Effect of Chilled Storage on Antioxidant Capacities and Volatile Flavors of Synbiotic Yogurt Made with Probiotic Yeast Saccharomyces boulardii CNCM I-745 in Combination with Inulin

Fermentation of available sugars in milk by yogurt starter culture initially and later by Saccharomyces boulardii (Probiotic yeast) improves the bioavailability of nutrients and produces bioactive substances and volatile compounds that enhance consumer acceptability. The combination of S. boulardii, a unique species of probiotic yeast, and inulin, an exopolysaccharide used as a prebiotic, showed remarkable probiotic and hydrocolloid properties in dairy products. The present study was designed to study the effect of fermentation and storage on antioxidant and volatile capacities of probiotic and synbiotic yogurt by incorporation of S. boulardii and inulin at 1%, 1.5%, and 2% (w/v), compared with the probiotic and control plain yogurt. All samples were stored at 4 °C, and during these four weeks, they were analyzed in terms of their antioxidant and volatile compounds. The synbiotic yogurt samples having inulin and S. boulardii displayed significantly higher DPPH (2,2-diphenyl-1-picrylhydrazyl) free radical activity values and more values of TPC (total phenol contents) than control plain yogurt. A total of 16 volatile compounds were identified in S5-syn2 and S4-syn1.5, while S3-syn1 and S2-P had 14, compared with the control S1-C plain yogurt samples, which had only 6. The number of volatile compounds increased with the increasing concentration of inulin throughout the storage period. Therefore, this novel synbiotic yogurt with higher antioxidant and volatile compounds, even with chilling storage conditions, will be a good choice for consumer acceptability.


Introduction
Functional foods contain one or more specific compounds that have a functional effect on improving the health and well-being of the consumer. These beneficial components can be naturally increased in the food or intentionally added during the production process to Inulin, one of the most common prebiotics, is mainly found in roots of chicory (Cichorium intybus), garlic (Allium sativum), wheat (Triticum spp.), oat (Avea sativa), and Dalia (Bulgar) [27]. It is known to be a storage polymer consisting of a β-2-1-linked fructosyl unit with a terminal glucosyl unit [28]. At present, there is an increasing interest in the addition of inulin and other oligofructoses to food products (e.g., yogurt) for their healthful effects, e.g., enhancing Lactobacillus and Bifidobacterium growth in the colon, boosting the bioavailability of a variety of minerals such as calcium and iron, increasing antioxidant activities, and boosting immune functions [29]. Inulin and oligofructose improved sensory quality and increased the viable probiotic count in functional dairy foods [30]. Supplementation of food fibers such as inulin could reduce wheying-off, thus improving the textural properties of the food matrix, and it was also found to remarkably elevate viscosity and shear thinning behavior of different dairy products [31].
Previously, different yogurt samples containing this probiotic yeast with different inulin concentrations showed the best physicochemical, microbiological, sensory properties, microrheology, and microstructure [32]. Owing to the interest of consumers and increasing demands of the food industry, in the present study, synbiotic yogurt having probiotic S. boulardii and prebiotic inulin in different concentrations were evaluated in terms of their antioxidants and volatile compounds during 28 days of storage.

Microorganisms and Culture Condition
Probiotic yeast (S. boulardii) was purchased as a lyophilized powder in the form of a sachet (Martin Dow). The yeast culture (8.22 ± 0.28 CFU/mL) of S. boulardii was prepared according to the method described by Eunice et al. (2017) and used in yogurt making. The number of colony-forming units of S. boulardii (CFU/g) was determined using Sabouraud dextrose agar in different dilutions made by dissolving a 250 mg sachet in 9 mL peptone water [33]. The yoghurt starter culture containing L. delbrueckii ssp. bulgaricus and Streptococcus thermophilus was purchased in powder form (DANISCO, France) and activated by consecutively transferring it three times to 10% (w/v) reconstituted skim milk at 37 • C for 24 h. Inulin was purchased in powder form (Digestive-Now).

