Effects of phosvitin phosphopeptide-Ca complex prepared by efficient enzymatic hydrolysis on calcium absorption and bone deposition of mice

Mengdie Zho, Dong Uk Ahn, Songming Li, Wei Liu, Shengwei Yi, Xi Hung,*

a National Research and Development Center for Egg Processing, College of Food Science and Technology, Huazhong Agricultural University, Wuhan 430070, China

b Animal Science Department, Iowa State University, Ames 50011, USA

c Chongqing Institute for Food and Drug Control, Chongqing 401121, China

ABSTRACT

Phosvitin (PV) was treated with high-temperature, mild pressure (HTMP), and enzyme combination, and then phosvitin phosphopeptides-calcium (PPP-Ca) complexes were prepared. The low-calcium specific pathogen free-Kunming (SPF-KM) mice were used to determine the effect of PPP-Ca complexes on intestinal calcium absorption and their utilization for bone formation. The serum calcium content was the highest with the HTMP-Enz-PPP-Ca treatment (2.19 mmol/L), and it significantly down-regulated the abnormal elevation of serum alkaline phosphatase (AKP) caused by calcium deficiency. The low-calcium control group had the lowest calcium deposited to the femur (80.41 mg/g) and the lowest femur bone mineral density (BMD)(0.17 g/cm3), while HTMP-Enz-PPP-Ca significantly improved bone calcium content (94.33 mg/g) and BMD (0.29 g/cm3). The micro-computed tomography (MCT) images showed that the femur with the normal control, PV-Ca, and HTMP-Enz-PPP-Ca treatments had a more compact, complete, and thicker trabecular network than the low-calcium and CaCl2 treatments. These results indicated that the organic calcium (HTMP-Enz-PPP-Ca) promoted calcium absorption and bone deposition, and the effect of HTMP-Enz-PPP-Ca was better than the inorganic CaCl2.

Keywords:

Phosvitin phosphopeptide (PPP)-Ca complex

SPF KM mice

Calcium absorption

Bone formation

1. Introduction

Calcium is the major divalent cation in the human body and accounts for about 1.5%-2.2% of total body weight [1]. The majority (99%) of calcium in the body is deposited in bones and teeth. At the same time, calcium also exists in an ionic form in the soft tissues, extracellular fluid, and blood, which acts not only as an intracellular messenger for muscle contraction and relaxation,neurotransmission, immune response, and cell proliferation but also maintain a dynamic balance between the serum and bone calcium content [2]. Thus, calcium deficiency can lead to metabolic bone diseases such as rickets and osteoporosis.

Dietary intake is the only way to supplement body calcium and is absorbed in the small intestine through the voltage-gated (active)transcellular and the passive paracellular pathways [3]. The active transcellular pathway is known as the main transport pathway in the duodenum that requires the combined actions of three calcium transport proteins: 1) the regulation of calcium influx by calcium transport proteins (mainly transient receptor potential vanilloid type 6,TRPV6) [4]; 2) calcium transfer mainly by the calcium-binding proteins (calbindin-D9k) [5]; and 3) the extrusion of calcium into the blood by the plasma membrane calcium ATPase1b(PMCA1b) [6]. In the paracellular transport, ionized calcium diffuses through the tight junctions into the basolateral spaces of the enterocytes and the blood.Paracellular calcium absorption mainly occurs in the jejunum and ileum, especially when the dietary calcium levels are high [7].

Because calcium is only absorbed in its ionic form, it should be solubilized or released from its sources. However, some of the solubilized calcium can form insoluble complexes with minerals or other dietary constituents such as oxalic acid and phytate in the alkaline pH of the small intestine, resulting in inadequate calcium absorption and utilization [8]. Recent research showed that casein phosphopeptides (CPPs) could bind calcium ions to form soluble peptide-calcium complexes, promote calcium absorption, and improve calcium accumulation in bones [9]. The CPPs produced from caseinpromoted calcium uptake in the Caco-2 cells by up-regulating the expression of TRPV6, a key calcium-transport protein in the duodenum, and increased serum Ca2+levels, femur length,and femur calcium in a Sprague-Dawley rat model by up-regulating the expression of TRPV6 [10]. Zhang et al. [11] reported that the phosphorylation of functional proteins or polypeptides is vital because the phosphate group’s multiple negative charges play an essential role in the binding of divalent metal ions.

Phosvitin (PV), a natural phosphoprotein in egg yolk, is the most phosphorylated protein in nature. PV accounts for 8%–11% of the egg yolk protein and consists of 217 amino acids, of which 128 amino acid residues (123 serines, 4 threonines, and 1 tyrosine)can be phosphorylated [12,13]. Almost all the serine and threonine residues in PV are phosphorylated, and many groups of serine residues are arranged in clusters of 15 consecutive residues [13].Although all the commercial phosphopeptides are currently prepared using casein, PV has a much higher phosphorylation level and is a much better source for the preparation of phosphopeptide.However, the preparation of phosvitin phosphopeptides (PPP) using PV is exceptionally challenging because PV has extreme negative charges that block the access of proteases to the cleavage site [13].Recently, our group developed a high temperature and mild pressure(HTMP, 121 °C at 0.1 MPa) pretreatment to improve the enzymatic hydrolysis of PV without losing phosphate groups in the PV. The result showed that HTMP pretreatment alone produced 310 peptides,but the HTMP and subsequent enzyme treatments further improved PV hydrolysis and produced up to 605 phosphopeptides from PV [12]. Because PPP would have similar structural characteristics to the CPPs, it is assumed that the PPP also would promote calcium absorption in animals.

