*Article* **Evaluation of Innovative Dried Purée from Jerusalem Artichoke—In Vitro Studies of Its Physicochemical and Health-Promoting Properties**

**Jan Oszmia ´nski <sup>1</sup> , Sabina Lachowicz 2,\* , Paulina Nowicka <sup>1</sup> , Paweł Rubi ´nski <sup>3</sup> and Tomasz Cebulak <sup>4</sup>**


**Abstract:** The present study aimed to evaluate the effect of Jerusalem artichoke processing methods and drying methods (freeze drying, sublimation drying, vacuum drying) on the basic physicochemical parameters, profiles and contents of sugars and polyphenolic compounds, and health-promoting properties (antioxidant activity, inhibition of the activities of α-amylase, α-glucosidase, and pancreatic lipase) of the produced purée. A total of 25 polyphenolic compounds belonging to hydroxycinnamic phenolic acids (LC-PDA-MS-QTof) were detected in Jerusalem artichoke purée. Their average content in the raw material was at 820 mg/100 g dm (UPLC-PDA-FL) and was 2.7 times higher than in the cooked material. The chemical composition and the health-promoting value of the purées were affected by the drying method, with the most beneficial values of the evaluated parameters obtained upon freeze drying. Vacuum drying could offer an alternative to freeze drying, as both methods ensured relatively comparable values of the assessed parameters.

**Keywords:** functional food; innovative food; drying; natural food; *Helianthus tuberosus*; pro-healthy properties

#### **1. Introduction**

The Jerusalem artichoke (*Helianthus tuberosus*; (JA)) is a species of sunflower from the genus Helianthus, belonging to the family Asteraceae, and derived from the North America. In Europe, it has been cultivated since the 17th century [1,2]. It is characterized by a fast growth rate, is tolerant to droughts, salinity, and frost, and is resistant to diseases and pests [2,3]. Due to its valuable chemical composition and scientifically proven healthpromoting properties, JA has spurred a growing interest as an edible plant [4]. Its tubers contain ca. 80% of water, 2% of protein, and ca. 20% of carbohydrates, ca. 90% of which are represented by inulin [2]. JA is also valuable considering its bioactive compounds, such as, e.g., polyphenolic compounds, including phenolic acids, which exhibit strong antioxidant properties, and has also been confirmed to elicit antiviral, antibacterial, antiinflammatory, and anti-carcinogenic effects [1,5]. In turn, as a prebiotic and soluble dietary fiber, inulin contained in JA tubers and stalks (considered to be its richest sources) ensures a hypoglycemic effect in diabetes treatment. In the gastrointestinal tract, inulin undergoes fermentation by the gut microbiota, affecting the state of eubiosis. In addition, it contributes to the increased availability of such minerals as Fe, Mg, and Ca, and influences lipid metabolism [1,6,7]. Furthermore, JA improves immunity and concentration, alleviates stress, and eliminates toxic metabolites from the body [8]. Its main applications include the

**Citation:** Oszmia ´nski, J.; Lachowicz, S.; Nowicka, P.; Rubi ´nski, P.; Cebulak, T. Evaluation of Innovative Dried Purée from Jerusalem Artichoke—In Vitro Studies of Its Physicochemical and Health-Promoting Properties. *Molecules* **2021**, *26*, 2644. https:// doi.org/10.3390/molecules26092644

Academic Editor: Celestino Santos-Buelga

Received: 6 April 2021 Accepted: 28 April 2021 Published: 30 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

production of inulin [9], feedstuff, fructose syrup, flour, French fries, biochemical materials, and bioethanol [10–12]. JA tubers processed with various cooking methods were evaluated for their sensory profiles [12]. Thus, taking into account their beneficial effects of providing valuable substances, it is necessary to develop a product with the lowest possible losses of these valuable substances and high storage stability.

Considering the above, we proposed JA purée preserved with a properly selected drying method. The use of the drying process will enable the preservation of the product, making it available all year round, and not only in the maturity period. The most common drying method is convective drying due to its low cost and relatively high efficiency [13]. In turn, sublimation drying (SD) requires high temperatures, is relatively long, which in turn leads to large losses of compounds valuable to the human body, resulting from the high access of oxygen [13]. On the other hand, freeze drying (FD) is the best method to obtain products with the lowest thermal degradation of bioactive compounds during water removal; however, it is relatively costly. Hence, vacuum drying (VD) can be an alternative to SD because it offers a shorter drying time, due to the reduced pressure, and heat supply by conduction [14]. In addition, it allows temperature control, which can reduce the thermal degradation of thermolabile compounds, such as polyphenolic compounds, and is relatively economical. However, different drying processes can affect the quality and induce different positive or negative changes in the finished product. Therefore, it is important to monitor these changes depending on the product being dried. Furthermore, the impact of the technological treatment and changes induced by drying on the content of inulin and polyphenols, and health-promoting values in the innovative dried purée from JA (raw and cooked tubers) has not been studied so far. Considering the above, the present study aimed to evaluate the effect of Jerusalem artichoke processing and drying methods (freeze drying, sublimation drying, vacuum drying) based on the analysis of physicochemical parameters, profiles and contents of sugars and polyphenolic compounds, and health-promoting properties (antioxidant, anti-diabetic, and anti-obesity activity) of the produced purée.

#### **2. Results**

#### *2.1. Chemical Parameters*

#### 2.1.1. Basic Chemical Parameters

The results of analyses of six variants of dry purées obtained from fresh and cooked JA are presented in Table 1. Both the drying methods and the purée preparation technology statistically significantly affected the ash and pectin contents, while they had no significant effect on the dry matter content, total acidity, and pH (*p* < 0.05). The FD products showed a higher content of dry matter, ash, and pectins. In turn, the lowest contents of pectin, ash, and dry matter were obtained in the products after SD. Therefore, this method was proved the most advantageous for the preparation of the innovative dried JA purée; however, due to its high costs, the VD can be used as an alternative. Taking into account the JA preparation technology, the greatest differences were noted in SD products, where the contents of dry matter and ash were 5% and 9% higher in the purée prepared from cooked JA, while pectin content was 41% higher in the purée made of fresh material. In the case of pectins, after VD and FD, their content was 12% and 28% lower in the purée made from the cooked material than from the raw one. This is because cooking causes plant tissues to break down into individual cells and pectins to leach out. This phenomenon is also characteristic of potatoes and JA, but it is not observed when cooking root vegetables due to their thicker and harder cell membranes [15]. In addition, pectin is a form of soluble fiber that helps prevent cardiovascular diseases, diabetes, and obesity [16]. Therefore, a purée prepared from raw JA will be a more desirable product, especially when designing products dedicated to obese and diabetic patients.


**Table 1.** The results of analyzes of purées and dried JA.

<sup>a</sup> Values are means <sup>±</sup> standard deviation, *<sup>n</sup>* = 3. Mean values within a row with different letters as a, b, c, d, e, f are significantly different at *p* < 0.05. Abbreviations: FDC, freeze drying of cooked material; FDR, freeze drying of raw material; SDC, sublimation drying of cooked material; SDR, freeze drying of raw material; VDC, vacuum drying of cooked material; VDR, vacuum drying of raw material.

#### 2.1.2. Determination of Sugar Changes

Table 1 presents the results of sugar content determination by ultra-efficient liquid chromatography coupled with an ELSD detector. The assessed material contained three types of sugars, i.e., inulin (accounting for 95.8% of total sugars on average from all analyzed sample) > sucrose (3.7%) > fructose in trace amounts (0.5%). The highest amount of total sugar was ranged from 41.4 g/100 g in sample made of raw material after the VD drying method to 45.3 g/100 g in sample made of raw material after the FD drying method. In turn, the highest content of inulin was determined in the purée from raw JA after FD (43.3 g/100 g), and the lowest one in the purée from raw JA after VD (38.9 g/100 g). Similar results concerning inulin content in the material subjected to FD were obtained by Michalska-Ciechanowska et al. [17] and Cieslik et al. [18]. In the case of fructose, it has been noted that the higher the inulin content was, the lower was the fructose content. This means that inulin has not been hydrolyzed to fructose, which usually occurs by an acid or by inulinase [19]. The authors state that JA contains the highest amounts of inulin during the harvest period from October to December, and that in the remaining months, inulin is hydrolyzed to a simple sugar [19]. However, it was noted that the purée made of raw JA contained from 29% to 78% more fructose after the FD and VD drying method, respectively, compared to the purée made of cooked JA. Thus, this may be related to inulin hydrolysis. According to Böhm et al. [20] the 1 h heat treatment at 100 to 135 ◦C after the acid hydrolysis of chicory inulin did not affect its degradation, which may also explain the slight differences between the purées prepared from the cooked and raw JA.

#### 2.1.3. Determination of Polyphenolic Compounds

The detailed identification of polyphenolic compounds in dried JA purée made of raw and cooked material dried with all drying methods using LC-PDA-MS-QTof and UPLC-PDA-FL showed the presence of 25 compounds, all of which belonged to the class of hydroxycinnamic phenolic acids (Table 2). Similar results were obtained by Kapusta et al. [21] and Michalska-Ciechanowska et al. [17]. The obtained results indicate that the drying method used had a significant (*p* < 0.05) effect on the content of polyphenolic compounds. The highest total content of polyphenols, reaching 923.5 mg/100 g d.m., was found in the purée made of raw JA after FD. It was about five times higher compared to the content of polyphenolic compounds determined by Kapusta et al. [21]. Depending on the drying method used, the result obtained after FD was 21% (made of cooked material) and 15% (made of raw material) higher compared to the results obtained after SD and VD drying, respectively. According to Michalska et al. [14], the drying method had a significant impact on the final content of bioactive compounds in the finished product; hence, it is essential to select the appropriate drying method that would allow the maintenance of a relatively high content of the tested compounds. In the present study, the content of polyphenols was also statistically significantly affected by the technological treatment of JA. The purée made of cooked JA tubers contained 62% for FD, 69% for SD, and 60% for VD less bioactive compounds than the purée obtained from the unprocessed raw material dried by FD, SD,

and VD, respectively. Similar results were reported in the study by Laib and Barkat [22], in which the content of polyphenolic compounds was much lower in cooked potatoes than those that were not heat-treated. This is probably due to the fact that these compounds are thermolabile, thus were destroyed and leached out to the solution.

**Table 2.** Analysis results of phenolic compounds (mg/100 g dm) in dry JA samples.


<sup>a</sup> The calibration curve of ferulic acid was used to quantify; <sup>b</sup> the calibration curve of 5-*O*-caffeoylquinic was used to quantify; <sup>c</sup> the calibration curve of caffeic acid was used to quantify; <sup>d</sup> the calibration curve of 3-*O*-caffeoylquinic acid was used to quantify; <sup>e</sup> values are means ± standard deviation, *n* = 3. Mean values within a row with different letters as a, b, c, d, e, f are significantly different at *p* < 0.05. Abbreviations: FDC, freeze drying of cooked material; FDR, freeze drying of raw material; SDC, sublimation drying of cooked material; SDR, freeze drying of raw material; VDC, vacuum drying of cooked material; VDR, vacuum drying of raw material; R.t., retention time; nd, not detected.