Preparation of Yogurt
High-quality fresh cow milk was purchased from a local dairy (Beijing SanYuan-Dairy Co., Ltd., Beijing, China). Five experimental groups of yogurt were arranged (Table 1)namely, control plain yogurt (S1-C), probiotic yogurt with 0.5% S. boulardii (S2-P), synbiotic yogurt with 0.5% S. boulardii + 1% inulin (S3-Syn1), 0.5% S. boulardii + 1.5% inulin (S4-Syn1.5), and 0.5% S. boulardii + 2% inulin (S5-Syn2)-which were made according to the method described by [32,33], with some modifications. After blending the fresh milk with 5% skim milk powder, the mixture was homogenized (performed with a homogenizer), pasteurized (85 • C, 30 min), and cooled to 43 • C. The mixture was inoculated with 3% (w/v) of the yogurt starter culture and mixed well. Then, the S. boulardii culture and inulin were added, as described in Table 1. The yogurt samples were incubated at 43 • C ± 2 • C until reaching approximately pH 4.5, and then they were stored at 4 • C for four weeks. Sampling was performed every week during the storage for subsequent analyses. Table 1. Experimental groups of yogurt samples.
The antioxidant activity of the yogurt samples was determined using DPPH free radical scavenging assay according to [34], with some modifications. Briefly, a 1/10 (w/v) dilution of the yogurt sample in water was prepared. Then, an aliquot of 0.1 mL of each sample was added to 4 mL of DPPH solution with a concentration of 0.004% in methanol. The mixture was shaken thoroughly and allowed to stand at room temperature for 30 min. The OD of the samples was read at 517 nm against water as blank. The inhibition percent (I%) of the DPPH free radical was calculated. The readings were recorded in triplicates.
The percentage of radical scavenging activity (RSA) was calculated as follows:

Total Phenol Content Determination
The total phenolic content of yogurt samples was measured using the Folin-Ciocalteu method. Briefly, 5 g of yogurt samples were mixed with 15 mL distilled water in centrifuge tubes and centrifuged at 4000 g for 10 min. A 0.1 mL aliquot of supernatant of each sample was mixed with 4 mL of 50% (v/v) Folin-Ciocalteu reagent, and 2 mL of 2% sodium carbonate solution was added to the mixture and allowed to stand for 2 h. The OD of the samples was read at 750 nm against a solution consisting of 4 mL Folin-Ciocalteu and 2 mL sodium carbonate as blank. Gallic acid was used as a standard to prepare the calibration curve. The total phenolic content of the yogurt samples is expressed as mg gallic acid/kg yogurt [35].

Extraction of Volatile Flavor Compounds
The volatile compounds in yogurt samples were analyzed using headspace solidphase microextraction-gas chromatography-mass spectrometry (HS-SPME-GC-MS) on days 1, 7, 14, 21, and 28 of chilled storage. Briefly, a 5 g sample was mixed with 2 µL of 1,2-dichlorobenzene as internal standard (I.S, 1500 ppm) in a 20 mL headspace vial, which was then firmly closed with a silicon septum. The vial was placed in a water bath for 60 min to equilibrate at 55 • C, after which the volatile compounds were extracted using a fiber manufactured from StableFlex divinylbenzene carboxen polydimethylsiloxane (DVB-CAR-PDMS) in 50/30 µm (Supelco, Bellefonte, PA, USA) at the same temperature. The absorbed volatile compounds were then desorbed in the injection port of a GC-MS for 5 min in a splitless mode at 250 • C.

GC-MS Analysis
A polar column (DB-WAX, 30 m 0.25 mm 0.25 m, Agilent Technology, Santa Clara, CA, USA) was used in a PerkinElmer GC-MS (GC, Clarus 680; MS, Model SQ8C) using helium as the carrier gas at a follow rate of 1 mL/min. The initial oven temperature was held at 35 • C for 5 min, then ramped up to 140 • C at 4 • C/min and kept for 5 min, before being elevated to 230 • C at 10 • C/min and kept for 5 min. An MS detector was used with a mass scan range of 33-450 amu (m/z) and a 70 eV electron ionization (EI) voltage. Both the ion source and line transfer were adjusted at 250 • C.

Identification and Quantitation of Volatile Flavor Compounds
The retention indices (RIs) and mass spectra (MS) of volatiles in the DB-WAX column were compared with those published in the National Institute of Standards and Technology library (NIST 14).
The internal standard method was used for quantitation analysis, in which the peak area of the detected volatiles was compared with the internal standard's matching peak area. The following equation was used to calculate the concentration of each compound: Con. (mg/100 g) = Peak area ratio volatile internal standard × con. of internal standard Sample weight × 100

Statistical Analysis
All measurements were performed in triplicate. The results were statistically analyzed with one-way ANOVA using Statistix 8.1 software. The volatile compounds data were also subjected to multivariate statistical techniques in terms of principal component analysis (PCA) and hierarchical cluster analysis (HCA).

