Synthesis and characterization of calcium phosphorylated inulin complex as a new source of enriched calcium supplement with prebiotic effect in food

Onulin is a non-branched polydisperse fructan that mainly consists of β(2→1) linked fructose units with a glucose unit at the reducing end (Morros et al., 2011; Lingyun et al., 2007). Ot belongs to a class of fibers known as non-digestible oligosaccharide that improve or maintain colon health as a prebiotic (Dewulf et al., 2013; Miremadi & Shah, 2012; Butts et al., 2016). Dver the last few decades, inulin has been modified by various methods to achieve functionalities suitable for various applications such as carboxymethylinulin, inulin-D-α-tocopherol succinate, p(inulin) microgel (Verraest et al., 1995; Mandracchia et al., 2014; Sahiner & Sagbas, 2014). Onulin was acylated with acetic anhydride, propionic anhydride, or butyric anhydride. Gallic acid (GA) was encapsulated with native (NOn), cross-linked (COn) and acetylated inulin (AOn). Gallic acid release profile data in yogurt for GA-NOn, GA-COn and GA-AOn were fitted to Peppas and Higuchi models (García et al., 2015). Acylation of inulin was beneficial for traveling through the upper gastrointestinal tract intact, delivering high concentrations of short chain fatty acid (SCFA) to more distal regions of the colon and remained bifidogenicity (Hartzell et al., 2013a). Onulin was esterified with 5-formylaminosalicylic acid (5-fASA). During in vitro digestion and fermentation of 5-fASA-inulin, conjugation of inulin with 5-fASA supported SCFA and especially butyrate-producing bacteria through fermentation in the distal colon (Hartzell et al., 2013b).


Introduction
Onulin is a non-branched polydisperse fructan that mainly consists of β(2→1) linked fructose units with a glucose unit at the reducing end (Morros et al., 2011;Lingyun et al., 2007).Ot belongs to a class of fibers known as non-digestible oligosaccharide that improve or maintain colon health as a prebiotic (Dewulf et al., 2013;Miremadi & Shah, 2012;Butts et al., 2016).Dver the last few decades, inulin has been modified by various methods to achieve functionalities suitable for various applications such as carboxymethylinulin, inulin-D-α-tocopherol succinate, p(inulin) microgel (Verraest et al., 1995;Mandracchia et al., 2014;Sahiner & Sagbas, 2014).Onulin was acylated with acetic anhydride, propionic anhydride, or butyric anhydride.Gallic acid (GA) was encapsulated with native (NOn), cross-linked (COn) and acetylated inulin (AOn).Gallic acid release profile data in yogurt for GA-NOn, GA-COn and GA-AOn were fitted to Peppas and Higuchi models (García et al., 2015).Acylation of inulin was beneficial for traveling through the upper gastrointestinal tract intact, delivering high concentrations of short chain fatty acid (SCFA) to more distal regions of the colon and remained bifidogenicity (Hartzell et al., 2013a).Onulin was esterified with 5-formylaminosalicylic acid (5-fASA).During in vitro digestion and fermentation of 5-fASA-inulin, conjugation of inulin with 5-fASA supported SCFA and especially butyrate-producing bacteria through fermentation in the distal colon (Hartzell et al., 2013b).Bioavailability of calcium in food is not high.Calcium ions can react with the organic acids such as oxalic acid and phytic acid to form insoluble salt (Weaver et al., 1987;Dendougui & Schwedt, 2004).Onulin was chemically modified with a charged group in order to chelate calcium.Onulin can promote the growth and lactic acid production of Bifidobacteria (Roberfroid, 2007).Lactic acid can facilitate the dissolution of calcium in colon (Coudray et al., 2005;Abrams et al., 2005;Griffin et al., 2003).A 12-month study with 100 adolescents ingesting 8 g/day shortand long-chain inulin fructans showed a significant increase in calcium absorption that led to greater bone mineral density (Slavin, 2013;Scholz-Ahrens et al., 2002).The aim of this work was the synthesis and the function of the derivatives based on inulin.The inulin derivatives were able to chelate calcium.Calcium ions were released out to be absorbed after the inulin derivatives were fermented by bacteria in colon (Jedrzejczak-Krzepkowska et al., 2011).Ot can effectively avoid that calcium ions directly reacted with the organic acids in food to form insoluble salt.The inulin derivatives will be a new source of enriched calcium supplement with prebiotic effect in food (Legette et al., 2012).