> 3-*O*-caffeoylquinic acid (with main ion at *m*/*z* 353) and 3,5-di-*O*-caffeoylquinic acid (with main ion at *m*/*z* 515) were the major compounds [17] and, on average, accounted for 18% (from 16% of FDR to 24% of FDC) and 11% (from 8% of SDC to 13% of FDR) of total phenolic hydroxycinnamic acids, respectively. Their contents were also significantly dependent on the drying method and technological treatment, with the highest ones determined in the purées prepared from raw JA after FD, while the lowest ones were in the purée made of cooked JA after SD. During the peeling of Jerusalem artichoke, sulfur dioxide was added to protect the color. The UPLC-PDA chromatograms (Supplementary Materials Figure S1) of purées made of raw JA tuber revealed additional peaks of compounds, which were identified as derivatives of quinones, i.e., phenolic acid oxidation products in combination with a sulfite molecule. Such a combination with quinone sulfite was noted in caffeoylquinic acid (with main ion at m/z 415), compounds, whose fragmentation ions were found at *m*/*z* 387, 258, 191, 179, and 161, and in dicaffeoylquinic acid compounds (with main ion at *m*/*z* 577) whose fragmentation ions were found at *m*/*z* 415, 387, 258, 191, 179, and 161. The contribution of derivatives of quinones on the total polyphenolic compound concentration ranged from 30% for FDR to 34% for SDR. In turn, these compounds were not identified in the purées made of cooked JA tubers. They were probably washed out from the root surface by the water solutions in which they were cooked. In addition, the high cooking temperature contributed to the inactivation of enzymes that could cause the oxidation of phenolic acids to quinones. The use of sulfur dioxide as a preservative had a significant impact on the protection of bioactive compounds in the purées made of raw rather than cooked material. Similar observations regarding the protection of polyphenolic compounds were noted upon the use of sulfur dioxide in white wines [23]. In addition, compounds such as caffeoylquinic acids (isomer of chlorogenic acid; with main ion at *m*/*z* 353) and 1,5-dicaffeoylquinic acid (with main ion at *m*/*z* 515) were significantly influenced by cooking, as their contents in the purées were on average 57% and 52% lower,

respectively, compared to those determined in the purées made of raw tubers. Similar observations were made by Laib and Barkat [22], who investigated the effect of heat treatment, including cooking, on the content of compounds in potato tubers. The analyzed compounds were probably released during heat treatment because some polyphenols, including phenolic acids, may be associated with non-digestible components of the cellular structure and may be released and/or solubilized during this structure's damage [22,24]. Moreover, the loss of the analyzed compounds is also strongly affected by their chemical structure, because the cooking process may differently influence compounds classified into one subclass. These losses can be influenced by the hydroxylation pattern, particle size, solubility, polarity, and sugar bonding [25]. Therefore, it is important to monitor changes during the selection of the most advantageous drying technique. The best method to obtain the dried purées turned out to be the freeze drying; however, due to its costs, the vacuum drying seems a fine alternative. In turn, depending on the chemical composition of the finished product, the raw material processing method can be used appropriately when designing new products. Considering the content of bioactive compounds, it is more advisable to produce the dried purée from raw JA tubers, irrespective of the compounds released during processing.

#### *2.2. Physical Parameters*

#### Color Parameters and Water Activity

The dried purées were found to differ significantly in their color parameters as affected by both the technological treatment and drying method (Table 3). The best in terms of brightest turned out to be the dried purée prepared from raw JA after FD compared to that made of boiled JA. In contrast, the use of VD and SD for product preservation caused 5% and 9% darkening of the purée made of the raw tubers and 5% and 4% darkening of the purée made of the cooked tubers. The a \* and b \* color parameters of the tested material indicated that JA cooking intensified the green color and darkened the yellow color of the purées, while purées made of the raw material were more yellow with a slight hue of green. Similar dependencies were observed in the measurements of the a\* and b\* color parameters depending on the drying method used; the FD purées were characterized by a light hue of green and a dark hue of yellow. An opposite tendency was noted in SD products, revealing a darker hue of green and a lighter hue of yellow. The results obtained for the purée after FD were comparable with the color measurement results reported by Antal et al. [2] for JA subjected to FD drying only. Those authors demonstrated a similar dependency; namely that the color of the dried material depended on the drying method used, and thus the brightest products were also obtained after FD [2].

The evaluation of the dried purées in terms of water activity (aw) showed statistically significant differences caused by both the drying method and the method of purée preparation (Table 1). The lowest aw, reaching 0.012 for FDC and 0.015 for FDR, was determined for the FD purée, and this value was on average 18 and 11 times lower compared to SD and VD purées, respectively. On the other hand, the a<sup>w</sup> value determined after VD was two times lower compared to the value determined after SD. According to the results reported by Antal et al. [2], the a<sup>w</sup> value recorded for freeze-dried JA without technological treatment was seven times higher compared to our study. A lower a<sup>w</sup> value of dried fruits of Saskatoon berry was also noted after FD, whereas there was a higher value after SD [26]. In turn, regardless of the drying method used, the purées made of raw material were characterized by an on average 40% higher a<sup>w</sup> value. This can be explained by the slight evaporation of water during the cooking process of [12]. However, regardless of the drying method and preparation technology used, the aw of all dried purées was below the critical level (a<sup>w</sup> = 0.60). This means that they meet the requirement of a product safe from microbiological spoilage, i.e., from contamination with bacteria and mold, because the a<sup>w</sup> value above 0.60 may cause microbiological spoilage of the finished product [2].


**Table 3.** Color measurement results.

<sup>a</sup> Values are means <sup>±</sup> standard deviation, *<sup>n</sup>* = 3. Mean values within a row with different letters as a, b, c, d, e, f are significantly different at *p* < 0.05. Abbreviations: FDC, freeze drying of cooked material; FDR, freeze drying of raw material; SDC, sublimation drying of cooked material; SDR, freeze drying of raw material; VDC, vacuum drying of cooked material; VDR, vacuum drying of raw material.

#### *2.3. Pro-Healthy Properties*

The study also determined the health-promoting properties of JA preserved using various drying methods. The purées were analyzed in terms of their antioxidant, antidiabetic (the ability to inhibit α-amylase and α-glucosidase), and anti-obesity properties (the ability to inhibit pancreatic lipase). Finding effective inhibitors of α-amylase and α-glucosidase would allow a delay in sugar absorption and a reduction in postprandial blood glucose. On the other hand, finding an effective inhibitor of pancreatic lipase activity by stimulating the cell membrane permeability would enable the apt functioning of the pancreas as a gland responsible for the proper insulin secretion. Additionally, pancreatic lipase is a key dietary fat-absorbing enzyme responsible for the hydrolysis of triglycerides to 2-monoacylglycerides and free fatty acids that can be absorbed by enterocytes. Its inhibition is used to reduce the rate of dietary fat absorption and, therefore, may offer an alternative approach to treating overweight and obesity [27].

The analysis of the antioxidant activity of the studied variants of JA purées showed that the greatest antioxidant effect was obtained in JA dried using the vacuum and sublimation methods (Table 4). It was proved that not only the drying method, but also the type of raw material used for drying, played a significant role in modeling the antioxidant properties of the tested material. Hence, a much better effect in developing the health-promoting properties was obtained by drying fresh than the previously cooked material.

A slightly different trend was observed in the JA ability to inhibit pancreatic lipase, α-amylase, and α-glucosidase (Table 4). A more effective inhibitor turned out to be SDC. An opposite effect was observed with the vacuum method. It was shown to be the most effective in modeling both anti-diabetic and anti-obesity properties. Moreover, as in the case of antioxidant properties, it appeared more effective to dry fresh raw material than the cooked one. In general, it can be concluded that the produced dried purées were characterized by a high anti-diabetic potential—similar results were obtained for both α-amylase (IC<sup>50</sup> values from 130 µg/mL to 736 µg/mL) and α-glucosidase (IC<sup>50</sup> values from 120 µg/mL to 898 µg/mL). A study by Wang et al. [28] has shown that the ability to inhibit α-glucosidase is stimulated by hydroxycinnamic acid derivatives. However, no such trend was observed in the present study. It has been shown that procyanidin polymers can also be involved in enzyme inhibition. This tendency was confirmed by Boath et al. [29], who suggested that fruit extracts rich in procyanidins were effective inhibitors of α-amylase because they had the ability to form tannin–enzyme complexes, which effectively inhibited the hydrolysis of polysaccharides to simple sugars. Other authors have shown that the effective blocking of diabetes-related enzymes may be due to high concentrations of inulin [30].


**Table 4.** The degree of metabolism and absorption of sugars derived from JA products.

<sup>a</sup> Values are means <sup>±</sup> standard deviation, *<sup>n</sup>* = 3. Mean values within a row with different letters as a, b, c, d, e, f, are significantly different at *p* < 0.05. Abbreviations: FDC, freeze drying of cooked material; FDR, freeze drying of raw material; SDC, sublimation drying of cooked material; SDR, freeze drying of raw material; VDC, vacuum drying of cooked material; VDR, vacuum drying of raw material.

> It should be emphasized, however, that JA was the most effective pancreatic lipase inhibitor. The IC<sup>50</sup> values determined for this enzyme ranged from 31 µg/mL (air-dried Jerusalem artichoke, previously cooked) to 58 µg/mL (air-dried JA, fresh). The observed trend may again be due to the high concentration of inulin in this product. Recently, it has been proved that inulin effectively prevents the occurrence of obesity and diabetes, i.a., by lowering the blood levels of triglycerides, cholesterol, and glucose [31].

> The conducted research has shown that JA is an interesting raw material with a wide spectrum of health-promoting properties that can be modulated in certain ranges by selecting appropriate processing and preservation methods.

#### **3. Materials and Methods**

#### *3.1. Materials*

Acetonitrile, formic acid, methanol, ABTS (2,20 -azinobis(3-ethylbenzothiazoline-6 sulfonic acid), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox), 2,4,6-tri(2 pyridyl)-s-triazine (TPTZ), 2,2-Di(4-tert-octylphenyl)-1-picrylhydrazyl (DPPH), methanol, acetic acid, 2,20 -azobis (2-amidino-propane) dihydrochloride (AAPH), fluorescein disodium (FL), potassium persulfate, TPTZ (2,4,6-tripyridyl-1,3,5-triazine), FeCl3, phloroglucinol, 3,5 dinitrosalicylic acid, potassium sodium tartrate tetrahydrate, sodium phosphate monobasic, starch from potato, α-amylase from porcine pancreas (type VI-8), dipotassium hydrogen orthophosphate dihydrogen, p-nitrophenyl-α-D-glucopyranoside, α-glucosidase from Saccharomyces cerevisiae (type I), sugar, and polyphenolic standards were purchased from Sigma-Aldrich (Steinheim, Germany). Acetonitrile for ultra-performance liquid chromatography (UPLC; gradient grade) and ascorbic acid were from Merck (Darmstadt, Germany).

