Swim Bladder Biochemical Composition
The swim bladder represented 2% of total body weight in totoaba and contained 67.43 ± 1.24% moisture. Crude protein content (dry weight) was 95.72 ± 1.07%, lipids were 2.46 ± 0.18%, and ash was 0.88 ± 0.06%. Moisture content was low compared to other species: 75.20% in bighead carp (Hypophthalmichthys nobilis) [11]; 83.33% in yellowfin tuna (Thunnus albacares) [10]; 78.83% in miiuy croaker [12]; and 82.8% in catla [8]. These differences may be attributed to variations in swim bladder water content during tissue processing and storage, as well as biological factors. The low total lipids and minerals contents of the totoaba swim bladder are comparable to those of the miiuy croaker [12]. Protein content was higher than reported for catla (83.0%) [8] and miiuy croaker (90.55%) [12].
Proximate composition in fish can depend on many factors, including seasonal variations in feeding behavior, age, and habitat. Nineteen amino acids were identified in the totoaba swim bladders and the amino acids profile showed collagen to be the predominant protein, which coincides with the swim bladders of other fish species [24]. Glycine was the most abundant amino acid, followed by alanine, proline, arginine, glutamic acid, hydroxyproline, and aspartic acid, which represented 85% of total amino acids (AA). Its amino acids composition showed the totoaba swim bladder to be nutritionally poor since it contained only 12% essential AA compared to 53% conditionally essential and 35% non-essential (35%) amino acids. However, it is rich in functional AA (71%), such as glycine, glutamic acid, aspartic acid, proline, alanine, and arginine, all of which participate in and regulate key metabolic pathways [25]. It also contains high levels of hydrophobic amino acids which are frequently found in antioxidant peptides [26]. Overall, totoaba swim bladder had high protein and low lipids contents, suggesting it potential use in collagen extraction, as a source of functional AA, or a substrate for bioactive peptide production.
Collagen Yield
Hydrolysis time for PSC extraction can be as long as 72 h [11, 27], so four extraction times were used in the present study (20, 24, 32, and 72 h) (Fig. 1a). During the extraction process, swim bladder collagen fibers were solubilized entirely in acetic acid upon proteolysis with pepsin (24 h) with good collagen yields (68.18 ± 1.62%, dwb). According to previous studies, intermolecular cross-links in the telopeptide region and triple helices formed via condensation of aldehyde groups cause a decrease in collagen solubility [10], and the pepsin cleaves specifically on the telopeptide region, leading to isolated tropocollagen molecules. For this reason, pepsin is the principal protease used for increasing collagen extraction efficiency and reducing the collagen antigenicity caused by telopeptides [14, 15]. Thus, the PSC yields from totoaba swim bladder are comparable to those of Gulf corvina (Cynoscion othonopterus) (69%) [20] and significantly higher than PSC yields from other species: miiuy croaker (8%), yellowfin tuna (12%), giant croaker (15%) (Nibea japonica), bester sturgeon (38%) (Huso x Acipenser ruthenus), catfish (40%) (Tachysurus maculatus), rohu (47%) (Labeo rohita), bighead carp (59%) and catla (61%) [8, 10–12, 14, 15, 22, 28]. Interestingly, collagen yield from totoaba and other fish species is lower than the 85.3% (dwb) ASC yield reported for seabass (Lates calcarifer) swim bladder [29], suggesting that seabass swim bladder may have less cross-linked collagen fibers. These differences in collagen yield are probably due to extraction conditions, swim bladder firmness (i.e., degree of cross-linking), animal age, nutrition, and development conditions (wild or farmed). Protein content in the totoaba swim bladder collagen (TSBC) was 96.34 ± 1.19%, ash content was 0.83 ± 0.09%, and no fat was detected, which is consistent with collagen from miiuy croaker [12].
