Formulation and storage properties of symbiotic rice-based yogurt-like product using polymerized whey protein as a gelation agent

ABSTRACT A symbiotic rice-based yogurt-like product (SRYP) using polymerized whey protein (PWP) as a gelation agent was developed. Rice milk was optimized and prepared by soaking rice for 3 h at room temperature, heating at 80°C for 10 min at then emulsified with 0.05% (w/v) lecithin. The finalized rice yogurt was formulated with 1% rice protein, 1% prebiotic (inulin), 1% soy oil, 7% sugar, 0.4% PWP and 0.03% xanthan gum, and then fermented at 43°C for 3.5 h with 0.01% (w/v) ABY-3 starter culture containing probiotics of Bifidobacterium and L. acidophilus through optimization of each composition and fermentation conditions. SRYP contained 2.0% protein and 1.1% fat and exhibited good sensory attributes with an average score of about 4.3 (5-scale). Microstructure analysis showed a compact and uniform network with small voids. Both probiotics remained above 106 CFU/mL after 8 weeks. Results indicated that the symbiotic rice-based yogurt-like product was a novel functional food high in protein and low in fat with good texture.


Introduction
Rice, seeds of the grass species Oryza sativa (Asian rice) or Oryza glaberrima (African rice); serves as a staple food for about half of the global populations (Miah et al., 2002). It is good sources of protein, carbohydrate, vitamins and minerals. Rice protein, accounting for 6 ~ 9% in raw rice, is of highquality with BV (biological value) of 77 and PER (protein efficacy ratio) value ranging from 1.36 to 2.56 (Rabiey & Britten, 2009). Also, rice protein is rich in L-arginine, which is reported to prevent atherosclerosis development (Ni et al., 2003). Rice starch content is about 75% and composed of 19% amylose and 81% amylopectin. Functional components of rice such as dietary fiber, tocopherols etc. have been reported for many biological activities such as: controlling blood sugar; lowering cholesterol; antioxidant and anticancer properties (Tecson-Mendoza, 2007). Rice and rice flour are commonly consumed in various forms such as porridges, biscuits, cakes, cookies, tortillas, and bread (Kaur et al., 2014). With recognition of the nutrition and health benefits of rice, it is now widely used in the development of functional foods. Rice beer is used by India people as a traditional medicine for many ailments (Das et al., 2014). A purple rice drink with probiotics is reported to have high short-chain fatty acid and reduce ammonia production in colon (Worametrachanon et al., 2014). Pasta formulated based on rice was a reliable alternative for gluten-free products (Albuja-vaca et al., 2020). Cross-linked rice starch has been added to low fat cream as a fat replacer which produced creams with good sensory and physicochemical properties (Bagheri et al., 2018). Waxy rice starch can improve the expansion, homogeneity, hardness, crispness and colour of microwave-puffed cheese chips (Liu et al., 2018). Fermented milk drink with red rice extract was an alternative for the consumer and has a great nutritional and functional benefits (Boêno et al., 2020).
Whey proteins are obtained by ultrafiltration or ultrafiltration/diafiltration of whey-byproduct during cheese making. The main component of whey protein is β-lactoglobulin, which is a globular protein (Ahmad et al., 2019). When heated at certain temperatures, three-dimensional structure of β-lactoglobulin unfolded and the embedded free sulfhydryl groups were exposed. The denatured whey protein can aggregate into polymerized whey protein (PWP) by forming intermolecular disulfide bonds. PWP is soluble and could be induced into gel in acidic environment (C.N. Wang et al., 2015). The induced gel could entrap water and small molecules, and can be used as a thickening agent in fermented foods (Li & Guo, 2006).
Fermented dairy products have been utilized since ancient times (Bansal et al., 2016) due to the benefits to the consumers such as hypolipidaemic effects (Hassanzadeh-Taheri et al., 2018;Sarfraz et al., 2019). The objectives of this study were to develop a symbiotic rice-based yogurt-like product (SRYP) using polymerized whey protein (PWP) as a major gelation agent and to analyze the physiochemical properties, microstructure and storage properties of the new product.

Preparation of rice milk
The symbiotic rice-based yogurt-like product (SRYP) was prepared by fermenting formulated rice milk at optimized conditions. Rice milk was prepared through soaking, grinding, gelatinization and emulsifying processes.