*Helianthus tuberosus* was harvested at the organic cultivation Gospodarstwo Rolno-Ogrodnicze Marek Strojs in Kalisz (Poland). Jerusalem artichoke (around 10 kg) "Albik" cultivar was collected in August 2020.

#### *3.2. Sample Preparation*

The production process of dried Jerusalem artichoke purée included 3 main stages:


Harz, Germany) for 24 h, (ii) sublimation drying (SD)—carried out in a (KC 100/200 Wytwórnia Aparatury Elektronicznej i Medycznej Warszawa) for 8 h at 80 ◦C, (iii) vacuum drying (VD)—carried out in a VACUCELL 111 ECO LINE vacuum dryer (MMM Medcenter Einrichtungen GmbH, Planegg, Germany) for 8 h at 80 ◦C.

All drying experiments were performed in duplicate. The samples obtained were milled by a laboratory mill (IKA A.11, Wilmington, NC, USA), and vacuum sealed. The powders were kept in a freezer (−20 ◦C) until the extracts' preparation.

#### *3.3. Chemical Parameters*

#### 3.3.1. Basic Parameters

The dry matter was determined by mixing the sample with diatomaceous earth, predrying, and final drying under reduced pressure. Total acidity was extracted by mixing with water in 100 mL flasks, cooking for 20 min, and cooling for 20 min. Total acidity (TA) was analyzed supernatant (8 mL) by the titration of products with 0.1N NaOH to pH 8.1 and the results were expressed as g malic acid/100 g. TA and pH were determined using an automatic pH titrator system (TitroLine 5000, Xylem Analytics GmbH, Weilheim in Oberbayern, Germany). The soluble solids content, dry matter, titratable acidity, and ash were taken according to European Standards, PN-EN 12143:2000, PN-EN 12145:2001, and PN-EN 12145:2000, PN-EN 450:1998, respectively. Pectin content (g/100 g) was measured according to the Morris method reported by Pijanowski et al. [32]. All measurements were taken three times.

#### 3.3.2. Determination of Sugar by HPLC

An analysis of sugar by the HPLC-ELSD method was performed according to the protocol described by Oszmia´nski and Lachowicz [33]. Calibration curves (*R* <sup>2</sup> = 0.9999) were created for glucose, fructose, sorbitol, and sucrose. All data were obtained in triplicate. The results were expressed as g per 100 g dm.

#### 3.3.3. Polyphenolic Compounds by UPLC

The extraction of the samples for phenolic compounds and their chromatographic analysis were performed exactly as described by Oszmia´nski and Lachowicz (2016). The samples were analyzed by an Ultra-Performance Liquid Chromatography Photodiode Array Detector (UPLC-PDA; Acquity UPLC System, Waters Corp., Milford, MA, USA). The study identified phenolic acids and their sums were calculated as chlorogenic acid, which is based on dominant compounds and compared with reference standards (Figure S1). All results were taken in triplicate and shown as mg/100 g dm (dry mass) of sample.

#### *3.4. Physical Parameters*

#### 3.4.1. Color Parameters by CIE Lab System

The color properties (L \*, a \*, b \*) of prepared products were determined by reflectance measurement with a Color Quest XE HunterLab colorimeter (Biosens, Warsaw, Poland). The samples were filled in a 1-cm cell, and L\*, a\*, b\* values were determined using Illuminant D65 and 10◦ observer angle. Samples were measured against a white ceramic reference plate (L \* = 93.92; a \* = 1.03; b \* = 0.52). The total change in color of powders (DE \*) [34] and also EP, and Dom WL were measured. The data were the mean of three measurements.

#### 3.4.2. Water Activity

Water activity (aw) was measured in triplicate (*n* = 3) at 25 ◦C ± 2 using an AQUA LAB DewPoint water activity meter (Pullman, WA, USA) [26].

#### *3.5. Pro-Helathy Properties*

#### 3.5.1. The Antioxidants Activity

The extraction procedure of the radical activity (ABTS), reducing potency (FRAP), and the oxygen radical absorbance capacity (ORAC) test was the same for all determinations and was carried out identically, as described by Lachowicz, Swieca, and Pejcz [ ´ 35], and Nowicka, Wojdyło, Laskowski [36]. The FRAP, ABTS, and ORAC tests were prepared as previously described by Benzie and Strain [37], Re et al. [38], and Ou et al. [39], respectively. The antioxidant capacity was expressed as millimoles of Trolox per 100 g of sample. The ORAC assay was carried out on an RF-5301 PC spectrofluorometer (Shimadzu, Kyoto, Japan). Measurements by means of ABTS, and FRAP method involved a UV-2401 PC spectrophotometer (Shimadzu, Kyoto, Japan).

#### 3.5.2. Activity of α-Amylase, α-Glucosidase, Pancreatic Lipase Inhibitors

The α-amylase and α-glucosidase inhibitory effect of the sample extracts was assayed according to the procedure described previously by Nowicka, Wojdyło, Samoticha [36] while the inhibition of lipase activity was determined according to Pods˛edek et al. [40], respectively. Acarbose was included in the case of α-amylase and α-glucosidase as a positive control, while the orlistat was used as a positive control for pancreatic lipase. The results were expressed as IC<sup>50</sup> value.

#### *3.6. Statistics Methods*

Statistical analysis was conducted using Statistica version 13 (StatSoft, Krakow, Poland). Significant differences (*p* ≤ 0.05) between means were evaluated by two-way ANOVA and Tuckey's multiple range test. All data included in this study are presented as the mean value ± standard deviation and were performed at least three times.

#### **4. Conclusions**

The analyses of the content of polyphenolic compounds and the health-promoting activity of selected purée variants showed that the best results were obtained in JA dried with FD. It has been proved that not only the drying method but also the type of raw material used for drying played a significant role in modulating the antioxidant, antidiabetic, and anti-obesity properties of the tested material. Hence, much better effects regarding the contents of pectin and health-promoting compounds as well as modeling health-promoting properties were obtained by drying the raw rather than the previously cooked material. The best effects of preserving the natural light color after production were obtained in the freeze-dried samples; however, due to the high costs of this method, it can be replaced by vacuum drying.

**Supplementary Materials:** Supplementary Materials Figure S1: chromatograms registered at 280 nm of dried purée from Jerusalem artichoke: with derivatives of SO<sup>2</sup> (1), without derivatives of SO<sup>2</sup> (2).

**Author Contributions:** Conceptualization, J.O., P.R., S.L., P.N.; material, P.R. methodology, J.O., S.L., P.N.; formal analysis, J.O., S.L., P.N.; T.C.; writing—original draft preparation, J.O., S.L., P.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** The research is financed under the Operational Program: Smart Growth: Priority Axis 2 Support for the environment and the potential of enterprises to conduct R & D & I activity, measure 2.3 Pro-innovative services for enterprises, sub-measure 2.3.2. The project is carried out by FNT Food and Technology Sp. z o.o.–Wrocław, Poland. Number of project: POIR.02.03.02-02-0027/18.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The authors declare that the data is available.

**Conflicts of Interest:** The authors declare no conflict of interest.

**Sample Availability:** Samples of the compounds as JA are available from the authors.

#### **References**


## *Communication* **Antioxidant Activity of Polyphenols, from** *Mauritia flexuosa* **(Aguaje), Based on Controlled Dehydration**

**Hichem Bensaada <sup>1</sup> , María Fernanda Soto-Garcia <sup>2</sup> and Juan Carlos Carmona-Hernandez 2,\***


**Abstract:** Plant polyphenols offer several benefits for the prevention of diverse illnesses. Fruit's edible and inedible parts (pulp, seeds, peels, stems, flowers) are important sources of polyphenols. Different industrial processes for fruit treatment and commercialization affect the total polyphenol content (TPC), and probably the biological activity. The purpose of the present work was to determine the TPC and antioxidant activity (by DPPH) of polyphenols extracted from the pulp and seeds of *Mauritia flexuosa* (aguaje), in fresh and dehydrated forms, in order to determine the possible connection with the quantity of polyphenols and their specific antioxidant activity. The highest phenolic content for *M. flexuosa* seeds in fresh form (non-dehydrated) was 270.75 mg GAE/100 g with a 96-h extraction. With respect to the dehydrated samples, the best yield was quantified in the 96-h dehydrated seed sample. For all pulp and seeds, dehydrated for 24, 48, and 96 h, TPC showed a slightly decreasing pattern. The DPPH results were the highest in the 96-h dehydrated samples and the differences among all dehydrated pulp and seed samples were minimal. More studies testing the presence of other antioxidant components could help in understanding the detailed antioxidant activity, and related more to the specific action, rather than only total polyphenol content.

**Keywords:** polyphenols; antioxidant activity; *Mauritia flexuosa* (aguaje); controlled dehydration

#### **1. Introduction**

The antioxidant benefit of fruit consumption is mainly connected to the polyphenol content; these types of phytochemicals display several other functions [1,2]. Multiple publications have reported improved lipid metabolism in overweight and obese humans due to a regular diet inclusion of plant polyphenols [3–5]. These phytochemicals offer a protective activity, leading to health benefits; several studies relate oxidative stress with the development of diseases such as cardiovascular diseases, neurodegenerative disorders, and cancer [6,7]. With respect to ovarian cancer, there is histological evidence of aortas showing promising anti-proliferative and anti-inflammatory effects and leading to a reduction of cancer cell viability [8,9]. Given the multiple health benefits of polyphenols, it is of general interest to know the phenolic content and major phenolic compounds, such as flavonoids, found in regularly harvested and consumed fruits [10–12].

The oxido-reducing activity of compounds containing phenolic rings is the most studied biological property in plant polyphenols. This type of chemical reaction is representative of a misbalance in oxidizing and reducing compounds; which at the cellular level can lead to molecular damage [13]. Plants and fruits can contribute to the reduction of negative side effects in different pathologies.

Commonly known as buriti plant, *Mauritia flexuosa* (Arecaceae) is a palm broadly cultivated in Colombia, Venezuela, the Guianas, Trinidad, Ecuador, Peru, Brazil, and Bolivia [2,14]. The fresh form of *M. flexuosa* fruits (Figure 1) has an orange, soft, watersoluble, and edible pulp and numerous small circular dark red and brown flat seeds. Several

**Citation:** Bensaada, H.; Soto-Garcia, M.F.; Carmona-Hernandez, J.C. Antioxidant Activity of Polyphenols, from *Mauritia flexuosa* (Aguaje), Based on Controlled Dehydration. *Molecules* **2022**, *27*, 3065. https:// doi.org/10.3390/molecules27103065

Academic Editors: Jan Oszmianski and Sabina Lachowicz-Wi´sniewska

Received: 3 March 2022 Accepted: 28 April 2022 Published: 10 May 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

South American aboriginal groups use this Amazonic fruit for medicinal purposes [15–19]. The mesocarp oil is used to treat respiratory symptoms, pneumonia, influenza, snake bites, and heart problems [20]. In Colombia, *M. flexuosa* is also named "aguaje" [2].