Collagen Characterization
Amino Acid Composition
All collagens have a general (Gly-X-Y)n sequence in their polypeptide chains, so glycine can be expected to be the main amino acid [6]. This agrees with the present results in that TSBC glycine content was 309 /1000 residues, followed by alanine (132 /1000 residues) and proline (122 /1000 residues) (Table 1). Low levels of tyrosine, histidine, isoleucine, hydroxylysine, and methionine were observed, and cysteine was not detected, which is reported for collagens [12,25,26]. Aromatic amino acids, mainly tyrosine, are generally found in low concentrations in PSC [22]. Compared to PSC isolated from Gulf corvina and miiuy croaker swim bladders (family, Sciaenidae) [12, 20], the TSBC had higher levels of aspartic acid, glutamic acid, proline, and alanine but lower levels of valine, threonine, isoleucine and leucine (Table 1). This variation in amino acid content could be due various factors, such as fish species biology (health state and age), environment (water temperature and feeding), and habitat (wild or farmed). Imino acid (proline and hydroxyproline) content in the TSBC was 205 /1000 residues (Table 1), which is consistent with miiuy croaker swim bladder collagen [12]. Imino acid content has been reported to positively affect collagen structural stability because the pyrrolidine ring imposes restrictions on polypeptide chain conformation, thus strengthening the triple helix structure [30, 31]. The degree of hydroxylation of proline (41%) and lysine (16%) also influences collagen self-assembly and stabilization [32].
Table 1. Amino acid composition of swim bladder (composition percentage) and swim bladder collagen from farmed totoaba (residues/1000 residues).
Amino Acids
|
Swim Bladder
|
TSBC1
|
Gulf corvina
PSC2
|
PSC Miiuy croaker3
|
Asp
|
5.13 ± 0.05
|
52 ± 1.48
|
38
|
39
|
Glu
|
9.61 ± 0.14
|
101 ± 1.95
|
79
|
85
|
Hyp
|
5.43 ± 0.05
|
83 ± 1.43
|
81
|
88
|
Ser
|
2.02 ± 0.08
|
23 ± 0.69
|
34
|
28
|
Gly
|
29.19 ± 0.31
|
309 ± 3.15
|
303
|
334
|
His
|
0.47 ± 0.02
|
5 ± 0.34
|
7
|
8
|
Arg
|
11.58 ± 0.37
|
59 ± 4.47
|
70
|
55
|
Thr
|
1.76 ± 0.06
|
13 ± 2.62
|
15
|
22
|
Ala
|
12.26 ± 0.13
|
132 ± 1.27
|
118
|
95
|
Pro
|
12.10 ± 0.13
|
122 ± 1.33
|
106
|
112
|
Tyr
|
0.47 ± 0.03
|
2 ± 0.19
|
2
|
2
|
Val
|
1.58 ± 0.04
|
16 ± 0.30
|
20
|
33
|
Met
|
1.36 ± 0.03
|
7 ± 0.47
|
9
|
5
|
Cys
|
0.03 ± 0.01
|
Not detected
|
|
0.4
|
Ile
|
0.63 ± 0.01
|
5 ± 0.22
|
8
|
13
|
Leu
|
1.94 ± 0.05
|
20 ± 0.08
|
31
|
27
|
Hyl
|
0.28 ± 0.02
|
5 ± 0.22
|
9
|
6
|
Phe
|
1.61 ± 0.03
|
19 ± 0.54
|
26
|
23
|
Lys
|
2.52 ± 0.04
|
31 ± 1.63
|
44
|
24
|
Imino acid3
|
|
205
|
187
|
199.5
|
Degree of Hydroxylation (%)
|
|
|
|
|
Pro
|
|
40.65
|
|
|
Lys
|
|
14.43
|
|
|
TSBC1: totoaba swim bladder collagen. Pepsin-soluble collagen (PSC) from Gulf corvina (C. othonopterus)2 [20] and miiuy croaker (M. miiuy)3 [12]. Imino acid: Proline + Hydroxyproline.
