Introduction

Calcium plays an important role in bone health. Calcium intake could increase bone mineral density (BMD) in children, and it is also essential among middle-aged individuals and the elderly for the prevention of osteoporosis [1]. Low bioavailability of calcium may negatively affect peak bone mass, result in hypertension and contribute to the development of osteoporosis and intestinal cancer [2–4]. With the increase in the aged population throughout the world, there is a growing interest in developing more effective calcium supplements to prevent and treat bone disease [5]. The weak binding of certain substances (especially proteins and peptides) to calcium can prevent it from precipitating and therefore effectively increase calcium absorption within the body [6]. For this reason, calcium-chelating peptides such as organic calcium supplements have started to become a popular topic of research [7, 8]. It has been reported that peptides from soybean [9], fish bone [10, 11], shrimp processing by-products [12] and whey protein [13] were found to bind calcium and to promote both in vitro and in vivo absorption of calcium.

Salmon, which is considered to be healthy due to the fish’s high nutritional value and pharmacological activity, is a popular food throughout the world. Salmon bone, the underutilized by-product from salmon processing, consists of considerable amounts of ossein, and thus, it may be developed into an important source of ossein. Presently, commercially used collagen is mainly obtained from cattle and pigs. Furthermore, its application seems to be restricted by safety and religious barriers [14]. In the last few years, marine food sources have gained attention as an alternative source of collagen as previous study has reported on the development and characteristics of such collagen resources [15]. As one kind of bioactive substance in functional foods, the calcium-chelating capacity of enzyme-hydrolyzed salmon ossein peptides is of great interest and should be investigated as a possible supplementary treatment for calcium deficiency. However, a systematic study of calcium-chelating peptides in salmon bone has never previously been undertaken.

In the present study, salmon ossein oligopeptides (SOOP) were prepared by hydrolyzing Atlantic salmon (Salmo salar L.) bone and calcium-chelating salmon ossein oligopeptides (SOOP-Ca) were obtained by calcium chelating these peptide reactions from SOOP. The calcium-chelating capacity, chemical composition, amino acid composition, molecular weight distribution and characterization of SOOP-Ca were determined. SOOP-Ca was separated by reversed-phase high-performance liquid chromatography (RP-HPLC) and subjected to mass spectrometry to identify component peptides.

Materials and methods

Materials and reagents

Atlantic salmon (Salmo salar L.) bones were donated by CF Haishi Biotechnology Ltd. Co. (Beijing, China). Acalase (2.4 L), papain, pepsin and trypsin were purchased from Novozymes Biological Co. (Tianjin, China). Anhydrous calcium chloride, ethanol, ammonium hydroxide, ammonium chloride, eriochrome black T, magnesium sulfate, triethanolamine, ethylenediaminetetraacetic acid (EDTA), hydrochloric acid and sodium hydroxide were bought from Beijing Chemical Reagent Co. (Beijing, China). Acetonitrile was purchased from Fisher Scientific (Pittsburgh, USA). Trifluoroacetic acid (TFA) was obtained from Alfa Aesar. All other reagents were of analytical or guaranteed reagent grade.

Preparation of SOOP-Ca

Atlantic salmon bones (100 g) were crushed to powder and dissolved in 1000 mL of distilled water at 90 °C for 10 min. The slurry was first digested by alcalase (2.4 L) using an enzyme-to-substrate protein ratio of 2:100 (w/w) at pH 8.5 and a temperature of 50 °C for 3 h. The resulting supernatant was further digested by papain (3:100, w/w) at pH 7 and 60 °C for 2 h. Hydrolysis was carried out in a thermostatically stirred-batch reactor (Dongding Machinery Co., Wenzhou, China). The pH was held constant during hydrolysis by continuous addition of 1 M NaOH. At the end of the reaction, the hydrolysis was stopped by heating at 100 °C for 10 min. The hydrolysate was centrifuged (LG10-2.4A, Beijing LAB Centrifuge Co. Ltd, China) at 1000×g for 30 min. The supernatant was passed through 10 and 1 ku molecular weight cutoff ceramic membranes (Filter and Membrane Technology Co. Ltd, Fujian, China) successively to obtain the permeate liquid containing low molecular weight peptides. Afterward, the concentrated liquid was spray-dried using an L-217 Lab spray dryer (Beijing Laiheng Lab-Equipments Co. Ltd, China) under an inlet temperature of 220 °C and an outlet temperature of 200 °C, to obtain the salmon SOOP powder.