Antioxidant Activities of Yogurt Samples
The antioxidant activities (DPPH and TPC) were determined in the analyzed yogurt samples throughout the storage period.

Radical Scavenging Assay during Storage
The DPPH radical is one of the rare stable organic nitrogen radicals and can take up an electron or hydrogen radical to become a stable diamagnetic molecule. The stable DPPH radical scavenging model is a broadly used method to evaluate antioxidant activity in a short period of time, compared with other methods. The effect of antioxidants on DPPH radical scavenging is related to their ability to donate hydrogen [36]. The DPPH activities of probiotic and synbiotic yogurt samples were lowest after day 1 storage, when values, including control (S1-C), were below 50%. However, after 14 days of storage, the activities of probiotic (S2-P = 54.65%) and synbiotic (S3-Syn1 = 61.01%, S4-Syn1.5 = 62.73%, and S5-Syn2 = 65.51%) yogurt samples were significantly increased, compared with that of the control sample (S1-C = 26.44%), as shown in Table 2. Increased antioxidant activity in probiotic and synbiotic yogurt may result from bioactive (antioxidative) peptides released from protein digestion via bacterial fermentation. Similarly, the pH of yogurt, because of the fermentation of available lactose by yogurt culture, becomes acidic, which creates a favorable environment for S. boulardii to utilize the remaining sugars available and produce more bioactive compounds. All the yogurt samples displayed radical scavenging capacities in descending order, i.e., S5-Syn2 < S4-Syn1.5 < S3-Syn1 < S2-P < S1-C. The addition of inulin in synbiotic yogurt also contributed to the higher antioxidant activities. In previous studies, Valentina et al. [37] reported that inulin had significantly higher antioxidant activity than other sugars, even after cooking and digestion. In another study, Jansen et al. [38] observed that probiotics in yogurt drinks stored at refrigerated temperature had higher antioxidant activity than those in yogurt drinks stored at room temperature. DPPH activity was increased in yogurt with the addition of an aqueous extract of Matricaria recutita [39]. Hydrolysis of milk protein or organic acid production can be another reason for yogurt's antioxidant activity because of microbial metabolic activity during fermentation and storage under refrigeration [40].

Total Phenol Content of Yogurt Samples during Storage
The total phenol content in all yogurt samples was significantly (p ≤ 0.05) different during the refrigerated storage period. All synbiotic samples showed higher phenol content with respect to the concentration of inulin used, as presented in Table 3. Among the samples, S5-Syn2 showed the best results, at 4.02 mg.GAE/g, while S1-C had the lowest, at 0.57 mg.GAE/g. A higher concentration of TPC up to 6.30 mg GAE/g on the 28th day of storage was reported by Adriana et al. (2018) [41] in yogurt supplemented with Marjoram (Origanum vulgare) extract. All synbiotic yogurt samples (S3-Syn1 = 3.54 mg.GAE/g, S4-Syn1.5 = 3.75 mg.GAE/g, and S5-Syn2 = 4.02 mg.GAE/g) supplemented with 1%, 1.5%, and 2% inulin were also rich in total phenol content at the end of cold storage (28th day), compared with probiotic (S2-P = 1.52 mg.GAE/g) and control plain yogurt (S1-C = 0.78 mg.GAE/g). The higher TPC in synbiotic yogurt is due to inulin, a fructan polysaccharide found in many different types of plants, such as chicory, artichoke, salsify, and Jerusalem artichoke. Senadeera et al. (2018) [42] observed the highest phenolic content of 15.53 ± 0.46 mg GAE/100 g in yogurt fortified with Annona species pulp. In another study, Chouchouli et al. [43] observed a higher TPC value for yogurt containing grape seed extracts. Moreover, the addition of strawberry pulp enhanced the phenolic content and antioxidant properties of yogurt [44]. Furthermore, green tea supplementation increased the TPC value of probiotic yogurt during refrigerated storage [45]. In a recent study, a higher TPC value was observed in yogurt supplemented with osmo-air-dried mulberry, compared to the control [46].