Soaking process
Effects of rice to water ratio (RWR), soaking temperature and time on water absorption rate (WAR) were studied. The three factors were studied as follows: RWR of 1:1 ~ 8, soaking time of 15 min ~ 6 h and the temperature of 4 and 25°C. Weighed rice sample (100 g) was soaked in container with various volumes of deionized water for appropriate time at 4 or 25° C. Subsequently, soaked rice strained through a cheese cloth for 1 h was weighed and recorded as W 1 . The WAR was calculated as the following equation:

Emulsifying agent types and levels
Three types of emulsifying agents at different levels were optimized as follow (w/v): 0.2 ~ 0.6% glyceryl monostearate, 0.05 ~ 0.4% sucrose ester and 0.01 ~ 0.15% lecithin, respectively. Samples with different emulsifying agents were stored at 4°C for up to 7 days. The stability was observed and recorded at 12 and 24 h, and on days 2 to 7.

Formulation of SRYP
Inulin, as a prebiotic, was added in rice milk at two levels (1% and 2%, w/v). Rice protein (RP), soy protein isolate (SPI) or whey protein concentrate (WPC) were added at levels of 1% or 2%. Different combinations of PWP and other gelation agents were studied as listed in Table 1. PWP was prepared as described by C. N. Wang et al. (2015).

Preparation of SRYP
Three batches of SRYP samples were prepared according to the preliminary results. Raw polished rice was soaked at 25°C for 3 h in RWR of 1:3, and then ground into slurry. Water was added in the proportion of 1:12 for the gelatinization at 80°C for 10 min; and then 0.05% lecithin was added as the emulsifier. The obtained rice milk was formulated with 1% inulin, 7% sugar, 1% rice protein, 0.4% PWP and 0.03% xanthan gum, and then fermented at 43°C for 3.5 h with 0.01% ABY-3. The SRYP samples (each in 150 mL) for a total of 180 cups were stored at 4°C for further analysis.

Physiochemical analysis of SRYP
A moisture meter (MJ33, Mettler Toledo, Zurich, Switzerland) was used to determine the total solids content. Protein content was measured by Kieldahl Azotometer (UDK-159, VELP Scientific Co., Ltd, Brianza, Italy) according to Chinese standard method (GB 5009.5-2016) using a conversion factor of 5.95. Fat and ash content were measured according to Chinese standard method (GB 5009.4-2016;GB 5009.6-2016). Carbohydrate content was obtained by subtract protein, fat and ash content from total solids. All measurements were performed in triplicates for three trials.

Texture profile and viscosity determination of SRYP
Texture profile of the samples was measured by a texture analyzer (CT-3, Brookfield Engineering Laboratories, Inc) using the following parameters: mode: TPA, probe: TA4/ 1000; distance: 10 mm; trigger: 4.5 g. Viscosity determination was conducted by Brookfield Viscometer (DV-3; Brookfield Engineering Laboratories, Inc., Middleboro, MA, USA) using T-D spindle at the revolution of 150 r/min.

Water holding capacity determination of SRYP
Water holding capacity of all was calculated by 100% minus syneresis values. The syneresis was determined as described by C. N. Wang et al. (2015). Sample was weighed (W 2 ) and fermented in a centrifuge tube. After fermentation, sample was centrifuged at 5000 × g for 10 min; the liquid supernatant was separated and weighed as W 3 . The syneresis was calculated as the following equation:

Sensory evaluation of SRYP
At least 10 graduate students in the Department of Food Science at Jilin University were trained for the sensory evaluation according to our previous study (C. N. Wang et al., 2015). The test parameters included texture, color, flavor, taste, and overall acceptability. The score scale for each index was 5-point ranging from dislike extremely to like extremely.

Microstructure analysis of SRYP
Rice milk was poured into an agar well (2.5%, w/v) with size of 1 cm × 1 cm × 1 cm and sealed with an agar slide for fermentation. After fermentation, the agar cube was carefully wiped off. The rice yogurt sample was fixed, washed, dehydrated and dried according to the method of C. N. Wang et al. (2015). Dried Samples were fixed on SEM stubs and sputter coated using 3 nm of Au/Pd (80/20) alloy. Microstructure was analyzed by SEM (S-3400, Hitachi, Japan) at 5 kV and 10 kV.

Shelf-life and probiotic survivability of SRYP
Samples were determined for the population of probiotics, pH, viscosity and hardness each week for total of 8 weeks. Samples at beginning and week 8 were also studied for SDS-PAGE analysis. Enumeration of Bifidobacterium and L. acidophilus were conducted according to the procedure of Chr. Hansen and expressed as log CFU/mL. Coliform, mold and yeast counts were carried out using Petrifilm plates (3M TM Petrifilm TM , St. Paul, MN, USA) (Hansen, 2005).