The objectives of this study were to determine the effects of the dietary PPP-Ca complex on intestinal calcium absorption and to elucidate the mechanisms involved in the use of absorbed calcium in bone formation using a low-calcium specific pathogen free-Kunming(SPF-KM) mice model.

2. Materials and methods

2.1 Materials

PV was prepared using the method of Lee et al. [14]. A standard feed for mice (AIN93) was purchased from Trophic Animal Feed High-tech Co., Ltd. (Jiangsu, China), trypsin (E.C.3.4.4.4,15 500 U/mg protein) and thermolysin from Bacillus thermoproteolyticus rokko (Thermoase PC10F, E.C. 3.4.24.27, 113 U/mg protein) were obtained from American Enzyme Co., Ltd. (Elgin, IL, USA), and serum calcium, phosphorus, and alkaline phosphatase (AKP) kit were from Nanjing Jiancheng Bioengineering Institute (Nanjing, China).

2.2 Preparation of PPP from phosvitin

The PPP were prepared following the method of Huang et al. [12]with some modifications. The temperature, pressure, and time conditions used for the HPMP pretreatment were 121 °C at 0.1 MPa for 30 min, and then HTMP pretreated PV was further hydrolyzed using trypsin and thermolysin successively to prepare HTMP-Enz-PPP. The same enzyme:substrate ratio (1:50, m/m) and incubation time (8 h) were used for both enzymes. The hydrolysis conditions for trypsin were pH 8.0 and 37 °C incubation temperature,and those of the thermolysin were pH 8.0 and 68 °C incubation temperature. The hydrolysis with the second enzyme (thermolysin)was performed after inactivating the first enzyme (trypsin) at the end of the 8 h incubation. Enzymatic digestion was arrested for both enzymes by keeping the sample in a boiling water bath for 10 min.The hydrolyzed solutions were lyophilized and stored in a -20 °C freezer until use.

2.3 Preparation and characterization of PPP-Ca complexes

The PPP-Ca complexes were prepared from the HTMP-Enz-PPP combinations. The HTMP-Enz-PPP were dissolved in deionized water (10 mg/mL), and then CaCl2was added with different mass ratios (PPP:CaCl2= 5:1–10:1). The calcium-binding rate was used to screen the peptide calcium ratio. The factors such as pH, temperature, and time in the chelating reaction process were obtained by a single factor experiment with calcium-binding rate as an index (specific data are not presented). Thefinal chelation method was that HTMP-Enz-PPP and CaCl2(at ratio of 7:1) were mixed with distilled water, first.Then the pH of mixture was adjusted to 9.5 and it was incubated for 60 min at room temperature for calcium-binding reactions. The absolute ethanol (9 volumes of the PPP solution) was added to the solution after chelation, held for 3 h at room temperature to precipitate the PPP-Ca complexes, and then centrifuged at 7 100 × g for 15 min at 4 °C. The precipitant was collected, lyophilized, and marked as HTMP-Enz-PPP-Ca complex. The natural PV was also used to prepare the PV-Ca complex using the same method as the HTMP-Enz-PPP.

The degree of hydrolysis (DH) and calcium-binding rate of PPP (HTMP-PV, HTMP-Trypsin-PPP, HTMP-Thermolysin-PPP,HTMP-Trypsin+Thermolysin-PPP) were determined using the ninhydrin method and the atomic absorption spectrometry [15]. Zeta potential was used to characterize the solution’s stability and pure charge of proteins and peptides [16]. The FTIR and fluorescence spectrometry was used to characterize the structure of PPP-Ca complexes.

2.4 Stability of PPP-Ca (HTMP-Enz-PPP-Ca) complexes

The gastrointestinal stability of the PPP-Ca complex was divided into two stages: the PPP-Ca complex was added to simulated gastric fluid (CZ0212, Leagene Biotechnology) and then incubated for 6 h in a water bath at 37 °C. During the incubation, an aliquot of samples was taken out at 0.5, 1.0, 2.0, 4.0, and 6.0 h of incubation,heated at 100 °C for 10 min to inactive pepsin, and then centrifuged.The contents of calcium ions in the supernatants were analyzed to determine calcium release from the PPP-Ca complexes under acidic stomach conditions. The remaining samples at each incubation time were taken out, and their pH was adjusted to 7.5 and then added to artificial intestinal juice (CZ0200, Leagene Biotechnology)for 8 h in a water bath at 37 °C. During the incubation, an aliquot of samples was taken out from each sample at 0.5, 1.0, 2.0, 4.0,and 8.0 h of incubation, heated at 100 °C for 10 min to inactive trypsin, alcohol sink, and then centrifuged. The content of calcium ions in the supernatant was analyzed to determine the stability of PPP-Ca complexes under small-intestinal conditions. The stability was expressed in terms of calcium retention rate:

Caafterrepresented calcium content after digest; Cabeforerepresented calcium content before digest.