The SOOP powder (3 g) and anhydrous calcium chloride (1 g) were dissolved in 100 mL of distilled water and incubated for 1 h at 50 °C in a water bath. After the incubation, the mixture was added to 400 mL of 95 % ethanol and left to stand for 1 h at room temperature. Then, the precipitate was collected by suction filtration and dried at room temperature. The SOOP-Ca that was obtained was stored in a desiccator for further analysis. The yield was calculated using M 1/M 2 × 100 %, where M 1 was the mass of SOOP-Ca obtained and M 2 were the total mass of SOOP and anhydrous calcium used.

Calcium-chelating capacity assay

The calcium-chelating capacity of SOOP-Ca was determined using EDTA complexometric titration. Briefly, 2 drops of chrome black T indicator and 2 drops of MgSO4 were added to the 15 mL of NH3·H2O-NH4Cl (pH 11), prior to the addition of 0.05 mol/L EDTA, until the solution turned blue. A volume of 5 mL of the mixture containing SOOP powder and anhydrous calcium chloride before chelating reaction (or SOOP-Ca obtained) was mixed with 20 mL distilled water and 3 drops of triethanolamine. Then, the titrated buffer solution was added to the reaction mixture, prior to the addition of 0.05 mol/L EDTA, until the solution turned blue. The volume of EDTA used was designated as V 1 (or V 2). The calcium-chelating capacity was calculated using V 2/V 1 × 100 %.

Chemical composition and amino acid composition assay

Crude protein content is expressed as total nitrogen (N) × 6.25, of which the nitrogen content was determined by the Kjeldahl method [16]. Amino acid composition was determined by a 835-50 automatic amino acid analyzer (Hitachi, Ltd., Tokyo, Japan) according to a pre-established method by Yang, Tao, Liu and Liu [17].

Determination of molecular weight distribution

The molecular weight distribution was determined using an LC-20A HPLC system (Shimadzu, Kyoto, Japan) equipped with a TSK-GEL G2000SWXL 300 × 7.8 mm column (Tosoh, Tokyo, Japan). The mobile phase used was acetonitrile/water (45:55, v/v) containing 0.1 % (v/v) trifluoroacetic acid. Samples were eluted at a flow rate of 0.5 mL/min and monitored at 220 nm at 30 °C. Tri-peptide GGG (189 u), tetrapeptide GGYR (451 u), bacitracin (1450 u), aprotinin (6500 u) and cytochrome C (12,500 u) (Sigma Chemical Co., St. Louis, USA) were used as molecular weight standards.

Characterization of SOOP-Ca

Scanning electron microscopy of SOOP and SOOP-Ca

The microstructures of SOOP and SOOP-Ca were analyzed by S-3400 N electron scanning microscopy (Hitachi, Ltd., Tokyo, Japan) under 500 and 1000 times magnification to observe the differences in the microstructures of SOOP and SOOP-Ca.

UV wavelength scanning of SOOP and SOOP-Ca

SOOP and SOOP-Ca solutions at the same concentrations were identified by UV spectra using UV752 ultraviolet–visible spectrophotometer (MAPADA Instrument Co. Ltd, Shanghai, China) within the range of 190–400 nm.

Infrared spectra of SOOP and SOOP-Ca

A Nicolet 6700 Fourier infrared spectrometer (PerkinElmer Co., Massachusetts, USA) was used to analyze SOOP or SOOP-Ca. The spectrograms were scanned under the condition of 4000–500 cm−1.

