2.8.2 Cell morphology

The morphology of MC3T3-E1 was assessed using an inverted microscope (ECLIPSE Ts2, Nikon, Japan) after 1, 3, and 5 d.

3.3.1 Protein type analysis

As shown in Fig. 1, SDS-PAGE protein patterns of DPC exhibited a similar pattern compared with rat tail type I collagen and mainly consisted of at least two α-chains (α₁ and α₂), β-chains (dimer) and γ-chains (trimer). Furthermore, this was consistent with silver carp collagen, Ujumuqin sheep bone collagen, and quail feet collagen reported by other researchers (Faralizadeh et al., 2021; Gao et al. 2018; Yousefi et al., 2017). Therefore, DPC is type I collagen. The molecular weights of the different chains were calculated using Quantity-One software. The SDS-PAGE pattern showed two different α-chains, α₁ and α₂, for DPC at 127 kDa and 117 kDa, respectively. Two α-chains, α₁ and α₂, of type I collagen from rat tail were 129 kDa and 118 kDa, respectively. The molecular weight of DPC was lower than that of rat rail type I collagen standard. The discrepancy may be due to the different resources of collagen, specifically, resources from mammalian animals versus aquatic animals. Furthermore, the molecular weight of DPC was similar to other collagen sources, such as Ujumuqin sheep ovine bones (α₁: MW 126.87 kDa; α₂: MW 110.18 kDa) (Gao et al., 2018), quail feet (α₁: MW 120 kDa; α₂: MW 116 kDa) (Yousefi et al., 2017), and scales of lizard fish (α₁: MW < 120 kDa; α₂: MW < 110 kDa) (Thuy et al., 2014). Most marine collagen such as walleye pollock collagen (Yan et al., 2007), quail feet collagen (Yousefi et al., 2017), and bigeye tuna collagen (Raju et al., 2019) have α₁ and α₂ subunits. The molecular weight ratio, α₁-chains/α₂-chain, was approximately 2:1. However, it was reported that in some collagen, α₃(I) migrated electrophoretically to the same position as α₁(I), and the molecular weight ratio, α₁-chains/α₂-chain/α₃-chain, was approximately 1:1:1 (Yousefi et al., 2017). Experimental results from LC-MS/MS and a UniProt database search showed that DPC consisted of α₁ and α₂ subunits but no α₃ subunit. The molecular weight ratio of DPC α₁-chain/α₂-chain determined by ImageJ software (VERSION 1.8.0, National Institute of Mental Health, Bethesda, MD, USA). was approximately 2:1, the same as rat tail collagen, suggesting that DPC is a type I collagen and consists of [(α₁)2α₂]. Additionally, the purity of DPC was estimated by SDS-PAGE using ImageJ (Vision 8.0, NIH, US) (Chen et al., 2019). The quantification of stained protein bands showed that α1, α2, β, and γ subunits contributed to 95.86 % ± 0.66 % of the total DPC, which exhibited a similar purity compared with the rat tail type I collagen standard (94.86 % ± 0.51 %).

Close

Fig. 1

SDS-PAGE pattern. Lanes 1: Marker26634; lane 2: Type Ⅰ rat tail collagen standard; lanes 3: DPC.

Export to PPT

3.3.2 Primary structure analysis

The primary structure of DPC was analyzed by using a combination of SDS-PAGE, LC-MS/MS methods, and UniProt database BLAST searches. The protein sequence coverage of collagen chains and the number of matching peptides between collagen chains and the UniProt database were the main parameters used to identify collagen. As shown in Table 1, the α₁ subunit score matching type I collagen subunit α₁ was 1642. There were 36 peptides in the UniProt database that matched the α₁ subunit of type I collagen with 37 % sequence coverage. The subunit α₂ score matching the α₂ (I) chain was 1626 and contained 27 matching peptides with 29 % sequence coverage. They were composed of 1448 and 1423 amino acids with a calculated molecular mass of 137,229 Da and 135,030 Da, respectively (Supplementary Materials 1 and 2). Thus, peptide sequences detected by LC-MS/MS matched the double-spotted pufferfish type I collagen.