2.5 Animal experiments

2.5.1 Experimental design and sample collection

The KM low-calcium mice model developed by Wei et al. [17] was used in this study. The mice experiment was approved by the Hubei Laboratorial Animal Administration Committee (certificate number:HZAUMU-2019-027), and all the procedures used followed the National Institute of Health Guidelines for the Care and Animals. A total of 90 SPF male KM mice (4 weeks of age) were purchased from Hubei Experimental Animal Research Center (Wuhan, Hubei, China)and used for this study. Upon receiving, the mice were acclimated under the laboratory conditions ((22 ± 2) °C temperature, (60 ± 5)% humidity, and 12 h dark/light cycle) with the control diet for 1 week, and then divided into 5 groups (normal-calcium control;low-calcium control; PV-Ca; HTMP-Enz-PPP-Ca; CaCl2, 15 mice per group). The normal-calcium control group was fed with AIN-93 standard animal feed (calcium content 0.5%) until the end of the experiment (8 weeks), and the low-calcium control group was continued to feed with the same low-calcium diet for 4 weeks. The other 4 groups were fed with a low-calcium AIN-93 diet (0.05% calcium)supplemented with PV-Ca, HTMP-Enz-PPP-Ca or CaCl2for 4 weeks in the form of solution through oral gavage once a day. The targeted calcium intake for the calcium sources (PV-Ca, HTMP-Enz-PPP-Ca,and CaCl2) was set at 19.95 mg Ca/(day·kg bw). The amount would be 19.95 × 8 mg bw/day for PV-Ca, and HTMP-Enz-PPP-Ca,and 19.95 × 2.775 mg bw/day for CaCl2. The calcium sources were dissolved in distilled water and fed by gavage separately from the feed.

2.5.2 Measurement of serum indicators

At the end of the 28-day feeding trial, the mice have fasted for 14 h, chloral hydrate was administered intravenously to sedate the mice, and then blood samples were collected from the eyeballs. The blood was centrifuged at 1 170 × g for 10 min, and then the serum was used to measure serum indicators (calcium and phosphate content and serum AKP activity). The mice were then sacrificed by cervical dislocation, and the duodenum and femur were collected and used to determine the expression of proteins associated with the TRPV6 Ca2+pathway and calcium deposition, respectively.

2.5.3 Determination of duodenal calcium pathway proteins

The duodenal calcium pathway-related proteins were determined using Western blotting. After sacrificing the mice, the duodenal tissues were collected and placed in 2 mL Eppendorf tubes to extract the total protein. Each of the Eppendorf tubes was added with 200 μL of decontamination lysis buffer (containing 2 μL phenylmethylsulfonyl fluoride (PMSF) and 2 μL phosphatase inhibitor), homogenized in an automatic homogenizer (T8-1, Zhongda Instrument Co., Ltd., Jintan,Jiangsu, China), and then placed on ice for 30 min for complete lysis.The lysed samples were centrifuged at 13 700 × g at 4 °C for 5 min.The proteins in the supernatant were separated using 10% SDS-PAGE gels, and the proteins were transferred to a PVDF membrane that was blocked for 1 h with Tris-buffered saline Tween (TBST) containing 5% skimmed milk (blocking solution: 10 mmol/L Tris-HCl,pH 8.0; 150 mm NaCl; 0.1% Tween 20) at room temperature. The PVDF membrane was immersed in the primary antibody incubation solution overnight at 4 °C. The PVDF membrane was thoroughly washed with TBST 5–6 times (5 min/time) to remove the excess primary antibody and then immersed in a secondary antibody solution.After incubation at 37 °C for 2 h, the PVDF membrane was washed with TBST 5 times (5 min/time) and then visualized.

2.5.4 Measurement of bone indicators

The left femur of mice was separated, dried to constant weight in a drying oven at 105 °C, and the weight was recorded. The length and diameter of the femur were measured with a vernier calliper. Then,the femur was mashed and placed in a digestive tube, added with 10 mL nitric acid and 0.5 mL perchloric acid, and then digested on an electric furnace until it was clear and transparent. After adjusting the volume, the contents of calcium in the digest were determined using an atomic absorption spectrophotometer.

2.5.5 Micro-computed tomography (micro-CT)

The right femur of mice was separated and used to determine the bone mineral density (BMD) and observe the structure of trabecular bone using the micro-computed tomography (micro-CT) imaging system (Skyscan 1276, Belgium). 200 layers of the femur were scanned (9 μm pixel size) from the growth plate, and the images were reconstructed using a pre-installed three-dimensional (3D) analysis software in the micro-CT. The area was measured three times for each mouse, and each group’s area mean was calculated.

2.6 Statistical analysis

All data were analyzed using the SPSS 16.0 software. All the chemical analyses were repeated three times, and the data were expressed as the mean ± SD. The data analysis was performed using the one-way analysis of variance (ANOVA) followed by Duncan’s posthoc test to find significance among the treatments. Statistical significance was set at P < 0.05. All Columnar patterns in this study were generated using the Origin 8.0 software.