Stability of SOOP-Ca

The stability of SOOP-Ca against temperature, pH and in vitro protease digestion treatment was evaluated according to the method developed by Wu and Ding, with some modification [18]. SOOP-Ca solutions (2 mg/mL) were incubated at various temperatures (20, 40, 60 and 80 °C) for 2 h. The solutions were also incubated at 37 °C and pH values of 3, 5, 7 and 9 for 2 h. After the solutions were acclimated to room temperature, the molecular weight distribution and calcium content were determined.

The stability against in vitro pepsin was assessed by treating 2 mg/mL of SOOP-Ca solution in 0.1 M HCl-KCl buffer (pH 2.0), with 3 % (w/w) pepsin for 3 h in a water bath at 37 °C. The reaction was stopped by boiling for 10 min. For stability during digestion with trypsin, 2 mg/mL SOOP-Ca solution in 0.1 M KH2PO4-NaOH buffer (pH = 6.8) was digested with 3 % (w/w) trypsin for 3 h in a water bath at 37 °C. The reaction was stopped by boiling for 10 min. For stability against in vitro pepsin digestion and further trypsin digestion, 2 mg/mL SOOP-Ca solution was first digested by 3 % (w/w) pepsin for 3 h and adjusted to pH 6.8 with the addition of 1 N NaOH solution. The solution was digested further by 3 % (w/w) trypsin at 37 °C for 3 h. The enzyme was inactivated by boiling for 10 min. After the solutions were acclimated to room temperature, the molecular weight distribution and calcium content were determined. Molecular weight distribution assay methods are described above. The calcium content was determined using the method of EDTA complexometric titration as described above. One milliliter of EDTA equaled 2.0039 mg of calcium.

Separation of SOOP-Ca with RP-HPLC

SOOP-Ca was separated using an RP-HPLC system (Shimadzu, Kyoto, Japan) equipped with an XBridge BEH130 C18 column (4.6 × 250 mm, Waters, USA). Gradient elution was performed using eluent A (Milli-Q water containing 0.1 % TFA) and eluent B (80 % acetonitrile containing 0.1 % TFA). The separation was performed at a flow rate of 0.6 mL/min with a nonlinear gradient as follows: 0–10 min, 0–5 % B; 10–20 min, 5–5 % B; 20–35 min, 5–9 % B; 35–45 min, 9–13 % B; 45–60 min, 13–13 % B; 60–70 min, 13–100 % B; 70–85 min, 100–100 % B. The effluent was monitored at 220 nm. The sample concentration was 10 mg/mL, and the injection volume was 100 µL. The separation procedure was repeated until enough samples were collected. The fractions from the RP-HPLC system were freeze-dried for further analysis.

Identification of active peptides

The molecular mass and amino acid sequence of peptides were analyzed by a quadrupole time-of-flight mass spectrometer (Q-TOF2; Micromass Co., Manchester, UK) equipped with an electrospray ionization (ESI) source.

Statistical analysis

All assays were carried out in triplicate. Data were expressed as mean ± standard deviation of the mean.

Results and discussion

Chemical composition and amino acid composition of SOOP-Ca

The yield of SOOP-Ca was 40.19 ± 0.11 %. The calcium-chelating capacity of SOOP-Ca was 52.47 ± 0.08 %. Chemical composition analysis showed that SOOP and SOOP-Ca contained 90.14 ± 1.15 % and 63.42 ± 1.02 % protein (on a dry basis, N × 6.25), respectively.