Volatile Compounds of Synbiotic Yogurt
The volatile compounds of synbiotic, probiotic, and control yogurt were determined using GC-MS during cold storage for up to four weeks. Flavor in yogurt is formed by the action of yogurt starter bacteria and originated from biochemical changes in carbohydrates, lipids, and proteins. The flavor is frequently the first indicator when consumers choose food. Consumers will not be interested in functional food consumption if biologically active ingredients lead to unpleasant flavor [47]. The volatile analysis is widely applied in the objective assessment of dairy foods, to determine consumers' acceptance of new functional products [48].
Storage conditions greatly influenced the formation of volatile compounds in the synbiotic yogurt made with probiotic S. boulardii and inulin compared with the control plain yogurt and the probiotic yogurt sample using yeast without inulin, as shown in Table 4. Analysis with HS-SPME-GC-MS showed that a total of 14 volatile compounds were identified in the probiotic yogurt (S2-P) in week 0, 13 in weeks 1 and 2, 10 in week 3, and 8 in week 4, compared with control (S1-C), having only 5 in week 0 and 6 in week 1, which was reduced to 4 in the refrigerated reaming storage. However, synbiotic yogurt samples showed a higher concentration of volatile compounds throughout the storage period. A total of 14 volatiles were identified in S3-Syn1 in week 0, 14 in week 1, 13 in week 2, 12 in week 3, and 11 in week 4. Similarly, in synbiotic samples (S4-Syn1.5, S5-Syn2), 16 volatiles were formed that were reduced to 13 and 12, respectively, in week 4 of cold storage.   These volatile compounds were from different chemical families, but limonene and L-limonene (Terpene family) were present in probiotic and synbiotic yogurt samples having S. boulardii in common. Among these samples, S3-Syn1 had the highest limonene concentration (162.28 ± 7.30 mg/100 g), while S5-Syn2 had more L-limonene concentration (102.29 ± 5.66 mg/100 g) in weeks 4 and 2, respectively. Limonene is generally considered safe as a food additive or flavoring and a fragrance additive. Previous studies in the literature have revealed that limonene can present antimicrobial, antifungal, antimalarial, and antitumoral activities [49].
Among alcohols, ethanol was the predominant compound present in all yogurt samples, but the concentration was lower in S1-C, ranging from 0.83 ± 0.15 at week 1 to 1.72 ± 0.66 mg/100 g at week 4, compared with probiotic S2-P, ranging from 28.13 ± 2.76 at week 1 to 59.89 ± 4.33 mg/100 g at week 4, and much higher concentration in synbiotic yogurt samples S3-Syn1, ranging from 93.32 ± 7.34 at week 1 to 306.81 ± 22.08 mg/100 g at week 4 of the refrigerated storage period. These results indicated that it is a low alcohol product, compared with kefir fermented with yeast and lactic acid bacteria. In terms of the final fermentation products, kefir lactic acid and ethanol were the main ones, whereas only traces of acetic acid were detected. The concentration of ethanol was~0.9% w/v (~9 g/L), as reported by Tzavaras et al. [50]. Similarly, Chen et al. [51] also reported that ethanol was one of the main components in volatiles, in their study of the effect of lactic acid bacteria and yeasts on the structure and fermentation properties of Tibetan kefir grains. The other alcohol (phenylethyl alcohol) was only found in the synbiotic yogurt samples (S4-Syn1.5 = 13.09 ± 0.98 and S5-Syn2 = 16.51 ± 2.11 mg/100 g at week 3), having a higher concentration of inulin. Yeast cells are capable of producing phenyl ethyl alcohols via normal metabolism. This can be synthesized through two independent routes in yeast cells, either de novo via the Shikimate pathway or via the Ehrlich pathway [52]. Alcoholic compounds also contribute to flavor improvement in fermented dairy products.
A total of four acids (acetic acid, butanoic acid, hexanoic acid, and octanoic acid) were identified in the yogurt samples having higher concentrations in the descending order of acetic acid ≤ hexanoic acid ≤ octanoic acid ≤ butanoic acid found in synbiotic samples, followed by probiotic and control. Acetic acid was one of the main acids in yogurt samples, and significant differences were observed among synbiotic, probiotic, and control samples in terms of its concentration. Acidity is a decisive factor for flavor acceptance of yogurt and is maintained near pH 4.5 [53]. Acid compounds are generally present in various fermented dairy products [54].
Two short-chain fatty acid compounds (propanoic acid and 2-methyl-butanoic acid) were present in probiotic and synbiotic samples having S. boulardii in common. However, they were not found in control plain yogurt. This might be credited to the ability of S. boulardii CNCM I-745 to secrete alkaline phosphatase, which causes dephosphorylation of phospholipid substrates to generate phosphate and other short-chain fatty acids [55,56].
Three esters (1-butanol-3-methyl-acetate, butanoic acid-3-methyl butyl ester, and octanoic acid ethyl ester) were identified in only probiotic and synbiotic yogurt samples. 1-Butanol-3-methyl-acetate ester was detected in all of the probiotic and synbiotic yogurt samples of this study up to four weeks of storage, but butanoic acid-3-methyl butyl ester was detected only up to the second week, while octanoic acid ethyl ester was only found in S4-Syn1.5 and S5-Syn2 having higher concentrations of inulin. Ester volatile compounds are generally produced at low concentrations in dairy products when lactose is fermented with lactic acid bacteria (LAB) [57].
Aldehydes are considered important aroma compounds contributing to the volatile profile of fermented dairy products with lactic acid bacteria [58]. The only aldehyde (3-hydroxy butanal) found was in probiotic yogurt (S2-P = 0.63 ± 0.11 mg/100 g) and synbiotic yogurt (S3-Syn1 = 13.95 ± 3.31 mg/100 g) samples of this study up to one week of storage. The formation of more volatile compounds in the synbiotic yogurt samples enriched its flavor, compared with the plain and probiotic yeast yogurts that contained fewer volatile compounds. Dan et al. (2017) [59] also showed that the aroma profile of yogurt made with pure culture was different from those made with the addition of probiotics alone or with a combination of prebiotics.