DPC is collagen with a triple helix structure formed by three α subunits intertwined with each other. The conserved sequence Gly-X-Y triplets of collagen α subunits is an important factor to ensure collagen stability. In Gly-X-Y triplets, Gly is glycine, and X and Y can be any amino acid residue except tryptophan. X and Y are usually proline (Pro) and hydroxyproline (Hyp), respectively (Sun et al. 2022).

The amino acid sequence of the DPC α₁ subunit included 378 Gly-X-Y triplets (1134 residues), accounting for 78.3 % of the total amino acids in the α₁ subunit chains. The conserved sequence Gly-X-Y triplets in the α₁ subunit chains were mainly in the center domains (G¹⁶³-G¹¹⁷⁹) (Supplementary Material 3). They comprised 339 Gly-X-Y triplets (1017 residues), accounting for 90.4 % of the total content of all the conserved sequences in the α₁ (I) chains. Additionally, the N-terminal domain of the α₁ subunit chains included 22 Gly-X-Y triplets distributed over nine regions, namely G²²-G²⁴, G³¹-C³³, G³⁷-L³⁹, G⁵⁹-V⁶¹, G⁸²-C⁸⁴, G⁹³-V⁹⁸, G¹⁰⁵-F¹⁴³, G¹⁴⁸-V¹⁵⁰, and G¹⁵⁶-G¹⁵⁸. The C-terminal domain of α₁ subunit chains contained 17 Gly-X-Y triplets distributed over 15 regions, namely G¹¹⁹⁷-G¹¹⁹⁹, G¹²³⁴-Q¹²³⁶, G¹²⁵⁶-Y¹²⁵⁸, G¹²⁶⁵-T¹²⁶⁷, G¹²⁸⁰-T¹²⁸², G¹³¹⁶-F¹³¹⁸, G¹³²³-D¹³²⁸, G¹³⁶⁶-L¹³⁶⁸, G¹³⁷⁶-N¹³⁷⁸, G¹³⁸⁶-S¹³⁸⁸, G¹³⁹⁸-T¹⁴⁰⁰, G¹⁴⁰⁴-T¹⁴⁰⁹, G¹⁴³²-P¹⁴³⁴, G¹⁴³⁹-E¹⁴⁴¹, and G¹⁴⁴³-G¹⁴⁴⁵.

The amino acid sequence of the DPC α₂ subunit had 372 Gly-X-Y triplets (1116 residues), accounting for approximately 78.4 % of the total amino acids in the α₂ subunit chains. As shown in Supplementary Material 4, the conserved sequence Gly-X-Y triplets in the α₂ subunit were mainly found in the center domains (G¹⁵⁵-G¹¹⁶⁸). They comprised 338 Gly-X-Y triplets (1014 residues), which accounted for 90.1 % of the total content of all the conserved sequences in the α₂ subunit chains. The N-terminal domain of the α₂ subunit chains contained 20 Gly-X-Y triplets distributed over eight regions, namely G¹⁴-P¹⁶, G²²-H²⁴, G⁵⁰-V⁵², G⁵⁴-S⁵⁶, G⁷⁴-A⁷⁶, G⁹⁹-F¹³⁷, G¹⁴³-K¹⁴⁵, and G¹⁵⁰-G¹⁵². The C-terminal domains of the α₂ subunit chains contained 14 Gly-X-Y triplets distributed over 13 regions, namely G¹¹⁷³-D¹¹⁷⁵, G¹²¹⁰-R¹²¹², G¹²³²-Y¹²³⁴, G¹²⁴¹-M¹²⁴³, G¹²⁸⁵-T¹²⁸⁷, G¹²⁹⁰-T¹²⁹², G¹³³⁷-M¹³⁴², G¹³⁵⁰-N¹³⁵², G¹³⁶⁰-S¹³⁶², G¹³⁷²-T¹³⁷⁴, G¹⁴⁰⁶-A¹⁴⁰⁸, G¹⁴¹³-D¹⁴¹⁵, and G¹⁴¹⁷-D¹⁴¹⁹.

These results indicated the structural stability of DPC, and that DPC showed promise as a raw material for biofunctional materials.