**Figure 1.** (**a**) *Mauritia flexuosa* (aguaje) palm and fruit (seedless pulp) [2]. (**b**) Dehydrated aguaje samples, seeds and pulp.

Different variables can affect the total polyphenol content (TPC) and the antioxidant capacity of fruits, such as passing from a fresh to a dehydrated long-shelf-life state. This is still an area of active research, while conflicting results have been reported in the literature [21–23]. Information is lacking with respect to the specific activity and bio-accessibility of polyphenols, especially due to structural alterations, such as ring modifications, and polyphenol interaction with other food matrixes based on high or low water contents [24,25]. Multiple studies have focused on testing total phenolic content and antioxidant activity in *M. flexuosa*, but a comparison of the influence in dehydration states in pulp or seeds and its connection to antioxidant activity is still lacking in information and research [26–29]. The contradicting results show that some studies highlight decreasing TPC and higher antioxidant activity, while others report the opposite situations tested in a different food matrix, such as grapes, onions, rice, and other vegetables [30–33]. The goal of the present study was to determine a comparative approach to polyphenols from *M. flexuosa*, with high and low water content (fresh and de-hydrated samples), under active interconversion, and their relation to effective antioxidant activity.

#### **2. Results and Discussion**

#### *2.1. Food Matrix Dehydration and Polyphenol Extraction*

Samples of *M. flexuosa* pulp and seeds were treated by extraction and a controlled dehydration process at four different times (24, 48, 72, and 96 h). The biggest weight difference was recorded for the first 24 h of dehydration. The water content in pulp and seeds was approximately 85% and 44%, respectively. Dehydrated weights were stable at 96 h at 50 ◦C. Considering the high difference in water content for pulp and seeds, the dehydration pattern (Figure 2) yielded similar values.

Polyphenols from fresh pulp and seeds were extracted in ethanol (80% *w*/*v*) and more polyphenol extractions were done with seeds and pulp at three dehydration intervals (24, 48, and 96 h). The comparative extraction looked for differences in polyphenol content, based on conflicting studies that reported higher phenolics and antioxidant activity in fresh fruit, seeds, or peels [34,35]. Figure 3 shows the comparative values for total polyphenol content from pulp and seeds in fresh or dehydrated form.

**Figure 2.** Dehydration curve for *M. flexuosa*, pulp and seeds at a constant 50 ◦C from 0 to 96 h.

**Figure 3.** TPC (mg of gallic acid equivalents per 100 g) of fresh and dehydrated pulp and seeds of (**a**) *M. flexuosa* (aguaje) fresh and dehydrated pulp and seeds for the initial experimental assay and (**b**) TPC in *M. flexuosa* dehydrated pulp and seeds in a second assay. Data are means (±standard deviation), and lower-case letters represent significant differences, for total polyphenol content in mg GAE/100 g of initial sample, based on ANOVA followed by Tukey test (*p* < 0.05).

#### *2.2. Total Polyphenol Content (TPC)*

Polyphenols that were initially extracted from the fresh pulp of aguaje at 0 h of dehydration and the samples of fresh pulp that continued with the extraction process for another 96 h, in the dark without stirring, registered the lowest TPC with statistical significance, with respect to the phenolic content in fresh and dehydrated (96 h) seeds, registering approximately 90 mg GAE/100 g. The highest TPC was detected in fresh and dehydrated aguaje seeds (dehydration at 96 h) yielding more than 270 mg GAE/100 g in fresh seeds, as shown in Figure 3a that represents the results for the first experimental assays. The best aqueous conditions for polyphenol detection and reactions (fresh pulp matrix) did not favor higher TPC. These results suggest, coinciding with other researchers, that higher phenolic presence is more connected to the food matrix and the solvent affinity due to polyphenol polarity [35]. Nevertheless, in comparison with other fruits, we registered a high TPC; other studies reported *M. flexuosa*, with TPC values of 435.08 ± 6.97 and 362.90 ± 7.98 mg GAE/100 g of the whole pulp in fresh form [29,36]. The results here

differed with studies where the TPC in fresh fruit, after food processing, was lowered or lost due to temperature or processing changes [31,32].

For the second experimental assay, concerning the TPC results for dehydrated *M. flexuosa* pulp and seeds, as shown in Figure 3b, a regular decreasing TPC pattern from 24 to 96 h of dehydration was registered. TPC values at the three different dehydration points were higher than the TPC in fresh pulp and seeds; the highest TPC was quantified at 24 h of dehydrating pulp and seeds at a constant 50 ◦C. The results at 48 and 96 h yielded slightly lower TPC values with statistical significance. The lowest phenolic content was measured in dehydrated pulp at 96 h (149.28 ± 0.81 mg GAE/100 g). This decreasing trend in phenolic concentration is consistent with research done with the purpose of evaluating TPC values in fruits processed with long effective storage time for exportation purposes [31].

#### *2.3. Antioxidant Activity, DPPH (2,2-Diphenyl-1-picrylhydrazyl) Radical Scavenging Assays*

The results for the antioxidant action of phenolic compounds, extracted in ethanol 80%, in a fresh and dehydrated matrix at different times for *M. flexuosa* (aguaje) samples are shown in Figure 4, showing the comparative results for two different experimental assays. The total inhibitory concentration, IC 50%, that yielded the best output was due to phenolics in dehydrated seeds at 96 h (78.28 ± 0.67 mg AAE/100 g of dehydrated sample). All DPPH results were higher in the extractions from fresh or dehydrated seeds. In both assays, Figure 4a,b, the antioxidant activity registered an increasing tendency. The best DPPH results, with statistical significance, were detected in *M. flexuosa* pulp and seeds dehydrated for 96 h.

**Figure 4.** Antioxidant activity (mg AAE/100 g FP) of extracts in two different experimental assays. (**a**) fresh and dehydrated *M. flexuosa*) and (**b**) dehydrated pulp and seeds at different times. Data are means (±standard deviation) and lower-case letters represent significant differences, for antioxidant activity of aguaje extracts according to DPPH (IC 50%), based on ANOVA followed by a Tukey test (*p* < 0.05).

These results for antioxidant activity coincide with the assays, in the same fruit, where other parts of the plant were evaluated [26,37]. The dehydration approaches in the present work led to improved DPPH values, with respect to the previously cited studies. The results are comparable and coincide with studies applying more complex dehydration processes [38,39]. Figure 5 represents a summary display of the dehydration process with respect to TPC and antioxidant activity seen in *M. flexuosa* pulp and seed extracts, as a comparative base for the various other studies.

**Figure 5.** TPC (mg of gallic acid equivalents) and DPPH (mg ascorbic acid equivalents) in *M. flexuosa* (aguaje) pulp and seeds with respect to a constant dehydration pattern from 0 to 96 h. The stable flattening dehydration pattern is connected to lower TPC values and higher redox activity. From 72 to 96 h, the weight differences were stable, and statistical significance was registered at different points in the dehydration process.

Studies evaluating the biological activity of polyphenols in red raspberries showed that dehydration led to a lowering polyphenol content and less antioxidant activity, while the rate of this reduction was connected to the dehydration method, and this showed that a high dehydration temperature was linked to polyphenol content loss [40,41]. Increasing temperature promotes higher solubility and diffusion coefficient of polyphenolic compounds into the extracting solvents, and higher temperatures also enhance the penetration of solvents into the cell matrix; hence, increasing the TPC of the extracts [41].

Furthermore, the results confirmed a positive correlation between the total polyphenol content and the antioxidant activity for the aglycone compounds undergoing the drying process (74.7%). These lowering TPC values may be due to oxidative and thermal degradation of the phenolic compounds [42,43]. The major phenolic content drop, during fruit processing at high temperatures in enzymatic reactions, could be related to the action of oxidative enzymes such as polyphenol oxidases (PODs) and polyphenol peroxidases (PPOs) [43,44]. PODs can enhance the degradation of phenols when coexisting with PPOs; both the PPO and POD enzymatic activities play a key role in determining the phenolic profile of olive oil. In contrast, low temperatures during the dehydration process decrease the oxidation of volatile compounds [45].

When Cabernet Sauvignon and Merlot grapes were dried using a constant low temperature (7 ◦C) for several weeks, there was an increase in TPC and antioxidant activity [46]. This could be due to the effect of the concentration of the phenolic compounds as a result of water loss caused by the dehydration, even if the constant temperature (40 ◦C) was higher [47]. The results in these two previous studies partially coincide with the findings in the present work, with respect to the higher antioxidant activity of polyphenols in dehydrated aguaje fruit samples, but differ with respect to the decreasing phenolic content during d the dehydration process.

In a study evaluating the redox activity of phenolics from goldenberry, the levels of TPC and antioxidant activity (determined with the ferric reducing antioxidant power FRAP method) increased in dehydrated samples [20]. Furthermore, the highest TPC was registered in samples that were dried at 90 ◦C, coinciding with the results in this study, where samples were dehydrated at lower temperature (50 ◦C). The interconversion of phenolic compounds at high temperatures might be caused by the availability of phenolic precursor molecules through the non-enzymatic rearrangement between phenolic molecules [48]. This higher phenolic content may originate from the disruption of cell walls during processing or the breakdown of insoluble phenolic compounds. Therefore, this could lead to a better extractability for these particular types of phenolic compounds [49].

These comparative studies suggest that the antioxidant activity could be due to the combined reactions of total phenolics rather that certain individual components or the action of polyphenols as a whole group. The results in this study, using fresh and dehydrated aguaje seeds and pulp, coincide with the findings of Gupta et al. (2021), testing different parts of pomelo fruit (*Citrus grandis* (L.) Osbeck) in similar experiments for antioxidant activity. They found that TPC was highest in the membrane of the fruit, and DPPH registered the highest activity in the pomelo juice [50]. Considering the total polyphenols in the present study, the improvement in TPC could be a result of the destruction of the covering structure and the release of more phenolic compounds, facilitating and increasing the extraction yield [48]. Moreover, the dehydration process could induce metabolic pathways that can generate, and increase the number of, precursors for different categories of phenolic compounds [49].

#### **3. Materials and Methods**

#### *3.1. Reagents and Chemicals*

Ethanol, sodium carbonate, and Folin–Ciocalteu reagent were purchased from PanReac AppliChem, ITW Reagents, (Darmstadt, Germany), and methanol, from Sigma Aldrich (St. Louis, MO, USA). Ascorbic acid (Sebion) and DPPH (2,2-Diphenyl-1-picrylhydrazyl, Sigma-Aldrich) were purchased from Merck KGaA (Darmstadt, Germany).