Principal Component Analysis of Volatile Compounds
Volatile compounds identified via GC-MS among the different probiotic, synbiotic, and control yogurt samples (S1-C: control plain yogurt; S2-P: yogurt with S. boulardii; S3-Syn1: yogurt with S. boulardii + 1% inulin; S4-Syn1.5: yogurt with S. boulardii + 1.5% inulin; and S5-Syn2: yogurt with S. boulardii + 2% inulin) during four weeks of storage were further analyzed via principal component analysis (PCA) and hierarchical cluster analysis (HCA). Both these multivariate statistical techniques were used for easy interpretation of the volatile compounds' data from GC-MS. Hierarchical cluster analysis is shown in Figure 1 hydroxy butanal) found was in probiotic yogurt (S2-P = 0.63 ± 0.11 mg/100 g) and synbiotic yogurt (S3-Syn1 = 13.95 ± 3.31 mg/100 g) samples of this study up to one week of storage. The formation of more volatile compounds in the synbiotic yogurt samples enriched its flavor, compared with the plain and probiotic yeast yogurts that contained fewer volatile compounds. Dan et al. (2017) [59] also showed that the aroma profile of yogurt made with pure culture was different from those made with the addition of probiotics alone or with a combination of prebiotics.

Principal Component Analysis of Volatile Compounds
Volatile compounds identified via GC-MS among the different probiotic, synbiotic, and control yogurt samples (S1-C: control plain yogurt; S2-P: yogurt with S. boulardii; S3-Syn1: yogurt with S. boulardii + 1% inulin; S4-Syn1.5: yogurt with S. boulardii + 1.5% inulin; and S5-Syn2: yogurt with S. boulardii + 2% inulin) during four weeks of storage were further analyzed via principal component analysis (PCA) and hierarchical cluster analysis (HCA). Both these multivariate statistical techniques were used for easy interpretation of the volatile compounds' data from GC-MS. Hierarchical cluster analysis is shown in Figure 1, in which the yogurt samples are divided into three groups, determined by the average linkage among different groups based on Euclidian distance. The principal component analysis explained a total of 96.7% variance, to which principal component 1 (PC1) contributed 90.3%, whereas principal component 2 (PC2) contributed 6.4%, respectively.  Two principal components were used in most cases of PCA analysis, appropriately explaining a great proportion of variations in original parameters. A bi-plot of different yogurt sample scores is shown in Figure 2, revealing the most important loadings and the percentage accounted for by the two first principal components (PC1 and PC2) after analysis. There were three different groups, and the samples were positioned based on the concentrations of volatiles and storage weeks. All yogurt samples were positioned on the left side of PC1, but group1 (S1-C and S2-P) were in the negative part of the biplot in week 0. The control samples remained in the negative zone throughout the four weeks of storage, having very few volatiles produced. Group 2 samples having probiotic and synbiotic yogurt samples were positioned on the positive side of PC1, especially after 1 week of storage. Samples in group 3 (synbiotic only having both S. boulardii and inulin) were positioned on the positive side of PC1 and located together in the bi-plot. Significant differences were observed among old yogurts of these groups at weeks 0 and 4. In summary, the bi-plot indicates that limonene, L-limonene, and carboxylic acids (octanoic acids, hexanoic acid, butanoic acid, and acetic acid), along with ethanol, were grouped on the positive side of PC1, strongly contributing to the flavor of probiotic and