3.3.3 Secondary structure analysis

3.3.3.1

3.3.3.1 FTIR

As shown in Fig. 2(a), DPC has characteristic collagen peaks including five amide bands: amide A, B, I, II, and III. These results are consistent with those of collagen extracted from other fish species (Wang et al., 2019; Gao et al., 2018). The amide A band measured at 3295.24 cm⁻¹ was associated with N—H stretching vibrations and was lower than the normal range of 3400–3440 cm⁻¹. As Chen et al., (2019) reported, when the N—H group of a peptide is involved in a hydrogen bond, the position is shifted to a lower frequency. The amide B band was found at 2947.61 cm⁻¹, associated with an asymmetrical stretch of CH₂.

Close

Fig. 2

The secondary and tertiary structure of DPC. (a) FTIR spectrum; (b) Spectral curve-fitting; (c) X-ray s spectrum.

Export to PPT

The infrared spectrum of amide I, amide II, and amide III was observed at 1662.98 cm⁻¹, 1534 cm⁻¹, and 1240.67 cm⁻¹, respectively, which were related to the degree of molecular order and the triple-helical structure of collagen (Wang et al., 2019). The amide I band was around 1600–1700 cm⁻¹, which was related to the stretching vibration of the carbonyl group (C⚌O) and the secondary structure of collagen (Yan et al., 2007). The shift of the amide II peak to a lower wavelength was related to the increased hydrogen bonding of the NH group (Yousefi et al., 2017). The amide II band usually appears at 1550–1600 cm⁻¹, which indicated N—H bending coupled with a C—N stretching vibration (Krimm and Bandekar, 1986). Therefore, the lower wavelength of the amide II band was related to the increased hydrogen bonding by NH groups. The absorption ratio of the amide III and 1450 cm⁻¹ band was 1.19, which was close to 1, indicating that the prepared collagen had a complete structure and retained the shape of natural collagen (Sun et al., 2017).

3.3.4 Tertiary structure analysis

As shown in Fig. 2(c), DPC had two characteristic collagen peaks (peak A, peak B) at diffraction angles (2θ) 7.74° and 21.41°, respectively, corresponding to the characteristic diffraction peaks of collagen that contains proteins having an ordered structure or ordered structure snippet. Peak A represents a relatively sharp diffraction peak, and peak B represents a rounded and broad diffraction peak. This was consistent with tilapia collagen (Sun et al., 2017). The value of d calculated by the Bragg equation (Equation 3) corresponding to peak A was 11.45 (Å), reflecting the distance between collagen molecular chains, which is related to the triple helix conformation. The d value of peak B was 4.22 Å, indicating the distance between skeletons. This was consistent with Pacific cod, tilapia scales, and grass carp swim bladders (Sun et al., 2017; Chen et al., 2016; Li et al., 2022), in which the d values corresponding to both sharp and broad peak pairs were 11.63 Å and 3.96 Å, 11.569 Å and 4.485 Å, and 11.39 Å and 4.17 Å, respectively. The d value of DPC was lower than the fish skin, indicating that the hydrogen bond between molecules was less weakened, and the collagen was a complete triple helix structure.

3.4.1 Td

The Td value of collagen is defined as the temperature at which the change in relative viscosity was half completed when the collagen triple helix structure dissociated into random coils in a solution (Chen et al., 2019). The denaturation temperature of DPC was 27.8 °C, which was close to the Td value of the ocellate puffer fish (Takifugu rubripes, Td = 28 °C) and the balloon fish (Diodon holocanthus, Td = 29.64 °C) determined by other researchers (Nagai et.al, 2002; Huang et al., 2011; Wang et al., 2019). The factors affecting the thermal stability of fish collagen are mainly the living environment of fish. Compared with other aquatic animals, the DPC Td was higher than the Td of the backbone of Baltic cod (14.4 °C, Zelechowska et al., 2010), deep-sea redfish (16.1 °C, Wang L et al., 2008), and red stingray skin (23.82 °C, Chen J D et al., 2019). Compared to other collagens, DPC had a relatively high thermal stability and a large content of ordered secondary structures (76.74 %) (Choi and Ma, 2005), which is an advantage for its application in the field of food or biofunctional materials in the future.