The amino acid compositions of SOOP and SOOP-Ca are listed in Table 1. SOOP was rich in glycine, alanine and glutamine acid, and low in histidine, phenylalanine, cysteine and methionine, which was similar to that reported by Bae et al. [19]. The contents of aspartic acid, glutamine acid, glycine and alanine in SOOP were 6.69 ± 0.13 %, 8.70 ± 0.26 %, 20.02 ± 0.78 % and 11.63 ± 0.28 %, respectively. After chelating calcium, the contents of these amino acids in SOOP-Ca were 11.92 ± 0.32 %, 15.22 ± 0.42 %, 11.82 ± 0.36 % and 10.33 ± 0.29 %, respectively. These results indicated that the calcium-chelating capacity of acidic amino acids, aspartic acid and glutamine acid was better than that of glycine and alanine. This may be due to the carboxyl groups on their side chain which might exert a stronger calcium-chelating capacity during the reaction. Previous studies have shown that soybean peptides can also bind calcium by the carboxyl groups of Glu and Asp residues [9, 20].

Table 1 Amino acid composition of SOOP and SOOP-Ca (%, w/w)
Full size table

Molecular weight distribution of SOOP-Ca

The molecular weight distributions of SOOP and SOOP-Ca were analyzed by size exclusion chromatography using a HPLC system. The result (Fig. 1) showed that SOOP and SOOP-Ca were dominated by peptides below 1000 u (79.07 and 72.32 %, respectively). The range of 500–140 u was the main molecular weight interval for both SOOP and SOOP-Ca, and the range of 500–140 u accounted for 43.17 and 34.00 %, respectively, in SOOP and SOOP-Ca. These peptides were mainly di- and tri-peptides. It has been reported that di- and tri-peptides are actively transported via a specific peptide transporter in intestinal epithelial cells and that amino acid residues are absorbed more rapidly from di- and tri-peptides than from free amino acids [21]. The average molecular weight distributions of SOOP and SOOP-Ca were 610.20 and 639.90 u, respectively. The results showed that after chelating calcium, the molecular weight distribution of SOOP-Ca was slightly higher than that of SOOP, which indicated that as for peptides below 1000 u, relatively larger peptides might have a stronger calcium-chelating capacity than low molecular weight peptides.

Fig. 1
figure 1

Molecular weight calibration curve of the standards (a) and size exclusion chromatography showing the molecular weight (MW) distribution of SOOP (b) and SOOP-Ca (c). The chromatography was carried out on a TSK-GEL G2000SWXL column (7.8 × 300 mm) eluted in 45 % acetonitrile with 0.1 % trifluoroacetic acid at a flow rate of 0.5 mL/min

Full size image

Characterization of SOOP-Ca

Scanning electron microscopy of SOOP and SOOP-Ca

As depicted in Fig. 2, the size, shape and surface traits of the powder particles of SOOP and SOOP-Ca showed different states. The microstructure of SOOP was spherical shaped with holes, while the microstructure of SOOP-Ca was mainly clump shaped with a rough surface. Therefore, it was confirmed that after the chelation reaction, SOOP and SOOP-Ca were two completely different substances.

Fig. 2
figure 2

Scanning electron microscopy of SOOP (a) under 500 (left) and 1000 (right) times magnification and SOOP-Ca (b) under 500 (left) and 1000 (right) times magnification

Full size image

UV wavelength scanning of SOOP and SOOP-Ca

UV wavelength scanning analysis (Fig. 3) showed that the absorbance peak was 230 nm, which conformed to the peak characteristics of peptides in the ultraviolet spectrum. After chelation with calcium, the maximum absorption peak moved to a shorter wavelength (221 nm). Obvious differences were also observed between the peak shapes of SOOP and SOOP-Ca before and after calcium chelation. This was mainly due to the transition of valence electrons corresponding to parent atoms changing after calcium chelation, which led to changes in light absorption properties of the ligands.

