3.1. Fish collagen hydrolysate extraction
Most reported methods for extraction of collagen hydrolysate from fish skin use enzymatic or acid hydrolysis through several stages, with the disadvantage that in the skin processing step, residues such as scales, bones, meat or fat are manually removed, a process that can take a long time and may be inefficient for the total removal of these residues. Other disadvantages reside in the cost of enzymes, the long hydrolysis time and the difficulty of purification of the final product, residues of acids or enzymes affecting the properties of the final product.
The method proposed in the present paper eliminates some of these disadvantages by using neutral hydrolysis, employing water in subcritical conditions at 135°C and 315 kPa. Several pretreatment procedures were developed for removing residues (scales, bones, meat or fat), non-colagenous proteins and pigments involving food and cosmetic compatible organic acids (lactic and citric acid), mild alkaline or saline treatment, diluted ethanol and hydrogen peroxide, respectively, as presented in Fig. 1. The solvent concentrations involved in the extraction process were chosen based both on our previous experience in obtaining collagen and on other studies reported in the literature5.
After the hydrolysis process, the samples were stored at 4° C for 24 hours. In most of the cases, stable solid hydrolysates were obtained except for the sample HO4P5, resulted after employing lactic acid and NaOH in the separation process which yielded gelatin, proving this treatment unsuitable for obtaining fish skin hydrolysate.
Collagen hydrolysates extraction yields for different process conditions were obtained as follows: 35.5 % for HO4P1 (3% lactic and citric acids, NaOH 1M), 28.1 % for HO4P2 (3% lactic acid, NaCl 1M), 24.6 % for HO4P3 (3% lactic acid, NaHCO3 1M ), and 29.3 % for HO4P4 (3% citric acid, NaOH 1M), respectively. The mixture of organic acids and NaOH 1M pre-hydrolysis treatment (sample HO4P1) proved to be the most efficient procedure. The obtained extraction yields are in good agreement with acid soluble collagen obtained from carp skin, Cyprinus carpio- 41.3% [16] and Ctenopharyngodon idellus -25.5% [17]. Literature data report different extraction yields for acid soluble fish skin collagen (red Snapper -5.71 %, milkfish -4.00 %, paper nautilus -55.2%, Japanese sea-bass -51.4%, chub mackerel -49.8%, bullhead shark -50.1%) explained as a difference in solubility in acid solutions due to the degree of cross-linking of collagen molecules [16].
3.2. Physico-chemical characterization
Physicochemical parameters and the estimations of collagen content are important data in evaluating the potential value of fish collagen hydrolysates. Thus, the obtained Cyprinus carpio skin collagen hydrolysates HO4P1- HO4P4 were subjected to a standard physico-chemical characterization, the results are presented in Table 1:
Examining the data from Table 1 we may conclude that all the obtained fish collagen hydrolysates samples present a high degree of purity according to the very high percentage of protein (about 98%) and undetected values for ash content, which represent two important advantages of the extraction method employed. By comparison, reported literature data for acid soluble Cyprinus carpio skin collagen hydrolysate show a percentage of protein of 27-27.9% and an ash content of 0.22- 1.21% [18]. The proposed procedure for the separation of collagen hydrolysates yielded samples with pH values in the range 6-7, which indicates the suitability of obtained carp skin hydrolysates for food supplements or cosmetics.
Additional information resulted by examining the values for total nitrogen, Nt and amino nitrogen (NH2) in order to obtain an estimation of the degree of hydrolysis, DH, according to literature data (eq 2) [19]:
As the degree of hydrolysis increases, the average molecular mass of the hydrolysate decreases, the latter being also correlated with the amine nitrogen content [20,21]. Thus, HO4P2 is the hydrolysate with the highest molecular weight (DH value 3.16%, estimated average molecular mass >25000 Da) whilst HO4P4 is the hydrolysate with the lowest molecular weight (DH value 26.49%, estimated average molecular mass < 3000 Da). This may be explained by the differences in the applied treatment, pKa value of the acid involved in the pretreatment step and the nature of non-collagenous protein removal agent: sample HO4P2 was obtained after treatment with lactic acid (pKa= 3.85) and by using NaCl 1M for removing non-collagenous proteins whilst HO4P4 resulted after pretreatment with citric acid (pKa1= 3.13) and with NaOH 1M, respectively which may cause a partial hydrolysis and by consequence a lower molecular weight of the fragments.