3. Results and discussion

3.1 Preparation of PPP

Huang et al. [12] reported that the HTMP pretreatment broke the peptide bonds of PV at random sites and facilitated the enzymatic hydrolysis (trypsin or thermolysin) of PV as well. Trypsin is a serine protease whose action site between lysine and arginine belongs to the endopeptidase. Thermolysin can hydrolyze the amino group of hydrophobic residues to form the peptide bond, which belongs to the outside peptidase. The DH of the HTMP pretreatment alone (HTMP-PV)and the HTMP-enzyme combination (HTMP-trypsin,HTMP-thermolysin, HTMP-trypsin+thermolysin) was 18.06%,21.95%, 26.69%, 45.55%, respectively (Fig. 1A). This result indicated that the treatment of the HTMP-PV with two enzymes successively(trypsin-thermolysin) was better than the single-enzyme treatments(trypsin, thermolysin) in producing small peptides. The calcium-binding rate of the PPP produced by HTMP-trypsin was 33.28%, and that of the HTMP-thermolysin was 39.03% (Fig. 1B). The calcium-binding rate of the PPP produced by HTMP-trypsin+thermolysin was 43.01%, which was significantly higher (10%, P < 0.05) than that of the HTMP-trypsin but was not different from that of the HTMP-thermolysin (Fig. 1B). Samaraweera et al. [18] and Huang et al. [12] showed that enzymes had powerful effects on the degree of hydrolysis of the PV and the HTMP-pretreated PV. The DH and calcium-binding rate suggested that the hydrolysis of HTMP-pretreated PV with trypsin followed by thermolysin would be the best enzyme treatment to produce PPP with a high calcium-binding rate (Fig. 1).

Fig. 1 Effects of enzyme hydrolysis on the (A) DH and (B) calcium-binding rate of PV hydrolysates. All data were expressed as mean values (mean ± SD,n = 3). Different letters indicate a significant difference (P < 0.05).

3.2 Preparation and characterization of PPP-calcium complexes

The mass ratio of peptides to calcium reflects the number of calcium ions bound to phosphopeptides. A high peptides/calcium mass ratio would result in the decreased utilization of the peptides, while a low mass ratio would result in poor stability of the peptide-calcium complexes [19]. The effect of peptides/calcium mass ratio on the amount of calcium-binding is shown in Fig. 2A. When the peptide/calcium mass ratio increased from 5:1 to 7:1, the calcium-binding rate increased from 73.30% to 98.44%, indicating that almost all the calcium added was used to make HTMP-Enz-PPP-Ca complexes. Although the calcium complexes remained stable as the PPP/Ca mass ratio continued to increase, the peptides/calcium ratio of 7:1 was considered as the best condition for the PPP-Ca complexes preparation, and the HTMP-Enz-PPP-Ca complexes prepared with the peptides/Ca ratio of 7:1 was used for the animal study.

The Zeta potential values of all samples were negative, indicating that the surface of the PV and the PPP were mainly negatively charged (Fig. 2B). The Zeta potential of PV was -42.47 mV, and the Zeta potential of PPP produced by the subsequent enzyme hydrolysis (HTMP-Enz-PPP) was -57.13 mV, which was much greater electronegativity than the PV. After binding with calcium,the electronegativity of the HTMP-Enz-PPP-Ca decreased to-24.2 mV, indicating Ca2+neutralized the negative charges on the surface of peptide molecules. The reduced surface potential of the molecules weakens the charge repulsion between molecules and forms a tight protein-calcium-protein bridge structure [20].

The FTIR peaks of PV showed no significant difference between the HTMP-Enz-PPP. However, after binding with Ca2+,the stretching vibration peak (N-H at 3 268.32 cm-1) had shifted to 3 294.84 cm-1, indicating that N-H was involved in the formation of HTMP-Enz-PPP-Ca (Fig. 2C). Hou et al. [21] reported that the absorption peak of -NH2moved from 3 296 cm-1to 3 405 cm-1when calcium was bound to the hydrolysate of shrimp proteins.The phosphate group (PO43-) has a characteristic bending vibrations peak at 580 cm-1[22]. For the PV and HTMP-Enz-PPP samples,the characteristic peaks at 980–920 cm-1area were ascribed to the symmetric stretching vibration of PO34-[2]. In HTMP-Enz-PPP-Ca,the absorption peak of HTMP-Enz-PPP moved from 973.41 cm-1to 990.77 cm-1, and the absorption intensity was enhanced. This result indicated that a large amount of calcium was bound to the phosphate groups of the HTMP-Enz-PPP.

The benzene ring of tryptophan and phenylalanine in protein or peptides can produce endogenous fluorescence, and thus the changes of fluorescence can reflect the structural change of protein/peptides [23].At the excitation wavelength of 295 nm, PV showed the maximum fluorescence emission at about 348 nm. The fluorescence intensity of the PV significantly decreased when the PV was pretreated with the HTMP + trypsin + thermolysin combination, indicating that HTMP + enzymatic hydrolysis changed the structure of PV. In HTMP-Enz-PPP-Ca, the binding of Ca2+to the phosphate groups of the HTMP-Enz-PPP weakened the endogenous fluorescence and caused a red-shift to 358 nm (Fig. 2D).