Materials
The purified inulin, Frutafit ® HD, with an average degree of polymerization of about 8, was supplied by Sensus Corporate (Roosendaal, the Netherlands).Sodium dihydrogen phosphate dihydrate and disodium hydrogen phosphate anhydrous were obtained from Sigma-Aldrich (Steinheim, Germany).All other chemicals used were of laboratory grade.

Synthesis of phosphorylated inulin (PInu)
Phosphorylation of inulin was performed by reacting inulin with a mixture of sodium dihydrogen phosphate dihydrate and disodium hydrogen phosphate anhydrous consisting of molar ratio (NaH 2 PD 4 •2H 2 D/Na 2 HPD 4 ) = 0.6393I:1 to keep the medium pH at 7. Different amounts of phosphate mixture were used in order to obtain phosphorylated inulin with different degrees of substitution of phosphate groups (DSp) (Sitohy et al., 2000).Prior to phosphorylation, inulin was dissolved in equal weight water at 90 °C.Accurate amounts of NaH 2 PD 4 •2H 2 D and Na 2 HPD 4 were calculated to prepare different molar ratios of phosphating agent to inulin, i.e., 0.3I:1, 0.4I:1, 0.5I:1, 0.6I:1, 0.7I:1, 0.8I:1, 0.9I:1, 1I:1, 1.1I:1.Phosphate mixture was added to the inulin solution and the resulting mixture was stirred for 2 hour at 80 °C.After cooling down to room temperature the reaction product was precipitated and stirred in methanol for 30 min to remove non-reacted phosphate and degradation products of inulin by vacuum filtration.The filtered product was dehydrated by washing with absolute ethanol, then lyophilized and grinded to obtain white powder.The reaction scheme was shown in Figure 1.

Colorimetric determination of the phosphorus content and calculation of the degree of substitution
The phosphorus content of phosphorylated inulin was determined using the vanadomolybdophosphoric acid method (Noguchi et al., 2017;Wongsagonsup et al., 2005).The degree of substitution of phosphate groups (DSp) was calculated according to the following Equation 1 (Passauer et al., 2010).With P being the colorimetric determined percentage of phosphorus content of the phosphorylated inulin, 162 the molar mass of the anhydrofructose unit, 3100 the atomic weight of phosphorus multiplied by 100 and 103 the molar mass of the phosphate substituent (-NaHPD 3 ).

Titration curve for phosphorylated inulin
20 mL of 1% phosphorylated inulin solution were pipetted into a 100 mL beaker.The pH probe was inserted.The magnetic stirrer was turned on and the stirring ratio was adjusted to a moderate speed without splashing.0.20 mL of 1 M NaDH were added at a time.The total volume of NaDH and pH value should be written down.Continue until the pH had reached approximately 12.The pH value versus mL NaDH was graphed to display the titration curve for phosphorylated inulin (Nishi et al., 1986).

Chelating ability of phosphorylated inulin for calcium
30 mL of the 0.04 M CaCl 2 solution were pipetted into a 250 mL Erlenmeyer flask. 2 drops of Calmagite indicator and 5 mL of pH 10 buffer were added.The initial color should be in wine red colour.The solution was titrated with the 0.04 M phosphorylated inulin with continuous swirling until the wine red colour turned into a blue endpoint.The titration was conducted in triplicate.The chelating ability of sample for calcium was calculated according to the relation (Equation 2)I: With V being the titration determined volume of the phosphorylated inulin, 162 the molar mass of the anhydrofructose unit, 40 the atomic weight of calcium and 103 the molar mass of the phosphate substituent (-NaHPD 3 ).Chelating ability of EDTA and citric acid was tested at the same time (Yu et al., 2006).

FTIR spectroscopy
The powder of inulin, phosphorylated inulin and CaPOnu were examined by Fourier transform infrared (FTOR) spectroscopy, on Nicolet Avatar 370 using a KBr pellet.

NMR spectroscopy
The 1 H NMR spectra and 13 C NMR spectra of phosphorylated inulin were recorded with a JEDL JMNGSX-400 MHz spectrometer in D 2 D. For those measurements, 10 mg of the sample was introduced into a NMR test tube, to which D 2 D was added, and finally the tube was kept at 70 °C to dissolve the sample in solution.