Statistical analysis
Data obtained from analysis were expressed as mean ± standard deviation (S.D.). The significant differences of data between samples and control were calculated using Version SPSS 19 (SPSS Inc. Chicago, IL). The significance level was set at P < .05 and P < .01. Data were checked for homogeneity by Leveneǐs test. When the data were homogeneous, analysis of variance (ANOVA) and then a least squared differences (LSD) model was used. All the figures were drawn by origin 8.0 (Origin Lab Corporation, Northampton, USA).

Preparation of rice milk
Soaking treatment is a necessary step and sufficient water content is crucial to gelatinization of rice starch and the sensory quality of the final product. Without soaking treatment, 24 ~ 28 h would be needed to complete the conventional parboiling of rough rice (Wambura et al., 2008). Effects of RWR, soaking time and temperature on WAR were investigated and the results are shown in Table S1. There was a significantly rapid increase in water content for the first 30 min (P < .05) with a WAR of ~23.28%, and then remained an almost same level from 30 min to 6 h without significant differences between samples. Differently, a higher WAR of 44% for rice sample was reported by Wambura et al. (2008). However, it should be noticed that the authors determined the moisture using an AOAC oven method which would certainly result a higher value. Water was absorbed into embryo which was mainly composed of protein with loose structure, and then transferred to cotyledon and endosperm. In endosperm, water permeated from out-layer protein to insider starch slowly. Therefore, although samples soaked at 30 min got a relatively high WAR, it would take a while for the rice kernels to get wet completely. Combined with the results of observation for cross sections of rice, 3-h soaking time was selected thereafter. A longer time more than 3 h would result in loss of some nutrients such as water-soluble vitamins and resultant lower WAR. The distances between rice grains would affect the water absorbing efficiency. With water proportion increased, the WAR showed a para-curve like trend with the highest value (23.56 ± 0.73%) in samples having RWR of 1:3, which was significantly higher than others (P < .05). Regarding effect of soaking temperature on WAR, commonly used storage temperatures of 4 and 25°C were tested. Results indicated no significant difference between the two samples (P > .05).
Starch is the main component in cereal grains including rice. The rice starch must undergo gelatinization before it exhibits its functionality, such as thickening, gelling, and ingredient-binding properties (Ie et al., 2012). Effects of RWR, temperature and time on DG of rice starch were determined and shown in Fig. S1. Uniformity is one of the most important indexes for gelatinization and an appropriate ratio of rice to water during gelatinization would give starch molecules enough space to swell. Therefore, soaked rice with RWR of 1:3 was further diluted with water to get the ratio of 1:8 to 1:16. When the RWR was 1:12, rice reached the significantly highest DG value (87.65 ± 3.02%) (P < .05). Gelatinization temperature and time are vital for the gelatinization process. A full combination of temperature from 65 to 80°C and time from 5 to 30 min were designed. When the rice was gelatinized at 80°C for 10 min, the gelatinization degree reached a significant higher value of about 87.65 ± 5.23% (P < .05) than those of other samples. This temperature and time are enough to ensure the gelatinization of rice, and also prevent from being over cooked (Wambura et al., 2008).
After gelatinization, 1% soy oil was added and then three food-grade emulsifiers at various levels were evaluated for the emulsion capacity (Fig. S2). The emulsifiers showed different characteristics to the resulting emulsion. Glycerin monostearate had the lowest stability, which remained stable for only 12 h even up to the level of 0.6%. Rice milk with soy lecithin showed the best stability for 6 days at the levels of 0.05 ~ 0.15%. Similar results were reported by Adepegba (2002). However, samples with 0.1% or 0.15% lecithin were too viscous for the further application. Lecithin, a zwitterionic surfactant, is an important natural emulsifier which is efficient on reducing interfacial tension (Mantovani et al., 2016). In the rice milk emulsion, the hydrophilic head group and lipophilic tail orientate themselves towards the water and oil phase, respectively. The prepared rice milk, similar to soy milk, could be consumed directly and may be a good choice for vegans/ vegetarians (Meharg et al., 2008).