Fig. 2 Prepared and Structural characterization of peptide-calcium chelate.(A) Effects of mass ratio of peptide/calcium on calcium-binding capacity.(B) Zeta potential, (C) FTIR spectra of PPP and its peptide-calcium chelate in the regions from 4 000 cm?1 to 500 cm?1. (D) Fluorescence spectra of PPP and its peptide-calcium chelate over the emission wavelength ranging from 300 nm to 500 nm at an excitation wavelength of 280 nm. Different letters indicate a significant difference (P < 0.05).

3.3 The stability of PPP-Ca complexes

The tolerance of PV-Ca and PPP-Ca complexes to pepsin and trypsin was used to evaluate their digestive stability. The calcium retention rate of the PV-Ca was 35.48% after 30 min of incubation in the simulated gastric juice (pH 2.0, stomach conditions), which is relatively low, and did not change much during the 6 h of incubation time. The calcium retention rates of the HTMP-Enz-PPP-Ca (approximately 65%–72%) were much higher than that of the PV-Ca (35.5%), which was unexpected the calcium retention rates of the HTMP-Enz-PPP-Ca was 64.91% after 30 min of incubation and did not change during the 6 h incubation time (Fig. 3). This result indicated that the calcium complexes of HTMP-Enz-PPP had a much higher tolerance to the simulated gastric juice than the PV,and higher amounts of calcium were released from the PV than the HTMP-Enz-PPP in the stomach. This result was unexpected and different from what was currently known: PV has a strong calcium-binding capability, and the calcium bound to PV complex (PV-Ca) was insoluble under stomach conditions [13,24], and thus PV inhibits the use of calcium by animals [25].

No calcium was released from the PV-Ca and HTMP-Enz-PPP-Ca in simulated pancreatic juice (pH 7.5) during the 8 h of incubation (data not shown). Wu et al. [26] reported that hydrogen ions in acidic conditions competed with the calcium ions for the binding sites of phosphate groups while no net hydrogen ions are available to compete for the calcium-binding site under neutral-alkaline pH conditions.Lee et al. [27] reported that the normal range for transit time for food includes the following: gastric emptying (2–5 h), small bowel transit (2–6 h), colonic transit (10–59 h), and whole gut transit (10–73 h).This result indicated that the calcium release during the incubation time in the gastric juice (pH 2.0 for 6 h) and pancreatic juice (pH 7.5 for 6 h) represents the general passage time of foods from the stomach to the small intestines. The patterns of calcium release from PV-Ca and PPP-Ca indicated that the calcium bound to PV could be absorbed more in the duodenum area while the calcium bound to PPP is absorbed in the lower part of the small intestine (jejunum and ileum), probably through the paracellular transport mechanisms.Samaraweera et al. [18] suggested that if the size of PPP is large and has many phosphate groups in their structure, the calcium bound to the PPP would not be released quickly. The high degree of DH in the HTMP-Trypsin + Thermolysin-PPP means their molecular sizes are smaller than those of HTMP-trypsin-PPP and HTMP-thermolysin-PPP,and thus the calcium bound to HTMP-Enz-PPP (HTMP-Trypsin +Thermolysin-PPP) releases calcium more quickly than others.However, the calcium ions bound to the hydrolyzed PV(HTMP-Trypsin + Thermolysin-PPP) were more stable than those bound to the natural phosvitin. The results of simulated digestion experiments showed that PV-Ca was digested into peptide fragments after entering the stomach and intestine, thus removing most of the calcium from PV (Fig. 3). In the gastric acid environment, free calcium ions will form precipitation, which is very unfavorable to the absorption and deposition of calcium ions. When the polypeptide binds to calcium ions, it forms a stable structure, which helps calcium ions to reach the intestine and be absorbed by the small intestine.At the same time, the peptide calcium complex can slow down the release of calcium ions, which is also conducive to calcium absorption and transport.

Fig. 3 The in fluence of simulated gastrointestinal (stomach at pH 2.0)digestion on the stability of protein/peptide-calcium complexes. a,b Values within a time with different letters differ significantly (P < 0.05), n = 3.

3.4 Weight and serum indicators of mice

A low calcium mice model was established to study the effect of different types of calcium supplements on mice’s weight and serum indicators [17]. After 4 weeks of the modeling process, the average daily gain among the dietary treatments was not different (Fig. 4A).During the experiment period, no mouse died, and the weight of the mice increased steadily. This result was consistent with Wei et al. [17],who found no difference in average daily weight gain in mice when using calcium phosphate-inulin compound as a calcium supplement with probiotic effects.

Serum calcium, phosphorus, and AKP can be used as biochemical indicators related to bone growth and critical indicators to evaluate intestinal calcium absorption [28]. No significant difference in the serum calcium content between the normal control and the low-calcium control groups was found (Fig. 4B), probably because the stress caused by the extremely low calcium content would have stimulated the secretion of AKP to break down bone calcium and maintained stable calcium content in the blood. The dietary PPP-Ca effectively increased the serum calcium content to 2.19 mmol/L,higher than the normal calcium (1.79 mmol/L) and low calcium (1.72 mmol/L) diet groups. Calcium deficiency resulted in lower serum phosphorus content than the normal control group (P < 0.05)(Fig. 4C), and supplementing PV-Ca, HTMP-Enz-PPP-Ca, and CaCl2did not alleviate the deficiency of serum phosphorus.