synbiotic yogurt samples. Similarly, esters (1-butanol-3-methyl-acetate, butanoic acid-3-methyl butyl ester, octanoic acid ethyl ester) were secondary in contributing to the flavor of probiotic and synbiotic yogurt samples and were also placed on the positive side of PC1. The remaining volatile compounds (propanoic acid and 2-methyl-butanoic acid, 3-hydroxy butanal) were grouped in the center of the bi-plot, with less contribution to the principal component, compared with the other compounds. According to these results, PCA, along with HCA, is a powerful tool to distinguish the samples based on treatment (probiotics, synbiotics) and magnitude of storage time.
Two principal components were used in most cases of PCA analysis, appropriately explaining a great proportion of variations in original parameters. A bi-plot of different yogurt sample scores is shown in Figure 2, revealing the most important loadings and the percentage accounted for by the two first principal components (PC1 and PC2) after analysis. There were three different groups, and the samples were positioned based on the concentrations of volatiles and storage weeks. All yogurt samples were positioned on the left side of PC1, but group1 (S1-C and S2-P) were in the negative part of the bi-plot in week 0. The control samples remained in the negative zone throughout the four weeks of storage, having very few volatiles produced. Group 2 samples having probiotic and synbiotic yogurt samples were positioned on the positive side of PC1, especially after 1 week of storage. Samples in group 3 (synbiotic only having both S. boulardii and inulin) were positioned on the positive side of PC1 and located together in the bi-plot. Significant differences were observed among old yogurts of these groups at weeks 0 and 4. In summary, the bi-plot indicates that limonene, L-limonene, and carboxylic acids (octanoic acids, hexanoic acid, butanoic acid, and acetic acid), along with ethanol, were grouped on the positive side of PC1, strongly contributing to the flavor of probiotic and synbiotic yogurt samples. Similarly, esters (1-butanol-3-methyl-acetate, butanoic acid-3-methyl butyl ester, octanoic acid ethyl ester) were secondary in contributing to the flavor of probiotic and synbiotic yogurt samples and were also placed on the positive side of PC1. The remaining volatile compounds (propanoic acid and 2-methyl-butanoic acid, 3hydroxy butanal) were grouped in the center of the bi-plot, with less contribution to the principal component, compared with the other compounds. According to these results, PCA, along with HCA, is a powerful tool to distinguish the samples based on treatment (probiotics, synbiotics) and magnitude of storage time.

Conclusions
Functional yogurt samples having S. boulardii and inulin in different concentrations showed that the antioxidant potential increased during the storage period and remained stable. Similarly, the probiotic and synbiotic yogurt samples showed that the number of volatile compounds also increased with chilled storage. The higher amounts of ethanol (S3-Syn1 = 306.81 ± 22.08 and S4-Syn1.5 = 304.44 ± 18.20 mg/100 g) indicated that S. boulardii could ferment the available sugars to produce more volatile compounds, such as limonene, acids, esters, and alcohols, in higher concentrations, while these were either absent or produced in very low concentrations in the control yogurt sample throughout the storage period. Thus, synbiotic yogurt with more antioxidant potential and higher concentrations of volatiles can strengthen consumer acceptability, representing a novel synbiotic dairy product with probiotic yeast.