3.3. Infrared spectroscopy (FT-IR)
The structure of collagen hydrolysates samples was confirmed by FT-IR spectroscopy as revealed by comparative IR spectrum for collagen hydrolysates in Fig. 2, similar with other fish collagen hydrolysates reported in the literature. The identified absorption bands are characteristic of the functional groups specific to collagen [22]:
- amide A, due to vibrations of the secondary amine group (-NH-), peaks from: 3282 cm-1 (HO4P1), 3285 cm-1 (HO4P2), 3286 cm-1 (HO4P3), 3289 cm-1 (HO4P4);
- amide B, due to the vibrations of the methylene group (-CH2-), peaks from: 3076 cm-1 (HO4P1), 3077 cm-1 (HO4P2), 3077 cm-1 (HO4P3), 3077 cm-1 (HO4P4);
- aliphatic groups -CH2 and -CH3, attributed to stretching vibrations, peaks from 2929 cm-1 (HO4P1), 2931 cm-1 (HO4P2), 2923 cm-1 (HO4P3), 2927 cm-1 (HO4P4);
- aliphatic groups -CH2 and -CH3, attributed to valence, symmetrical and asymmetric vibrations, located at: 1440 cm-1 (HO4P1), 1442 cm-1 (HO4P2), 1442 cm-1 (HO4P3), 1442 cm-1 (HO4P4);
- amide I, due to the stretch vibrations of the carbonyl group (= CO), peaks from: 1629 cm-1 (HO4P1), 1631 cm-1 (HO4P2), 1631 cm-1 (HO4P3), 1635 cm-1 (HO4P4);
- amide II, due to the deformation vibrations of the secondary amine group (-NH) and the valence vibrations of the nitrile group (-CN), the peaks from: 1532 cm-1 (HO4P1), 1534 cm-1 (HO4P2), 1534 cm-1 (HO4P3), 1538 cm-1 (HO4P4);
- amide III due to the same vibrations as in the case of amide II, peaks from: 1235 cm-1 (HO4P1), 1238 cm-1 (HO4P2), 1238 cm-1 (HO4P3), 1238 cm-1 (HO4P4).
Deconvolution of amide band I (Figure S1, supplementary information) and quantitative band fitting analysis for the main component peaks (Table 2) provided interesting information on protein conformational changes in the obtained fish skin hydrolysates. According to literature data [23, 24] the component peaks obtained after deconvolution of amide band I for the examined collagen hydrolysate samples may be attributed as follows:
- Peak 1, 1625- 1628 cm-1 - intramolecular beta- sheets
- Peak 2, 1651-1656 cm-1 - random coil conformations, imide residues and α-like helix conformations
- Peak 3, 1665-1681 cm-1 - β-turns of the C- and N-telopeptides in collagen
- Peak 4, 1691- 1694 cm-1 - gelatin and helices of aggregated collagen-like peptides
An intense absorption at 1660 cm-1 is generally attributed to the triple helix state of collagen22, a decrease of band intensity accompanied by enhancement of bands around 1630 cm-1 being generally associated with heat denaturation of collagen. Examining the deconvolution component peaks of amide I for the fish collagen hydrolysates samples, no triple helix characteristic absorption is observed, the most intense component being the one due to random coil conformations, imide residues and α-like helix conformations (Peak 2 at 1651-1656 cm-1). This conclusion was also sustained by the circular dicroism and XRD experiments, no triple helix structure being evidenced for all the obtained fish collagen hydrolysates.
3.4. Circular Dichroism Spectroscopy (CD)
Mechanical properties of collagen are greatly influenced by the existing triple helical structure and maintaining the integrity of this structure influence the further applications of the fish collagen hydrolysates. Information about the existence of triple helix structure were obtained by Circular Dichroism Spectroscopy, a spectroscopic method based on the differential absorption of left-right polarized circular light of chiral macromolecules, typically used to evaluate the secondary structure of proteins [25]. CD spectra of the analyzed fish collagen samples are presented in Fig. 3a.
All analyzed fish collagen hydrolysate samples showed a pronounced negative band around 196 to 200 nm, typical of a random coil structure and no positive maximum at the wavelength 220 nm, characteristic for the triple helix was observed. Thus, a specific structure characteristic for the denatured collagen (hydrolyzed collagen) was attributed to all fish collagen extracts HO4P1- HO4P4 [26]. Prevalence of random coil structure and no triple helix characteristic band are in good agreement with FT-IR spectra and intensity of component bands measured after deconvolution of amide I band, the highest intensity being observed for band 2, characteristic for random coil conformations. (Fig. S1).