Serum AKP is mainly derived from osteoblasts in the bone when the body’s calcium intake is seriously insufficient, which results in higher serum AKP than the normal healthy body [29]. The abnormally high AKP activity in serum can lead to bone calcium loss. The serum AKP content of the low calcium control group was 18.93 gold units/100 mL and was significantly (P < 0.05) higher than that of the normal calcium diet (10.89 gold units/100 mL) (Fig. 4D). These abnormal conditions are found when the body is in low calcium intake for a long time. When the low calcium control diet was switched to PV-Ca and HTMP-Enz-PPP-Ca diets, the serum AKP content of the mice in the switched groups significantly (P < 0.05) decreased. However,the decrease of serum AKP content in the HTMP-Enz-PPP-Ca group was the greatest (11.09 gold units/100 mL), and the AKP value reached a similar level to the normal control (Fig. 4D). In this experiment, chronic calcium deficiency due to the long-term low-calcium diet resulted in abnormal osteoblasts activity, and a large amount of AKP was released into the serum, as reported by Deng et al. [30].

3.5 Expression of duodenal calcium pathway proteins

The TRPV6 Ca2+pathway is the most crucial calcium transport channel in the duodenum, which is energy-dependent and is positively regulated by 1,25(OH)2D3genomic actions [31]. The expressions of TRPV6 proteins in mice were significantly increased from 0.53 ± 0.13 in the normal control group to 0.64 ± 0.12 in the low-calcium control group (Fig. 5B). The calcium homeostasis in the low-calcium control diet group was lost because of insufficient calcium intake, which led to the activation of the calcium-absorption-related pathway [32]. After 4 weeks of PV-Ca and HTMP-Enz-PPP-Ca gavage, the expression of TRPV6 decreased somewhat, but the difference was not significant compared with the low-calcium control group (P > 0.05) (Fig. 5B).This result indicated that the PV-Ca and PPP-Ca complexes could not directly regulate the expression of the TRPV6 proteins to improve the calcium balance in the mice. Benn et al. [33] showed that the TRPV6 and calbindin-D9kwere not essential for the Ca2+absorption in the duodenum via the active transport of the TRPV6 knockout mice and suggested that there would be other absorption pathways exist.

Fig. 4 (A) Weight change, (B) serum calcium content, (C) serum phosphorus content, and (D) serum AKP of mice after intragastric administration. The capital letters indicate comparison between normal- and low-calcium control groups (P < 0.05); the lowercase letters indicate the comparisons between low calcium control and other calcium supplement groups (P < 0.05); n = 6.

Similarly, the expression of another protein in the TRPV6 Ca2+pathway (duodenal calcium transporter calbindin-D9k) in the low calcium control group was 0.17 ± 0.01, which was significantly higher than that in the normal control group (0.09 ± 0.04). After switching the mice to diets containing PV-Ca and PPP-Ca complexes,however, the expressions of calbindin-D9K(0.12 ± 0.02) decreased to the normal level (Fig. 5C). These results indicated that PV-Ca and PPP-Ca complexes could regulate the expression of calbindin-D9K.Choi and Jeung [34] reported that calbindin-D9Kis the rate-limiting protein in the TRPV6 Ca2+pathway and played an important role in calcium absorption because the binding of calcium drove the calcium diffusion and vesicle transmission in the cytoplasm to calbindin-D9k.The calbindin-D9kbound calcium is transferred to PMAC1band then released into extracellular [35]. The PMAC1bprotein is also important in intestinal calcium absorption, but no significant difference in the expression of PMAC1bbetween all the calcium supplementing groups and the low-calcium control groups was found (Fig. 5D).Liu et al. [10] reported that the increased calcium uptake in the small intestine was accomplished by enhancing Ca2+transport in Caco-2 cells, and the transcellular pathway rather than the paracellular pathway was the main mechanism involved in intestinal absorption of Ca2+. However, our results indicated that the expressions of the transcellular calcium transport proteins (TRPV6, calbindin-D9K, and PMAC1b) in the CaCl2group was significantly higher (1.5–2 times)than other treatment groups (Figs. 5C, D), while serum calcium level was not affected (Fig. 3B). This indicated that more free calcium ions from CaCl2were absorbed in the duodenum and then transported to the blood, but a significant amount of them may have been excreted quickly by the kidney. The main parts of the digestive tract that absorb calcium are the duodenum and jejunum, but the ileum is also significantly involved in intestinal calcium absorption through the paracellular calcium pathway [36]. Karbach [37] reported that about 60%–70% of the mucosa-to-serosa calcium flux (absorbed) in the duodenum, jejunum, and ileum are paracellular probably because of the short passage time of foods through the duodenum, and only 30%–40% of the calcium transport is through the transcellular.Khanal et al. [7] also reported that paracellular transporters of the small intestine and the kidney are responsible for the bulk of Ca2+reabsorption. Therefore, the calcium absorption mechanisms from the inorganic (CaCl2) and the organic sources (PV-Ca and PPP-Ca complexes) could be different.