#### *3.2. Plant Material, Sample Preparation, and Polyphenol Extractions*

*M. flexuosa* pulp with seeds was obtained in the Colombian city of Leticia. All samples were refrigerated before laboratory analyses. The extraction of total polyphenol compounds was performed following a previous method from this research group [2,51]. Different amounts of fresh and dehydrated pulp and seeds of *M. flexuosa* were placed in ethanol (80% *w*/*v*) at proportions of 1:5 per volume, stirred for 15 min at 500 rpm, homogenized, and stored at room temperature, in the dark, for 24 h without stirring. Two different extractions times (—*al fresco*—0 h and 96 h of extraction) for the first experimental assay, as an explorative comparison, and three different dehydration times (24, 48, and 96 h) at a constant of 50 ◦C, including only samples in dry state as a second assay, were considered in this experimental process. The extracts were centrifuged for 10 min at 3500 rpm, and the supernatant was recovered for polyphenol quantitation and antioxidant activity evaluation.

#### *3.3. Total Polyphenol Content (TPC)*

The total polyphenol content in *M. flexuosa* pulp and seeds (fresh and dehydrated matrix) was quantified following the Folin–Ciocalteu (F–C) assay [51]. Samples of 1 mL of each extract were mixed with 1 mL F–C reagent (10% *w*/*v*), allowed to react for 2 min, and mixed with 2 mL sodium carbonate, Na2CO3, (3.5% *w*/*v*). Reactants were kept in the dark at room temperature for 90 min. All runs were performed in triplicate. Absorbance was read at 655 nm in a UV-Vis spectrophotometer (Mecasys Optizen POP, Daejeon, Korea). All data were calculated based on a gallic acid standard calibration curve, with a range of 0–4.0 mg/L and r<sup>2</sup> of 0.9982). TPC is expressed as milligrams gallic acid equivalent (GAE) per 100 g of fresh or dehydrated sample (mg GAE/100 g).

#### *3.4. DPPH Assay for Radical Scavenging Antioxidant Activity*

The DPPH radical scavenging test is one of the most useful techniques to evaluate the antioxidant activity in polyphenols extracted from natural products. The DPPH compound is a stable free radical in methanol. The DPPH assay was performed following previous work from this research group [2]. Volumes of 1900 µL of DPPH (100 µM) prepared in pure methanol were mixed with 100 µL of each diluted (1:5) extract and left to react in the dark at room temperature for 30 min. The antioxidant activity from phenolics, in fresh and dehydrated pulp and seeds, of *M. flexuosa* was measured via spectrophotometry at 517 nm, comparing against a methanol blank. A positive control of ascorbic acid based on a calibration curve, and in triplicates for each colorimetric reaction, was applied in this methodology. The control curve was prepared with concentrations of comparable reference ascorbic acid (Merck KGaA, Darmstadt, Germany) in a concentration range from 50 to 600 µg/mL (r<sup>2</sup> of 0.9945). All dilutions followed the same DPPH reaction conditions for the antioxidant activity evaluated in fruit extracts. The slope taken from the calibration curve served as the calculation of the inhibition concentration (IC 50%), when 50% of the antioxidant component was reduced. The results for IC 50% were determined based on the equation:

% scavenging DPPH free radical = (ABSControl − ABSExtracts/ABSControl) × 100%

The antioxidant activity of *M. flexuosa* extracted phenolics is expressed as mg of ascorbic acid equivalents per 100 g of fresh or dehydrated pulp or seeds (mg AAE/100 g).

#### *3.5. Statistical Analysis*

All analyses were carried out in triplicate, and TPC and DPPH values are expressed as mean ± standard deviation (SD). Means were tested for normality and homogeneity. Data were analyzed based on a ANOVA test followed by Tukey test (*p* < 0.05) with the IBM SPSS Statistics software version 20.0 (IBM Corp., Armonk, NY, USA).

#### **4. Conclusions**

The total polyphenol content and antioxidant activity of *M. flexuosa* pulp and seeds, in fresh and dehydrated form, were tested in this work. The water content in a food matrix allows for a specific polyphenol oriented chemical reaction, yielding better results in some cases where the water content is higher. A controlled dehydration process was considered in this experimental approach, with the purpose of evaluating the polyphenol availability and antioxidant action. TPC values were not directly proportional to antioxidant activity, suggesting that the polyphenol reactions for radical scavenging in pulp and seeds of *M. flexuosa* do not depend directly in the specific quantity of phenolic compounds, but rather on the specific chemical structure or on its re-accommodation or interconversion. More studies based on the specific polyphenol/flavonoid content and the presence of other antioxidants, such as C vitamin, in *M. flexuosa* could lead to understanding more of the specific antioxidant activity of this fruit with multiple processes of it edible parts.

**Author Contributions:** Conceptualization, formal analysis, data curation and writing—review and editing, H.B., M.F.S.-G. and J.C.C.-H.; methodology, H.B., M.F.S.-G. and J.C.C.-H.; supervision, J.C.C.- H.; project administration, H.B. and J.C.C.-H.; funding acquisition, J.C.C.-H. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** The authors thank all laboratory members, especially Carmen Serna Hurtado, at Universidad de Manizales (Colombia) for their constant support and help.

**Conflicts of Interest:** The authors declare no conflict of interest.

**Sample Availability:** Samples of the compounds are not available from the authors.

## **References**


## *Article* **Isoflavone Changes in Immature and Mature Soybeans by Thermal Processing**

**Shanshan Qu 1,†, Soon Jae Kwon 2,† , Shucheng Duan <sup>1</sup> , You Jin Lim <sup>1</sup> and Seok Hyun Eom 1,\***


**Abstract:** The isoflavone changes occurring in mature soybeans during food processing have been well studied, but less information is available on the changes in immature soybeans during thermal processing. This study aimed to determine the effect of thermal processing by dry- or wet-heating on the changes in the isoflavone profiles of immature and mature soybeans. In the malonylglycoside forms of isoflavone, their deglycosylation was more severe after wet-heating than after dry-heating regardless of the soybean maturity. The malonyl forms of isoflavones in the immature seeds were drastically degraded after a short wet-heating process. In the acetylglycoside forms of isoflavone, dry-heating produced relatively low amounts of the acetyl types in the immature soybeans compared with those in the mature soybeans. These results were explained by the content of acetyldaidzin being relatively less changed after dry-heating immature soybeans but increasing four to five times in the mature soybeans. More of the other types of acetylglycoside were produced by dry-heating soybeans regardless of their maturity. Acetylgenistin in wet-heating was a key molecule because its content was unchanged in the immature soybeans during processing but increased in the mature soybeans. This determined the total acetylglycoside content after wet-heating. In contrast, most of the acetyl forms of isoflavone were produced after 90 to 120 min of dry-heating regardless of the seed maturity. It can be suggested that the pattern of isoflavone conversion was significantly affected by the innate water content of the seeds, with a lower water content in the mature soybeans leading to the greater production of acetyl isoflavones regardless of the processing method even if only applied for a relatively short time. The results suggested that the isoflavone conversion in the immature soybeans mainly follows the wet-heating process and can be promoted in the application of stronger processing.

**Keywords:** isoflavone conversion; thermal process; immature seeds; mature seeds; internal water content

#### **1. Introduction**

Soybeans (*Glycine max* L.) are one of the most widely consumed legumes in the world. As well as their main role in providing protein, carbohydrates, and oil, soybeans are also a rich source of phytochemicals, particularly isoflavones [1,2]. The content of isoflavones, a type of flavonoid, is greater in soybeans than in other legumes [3]. The 12 major isoflavones in soybeans can be classified into four main forms: aglycones (daidzein, glycitein, and genistein); β-glycosides (daidzin, glycitin, and genistin); acetylglycosides (acetyldaidzin, acetylglycitin, and acetylgenistin); and malonylglycosides (malonyldaidzin, malonylglycosides, and malonylgenistin) [4,5]. Of these isoflavone groups, malonylglycosides are the predominant form in raw soybeans, followed by β-glycosides and acetylglycosides, with aglycones rarely observed [6]. Epidemiological studies have reported that the presence of different types of isoflavone in soybeans contributes to various biological activities,

**Citation:** Qu, S.; Kwon, S.J.; Duan, S.; Lim, Y.J.; Eom, S.H. Isoflavone Changes in Immature and Mature Soybeans by Thermal Processing. *Molecules* **2021**, *26*, 7471. https:// doi.org/10.3390/molecules26247471

Academic Editors: Jan Oszmianski and Sabina Lachowicz

Received: 23 November 2021 Accepted: 8 December 2021 Published: 10 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

such as reducing the risk of cancer, cardiovascular disease, and osteoporosis, and relieving menopausal symptoms [7–12]. Among the four forms of isoflavones, the bioavailability of malonyl-conjugated isoflavones was lower than that of corresponding non-conjugated β-glycoside isoflavones [13], and that of isoflavone aglycons was highest because aglycones were more easily and quickly absorbed by the intestine [14]. Furthermore, some studies reported that non-conjugated glycoside isoflavones also possessed high-quality antioxidant activity similar to aglycones. Thus, soybeans with a high content of non-conjugated glycosides and aglycones had high-quality antioxidant activity [15].

Because of their grassy-beany flavor and bitter taste [16,17], raw mature soybeans are mainly consumed after thermal processing, such as boiling or roasting, which greatly improves the flavor of the soybeans and soy products [18–20]. It has also been reported that thermal processing causes the conversion or degradation of isoflavones [21–24]. The predominant isoflavone form (malonylglycosides) is usually converted into acetylglycosides and β-glycosides by thermal processing [25–30], with the patterns of conversion depending significantly on the severity of heating. In general, the increase in the contents of isoflavone acetylglycosides and β-glycosides occurs through the decarboxylation and deesterification of malonylglycosides, respectively [28,31]. The wet-heating method promotes deesterification more than dry-heating as it causes the rapid conversion of isoflavone malonylglycosides to β-glycosides rather than acetylglycosides [32,33]. Chien et al. [34] reported no changes in the glycosides and aglycones during dry-heating below 150 ◦C, but Huang and Chou [35] reported a decrease in the aglycones in soybeans steamed for 30 min at temperatures above 60 ◦C.

The changes in isoflavone content during thermal processing in mature soybeans have been widely reported, but those in immature soybeans have rarely been studied. Immature soybeans, also known as edamame or maodou, are harvested when the green seeds fill the pod and have a water content of 60 to 65% [36,37]. Immature soybeans are also rich in isoflavones. The difference in the isoflavone content between immature and mature soybeans is significantly affected by the cultivar. Simonne et al. [38] investigated the iso-flavone contents of immature and mature beans of five soybean cultivars and found that the immature beans contained twice the isoflavone content of the mature beans in four of the cultivars, whereas Kim et al. [39] reported that the mature soybeans of several cultivars contained more isoflavone than the immature soybeans. Immature soybeans also exhibit a beany off-flavor and unique taste compared with mature soybeans [40]. Traditionally, in East Asia, immature soybeans have also been consumed as vegetables and snacks after thermal processing, such as boiling [40]. Simonne et al. [38] studied the influence of processing methods, such as boiling and freeze-drying, on the distribution of isoflavones in immature soybeans, but the patterns of variation in isoflavone content in immature soybeans during thermal processing are still unclear, particularly compared with those in mature soybeans.