Optimization for the formulation of SRYP
Inulin, a reserve carbohydrate presenting in many plants such as wheat, onion, garlic and chicory, is well known for the benefit of promoting friendly bacteria growth in human colon (Gibson et al., 2004;Valero-Cases & Frutos, 2015). A product in which a probiotic and a prebiotic are combined is called symbiotic (Ribeiro et al., 2019). To make a symbiotic product, inulin (1% or 2%) was incorporated into the rice yogurt as a prebiotic. Compared with the control, significant increases of 0.2 and 0.8 log CFU/mL for L. acidophilus and Bifidobacterium, respectively, in samples with 1% inulin were observed (P < .05). However, there was no significant difference between samples with 1% and 2% inulin (P > .05). Similar results were reported by C. N. Wang et al. (2015) who found that there was no significant difference between 1% and 2% inulin addition on the probiotics population in Chinese Laosuanna (Greek-yogurt like). Inulin addition at 1 or 2% can also enhance L. plantarum growth in fermented juice with no significant difference between these two levels (Valero-Cases & Frutos, 2017).
Three types of protein fortifier were studied and the results are showed in Figure 1. Soy protein and whey protein are commonly used for the fortification of protein and nutrition (Molina-Rubio et al., 2010). However, whey protein concentrate gave the sample a soft texture (Figure 1(a)) and soy protein resulted in an undesirable texture and taste (Figure 1(b)). Samples fortified with rice protein (RP) showed higher or at least same hardness and viscosity among all protein fortifiers. Samples with 1% RP had better taste, color, and texture values than those of the samples with 2% RP. Therefore, 1% RP was finally selected for fortification.
The acid-induced gel formation property of polymerized why protein made it a suitable gelation agent for fermented functional foods. Polymerized whey protein has been successfully used as a co-thickening agent in goat milk yogurt (W. Wang et al., 2012), and yogurt-like symbiotic fermented oat product (Walsh et al., 2010). Combinations of PWP with other gelation agents were studied and the results are shown in Figure 2. Rice yogurt-like samples with PWP and xanthan gum exhibited the highest hardness, viscosity and water holding capacity among all the combinations (P < .05). Xanthan gum is an anionic polysaccharide secreted by the bacterium xanthomonas campestris and has been commonly used in vegetable and plant-based drinks such as fermented Hazelnut milk (Bernat et al., 2014). Anionic polysaccharide could interact with positive domain of whey protein and affect the gelation properties of whey proteins. Zhang et al. (2014) reported that xanthan gum could decrease the gel strength of whey protein at ~ pH 6.5, but the gel strength was increased at the initial pH of rice milk (pH 6.48) in the present study. However, it should be noticed that the effect depends on the fermentation conditions, such as ionic concentration, and incubation temperature and others.
Suitable levels of PWP and xanthan gum were studied. As for water holding capacity, there was no significant difference between the levels tested for combinations of PWP and xanthan gum (Figure 2(c), P > .05). Since hardness and viscosity are the quantitative reflection of sensory properties. The final level of the thickening agents depends on the results of sensory evaluation. Even though rice yogurt samples X 3 -X 5 (0.3%~0.1% PWP and 0.05%~0.1% xanthan gum) displayed higher hardness and viscosity, formulation X 2 (0.4% PWP and 0.03% xanthan gum) was chosen based on the higher organoleptic scores (Figure 2(d)).

Optimization for fermentation conditions of SRYP
Starter culture is the most important factor during the fermentation process. A good starter culture would give the product a good texture and flavor. Acid producing rates (expressed as the changes in pH during fermentation) of L. plantarum, L. casei and ABY-3 were investigated and the results are displayed in Table S2. There was a 2.46 pH decrease for ABY-3 during 6 h fermentation, while the others were 1.14 and 0.8 for L. plantarum and L. casei, respectively. The initial pH of the rice milk was 6.46 ± 0.04 and the final pH was 4.00 ± 0.01, the decrease of 2.46 in pH was a similar pH change in cow milk which would be 6.86 to 4.50 (C. N. Wang et al., 2015). ABY-3 would be a suitable starter culture for the symbiotic rice yogurt production.
Fermentation conditions including inoculation level, fermentation temperature and time were optimized and the results are shown in Table 2. Inoculation level affects probiotic population, flavor and texture of the product. Low inoculant level resulted in long fermentation time and high level may lead to a short exponential phase, a bitter taste of the yogurt and the shrinkage of the curd. As shown in Table 2, inoculant level had significant effects on pH values and populations of L. acidophilus (P < .05). The rice yogurt inoculated with 0.008% starter culture resulted in highest number of L. acidophilus (6.3 ± 0.06 log CFU/mL). There was no significant difference in both hardness and Bifidobacterium population among samples with different inoculated levels.
Temperature affects enzyme activity of microorganisms and thus influences growth of the bacteria. Bifidobacterium population of samples fermented at 43°C was significantly higher than those of other samples (P < .05), which would be the optimum growth temperature for this culture. Furthermore, samples incubated at 43°C displayed the highest hardness value (P < .05). To reach the largest vitality of the probiotics, termination point of the fermentation should be at the stable phase where the probiotics propagated to the greatest population. Both L. acidophilus and Bifidobacterium populations started to decline from 3.5 h, which might be the beginning of decline phase. Therefore, the fermentation was stopped after 3.5 h by rapid cooling.