3.4.2 Isoelectric potential (pI)

The zeta potential, representing the surface charge of DPC, is shown in Fig. 6 at different pH values, ranging from 2 to 10. The zeta potential result indicated that a zero net charge of DPC was found at pH 6.7, which was its pI.

3.5.1 Dynamic sweep tests

The frequency dependence of the storage modulus (elastic modulus) G', loss modulus (viscous modulus) G'', complex viscosity η*, and loss tangent tanδ is shown in Fig. 3. Fig. 3(a–c) shows that at the same frequency, G', G'', and η* increased as the DPC concentration increased. As the concentration increased from 5 mg/mL to 25 mg/mL at 0.01 Hz, G' increased from 0.0199 Pa to 5.87 Pa, and G'' increased from 0.0388 Pa to 7.48 Pa. These results suggest that the viscoelasticity (G', G'') of DPC solutions are greatly affected by concentration, which is similar to that of the swim bladder of grass carp collagen (Li et al., 2022). This is because the entanglement and friction of collagen molecules increase as the concentration increases (Ren et al., 2012), which affect viscoelasticity. At the same time, it can also be observed from the results that G' and G'' increased with increasing frequency (0.1–10 Hz). Using 10 mg/mL of collagen solution as an example, the storage modulus (G') continuously increased from 0.225 Pa to 29.9 Pa with increasing frequency (0.1–10 Hz). Additionally, the loss modulus G'' also increased from 0.617 Pa to 23.2 Pa with increasing frequency (0.1–10 Hz) but was lower than the storage modulus (G').

Close

Fig. 3

Effect of concentration on the rheological behavior of DPC solution. (a) The storage modulus G'; (b) The loss modulus G“; (c) The complex viscosity η*; (d) Loss tangent (tan δ).

Export to PPT

It can be observed from Fig. 3(c) that η* decreased as the shear frequency increased in the range of 0.1–10 Hz, which showed that the collagen solution exhibited shear thinning, pseudoplastic fluid behavior. These results suggest that a high viscosity at low shear rates could contribute to excellent structural stability, whereas a highly pseudoplastic flow could provide good mouthfeel. As Brummer and Godersky (1999) reported, the initial skin feel is related to shear rate and dynamic viscosity. Therefore, DPC has excellent spreadability properties and can be applied in the cosmetics industry.

The loss tangent, tanδ, is defined as the ratio G''/G'. If tan δ < 1, the solution system mainly exhibits viscous behavior (liquid-like behavior), and if tan δ greater than 1, the solution system exhibits elastic behavior (solid-like behavior). Fig. 3(d) shows that in the frequency range of 0.1–10 Hz, the tanδ of the solution collagen gradually decreased as the shear frequency increased. Furthermore, tanδ ranged from more than one to less than one, which illustrated that the rheological behavior changed from viscous to elastic characteristics (Chen et al., 2019). The results suggest that as a dynamic viscoelastic fluid, DPC solutions might be developed as injectable materials for drug delivery, tissue engineering, and repair of dermatological defects because of their excellent viscoelasticity properties.

It can be observed from Fig. 4(a–c) that the G', G'', and η* of DPC decreased as the temperature increased. This showed that the temperature had a significant effect on the rheological properties of DPC. The G' and G'' decreased from 0.156 to 0.125 Pa and from 0.552 Pa to 0.369 Pa, respectively, and η* decreased from 9.13 Pa.s to 6.2 Pa.s when the temperature increased from 20 °C to 25 °C at 0.01 Hz. The viscoelasticity properties of DPC changed slightly because the temperature was lower than the Td (27.8 °C) of DPC. However, the rheological properties of DPC decreased significantly when the temperature increased from 25 °C to 30 °C. G' and G'' decreased from 0.125 to 0.00703 Pa and from 0.369 Pa to 0.00392 Pa, respectively, and η* decreased from 6.2 Pa.s to 0.128 Pa.s. This was mainly because at temperatures above the Td of DPC, the molecular movements of DPC were accelerated, which led to the destruction of the network structure of DPC (Ren et al., 2012). When the temperature was greater than 30 °C, G' and G'' lost their regularity and η* decreased significantly, which reflected the collapse of the DPC triple helix to a random coil (Ren et al., 2012), and the hydrogen bonds between the CO and NH groups of two neighboring chains were broken.