Animal experiment
Thirty male KM mice were obtained from the Center for Animal Experiment, Wuhan University at age of 3 weeks (initial body weight, 11.1-12.7 g), and were randomly assigned into one of 3 treatments (10 mice/treatment).The mice were allowed free access to pellet diets and distilled deionized water during the experiment.The diet composition was presented in Table 1.The level of Ca in control diet was 1.5 g/kg dry matter (DM) basis.The levels of Ca in calcium citrate diet and CaPOnu diet were 5 g/kg.The ambient temperature during the study was maintained between 21°C and 23°C, and the mice were exposed to a 12-h light, 12-h dark cycle.Animal studies were performed in compliance with relevant regulations relating to the guidelines of animal experiments and all procedures were approved by the local ethics committee of Wuhan University (No.I: 2015102702).The experiment lasted for 31 days in total.The mice were sacrificed after the dark period (between 08I:00 and 10I:00 AM) with 12 hours of fasting.The animals were weighed before and after the experiment and their food consumption was measured daily.Calcium and phosphate concentrations in the serum were determined (Bogden et al., 2008).The femurs and tibias were isolated, cleaned of soft tissues.The bones were then dried to constant weight and measured, weighed (Broulík & Broulíková, 2007).

Statistical analyses
The determinations of DSp of phosphorylated inulin, pH in titration, log CFU/g and pH values for Bifidobacterium animalis growth were done in three repetitions.The results were expressed as mean ± SD (standard deviation).Statistical analyses were performed with SPSS 16.0 software for windows (SPSS Onc., Chicago, OL, USA).A one-way ANDVA was used for comparisons of the single variables to identify significant differences.A Tukey's test was conducted when a significant difference was found (p < 0.05).

Degree of substitution of phosphorylated inulin
The colorimetric determined degree of substitution of phosphorylated inulin ranged between 0.047 and 0.215 (Figure 2), corresponding to different amounts of phosphate mixture which had been used for inulin phosphorylation.The DSp increased rapidly with the molar ratio (phosphate/inulin) extended from 0.4 to 0.8 in the chemical reaction process, but increased slightly then leveled off from 0.8 to 1.1.The results suggested that the molar ratio (phosphate/inulin) 0.9 in the chemical reaction process was high DSp (0.209) yield and feasible.Reports on the phosphorylation of inulin have been hardly seen so far.Starch was phosphorylated to different degrees of substitution using monosodium and disodium hydrogen orthophosphate at 160°C under vacuum.The DSp of phosphorylated starch could reach 0.33 (Sitohy et al., 2000).The reaction temperature at 160°C was so high for inulin because the phosphorylation of oligosaccharides had more side reactions than starch.This synthetic route of phosphrylated inulin had a mild reaction condition and was easy to control.

Titration curve
Titration curve for phosphorylated inulin (DSp = 0.209) was shown in Figure 3. Potentiometric titration was performed slowly with 3 h to obtain accurate titration curve.The titration curve (pH vs vol. of titrant) with 1 M NaDH was determined Ca-free Mineral mix (AON-93G) ensures the following mineral levels in the diets (per kg dry matter)I: P, 1.56 g; K, 3.6 g; S, 0.3 g; Mg, 0.5 g; Na, 1.0 g; Cl, 1.6 g; Cu, 6.0 mg; O, 0.2 mg; Fe, 37 mg; Mn, 10.5 mg; Se, 0.2 mg; Zn, 30 mg; CaPOnu = Calcium phosphorylated inulin.using a pH meter, with the inflection points being used as the equivalence points.The first equivalence point was approximately 3.85.The second equivalence point was approximately 8.76.Two equivalence points were clearly indicated in the titration curve for the phosphate group of phosphorylated inulin.
The phosphorylated inulin was found to give fairly broad titration curves and maintain a dissolved state over a wide pH range for its buffer capacity probably owing to its phosphate group.At the same time, the solubility of phosphorylated inulin in water could be increased by adding the base.The solubility of phosphorylated inulin in the acidic environment of the stomach can be decreased so that the speed of its hydrolysis reaction is slow.With pH increasing of the intestine, the solubility and digestibility of phosphorylated inulin will be accelerated.

Chelating ability of phosphorylated inulin for calcium
Onulin was chemically modified by esterification with phosphate mixture, which can conjugated with calcium.The chelating ability of phosphorylated inulin (DSp = 0.209) for calcium was 103.6 ± 1.1 mg/g.On this assay method, the chelating ability of EDTA for calcium was 269.2 ± 0.7 mg/g and citric acid was 90.3 ± 0.2 mg/g.The chelating ability of EDTA was significant stronger (p < 0.01) than that of phosphorylated inulin and citric acid.EDTA can be used to remove unwanted minerals and metals from the blood.The chelating ability of phosphorylated inulin for calcium was higher (p < 0.01) than that of citric acid.Phosphorylated inulin was stable in the stomach, small intestine and its phosphate groups could bind calcium strongly.