Fig. 3
figure 3

UV wavelength scanning of SOOP (a) and SOOP-Ca (b) with wavelengths in the range of 190–400 nm

Full size image

Infrared spectra of SOOP and SOOP-Ca

Figure 4 shows the infrared spectra of SOOP and SOOP-Ca. The sites at which peptides chelate calcium are mainly amide bonds, side chains and some carboxyl groups and terminal side amino groups. Therefore, carboxyl and amino groups had peak absorption bands in the infrared spectrum of SOOP-Ca. The absorption peaks of the infrared spectrum deviated after calcium was chelated by the peptides. The peak at 3066 cm−1 in the infrared spectrum of SOOP was narrow, which resulted from the stretch vibration band of N–H. In the infrared spectrum of SOOP-Ca, the wavelength changed to 3305 cm−1, which moved toward the high band for 239 wave bands. The absorption peak was broadened, which is the characteristic frequency of ammonium salt. It could be inferred that −NH2-chelated calcium and ammonium salt was generated. In the infrared spectrum of SOOP, there was a strong absorption peak at 1395 cm−1 indicating that there was -COOH in SOOP. In the infrared spectrum of SOOP-Ca, this peak moved toward the high band for 23 wave bands and changed to 1418 cm−1, which is the characteristic frequency of the carboxyl group, suggesting the strong combination of Ca2+ and -COOH. These results indicated that SOOP-Ca is a new type of marine fish ossein peptide-chelated calcium.

Fig. 4
figure 4

Infrared spectra of SOOP (lower line) and SOOP-Ca (upper line). The spectrograms were scanned under the condition of 4000–500 cm−1

Full size image

Stability of SOOP-Ca

With the great interest that has risen in preparing foods with these “functional ingredients,” it is important to test their processing stability. As depicted in Fig. 5a, these peptides retained molecular weight distribution and calcium content after various temperature treatments, which indicated that SOOP-Ca is relatively stable at a wide range of temperatures.

Fig. 5
figure 5

Stability of SOOP-Ca after 3-h incubation at various temperatures (a), pH values (b) and digestion with various proteases (c). The figures on the left represent size exclusion chromatography showing the molecular weight distribution of SOOP-Ca (a treated at 20, 40, 60 and 80 °C from bottom to top; b treated at pH 3, 5, 7 and 9 from bottom to top; c treated by pepsin digestion, trypsin digestion and pepsin digestion plus further trypsin digestion from bottom to top). The right panels show relative calcium content of SOOP-Ca. The relative calcium content was calculated as the ratio of calcium content between the control and the treatment groups

Full size image

After various pH treatments, no distinct changes of peak height were observed on the size exclusion chromatograms of SOOP-Ca (Fig. 5b). In acidic and alkaline conditions (pH 3, pH 5 and pH 9), SOOP-Ca molecular weights below 1000 u were slightly elevated, but the content did not exceed 6 %. This may be the case because the macromolecular ring structures formed by the peptides and calcium were broken to micromolecular structures by the alkaline and/or acidic conditions. The result indicated that SOOP-Ca showed some stability against changes to pH (depending on molecular weight). After exposure to different pH values, the calcium contents of SOOP-Ca decreased into the range of 65.9 ± 6.9 % to 88.7 ± 7.3 %, relative to the control. Under strong acidic or alkaline conditions, the calcium content decreased significantly (p < 0.05). However, the calcium content changed very little under neutral conditions. The remaining calcium-chelating peptides could be absorbed rapidly into the intestinal tract with the help of small peptides.

For stability against in vitro gastric proteases digestion, neither distinct new peak nor no distinct changes of peak height were observed on the size exclusion chromatograms of SOOP-Ca (Fig. 5c). After protease digestion, the SOOP-Ca molecular weights below 1000 u were slightly elevated, but they did not exceed 8 %. Compared to the control, after protease digestion, molecular weights of 500 u or below were slightly elevated indicating that peptides with larger molecular weights could be digested to smaller peptides or amino acids. Previous studies have also reported that small peptides have low susceptibility to hydrolysis by gastric proteases [18, 22]. These results suggest that SOOP-Ca may be resistant to digestion in the gastrointestinal tract. With respect to molecular weight, SOOP-Ca exhibited certain resistance against protease digestion by pepsin and trypsin. After being digested with different proteases, the calcium content of SOOP-Ca decreased (p < 0.05) to 58.9 ± 7.2 %, 54.3 ± 6.3 % and 41.2 ± 6.8 %, respectively, relative to the control. The structure of SOOP-Ca is an important factor that may influence its calcium content. As demonstrated by various pH and protease treatments, small changes of molecular weight (no more than 6–8 %) could lead to large changes in calcium content.