Therefore, the objectives of this study are: to compare the effects of thermal processing (dry- and wet-heating) on variations in the isoflavone profiles of immature and mature soybeans, and to determine the effect of the internal water content of soybeans on the isoflavone content during thermal processing by varying the soaking time of mature seeds before dry-heating. This study will provide basic information on utilizing isoflavones in soybeans with different levels of maturity.

#### **2. Results and Discussion**

#### *2.1. Physiological Characteristics of Immature and Mature Soybeans*

The physical characteristics of the raw immature and mature soybeans are shown in Table 1. The water contents of the immature and mature seeds were 66.87% and 13.04%, respectively. Takahashi et al. [37] reported that, in immature soybeans, the water contents of four different cultivars ranged from 61.5% to 76.8%, similar levels to those of the present study. It has also been reported that the moisture content of mature soybeans was 10– 15%, a similar content to the present study [41,42]. The dry weight of 10 immature seeds

(0.66 g) was significantly lower than that of 10 mature seeds (0.88 g) (Table 1). As expected, a remarkable difference in the SSC of the immature (2.75 ◦Brix) and mature seeds (15.76 ◦Brix) was also observed (Table 1). Sale and Campbell [43] reported that a dramatic reduction in the water content of soybeans was accompanied by a steady accumulation of dry matter and SSC during seed maturation from the R6 (immature) to the R8 (mature) stages. Figure 1 shows the morphologies of the immature and mature soybeans before and after thermal processing. After dry-heating, the immature and mature soybeans had shrunk, particularly the immature soybeans, whereas they had both swelled after wet-heating, particularly the mature seeds. A significant variation in seed color (green to yellow) was also observed, particularly for the immature soybeans, possibly because of the Maillard reaction [44]. a remarkable difference in the SSC of the immature (2.75 °Brix) and mature seeds (15.76 °Brix) was also observed (Table 1). Sale and Campbell [43] reported that a dramatic reduction in the water content of soybeans was accompanied by a steady accumulation of dry matter and SSC during seed maturation from the R6 (immature) to the R8 (mature) stages. Figure 1 shows the morphologies of the immature and mature soybeans before and after thermal processing. After dry-heating, the immature and mature soybeans had shrunk, particularly the immature soybeans, whereas they had both swelled after wet-heating, particularly the mature seeds. A significant variation in seed color (green to yellow) was also observed, particularly for the immature soybeans, possibly because of the Maillard reaction [44].

respectively. Takahashi et al. [37] reported that, in immature soybeans, the water contents of four different cultivars ranged from 61.5% to 76.8%, similar levels to those of the present study. It has also been reported that the moisture content of mature soybeans was 10%– 15%, a similar content to the present study [41,42]. The dry weight of 10 immature seeds (0.66 g) was significantly lower than that of 10 mature seeds (0.88 g) (Table 1). As expected,

**Table 1.** Physical characteristics of immature and mature soybean seeds. **Table 1.** Physical characteristics of immature and mature soybean seeds.

*Molecules* **2021**, *26*, x FOR PEER REVIEW 3 of 15


SSC means soluble solids content. Different letters (a, b) in a column indicate significant differences at *p* < 0.05 by Tukey's studentized range (HSD) test. Results are given as mean ± SE (*n* = 10). SSC means soluble solids content. Different letters (a, b) in a column indicate significant differences at *p* < 0.05 by Tukey's studentized range (HSD) test. Results are given as mean ± SE (*n* = 10).

**Figure 1.** Morphologies of immature and mature soybeans after dry- and wet-heating for different times. **Figure 1.** Morphologies of immature and mature soybeans after dry- and wet-heating for different times.

#### *2.2. The Variation in the Total Isoflavone Content and Four Isoflavone Forms of Soybeans during 2.2. The Variation in the Total Isoflavone Content and Four Isoflavone Forms of Soybeans during Thermal Processing*

*Thermal Processing* Many studies have reported the effect of thermal processing on the isoflavone profiles of mature soybeans or soy products [25,28,30,32]. However, studies comparing the thermal transformation or degradation of isoflavones in soybeans at different maturity levels are still limited. The variations in the total isoflavone content (TI) and in the other four forms of isoflavone (total isoflavone malonylglycosides (TIMG); total isoflavone acetylglycosides (TIAG); total isoflavone β-glycosides (TIG); and total isoflavone agly-Many studies have reported the effect of thermal processing on the isoflavone profiles of mature soybeans or soy products [25,28,30,32]. However, studies comparing the thermal transformation or degradation of isoflavones in soybeans at different maturity levels are still limited. The variations in the total isoflavone content (TI) and in the other four forms of isoflavone (total isoflavone malonylglycosides (TIMG); total isoflavone acetylglycosides (TIAG); total isoflavone β-glycosides (TIG); and total isoflavone aglycones (TIA)) in immature and mature soybeans during thermal processing are shown in Figure 2.

cones (TIA)) in immature and mature soybeans during thermal processing are shown in Figure 2. Before thermal treatment (freeze-dried, FD), the TI of the mature soybeans (304.51 mg/100 g DW) was 1.6 times higher than that of the immature soybeans. The isoflavone Before thermal treatment (freeze-dried, FD), the TI of the mature soybeans (304.51 mg/ 100 g DW) was 1.6 times higher than that of the immature soybeans. The isoflavone form with the highest content in both the immature and mature soybeans was TIMG, followed by TIG, with small amounts of TIAG and TIA (Figure 2). These results were consistent with those of Kim et al. [39], who reported that the TI of mature soybeans was higher than that of immature soybeans, with malonylglycoside isoflavones being the predominant form in both immature and mature soybeans, followed by the β-glycoside, acetylglycoside, and aglycone isoflavone forms.

form with the highest content in both the immature and mature soybeans was TIMG, followed by TIG, with small amounts of TIAG and TIA (Figure 2). These results were consistent with those of Kim et al. [39], who reported that the TI of mature soybeans was higher than that of immature soybeans, with malonylglycoside isoflavones being the predominant form in both immature and mature soybeans, followed by the β-glycoside,

acetylglycoside, and aglycone isoflavone forms.

**Figure 2.** Changes in the total isoflavone (TI), total isoflavone malonylglycosides (TIMG), total isoflavone acetylglycosides (TIAG), total isoflavone β-glycosides (TIG), and total isoflavone aglycones (TIA) contents of immature and mature soybeans during dry-heating (**A**) and wet-heating (**B**). Data are shown as mean ± SE (*n* = 3). Different letters indicate significant differences at *p* < 0.05 by Tukey's studentized range (HSD) test. **Figure 2.** Changes in the total isoflavone (TI), total isoflavone malonylglycosides (TIMG), total isoflavone acetylglycosides (TIAG), total isoflavone β-glycosides (TIG), and total isoflavone aglycones (TIA) contents of immature and mature soybeans during dry-heating (**A**) and wet-heating (**B**). Data are shown as mean ± SE (*n* = 3). Different letters indicate significant differences at *p* < 0.05 by Tukey's studentized range (HSD) test.

After thermal processing, the TI content tended to decrease regardless of the seed maturity and processing method. The TI content decreased significantly after wet-heating: in mature soybeans from 304.51 to 163.49 mg/100 g DW; in immature soybeans from 199.34 to 44.63 mg/100 g DW; compared with dry-heating: in mature soybeans to 200.32 mg/100 g DW; and in immature soybeans to 147.15 mg/100 g DW). This may have been caused by the difference in humidity between wet- and dry-heating affecting the transfer of thermal energy [34,45], which accelerates the decrease in TI content. This decrease in the TI content of thermally treated soybeans or soy products has been widely reported [13,29,38]. However, the present study has been the first to compare changes in TI content in soybeans at different maturity levels and after different processing methods. The significant decrease in the TIMG content was considered to be the main reason for the reduction in the TI content (Figure 2). As mentioned previously, malonylglycoside isoflavones are the predominant form in soybeans. Similar to the TI content, the content of TIMG in both the immature and mature soybeans tended to decrease after thermal treatment, and soybeans with a higher internal water content under wet-heating conditions promoted this decreasing trend (Figure 2(A-2,B-2)). The relative thermal instability of malonyl-conjugated β-glycoside isoflavones has been reported previously [34]. Wet-heating degraded malonyl isoflavones more than dry-heating because of the higher moisture content in wet-After thermal processing, the TI content tended to decrease regardless of the seed maturity and processing method. The TI content decreased significantly after wet-heating: in mature soybeans from 304.51 to 163.49 mg/100 g DW; in immature soybeans from 199.34 to 44.63 mg/100 g DW; compared with dry-heating: in mature soybeans to 200.32 mg/ 100 g DW; and in immature soybeans to 147.15 mg/100 g DW). This may have been caused by the difference in humidity between wet- and dry-heating affecting the transfer of thermal energy [34,45], which accelerates the decrease in TI content. This decrease in the TI content of thermally treated soybeans or soy products has been widely reported [13,29,38]. However, the present study has been the first to compare changes in TI content in soybeans at different maturity levels and after different processing methods. The significant decrease in the TIMG content was considered to be the main reason for the reduction in the TI content (Figure 2). As mentioned previously, malonylglycoside isoflavones are the predominant form in soybeans. Similar to the TI content, the content of TIMG in both the immature and mature soybeans tended to decrease after thermal treatment, and soybeans with a higher internal water content under wet-heating conditions promoted this decreasing trend (Figure 2(A-2,B-2)). The relative thermal instability of malonyl-conjugated β-glycoside isoflavones has been reported previously [34]. Wet-heating degraded malonyl isoflavones more than dry-heating because of the higher moisture content in wet-heat process [32,45].

heat process [32,45]. The decrease in the TIMG content, usually accompanied by an increase in the TIAG content in soybeans processed by thermal treatments, such as roasting [46] and baking [47], or in processed soy products, such as cooked soybeans [48] and soy milk [31], has been widely investigated. The increase in the TIAG content was caused by the decarboxylation of malonyl isoflavones during heat processing. Similarly, in mature soybeans, significant increases in the TIAG content after thermal processing were also observed (Figure 2(A-3,B-3)), with dry-heating increasing its content from 3.09 to 41.47 mg/100 g DW, thus being more effective than wet-heating, which increased its content from 3.09 to 22.23 mg/100 g DW. The difference in the effect of dry- and wet-heating on TIAG content can The decrease in the TIMG content, usually accompanied by an increase in the TIAG content in soybeans processed by thermal treatments, such as roasting [46] and baking [47], or in processed soy products, such as cooked soybeans [48] and soy milk [31], has been widely investigated. The increase in the TIAG content was caused by the decarboxylation of malonyl isoflavones during heat processing. Similarly, in mature soybeans, significant increases in the TIAG content after thermal processing were also observed (Figure 2(A-3,B-3)), with dry-heating increasing its content from 3.09 to 41.47 mg/100 g DW, thus being more effective than wet-heating, which increased its content from 3.09 to 22.23 mg/100 g DW. The difference in the effect of dry- and wet-heating on TIAG content can be explained by the maximum degradation rate of malonyl to acetyl isoflavones under the dry-heat condition, while a low conversion rate of malonyl to acetyl isoflavones was found under the wet-heat condition [34]. It indicated that wet-heating tended to convert or degrade malonylglycosides to other isoflavone derivatives instead of acetylglycosides, while the decarboxylation of malonyl isoflavones also occurred simultaneously. However, immature soybeans showed the reverse behavior in response to dry- and wet-heating: like the mature soybeans, dry-heating increased the TIAG content of immature beans from 5.20 to 19.36 mg/100 g DW, whereas wet-heating significantly decreased the TIAG content from 5.20 to 2.04 mg/100 g DW. As there has been no comparative study on the variation

in the TIAG content in immature soybeans under dry- and wet-heating conditions, the present study is the first to report that the variation in the TIAG content of immature soybeans after thermal processing is different from that of mature beans. This may have been caused by the higher internal water content of immature soybeans significantly promoting the transfer of energy under wet-heating compared with dry-heating conditions. This led to the rapid conversion of acetylglycosides to β-glycosides, or the degradation of acetylglycosides [34], a speculation also supported by Huang and Chou [35], who reported that the acetyl isoflavones content of soaked black soybeans decreased after steaming at 100 ◦C for 30 min. Under dry-heat conditions, the TIAG content of the mature soybeans increased significantly compared with the immature soybeans (Figure 2(A-3)). This could be interpreted as the variation in the TIAG content of immature soybeans with their higher internal water content under dry-heating being similar to that of mature soybeans under wet-heating.