Physiochemical and sensory properties of SRYP
Milk alternative fermented foods are gaining more popularity (Seleet et al., 2016). A rice-based yogurt-like product was developed in this study and the physiochemical and sensory properties of this product are shown in Table 3. The total solids of yogurt were 16.23 ± 0.22%. The protein content was 2.00 ± 0.15%, which was similar to that of Makgeolli, a traditional alcoholic beverage prepared from rice, barley, wheat or malt grains by fermentation using a natural starter called nuruk (Nile, 2015), and higher than that of other cereal milk, such as quinoa milk developed by Pineli et al. (2015). It had a good water holding capacity of about 94.92%, which was much a b Figure 1. Effects of protein fortification on the texture (a) and sensory properties (b) of symbiotic rice-based yogurt-like product. Different letters or * denote significance between samples at P < .05 Figura 1. Efectos del enriquecimiento proteico en la textura (a) y las propiedades sensoriales (b) del producto simbiótico tipo yogur a base de arroz. Nota: las letras distintas o el * denotan diferencias significativas entre las muestras a P < .05.
greater than that of cow yogurt stabilized by gelatin (43~64%) (Kiros et al., 2016), and similar to a strained yogurt using different whey fortifier minimizing the syneresis (Patroklos et al., 2016). The product received a good response from evaluators with average scores ranging from 3.38 to 4.85. The new product had high hardness value of 99.40 ± 6.20 g and viscosity of 1573.20 ± 155.45 mPa.s.
The rice-based yogurt-like product may be a good milk alternative fermented food. The high-protein and low-fat content and inclusion of probiotics and prebiotic may confer health benefits to people (Chetachukwu et al., 2019). Sensory properties of this product based on consumer test should be evaluated in further study.

Microstructure and microbiological properties of SRYP
The microstructure of the rice-based yogurt-like product at different magnifications is showed in Figure 3. The rice a b c d Figure 2. Effects of different combinations of thickening agents on hardness (a), viscosity (b), syneresis (c) and sensory score (d) of symbiotic rice-based yogurtlike product. Note: Different lower-case letters denote significance between samples at P < .05.

Figura 2.
Efectos de diferentes combinaciones de agentes espesantes en la dureza (a), la viscosidad (b), la sinéresis (c) y la puntuación sensorial (d) del producto similar al yogur simbiótico a base de arroz. Nota: las distintas letras minúsculas denotan diferencias significativas entre las muestras a P < .05. yogurt sample with PWP (Figure 3(a)) had a denser network with small pores compared with the sample without PWP (Figure 3(b)). The denser protein network might be responsible for the higher water holding capacity and the desirable texture properties of the new product. Shelf-life tests result of rice yogurt samples are shown in Figure 4. There was no yeast, mold, or coliform detected throughout the whole storage time. Storage time had a negative impact on the probiotic's population (L. acidophilus and Bifidobacterium). Similar results were reported by Ozcan et al. (2010) who studied viability of L. acidophilus and Bifidobacterium in rice pudding. Even though there was a reduction in the probiotic's population, the total count was still sufficient to obtain the desired therapeutic impact in the first 3 weeks (Figure 4  (a)). It has been recommended that viability of a probiotic should be at least 10 7 CFU/mL at the time of consumption to provide various health benefits (Rajam et al., 2012). The cereal based fermented foods are earning more and more consumers, however, the main drawback was the active metabolism of probiotics during storage (Bevilacqua et al., 2016). The pH of the product was relatively stable (Figure 4(b)), indicating the attenuation of the metabolism of the LABs. The poor post-acidification ability of the microorganisms indicated the storage stability. Figure 4(c,d) reveals significant decreases for both hardness and viscosity during the first 6 weeks of storage (P < .05). The possible explanation may be that there was a slight collapse in the network with storage time processed, which resulted in whey separation and then the decreased hardness and viscosity. Changes in the protein profile during the storage were evaluated by SDS-PAGE analysis ( Figure 5). There was no proteolysis occurred during storage.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The financial support for this project was provided by the Ministry of Science and Technology of China [Project # 2013BAD18B07].