Close

Fig. 4

Effect of temperature on the rheological behavior of DPC solution. (a) The storage modulus G'; (b) The loss modulus G“; (c) The complex viscosity η*; (d) Loss tangent (tan δ).

Export to PPT

Fig. 4(d) shows that when the temperature was lower than 30 °C, tanδ was greater than one, which demonstrated that the collagen solution exhibited a viscous behavior. The tanδ increased sharply but irregularly when above 30 °C. In the case of cosmetic and food applications, appropriate rheological behavior at different temperatures is essential to design a proper formulation and confirm its applicability under several conditions. Additionally, when the frequency was greater than 1 Hz, the collagen solution exhibited viscous behavior in the low frequency region and elastic rheological behavior in the higher frequency region. This was mainly due to the sharp reduction of G' from the breakdown of the hydrogen bonds that maintained the triple helical structure of collagen (Chen et al., 2019).

3.5.2 Steady shear tests

Fig. 5(a,b) shows the plots of shear stress and viscosity versus shear rate for DPC solutions at various concentrations at 20 °C. The viscosity significantly decreased with increasing shear rate, as shown in Fig. 5(b), which was consistent with previous studies (Li et al., 2022; Zhu et al., 2017). DPC solutions at different concentrations all exhibited typical flow behavior of non-Newtonian shear thinning, which became more conspicuous at higher concentrations. This was mainly because shearing led to the orientation or deformation of the DPC macromolecular networks in the direction of flow. DPC is a type I collagen, which has a right-handed triple helix of three peptide chains. The three strands are supercoiled around each other in a right-handed manner, and a ladder-type intermolecular backbone of N—H···O⚌C hydrogen bonds links adjacent strands. From the point of view of the DPC structure, at higher concentration solutions, the low-energy intra- or intermolecular interactions (including hydrogen bonds, hydrophobic interactions, and Van derWaals forces) were enhanced. The degree of intra- or intermolecular entanglements may offer a high initial resistance to shear, leading to more conspicuous shear-thinning behavior at higher concentrations (Huang et al., 2016). In practice, based on the shear-thinning phenomenon exhibited by the DPC solution, the shear rate can be increased depending on the requirements of the production process, such as by increasing the machine speed or propulsion speed within a certain range of shear rates, in order to reduce the viscosity to increase fluidity and reduce energy consumption, thus improving production efficiency.

Close

Fig. 5

Steady-state rheological properties of DPC. (a) plot of shear stress versus shear rate at various concentrations; (b) plot of viscosity versus shear rate at various concentrations; (c) plot of shear stress versus shear rate at various temperatures; (d) plot of viscosity versus shear rate at various temperatures;

Export to PPT

Fig. 5(c, d) shows the plots of shear stress and viscosity against shear rate for the DPC solution (10 mg/mL) at different temperatures. The viscosity significantly decreased with increasing shear rate at the tested temperatures, as shown in Fig. 5(d). Similarly, DPC samples with different temperatures all behaved as typical non-Newtonian shear-thinning fluids. The viscosity significantly decreased with gradually increasing temperatures, especially when the temperature was above 30 °C (Fig. 5(d)). As a macromolecule with molecular weight of approximately 300 kDa, the migration of DPC molecules in solutions is mainly achieved by movement of polypeptide chains, which is further ascribed to its molecular flexibility imparted by glycine. As the temperature increased, the random thermal motion of DPC molecules intensified, and the molecular chains of DPC moved more easily, resulting in decreased viscosity of the solution. Solutions of DPC perhaps had a greater random coil secondary structure composition caused by weakening of hydrogen bonds when the temperature was above 30 °C. These results suggested that to better preserve the natural structure of collagen and to obtain the desired medical properties when DPC is used as a raw material for the preparation of fibers, membranes, and other materials, the temperature should be maintained below 30 °C.