Characterization of phosphorylated inulin by FTIR
The purified samples of inulin, phosphorylated inulin and CaPOnu were characterized by the functional groups using FTOR technique.The OR spectra were presented in Figure 4.The inulin exhibited a very broad band in the region 3500-3200 cm −1 , centered at about 3350 cm −1 due to various D-H stretching in sugar rings (Abou-Arab et al., 2011).The absorption peak at 2933 cm -1 represented the C-H stretching in sugar rings (Grube, 2002).The weak peak at 1639 cm -1 was attributed to C=D stretching of fructose.The strong peak at 1032 cm -1 was attributed to C-D stretching of primary alcohol in sugar rings.Phosphorylation of inulin led to an emerging peak at 1662 cm -1 , which was attributed to the P=D asymmetric stretching from phosphates.An increase in the absorbance of the peak at 1074 cm −1 was also observed, and was assigned to the C-D-P stretching in phosphate esters, overlapping the C-D stretching vibrations in inulin ether groups, in accordance to the substitution of a hydroxyl group by a phosphate one (Jayakumar et al., 2008).Furthermore, the enlarged band between 3600 and 3100 cm -1 showed that the peaks of D-H stretching in sugar overlapped those in P-DH group.The absorptions due to P-D bond in phosphate group at 1000-1250 cm -1 changed remarkably by the binding of Ca 2+ ion (Nishi et al., 1987).The presence of sharp peak at 1236 cm -1 and the clear peak at 862 cm -1 were attributed to formed ionic bond (Ca-D) into this derivative (Shnawa, 2011).The FTOR spectra confirmed that the phosphorylation occurred in inulin and the phosphate group worked cooperatively for the calcium-binding.

1 H NMR and 13 C NMR spectra of phosphorylated inulin
The 1 H-NMR spectra of phosphorylated inulin were shown in Figure 5A.The 1 H of H 2 D had a high peak at 4.79 ppm indicating the powder of phosphorylated inulin were easy to absorb moisture.The resonance signals in the range between 4.26 and 3.60 ppm were caused by hydrogen atoms in the sugar rings of inulin (Wu & Lee, 2000).On contrast, the 1 H of P-DH appeared at a chemical shift of 1.15 ppm.The 1 H NMR spectra confirmed the successful incorporation of the phosphate group into the inulin.
The 13 C NMR spectra of phosphorylated inulin was given in Figure 5B.The spectra-showed that all the signals from each carbon were well separated from each other, and the carbons attached to the substituted hydroxyl group were clearly distinguished from the non-substituted one (Jayakumar et al., 2009).The C-2 resonance (δ = 103.16ppm) was attributed anomeric carbons and indicated the presence of β-fructofuranose (van Hijum et al., 2001).The peaks at 57.46, 62.16, 69.26, 74.34 and 81.10 ppm  were attributed to C-1, C-6, C-4, C-3 and C-5 respectively, of the furanose ring (Anwar et al., 2008;Shiomi, 1993).The spectra of phosphorylated inulin was complex because of the substituents at various positions on the fructose units.There was no free hydroxyl group to react with phosphate group at the C-1, C-2 and C-5 position in furanose ring of inulin.Chemical shift for C-6 moved from 62.16 ppm to 60.95 ppm; for C-3 from 74.34 ppm to 77.05 or 72.48 ppm; and for C-4 from 69.26 ppm to 71.20 ppm by substitution with phosphate group (Jayakumar et al., 2009;van Hijum et al., 2001).A clear contrast with the phosphorylation reactions of inulin proved the hydroxyl group at the C-6 and C-3 position was selectively substituted prior to the C-4 position.C 6 -DH in the chemical structure of phosphorylated inulin was located in the outside of the sugar ring and its steric hindrance was relatively small and esterification was performed with phosphate easily.The weak peak at 103.72 ppm compared to 103.16 ppm was related to C 2 -DH by substitution in the terminal glucose ring of inulin.These results indicated that the phosphorylation reaction was occurred into inulin.