Fig. 5 (A) The Western blot of proteins associated with TRPV6 Ca2+ pathways in mice duodenum. (B) TRPV6; (C) calbindin-D9K; (D) PMAC1b. The capital letters indicate a comparison between normal- and low-calcium control groups (P < 0.05); the lowercase letters indicate the comparisons between low calcium control and other calcium supplement groups (P < 0.05). n = 3.

The calcium from the inorganic source (CaCl2) would be absorbed fast from the duodenum and the anterior portion of the small intestine (jejunum). However, the PPP-Ca complexes will be slowly absorbed throughout from the stomach to the large intestine. Because of the abrupt increase of calcium content in the blood, the calcium homeostasis mechanisms in the body could have been kicked in and removed excessive blood calcium through the kidney. Also, some calcium influx to the enterocytes should have been outflux into the jejunum and ileum and then excreted.

All CaCl2calcium (100%) and 65% of the PV-Ca calcium could be released in the stomach but primarily absorbed in the duodenum.Thus, the duodenum would be the main absorption area for calcium from the CaCl2and PV-Ca. About 28%–35% of the PPP-Ca calcium was absorbed in the duodenum, and the rest in the jejunum and ileum.The high stability of PPP-Ca suggested that most of the calcium ions could have been absorbed as a whole peptide-Ca complex in the jejunum and ileum areas, or calcium was slowly released from the PPP-Ca complex and absorbed throughout the small intestines,which increased its utilization or efficiency. Xu et al. [38] suggested that functional peptides such as phosphopeptides could be directly transported from the gastrointestinal tract to the circulation via 4 possible peptide transport pathways, which include 1) peptide transportermediated permeation, 2) paracellular transport through tight junctions,3) transcytosis, and 4) passive transcellular diffusion. Linnankoski et al. [39]reported that the average pore radius in the epithelial membrane of the Caco-2 was range in 5-10 ?, which is large enough for the peptides with up to 27 AAs to pass through the pores via the passive permeation and paracellular transport mechanisms, while Roberts et al. [40] reported that the peptides with 10–51 amino acids could be absorbed through the intestines and produced biologic effects at the tissue level. Also,some food components such as chitosan and its derivatives widened the gap in tight junctions of the intestinal epithelia and increased the permeability of peptides [41].

Also, the results suggested that the calcium release and the absorption mechanisms between CaCl2and PPP-Ca are different:the former mainly uses the active transcellular calcium transport system and rapid calcium absorption in the duodenum, while the latter relies more heavily on the paracellular transport system and slow calcium absorption throughout small intestines, which minimizes calcium excretion through the kidney, which partially explains why the efficiency of calcium use was much better with the PPP-Ca than the CaCl2.

3.6 The calcium deposition to and BMD of the femur

Balance of maintaining plasma calcium can lead to loss of bone calcium and eventually damage bone quality. There was no significant difference in femur length and diameter of mice between the low calcium control group and all other diet groups (Table 1). The main reasons would be that the mice used in this experiment were adults,and the femur length and diameter have already beenfixed and could not be changed by the dietary calcium supplementation at the late stage of the growth [42].

BMC is also an important indicator of bone quality. Insufficient calcium intake causes bone calcium loss, which easily leads to body fractures and other diseases. The amount of bone calcium deposited in the femur of the mice fed with the low-calcium control diet for 28 days was the lowest, and the HTMP-Enz-PPP-Ca and CaCl2groups(Table 1) were the highest. However, no significant differences were found between the low-calcium control and the normal calcium and PV-Ca groups. The results showed that calcium supplementation using dietary PV-Ca, HTMP-Enz-PPP-Ca, and CaCl2in the later stage of the growth was not significant for the length and diameter of the femur but alleviated the loss of bone calcium. The femur index (Table 1) indicated that HTMP-Enz-PPP-Ca was the best for the calcium absorption in the small intestine and the calcium deposition to bones among the tested. Chen et al. [43] reported similar results when they fed mice with calcium-peptide complex produced using tilapia scale protein hydrolysate (TSPH-Ca). Liao et al. [44]reported that feeding calcium-binding bone collagen peptide (BCP-Ca)improved the maximum load, elastic load, and bone mineral content in the femur of rats. Our results were consistent with the previous studies showing that dietary HTMP-Enz-PPP-Ca improved calcium absorption and deposition in the femur of the mice.

BMD is another critical indicator to characterize the trabecular bone structure. The BMD of mice with the low-calcium control diet was 0.17 g/cm3, which was significantly lower (P < 0.05)than that of the normal control diet (0.29 g/cm3) (Table 1). After 4 weeks of calcium supplementation, the BMD of PV-Ca and HTMP-Enz-PPP-Ca groups were 0.23 and 0.27 g/cm3, respectively.However, the BMD of the mice with the CaCl2diet was only 0.19 g/cm3, which is significantly lower than that of the normal control((0.29 ± 0.04) g/cm3) and HTMP-Enz-PPP-Ca ((0.27 ± 0.01) g/cm3)groups. The results in Table 1 (calcium deposition and BMD)indicated that the HTMP-Enz-PPP-Ca was better than inorganic calcium (CaCl2) in bone remineralization.