3.6.1 WHC and OAC

WHC is regarded as an indicator of the ability of a protein to physically hold water against gravity, and is closely correlated to the viscosity, texture, and mouth feel of a food product (Çelik et al., 2019). The OAC of proteins generally reflects their capacity to interact with lipids and is strongly associated with emulsifying properties, flavor retention, and shelf life (Deng, et al. 2019). WHC and OAC are expressed as the amount of oil or water absorbed per gram of protein. The WHC of DPC was 23.93 ± 0.47 mL/g, which was higher than chicken feet collagen (1.9 ± 0.1 mL/g), red stingray collagen (20.76 ± 0.55 mL/g), and sour cherry kernel protein concentrate (2.42 ± 0.09 mL/g) (Dhakal et al., 2018; Chen et al., 2019; Çelik et al., 2019). In general, the high WHC of DPC can benefit aqueous food formulations by maintaining juiciness and softness in foods, for potential applications such as sausages, salami, meat analogues, and ready-to-eat meals. In addition to food applications, DPC can be used as a moisturizer in cosmetic creams for dry skin therapy.

OAC is related to flavor and texture retention in meat and confectionery products, and is affected by nonpolar side chains and hydrophobic amino acids (Zou et al., 2017). In our current study, the OAC of DPC was 31.93 ± 0.085 mL/g, which was very high when compared to that of chicken feet (5.3 ± 0.3 mL/g) and red stingray (Dasyatis akajei) (28.48 ± 0.69 mL/g) (Dhakal et al., 2018; Chen et al., 2019). The higher OAC of DPC may be due to its higher surface hydrophobicity, leading to increased oil absorption by the binding of proteins to the hydrocarbon chains of lipids (Wu et al, 2019). The higher OAC of DPC was also shown to be related to the higher β-sheet content of DPC (Zhu et al., 2020). Additionally, the OAC of DPC was greater than the WHC, indicating that the content of hydrophobic groups in DPC was higher than the content of hydrophilic groups.

In summary, the results show that DPC had a high water and oil absorption capacity, which is useful in the food processing industry. Specifically, the high OAC of DPC makes it a suitable ingredient for bread, cake batters, mayonnaise, sausages, and salad dressings.

3.6.2 Foaming properties

As shown in Fig. 7(a, b), the FC and FS of DPC at a range of pH values (2, 4, 6, 6.7, 8, and 10) generated U-shaped curves. The FC ranged from 8.26 % ± 0.38 % to 20.4 ± 0.57 % and reached its lowest value at pH 6.7 (8.26 % ± 0.38 %). These results could be due to the fact that the lowest foam properties coincided with the lowest solubility, when the DPC molecules aggregated at the DPC pI. When the pH shifted away from 6.7, the solubility and net charge of DPC increased, which led to a corresponding increase in the FC of DPC. The FC of DPC was found to be higher than the FC previously reported for rice bran protein concentrates of Basmati 370 (1.33 ± 0.05–2.50 ± 0.13) (Chandi and Sogi, 2007), casein (3.95 ± 0.07–10.15 ± 0.21 %) (Chandi and Sogi, 2007), and black ruff collagen (5.0 ± 0.01 %) (Bhuimbar et al., 2019). The excellent FC of DPC was related to the higher number of β-sheet structures of DPC. DPC may have a more open conformation due to more β-sheet structures, which promoted water–protein interactions, causing rapid conformational changes at the air–water interface with reduced surface tension, and then increased protein solubility and flexibility, thus increasing foam formation (Celik et al., 2019). The FS was similar to the FC, and ranged from 12.3 ± 0.38 % to 24.5 ± 0.38 %, and reached the lowest value (12.26 % ± 0.38 %) at pH 6.7. The FS increased when the pH was gradually shifted above or below 6.7, which was attributed to surface-active properties and adequate intermolecular interactions at the interface (Çelik et al., 2019). The FS of DPC was higher than that of sodium caseinate (21.75 ± 0.67) (Çelik et al., 2019), casein (0.17 ± 0.002–0.54 ± 0.61 %) (Chandi and Sogi, 2007), and black ruff collagen (0 %) (Bhuimbar et al., 2019).

Close

Fig. 6

Zeta potential of DPC.

Export to PPT

Close

Fig. 7

Functional properties of DPC. (a) FC; (b)FS; (c) EAI; (d) ESI. Different letters (a-d) indicate significant differences between samples.