Thermal processing also increased the TIG content by deesterifying malonylglycoside and acetylglycoside isoflavones [46–50]. The patterns of variation in the TIG content were significantly affected by the processing methods and level of seed maturity (Figure 2(A-4,B-4)). In mature soybeans, the effect of wet-heating, increasing the TIG content from 25.16 to 91.64 mg/100 g DW, was better than that of dry-heating, where it increased from 25.16 to 59.00 mg/100 g DW. This can be partly explained by the deesterification of acetylglycoside isoflavones to form β-glycoside isoflavones under wet-heating (Figure 2(A-3,B-3)). Huang and Chou [35] have also reported that the TIG content of mature soybeans increased as the temperature of the thermal treatment increased. In contrast, the opposite patterns were observed for immature soybeans. The TIG content after dryheating, ranging from 23.40 to 57.30 mg/100 g DW, was higher than that after wet-heating, which ranged from 23.40 to 45.26 mg/100 g DW. As mentioned earlier, the high internal moisture content of the soybeans under wet-heating may have led to excessive energy transfer, leading to the further conversion or degradation of the β-glycoside isoflavones. Thermal processes, such as oven drying [46], baking [47], frying [48], steaming [35], and autoclaving [50], have been reported as important methods for increasing the TIG content of mature soybeans. However, the effect of dry- and wet-heating on the variation in the TIG content of immature soybeans has not been studied. The results suggest that, unlike mature soybeans, dry-heating is more suitable for processing immature soybeans and leads to a higher TIG content than wet-heating.

As mentioned earlier, TIA accounts for only a small proportion of the TI in soybeans. Figure 2(A-5,B-5) shows the variation in the TIA content at different maturity levels of soybeans after thermal processing. The TIA content remained stable in the mature soybeans during dry-heating but decreased significantly during wet-heating. Similar results on mature soybeans have also been observed by Kao et al. [51] and Aguiar et al. [50] for dry- and wet-heating, respectively. The isoflavone deglycosylation from glycoside form to aglycone was observed only under high temperature, possibly because it was difficult to break down the glycoside groups to form aglycones at relatively low temperatures [52]. The present study has shown that the variation in the TIA content of the immature soybeans was similar to that of mature soybeans. In this study, because we used only one cultivar, further studies using numerous cultivars are required to understand the isoflavone deglycosylation patterns of soybeans regarding whether the isoflavone changes among immature soybean cultivars by thermal processing are presenting the same patterns or not.

#### *2.3. Correlation Analysis between Isoflavone Form and Corresponding Individual Isoflavones in Immature and Mature Soybeans*

To clarify the patterns of how isoflavones changed for different maturity levels of soybeans during thermal processing, the correlations between the contents of the isoflavone form and corresponding individual isoflavones were analyzed. Tables 2 and 3 show the variations in the content of 12 individual isoflavones (IMG (MDZI, MGLI, and MGNI), IAG (ADZI, AGLI, and AGNI), IG (DZI, GLI, and GNI), and IA (DZE, GLE, and GNE)) in immature and mature soybeans, respectively, after dry- and wet-heating. The profiles

Aglycones

GLI 7.53 ± 1.96 a,b 7.71 ± 0.86 <sup>b</sup>

GNI 8.64 ± 1.99 <sup>e</sup> 16.83 ± 0.58 <sup>d</sup> 29.26 ± 1.04 <sup>c</sup>

of 12 individual isoflavones detected by the reversed-phase high-performance liquid chromatography (HPLC) in immature and mature soybeans before treatment and after 120 min of thermal treatment are shown in Figure 3. GNE 0.94 ± 0.14 <sup>a</sup> 0.61 ± 0.08 a,b,c 0.70 ± 0.07 a,b,c 0.90 ± 0.12 <sup>a</sup> a,b 0.27 ± 0.06 <sup>c</sup> 0.30 ± 0.02 b,c 0.40 ± 0.10 b,c 0.45 ± 0.10 b,c Data are shown as mean with standard error (*n* = 3). Different letters (a–e) next to data in a row indicate significant differences at *p* < 0.05 by Tukey's studentized range (HSD) test. N.D. indicates not detected; Tr. Indicates trace amount.

21.07 ± 0.18 <sup>d</sup>

14.82 ± 1.82 a,b 17.30 ± 2.20 <sup>a</sup>

30.52 ± 0.71 b,c

0.38 ± 0.03 <sup>b</sup> 0.29 ± 0.15 <sup>b</sup> 0.31 ± 0.16 <sup>b</sup> 0.57 ± 0.13 a,b

13.97 ± 0.97 a,b

36.74 ± 0.91 <sup>b</sup> 47.44 ± 1.62 <sup>a</sup>

16.75 ± 0.57 <sup>a</sup>

*Molecules* **2021**, *26*, x FOR PEER REVIEW 7 of 15

10.37 ± 0.98 a,b

31.11 ± 2.08 b,c

0.59 ± 0.09 a,b

0.72 ± 0.04

GLE Tr. N.D. N.D. N.D. N.D. 0.38 ± 0.02 <sup>a</sup> N.D. N.D. N.D.

14.47 ± 1.09 a,b

33.99 ± 0.87 b,c

15.05 ± 0.43 a,b

**Figure 3.** Twelve isoflavones of immature and mature soybean seeds after 120 min of dry- and wetheat treatment determined by HPLC. 1, daidzin (DZI); 2, glycitin (GLI); 3, genistin (GNI); 4, malonyldaidzin (MDZI); 5, malonylglycitin (MGLI); 6, acetyldaidzin (ADZI); 7, acetylglycitin (AGLI); 8, malonylgenistin (MGNI); 9, daidzein (DZE); 10, glycitein (GNE); 11, acetylgenistin (AGNI); 12, genistein (GNE). **Figure 3.** Twelve isoflavones of immature and mature soybean seeds after 120 min of dry- and wet-heat treatment determined by HPLC. 1, daidzin (DZI); 2, glycitin (GLI); 3, genistin (GNI); 4, malonyldaidzin (MDZI); 5, malonylglycitin (MGLI); 6, acetyldaidzin (ADZI); 7, acetylglycitin (AGLI); 8, malonylgenistin (MGNI); 9, daidzein (DZE); 10, glycitein (GNE); 11, acetylgenistin (AGNI); 12, genistein (GNE).

Of the three types of malonylglycoside isoflavones, MGNI (78.18 mg/100 g DW in immature soybeans; 141.25 mg/100 g DW in mature soybeans) and MDZI (62.32 mg/100 g DW in immature soybeans; 119.20 mg/100 g DW in mature soybeans) were the main isoflavones in the FD samples (Tables 2 and 3). The deglycosylation of MGNI, MDZI, and MGLI was more severe after wet-heating than after dry-heating regardless of the level of the seed maturity. Significantly positive correlations (*p* < 0.001) between the contents of MGNI, MDZI, and MGLI individually and of TIMG were observed in all the samples (Table 4). The contents of the three types of malonyl isoflavone in the immature soybeans did not change after 30 min of dry-heating but drastically decreased after 30 min of wet-heating. Similar results have also been observed by Chien et al. [34], where moist-heating at 100 °C reduced the MGNI (standard compound) content more than dry-heating at 100 °C. Of the three types of malonylglycoside isoflavones, MGNI (78.18 mg/100 g DW in immature soybeans; 141.25 mg/100 g DW in mature soybeans) and MDZI (62.32 mg/ 100 g DW in immature soybeans; 119.20 mg/100 g DW in mature soybeans) were the main isoflavones in the FD samples (Tables 2 and 3). The deglycosylation of MGNI, MDZI, and MGLI was more severe after wet-heating than after dry-heating regardless of the level of the seed maturity. Significantly positive correlations (*p* < 0.001) between the contents of MGNI, MDZI, and MGLI individually and of TIMG were observed in all the samples (Table 4). The contents of the three types of malonyl isoflavone in the immature soybeans did not change after 30 min of dry-heating but drastically decreased after 30 min of wet-heating. Similar results have also been observed by Chien et al. [34], where moist-heating at 100 ◦C reduced the MGNI (standard compound) content more than dry-heating at 100 ◦C.

*Molecules* **2021**, *26*, 7471


**Table 3.** Isoflavone content (mg/100 g DW) in mature soybeans during thermal processing.


65

trace amount.


**Table 4.** Correlation coefficients between isoflavone forms and individual isoflavones in thermally treated soybeans.

<sup>z</sup> and <sup>y</sup> indicate dry-heating and wet-heating, respectively. \*, \*\*, and \*\*\* indicate significances at *p* < 0.05, *p* < 0.01, and *p* < 0.001 in Tukey's HSD test. ns indicates no significance at the test. Data of 12 individual isoflavones, TIMG, TIAG, TIG, and TIA, for correlation analysis were calculated by the time-dependent dry-heating and wet-heating values, ranging from 0 to 120 min treatments.