The micro-CT has been widely used to study the changes in the bone trabecular microstructure of the femur in experimental animals [45]. The micro-CT images of the distal femurs showed large cavities in the middle of the femur from the low-calcium control and CaCl2mice. This result indicated that the trabecular structure of the femur from those two diet groups was severely damaged (Fig. 6B).The whole femur (upper picture of Fig. 6) and the trabecular of the femur (bottom picture of Fig. 6) with the PV-Ca and HTMP-Enz-PPP-Ca diets was similar to that of the normal control diet and showed compact and complete network structures with thick bone trabeculae (Figs. 6A, C, and D). The results showed that the calcium bound to PV and HTMP-Enz-PPP (PV-Ca and HTMP-Enz-PPP-Ca)significantly improved bone microstructure and increased the integrity of trabecular bones. However, HTMP-Enz-PPP-Ca was the best among the three calcium supplementing agents when their effects on all other parameters, including the amount of serum calcium content,serum AKP activity, the expression of proteins associated with TRPV6 Ca2+pathways, the amount of calcium deposited in the femur,micro-CT, and the BMD index were considered (Figs. 4-6, Table 1).

The duodenum is located at the end of the stomach. Thus, most of the calcium released from the stomach would be absorbed by the active transporters in the duodenum. Calcium is absorbed across the intestinal through the brush border membrane of epithelial cells after binding to calbindin, a calcium-binding protein. The calcium-calbindin complex is transferred into the endoplasmic reticulum of the epithelial cells and then to the basal membrane and finally transported to the body by active transport. Under the acidic conditions of the stomach(pH 2.0 and pepsin), about 100%, 28%–35%, and 65% of the calcium bound to CaCl2, PV-Ca, and PPP (HTMP-Enz-PPP), respectively, were released within 30 min of incubation (Fig. 4C). In other words, about 1/3 of the PPP-Ca complexes (HTMP-Enz-PPP) Ca were released in the stomach and absorbed in the duodenum, and the rest of the tightly bound calcium to the PPP-Ca complexes should have been absorbed through the paracellular transport pathway when the PPP were passing through the jejunum and ileum part of the small intestine. However,the subsequent passage of the PPP-Ca complexes to the alkaline environment of the small intestine is not a favorable condition for the release of calcium from the PPP-Ca complexes (Fig. 4C).Thus, the release rate of Ca from the PPP-Ca complexes and the paracellular transport of PPP-Ca complexes in the jejunum and ileum would significantly impact the amount of calcium absorbed and utilized. For the PPP-Ca complexes to be used as efficient calcium supplementing agents, they should have relatively high calcium releasing characteristics under stomach conditions. Also, it is assumed that the production of smaller PPP with one or more phosphategroups but not too many phosphate groups in their structure would be ideal for the calcium supplementing agent for animals. In this study, HTMP-Enz-PPP-Ca has smaller peptide sizes than the PV-Ca,and the HTMP-Enz-PPP-Ca showed a better effect than PV-Ca and HTMP-PV-Ca on the re-calcification of the femur bones (Table 1)that lost their calcium due to the low calcium in the die. Initially, it was assumed that the duodenum and the active transport systems would play a significant role in calcium absorption, but their importance varied depending upon the calcium sources.

Fig. 6 Three-dimensional micro-CT images of the distal femur of each group. A, Normal control; B, low calcium control; C, PV-CaCl2;D, HTMP-Enz-PPP-CaCl2; E, CaCl2.

Table 1The length and diameter of the femur and the calcium and iron content of femur in mice.

4. Conclusion

The HTMP pretreatment combined with double enzyme hydrolysis was an excellent method for producing PPP with excellent calcium-binding capacity, and the PPP-Ca complexes were relatively stable under the stomach and very stable in the small intestine conditions. The calcium in the PPP-Ca complexes was absorbed in the gut and ameliorated the calcium deficiency caused by the low-calcium diet, effectively replenished calcium, and improved the BMD of the femur. The main mechanisms of the PPP-Ca complexes in improving the calcium absorption in the gut and their utilization of calcium in the bone restoration were 1) through the partial release of the calcium bound to the PPP in the stomach and the transcellular absorption of the released calcium in the duodenum, and 2) the absorption of PPP-Ca complexes in duodenum and jejunum, and 3) the release and absorption of calcium from the PPP-Ca complexes throughout the small intestine. The regulation and the expression of TRPV6 channel proteins in the duodenum would be more critical for the inorganic calcium than the PPP-Ca complexes. The PPP-Ca complexes prepared from PV using the HTMP-enzyme combinations could be excellent calcium supplementing agents for bone restoration, especially for those with osteoporotic conditions.

Declaration of competing interest

The authors declare that they have no known competingfinancial interests or personal relationships that could have appeared to in fluence the work reported in this paper.

Acknowledgments

This research was supported by the General Program of National Natural Science Foundation of China (32072237), the Hubei Provincial Natural Science Foundation of China (2020CFB583)and Fundamental Research Funds for the Central Universities(2662020SPPY006).