Export to PPT

Foaming properties contribute to smoothness, lightness, flavor dispersion, and palatability of foods. In summary, the results revealed that DPC had excellent foaming properties, therefore, could serve as a foaming agent in food systems requiring a high foaming capacity and foaming stability, such as baking, beverages, whipped toppings, and minor food ingredients.

3.6.3 Emulsifying properties

The EAI and ESI of DPC showed V-shaped curves within the tested pH range (2, 4, 6, 6.7, 8, and 10), similar to the relationship between foaming properties and pH. As shown in Fig. 7(c, d), EAI ranged from 44.36 ± 0.001 to 12.99 ± 0.0014 m²/g and reached the lowest value of 12.99 ± 0.0014 m²/g at pH 6.7. The ESI ranged from 45.23 ± 0.0033 min to 13.68 ± 0.0008 min and also reached the lowest point (13.68 ± 0.0008) at pH 6.7. When the pH value of the solution shifted away from the pI of DPC, the electrostatic repulsion between the molecules increased. The folding structure of DPC, determined by the a-helices, β-sheets, β-turns, hydrogen bonds, ionic bonds, and hydrophobic bonds, expanded and the hydrophobic groups inside the protein were exposed to the surface, thus accelerating emulsification (Zhang et al., 2020). These results could be because as the DPC molecules aggregated and precipitated, they were unable to unfold at the oil–water interface, and emulsification performance was minimized at the pI. The EAI of DPC was higher than previously reported for field pea protein isolates (11.8 m²/g) and protein isolates from soy (12.2 m²/g) (Shevkani et al., 2015). The ESI of DPC was higher than that of collagen of black ruff (10.02 ± 1.07 min) (Bhuimbar et al., 2019) and sharpnose stingray collagen (15.01 ± 0.09 min) (Shaik et al., 2021). Compared to other proteins, DPC has a higher EAI and ESI due to its greater β-sheet content (26.51 %), which resulted in increased flexibility and surface hydrophobicity of DPC, and increased interactions between DPC and corn oil as well as DPC molecular rearrangements at the oil–water interface (Zhu et al., 2020). These results indicate that DPC is a potential ingredient for many food formulations, such as salad dressing, sausages, comminuted meats, ice cream, cake batters, and mayonnaise. Additionally, a greater degree of unfolding of polypeptides might have occurred in the KBIs with a lower protein content during the emulsification process; consequently a much greater surface area of the protein was made available, which might have also enhanced the emulsifying activity.

3.7.1 Cell viability and proliferation

Cellular viability and proliferation of MC3T3-E1 cells were assessed to determine the cytocompatibility of DPC. It can be observed from Fig. 8 that relative cell proliferation rates were 110.35 % ± 5.16 %, 110.43 % ± 3.6 %, and 122.73 % ± 1.44 % on the first, third, and fifth days, respectively. This was higher on the fifth day than that of lizardfish skin collagen (113.43 % ± 2.40 %) (Chen et al., 2021), silver carp skin collagen (97 %) (Faralizadeh et al., 2021), and grass carp swim bladder collagen (115.98 %) (Li et al., 2022). Additionally, the cell viability of the experimental group was higher than that of the control group (p < 0.05), indicating that DPC was biocompatible and positively affected cell viability. The cell viability was greater than 70 %, indicating that DPC was nontoxic to the cells, according to ISO 10993–5 (Faralizadeh et al., 2021). These results indicate that DPC can be incorporated into biomaterials for wound healing and other clinical and cosmetic applications.

Close

Fig. 8

Cytocompatibility. (a) Cell viability of MC3T3-E1 cells treated after 1, 3, 5 days; (b) Morphological changes of MC3T3-E1 cells treated after 1, 3, 5 days. Different letters (a-b) indicate significant differences between samples.

Export to PPT

3.7.2 Cell morphology

The cell morphology of MC3T3-E1 cells with DPC is shown in Fig. 8. Microscopic observation of MC3T3-E1 fibroblasts shows normal and complete morphology and regular growth, indicating that cell proliferation was promoted by DPC. This phenomenon was consistent with the CCK-8 results, which demonstrated that collagen from pufferfish skin had good biocompatibility and induced integrin-mediated cellularity. In summary, DPC could be used as a nontoxic and suitable material for biomedical and tissue engineering applications, such as scaffolds.