Protein Patterns
Electrophoretic analyses of the TSBC showed it to be composed mainly of two different α chains (α1 and α2), in a 2:1 proportion, and a β chain (Fig. 1b). High molecular weight bands were also observed which correspond to β-chains (dimers) and γ-chains (trimer). This suggests that the TSBC is type I collagen, consisting of heterotrimer ([α1(I)]2α2(I)) chains. Swim bladder from other fish species has been reported to contain type I collagen [10,12,30]. Using the GelAnalyzer software, the apparent molecular weights of the TSBC α1 (142 kDa) and α2 chains (134 kDa) were calculated based on migration distance. The extraction process was clearly effective because the collagen preserved its native structure. Moreover, no low molecular weight (<100 kDa) components were observed, suggesting the pepsin cleaved specifically to the telopeptide region, as previously reported [33].
UV-Vis and FTIR Spectroscopy
Maximum absorption for collagen is near 230 nm, due to the peptide bond (R-CONH-R, amide group) of the polypeptide chains [12]. The UV-vis spectrum is therefore an essential parameter for detecting purified collagen [12]. In this spectrum the TSBC exhibited a maximum absorption peak at 228 nm (Fig. 2a), which was similar to collagen from calf skin, grass carp (Ctenopharyngodon idella) [34], and miiuy croaker [12]. As expected, neither the TSBC nor the bovine serum albumin (BSA) reference exhibited a peak at 280 nm; in the TSBC this was due to its low aromatic amino acid (tyrosine and phenylalanine) content (Table 1). This result indicates efficient non-collagen protein elimination, and consequent high TSBC purity.
In the FTIR spectra, TSBC showed characteristic bands of amide A, B, I, II, and III (Fig. 2b). Amide absorption bands A and B, which correspond to the stretching vibration of group N-H and asymmetric stretching of CH2, were observed in wave numbers 3280 and 3071 cm-1, respectively. Amide I (C ═ O stretching), amide II (N-H bending and C-N stretching), and amide III (C-N stretching and N-H bending) appeared in frequencies 1629, 1543, and 1237 cm-1, respectively. The absorption ratio between amide III and the 1454 cm−1 wavelength was 1.05, indicating preservation of the collagen’s triple helix structure. These results coincide with those reported for collagens isolated from other fish species [8, 10–12, 14, 34].
Structural Integrity
The TSBC x-ray diffraction (XRD) spectrum exhibited peaks at 7.7° and 20.02° (Fig. 2c). The former was sharp and corresponded to the triple helix arrangement and distance between molecular chains; the latter was broad and corresponded to the distance between the amino acid residues along the helix [35]. Both peaks were consistent with the characteristic diffraction pattern of the collagen triple helicoidal structure [36]. The circular dichroism (CD) analysis showed the TSBC to have a weak positive absorption peak at 222 nm and a negative one at 197 nm with a crossing point (zero rotation) at 215 nm (Fig. 3a). This CD spectrum pattern is characteristic of the collagen triple helix structure and consistent with previous reports [28]; the 222 nm peak disappears after thermal denaturation [28, 37]. The results confirm the helix structure of TSBC remained in its native form, and therefore that the isolation process did not affect its molecular integrity.
Thermal Behavior
Measurements of CD molar ellipticity (θ) as a function of temperature have been used to determine denaturation temperature (Td) [28]. The present CD (222) values decreased by approximately 34.5°C, indicating decomposition of the collagen triple helix structure (Fig. 3b). Specifically, the intramolecular hydrogen bonds that stabilized the secondary structure of the collagen broke, leading to collapse of the triple helix into a random coil [38]. The present results were similar to those reported for collagen isolated from yellowfin tuna swim bladder (33.9°C) [10] and Gulf corvina (32.5°C) [20]. The Td of TSBC was higher than for collagen from a cold-water fish such as cod (29.6°C) [37], and for a temperate water fish such as miiuy croaker (26.7°C) [12]. Swim bladder collagen from marine fish remains thermostable below 35°C whereas in freshwater fish the threshold is higher: 38°C in grass carp [33] and 39.38°C in catla [8]. Indeed, PSC isolated from the swim bladder of the freshwater fish rohu [14] retains thermal stability at up to 42.16°C, higher than pork skin collagen (37°C) and similar to calfskin collagen [35]. Collagen thermal behavior depends heavily on imino acid content [22, 30], as well as species optimum physiological temperature, which is closely related to its habitat [31, 39]. For the farmed totoaba from UMA, average water temperature is 27 ± 1°C, while under natural conditions surface temperatures in the upper Gulf of California, Mexico, can range from 16 to 31°C on the surface and 13 to 19°C in deep waters (100 to 200 m) [40]. The present TSBC thermal stability result (34.5°C) is probably linked to water temperature in its natural habitat. Possible use of a collagen depends heavily on its thermal stability [41], and the fact that the studied TSBC has thermal stability close to that of terrestrial mammal collagen makes it a promising alternative.