3.5. X-ray diffraction (XRD) analysis
X-ray diffraction is often used to assess collagen fibril distribution and orientation in fish mineralized tissues. There are two peaks characteristic of collagen molecule and considered as a signature. The first one (~7º) and sharpest is related to the triple helix conformation and distance between molecular chains, and the second peak (~20º) is related to the distance between the skeletons. The X-ray spectra of lyophilized fish collagen hydrolysates samples presented in Fig. 3b reveal no sharp peak associated with the triple helical structure, due to the loosing of the native conformations of collagen during the hydrolysis process as confirmed also by FT-IR characteristic amide I band analysis and CD spectra. The second peak is present at 21.19º (HO4P1), 19.18º (HO4P2), 19.28º (HO4P3) and 19.84º (HO4P4), respectively. The Bragg equation d(Å) = λ/2sin θ (where λ is the X-ray wavelength (1.54 Å), and θ is the Bragg diffraction angle), was used to calculate the minimum value of the repeated spacings. The distance between skeletons were 4.18 Å (HO4P1), 4.62 Å (HO4P2), 4.59 Å (HO4P3) and 4.47 Å (HO4P4) respectively, in good agreement with the reported values for collagen extracted from other fish species [27].
3.6. Determination of isoelectric point (pI)
Possible applications of fish collagen hydrolysates are influenced by their solubility and by the effect of pH on protein solubility determined by its isoelectric point (pI) value. Isoelectric point (pI) is the pH of the collagen molecule at zero charge. Zeta potential measurements for each fish collagen hydrolysate sample at variable pH were performed (Fig. 4) allowing determination of the isoelectric point considered as the pH value at which PZ intersects the x axis (equals zero). The calculated values for pI vary between 3.9 and 2.9 as follows: pI=2.9 for sample HO4P1, pI=2.9 for sample HO4P2; pI=3.7 for sample HO4P3 and pI=3.9 for sample HO4P4, respectively. Similar values of pI for acid and pepsin-soluble collagen were reported in the literature on the extraction of collagen from fish skin and bones. The pI value differences of collagen hydrolysates samples may be attributed to the ratio of acidic and basic amino acid residues, which was more likely governed by the removal of telopeptides after acidic treatment. Hydrolyzed collagen is an amphoteric macromolecule [28] composed of both acidic (COOH) and basic (NH2) functional groups and the pI decrement could be due to the deamination process. When hydrolyzed collagen was treated at high temperature, the asparagine groups transformed to aspartic acid and the glutamine groups into glutamic acid. This leads to a loss of amino groups and a large relative increase in the carboxyl groups, or a higher content of acidic amino acids, which become dominant, shifting the pI to lower values.
3.7. Scanning electron microscopy
Morphology of the obtained fish collagen hydrolysates was investigated using Optical microscopy and Scanning electron microscopy. The optical microscopy results (Fig. S2 supplementary information) confirm the presence of fibrils structure of collagen similar to other reported by literature data [29]. More detailed visual insights were obtained from SEM micrographs (Fig. 5), for all collagen hydrolysate samples a characteristic highly entangled structure, like sheets folded and wrinkled was identified. Both SEM and optical microscopy observations revealed a randomly distributed fibril pattern and confirmed the absence of undesired aggregates (cell debris, skeletal parts or undissociated collagen fibers) between the fibrils of collagen [30].
3.8. Biocompatibility assessment
Human adult keratinocytes, HaCaT cell line proliferation experiments cultured for 6 days on medium supplemented with 1% fish collagen hydrolysates and investigated using phase contrast microscopy (Fig. 6a) revealed no obvious differences in cell morphological aspect for the fish collagen hydrolysates and control, with the exception of the cells cultured in the presence of sample HO4P2 which was not associated with a confluent layer at the end of the incubation period. This result suggests that, at 1% concentration in culture medium, most of the fish collagen hydrolysates samples sustained cell proliferation.
This first observation was supported by the quantification of cell viability using the XTT assay (Fig. 6 b) at the end of the 6 days cultivation period. The percentage of viable cells for the fish collagen hydrolysates was above 80%, except for the sample HO4P2. Furthermore, all 3 samples - HO4P1, HO4P3 and HO4P4, showed a higher viability compared to the control bovine hydrolysate, suggesting a superior effect for biomedical applications. An explanation for low biocompatibility of fish sample hydrolysate HO4P2 may be the highest molecular mass of its component polypeptides as compared to other samples (as estimated from amine nitrogen content) assumption sustained by other literature reports.