Purification and identification of peptides from SOOP-Ca

RP-HPLC fractionation was used to identify the peptides from SOOP-Ca. Figure 6 shows the elution profile of SOOP-Ca. As shown in this figure, many peaks were detected. Ten major fractions, designated as fractions 1–10, were selected for further identification of active sequences. These fractions were pooled, freeze-dried and subjected to a Q-TOF2 mass spectrometry. Overall, 22 peptides were identified from SOOP-Ca and the amino acid sequences and molecular masses of the peptides are summarized in Table 2. All peptides contained two to ten amino acid residues, which was consistent with the molecular weight distribution analysis. The obtained peptide sequences were compared with the complete sequence of proteins that constitute ossein, such as collagen, in the website of NCBI (National Center for Biotechnology Information). Some of the peptide sequences could be located in salmon collagen. For example, the peptide Ser-Arg (SR) could be located in the position of 66–67, 963–964 and 1414–1415 of collagen type XI [Salmo salar] (ABQ59674), and Leu-Arg (LR) could be located in the position of 56–57 of collagen alpha-1XIII chain [Salmo salar] (NP_001167092).

Fig. 6
figure 6

Elution profile of SOOP-Ca by RP-HPLC. Separation was carried out using a nonlinear gradient of eluent A (0.1 % TFA in Milli-Q water, v/v) and eluent B (80 % acetonitrile with 0.1 % TFA, v/v) at a flow rate of 0.6 mL/min. Collected fractions are designated as fractions 1–10. Each fraction was collected according to the elution time, condensed and free-dried

Full size image
Table 2 Identification of peptides included in the RP-HPLC fractions of SOOP-Ca by Q-TOF2 mass spectrometry
Full size table

Previous studies have reported that soybean peptides can bind calcium via Glu and Asp residues [9, 20]. The presence of Glu and Asp, especially densely distributed in peptides or protein, could create a favorable environment for divalent metal ions to coordinate, most likely due to the modification of the local charge density by the carboxylate groups in the side chains [23]. Among our 22 identified peptides, 17 peptides had Glu and Asp residues. Therefore, we speculate that the Glu and Asp residues may be important to their role as calcium-chelating peptides. Also, it was reported that other amino acids, such as Ser, Thr, Met and Tyr, could contribute to calcium-binding capacity [24, 25]. Among the 5 peptides that did not have Glu and Asp residues, 4 peptides had Ser or Thr residues. These preferable amino acids were not found in the peptide Leu-Arg. In addition to the amino acid composition, the spatial structure formed through mutually interaction among amino acid residues in the peptide may affect calcium affinity [26]. Further study will be required to test this possibility.

Conclusions

In the present study, SOOP-Ca was prepared by hydrolyzing Atlantic salmon (Salmo salar L.) ossein prior to chelating calcium. Scanning electron microscopy, UV wavelength scanning and infrared spectra all showed that the structure of SOOP-Ca is different from that of SOOP. SOOP-Ca was stable and maintained certain resistance against temperature, pH and in vitro digestion with gastric proteases. SOOP-Ca was purified, and the amino acid sequences of the peptides were identified. Based on these results, it could be concluded that the production of SOOP-Ca is a practical way to utilize salmon bone. This study suggests that SOOP-Ca might be useful as food additives, dietary nutrients and pharmaceutical agents which could promote calcium absorption. Further studies should be carried out to evaluate the calcium-absorption efficiency in vitro and in vivo of SOOP-Ca and investigate the calcium-peptide chelating reaction using the purified peptides identified from SOOP-Ca.