Protein Solubility and Zeta Potential
Acid pH (2.0-4.0) caused higher solubility in the TSBC, but this parameter decreased from pH 5.0-6.0, resulting in protein precipitation (Fig. 3c). Collagen solubility was lowest at around pH 6, but increased slightly in the pH 7.0–10.0 range. This may be due to increased repulsion of collagen molecules as the negative charge increases [22]. Similar results have been reported for PSC from the swim bladder of grass carp [34], miiuy croaker [12], Gulf corvina [20] and giant croaker [22]. Zeta potential is a key marker of colloidal dispersion stability and varies in response to pH [12]. As pH increased in the TSBC suspension, the zeta potential progressively decreased from +27 mV (pH 2) to less than -24 mV at pH 10 (Fig. 3d). At a high magnitude of potential (positive or negative) a solution will resist aggregation, whereas low potential tends to formation of aggregates. For TSBC the zero surface net charge occurred at pH 5.4, this is considered the isoelectric point (pI) and is consistent with the protein solubility results (Fig. 3c). Since the pI occurred at an acid pH, it may be associated with higher contents of glutamic acid and aspartic acid rather than of basic amino acids, such as histidine, lysine, and arginine (Table 2). The pI value was lower than reported for swim bladder collagen from miiuy croaker (6.85) [12] but similar to that of yellowfin tuna (5.93) [10]. In collagen, the pI is generally closely linked to amino acid composition distribution on its surface.
Collagen Hydrolysate
Totoaba swim bladder has putative positive therapeutic effects in traditional Chinese medicine [5]. Peptides and collagen from croaker swim bladders have been shown to remove free radicals [12,17,20,23]. The peptide profiles of the hydrolysates produced from the TSBC using Alcalase® (HCA) and papain (HCP) showed the HCA to have more hydrophilic peptides than the HCP (Fig. 4a). In contrast, the HCP had more hydrophobic peptides when eluted from 10 to 20 minutes. Protein hydrolysates with antioxidant activity frequently contain mainly hydrophobic amino acids, which play a significant role in free radical elimination [26]. Based on this and the present peptide profiles, the HCP was expected to exhibit higher antioxidant activity than the HCA.
The DPPH radical scavenging assay is a popular and efficient way of predicting antioxidant activity since the DPPH radical is more stable than hydroxyl and superoxide radicals [17]. Using the DPPH assay, the antioxidant activity of ultrafiltered fractions (<3 kDa) of the TSBC hydrolysates was tested at 3.2 mg mL-1. Antioxidant activity was 37% higher (p<0.05) with the HCP than the HCA, although ascorbic acid far exceeded both (Fig. 4b). This contrasts with the antioxidant activity results of a study of hydrolysates from the swim bladder of croceine croaker and miiuy croaker in which, at 15-25 mg protein mL-1, the Alcalase® hydrolysate had significantly higher activity than hydrolysates prepared with papain, pepsin, neutrase and trypsin [26]. Of note is that, after ultrafiltration, a lower concentration of HCA and HCP (3.2 mg mL-1) produced higher antioxidant activity than in the above study. Overall, the present antioxidant activity indicates that this parameter depends strongly on the enzyme used for hydrolysis, suggesting further research is needed to isolate active peptides and clarify their antioxidant activity.