Comparison of prebiotic effects of various carbohydrates
The growth of intestinal bacteria varied considerably depending on the carbohydrate substrates (Krupa-Kozak et al., 2016;Sims et al., 2014).A decrease in pH of the culture media, as a result of short chain fatty acid production and an increase in log CFU/g of the plates reflected the ability of the bacteria to grow on the substrates (Bedani et al., 2013;Salazar et al., 2008) (Figure 6).The control culture without the source of carbon showed concentrations of Bifidobacterium animalis at 48 h of incubation was 7.83 log CFU/g.After 48 h incubation, all samples promoted higher increases in levels of Bifidobacteria populations than those occuring in control cultures without carbohydrates added, which was indicative of a stimulatory effect of these substrates on Bifidobacteria.The highest Bifidobacteria concentration of 8.92 log CFU/g was due to the glucose being used as carbon source since glucose was a reducing sugar providing antioxidant activity in anaerobic systems.Bifidobacteria concentration of 8.83 log CFU/g and pH4.38 of fermentation liquor were related to inulin as carbon source.CMC (carboxymethylcellulose) could be fermented by Bifidobacterium animalis because the carboxymethyl groups render the CMC (DS = 0.71) soluble and biodegradable.The pH value of fermentation liquor with 10 g/L POnu (Phosphorylated inulin, DSp = 0.209) as carbon source was 3.92.POnu carried many negative charges so that bacteria were unable to contact and utilize POnu easily.Negative charges in POnu could be neutralized by calcium.CaPOnu could stimulate the growth of Bifidobacteria and decrase pH in vitro fermentation similar to inulin.

Animal experiment
Onitial and final body weight of mice were not affected by CaPOnu diet or calcium citrate diet (Table 2).As a result, average daily gain was not different among the treatments.Compared to the control fed the low calcium diet, the serum calcium concentration in mice fed calcium citrate diet was significantly higher (p < 0.05), reaching 99.14 mg/L whereas the serum phosphate was not higher.CaPOnu diet increased the calcium concentration (p < 0.01) and phosphorus concentration (p < 0.05) in the serum compared with the control diet.Serum calcium x phosphate has been regarded as a risk factor for extraskeletal calcification with the general consensus that it should not exceed 7000 mg 2 /L 2 (Hawley, 2006).A high phosphate (8 g/kg) diet was noted to accelerate calcification of the heart, kidney and tongue in healthy DBA/2 mice (van den Broek & Beynen 1998;Lau et al., 2013).On the present study, the level of phosphate in the CaPOnu diet (4.3 g/kg) was a safe amount and serum calcium x phosphate in mice fed CaPOnu diet was a relatively modest value (4788.4mg 2 /L 2 ) within the appropriate target range.The content of serum calcium in the mice fed the CaPOnu diet had no difference compared with the calcium citrate diet.There were no differences in femur length and tibia length among the treatments.Moreover, the femur weight increased (p < 0.05) in the groups fed the calcium citrate diet or the CaPOnu diet compared to the group fed the control diet.The tibia weight tended to be increased in the groups fed the calcium citrate diet (p = 0.07) or the CaPOnu diet (p = 0.08) compared to the group fed the control diet.There were no differences in femur weight, femur length, tibia weight, tibia length between the mice with calcium citrate diet and CaPOnu diet.Of there is not enough calcium in the blood, then the body will take calcium from bones, thereby weakening bones.The stimulating effect of CaPOnu on the absorption and balance of calcium for growing mice was observed.CaPOnu is potential substances that could help to improve the supply with available calcium in human nutrition and by this contribute to bone health.Calcium citrate is a kind of nutrient supplement in food industry.Ot is speculated that CaPOnu has a good effect of calcium supplementation as calcium citrate.Ot can increase the mineral absorption and will be used in the treatment of calcium deficiency.

Conclusions
This study proposes to use a phosphorylated derivative of inulin as a potential functional food that contributes to enhance the bioavailability of calcium in the colon, propitiated by lactic acid production and after fermentation by Bifidobacteria.This work opens new perspectives by taking advantage of the introduction of phosphonic groups into inulin, providing the chelating ability of calcium ions for phosphorylated inulin.CaPOnu had positive effects on intestinal microflora.Calcium plays a very important role in the body.Compared to controls fed the low Calcium diet, mice fed the CaPOnu diet had a higher serum calcium (p < 0.01) and femur weight (p < 0.05).Study findings will aid future investigations in ascertaining the factors related to potential bone health benefits of CaPOnu which will aid in developing an effective prebiotics food supplement.Additional work is needed to examine potentia of the ability of CaPOnu to increase bone formation rates.

Figure 1 .
Figure 1.Reaction scheme for the preparation of phosphorylated inulin.

Figure 2 .
Figure 2. Effect of molar ratio (phosphate/inulin) on the DSp of phosphorylated inulin.Each value was expressed as mean ± SD (n = 3).

Table 1 .
Diet composition (g/kg DM) during the experiment.