Of the acetylglycoside isoflavones, ADZI (5.20 mg/100 g DW in immature soybeans; 3.09 mg/100 g DW in mature soybeans) was the main acetyl isoflavone in the FD samples (Tables 2 and 3). Only a trace amount of AGNI was detected in both the immature and mature soybeans, with no AGLI being detected. The three types of acetyl isoflavone in the soybeans of different maturity levels responded differently to thermal processing. The ADZI content changed relatively little in the dry-heated immature soybeans but increased by four to five times in the dry-heated mature soybeans. The ADZI content also decreased to undetectable levels in the wet-heated immature soybeans, but not in the wet-heated mature soybeans. More AGLI and AGNI were produced in the dry-heated soybeans regardless of the maturity level. The AGLI content of the immature soybeans did not change but increased slightly after wet-heating. The AGNI in wet-heating was a key molecule because its content was unchanged in the immature soybeans after processing but increased in the mature soybeans, thus determining the amount of total acetylglycosides after wetheating. The increase in the AGNI content was greater than that in the contents of ADZI and AGLI, possibly because of the varying thermal stability of the three acetylglycoside isoflavones and the corresponding malonylglycoside isoflavones. In contrast, the content of most of the types of acetyl increased up to 90 to 120 min of dry-heating regardless of the seed maturity. Significant positive correlations were found between the contents of ADZI, AGLI, AGNI, and that of TIAG for all the treatment groups except for the wet-heated immature soybeans (Table 4). A high negative correlation between the contents of AGNI and TIAG was observed for wet-heated immature soybeans, unlike the significant positive correlations for the other samples. This indicated that the amount of total acetylglycosides depended mainly on the content of AGNI during thermal processing.

The β-glycosides, the non-conjugated form of isoflavones, are the second major group after malonylglycosides in raw soybeans [52,53]. Of the β-glycosides, the content of

GLI (11.62 mg/100 g DW) was higher than that of DZI (3.71 mg/100 g DW) and GNI (8.08 mg/100 g DW) in the FD immature soybeans (Table 2), similar to results from Simonne et al. [38]. The content of GLI (7.53 mg/100 g DW) was relatively lower than that of DZI (7.56 mg/100 g DW) and GNI (8.64 mg/100 g DW) in the FD mature soybeans (Table 3), results that are consistent with those of Kim et al. [54]. During thermal processing, different patterns of variation in the contents of GLI, GNI, and DZI arose. In the mature soybeans, the increase in the GLI content was small after heating compared with a significant increase in the DZI and GNI contents, similar to results reported by Toda et al. [48]. In the immature soybeans, no significant differences (*p* > 0.05) in GLI content were observed between FD and thermally treated soybeans. The contents of DZI and GNI both increased significantly after thermal treatment regardless of the seed maturity and processing method. Wetheating was also more efficient in increasing β-glycoside isoflavones than dry-heating. The contents of GNI and DZI of all the treated samples were significantly positively correlated (*p* < 0.001) with the TIG content (Table 4). A good correlation (*p* < 0.01) between the GLI and TIG contents during thermal processing was found in the mature soybeans but not in the immature soybeans (*p* > 0.05). This indicated that the patterns of variation in the TIG content of soybeans were dominated more by the contents of DZI and GNI than the content of GLI during thermal processing even though the GLI content was relatively high in both FD samples.

In the FD soybeans, only small amounts of DZE (immature soybeans, 0.77 mg/100 g DW; mature soybeans, 1.21 mg/100 g DW) and GNE (immature soybeans, 0.22 mg/100 g DW; mature seeds, 0.94 mg/100 g DW) were detected, with a trace amount of GLE. The contents of the three aglycones in the immature soybeans were relatively stable under dry-heating but decreased to undetectable levels after 90 min of wet-heating. Both types of thermal processing decreased the contents of the three aglycones in the mature soybeans, particularly wet-heating. The variations in the contents of the three aglycones with the temperatures of thermal processing have been contradictory: Xu et al. [52] reported that aglycones in soybean flour extracts were generated with heat treatments above 135 ◦C, but Huang and Chou [35] reported that the contents of GNE, DZE, and GLE in black soybeans decreased at steaming temperatures of 60 ◦C or above for 30 min. A good correlation between the contents of DZE and TIA was found in all the treatments except for the dryheated immature soybeans (Table 4). However, only the GLE content was well correlated with the TIA content in the immature soybeans under dry-heating with the content of GNE being well correlated with the TIA content in the mature soybeans under wet-heating.

The three isoflavone types (daidzein, glycitein, and genistein) showed different conversion patterns under heat processing. The MDZI and MGNI decreased more rapidly in the initial 30 min than the MGLI. Moreover, the production of AGNI and GNI by heat processing were higher than that of other isoflavone types. Glycitein conjugate types had relatively low thermal-change compared to daidzein and genistein types regardless of the seed maturity. Similar results have been reported by Stintzing et al. [55], where glycitein carrying a meth-oxy group at 6 position of A-ring has higher stability upon dry-heating. Moreover, Mathias et al. [56] reported that the heat-induced loss of daidzein glycosides was higher than that of genistein glycosides. These results indicate that different deglycosylation rates among isoflavone types occur during different thermal process methods.

#### *2.4. Verification of the Relationship between Soybean Water Content and Changes in Patterns of Isoflavone Contents*

It is important to note the different patterns of variation in the contents of acetylglycoside and β-glycoside isoflavones during the dry- and wet-heating of soybeans at two maturity levels. The internal moisture content of the soybeans affected the composition of isoflavones during thermal processing. To confirm this assumption, fully mature soybeans were soaked in distilled water for 0, 1, 2, 4, and 8 h to obtain different internal water contents, then dry-heated, followed by further observations of the patterns of variation in isoflavone content after heating for 1 h. Figure 4A shows the variations in the moisture content of the soybeans after soaking. The water content of the soybeans gradually increased

from 6.34% to 57.04% as the soaking time increased. The contents of TI and of the four forms of isoflavone, TMIG, TAIG, TIG, and TIA, in the fully mature soybeans before and after the 1-h dry-heat treatment are shown in Figure 4B, C, D, E, and F, respectively. creased from 6.34% to 57.04% as the soaking time increased. The contents of TI and of the four forms of isoflavone, TMIG, TAIG, TIG, and TIA, in the fully mature soybeans before and after the 1-h dry-heat treatment are shown in Figure 4B, C, D, E, and F, respectively.

*2.4. Verification of the Relationship between Soybean Water Content and Changes in Patterns of* 

It is important to note the different patterns of variation in the contents of acetylglycoside and β-glycoside isoflavones during the dry- and wet-heating of soybeans at two maturity levels. The internal moisture content of the soybeans affected the composition of isoflavones during thermal processing. To confirm this assumption, fully mature soybeans were soaked in distilled water for 0, 1, 2, 4, and 8 h to obtain different internal water contents, then dry-heated, followed by further observations of the patterns of variation in isoflavone content after heating for 1 h. Figure 4A shows the variations in the moisture content of the soybeans after soaking. The water content of the soybeans gradually in-

*Molecules* **2021**, *26*, x FOR PEER REVIEW 10 of 15

*Isoflavone Contents*

**Figure 4.** Water content of mature seeds during serial soaking (**A**), and isoflavone contents in soaked seeds after freeze-drying (FD) and dry-heating (**B**–**F**). A, water content; B, total isoflavone (TI); C, total isoflavone malonylglycosides (TIMG); D, total isoflavone acetylglycosides (TIAG); E, total isoflavone β-glycosides; F, total isoflavone aglycones. Data are shown as mean with standard error (*n* = 3). Asterisks indicate statistically significant differences (\*, *p* < 0.05; \*\*, *p* < 0.01; \*\*\*, *p* < 0.001; ns indicates no significance). **Figure 4.** Water content of mature seeds during serial soaking (**A**), and isoflavone contents in soaked seeds after freeze-drying (FD) and dry-heating (**B**–**F**). (**A**), water content; (**B**), total isoflavone (TI); (**C**), total isoflavone malonylglycosides (TIMG); (**D**), total isoflavone acetylglycosides (TIAG); (**E**), total isoflavone β-glycosides; (**F**), total isoflavone aglycones. Data are shown as mean with standard error (*n* = 3). Asterisks indicate statistically significant differences (\*, *p* < 0.05; \*\*, *p* < 0.01; \*\*\*, *p* < 0.001; ns indicates no significance).

Before heating, no significant differences in the TI content were observed between the unsoaked and soaked soybeans. Wang and Murphy [24] have also reported that soaking at room temperature for 10 to 12 h significantly increased the moisture content from 11.03% to 63.23% and retained the TI of the soaked soybeans. Heating significantly decreased the content of TI and TIMG and increased the content of TIAG and TIG of the soybeans compared with the FD samples (Figure 4B, C, D, and E). The highest amount of TIAG generated was found in the unsoaked soybeans (0 h), and the lowest amount in soybeans soaked for a long time. There were no significant variations in the TIG content between the unsoaked and soaked soybeans after dry-heating for 1 h. These results were consistent with this report that the immature and mature beans had a similar TIG content after a long period of dry-heating, with even immature soybeans showing a lower TIAG content than mature soybeans (Figure 2). In contrast, the TIA content of the unsoaked soybeans was reduced by dry-heating, but, the longer the soaking time, the more TIA was generated after heating (Figure 4F). These results were different from the results we reported before, which may have been caused by differences between natural soybeans with a higher internal water content and artificially made soybeans with a higher water content. This also confirmed the assumption that the internal moisture content of soybeans Before heating, no significant differences in the TI content were observed between the unsoaked and soaked soybeans. Wang and Murphy [24] have also reported that soaking at room temperature for 10 to 12 h significantly increased the moisture content from 11.03% to 63.23% and retained the TI of the soaked soybeans. Heating significantly decreased the content of TI and TIMG and increased the content of TIAG and TIG of the soybeanscompared with the FD samples (Figure 4B–E). The highest amount of TIAG generated was found in the unsoaked soybeans (0 h), and the lowest amount in soybeans soaked for a long time. There were no significant variations in the TIG content between the unsoakedand soaked soybeans after dry-heating for 1 h. These results were consistent with this report that the immature and mature beans had a similar TIG content after a long period of dry-heating, with even immature soybeans showing a lower TIAG content than mature soybeans (Figure 2). In contrast, the TIA content of the unsoaked soybeans was reduced by dry-heating, but, the longer the soaking time, the more TIA was generated after heating (Figure 4F). These results were different from the results we reported before, which may have been caused by differences between natural soybeans with a higher internal water content and artificially made soybeans with a higher water content. This also confirmed the assumption that the internal moisture content of soybeans was an important factor affecting the different patterns of variation in isoflavone content in soybeans of different maturity.

It is notable that, the longer the soaking time, the less TIAG was produced after heating. Lee and Lee [46] reported that the content of acetyl isoflavones in soybeans soaked for 12 h did not change during 120 min of oven drying but that, in unsoaked soybeans, it increased significantly after roasting at 200 ◦C. This indicated that soybeans with a higher internal water content produced a lower amount of acetyl isoflavones after heating, which confirmed the previous assumption that the water content of soybeans significantly affected the pattern of isoflavone conversion. Therefore, the differences between the content of acetyl isoflavones in mature and immature soybeans after heating were caused by the difference in the internal water content. The increase in the content of aglycone isoflavones was highly related to soaking and heating, results similar to those of Lima et al. [57], who found no significant difference in the content of aglycones in soybeans soaked at 25 ◦C but a significant increase after soaking for 1 h at 70 ◦C.

#### **3. Materials and Methods**

#### *3.1. Chemical Reagents*

The HPLC-grade acetonitrile and water (Daejung Chemical & Metals Co., Siheung, Korea) were used as mobile phases for isoflavones analysis. Standards of isoflavone aglycones (daidzein, glycitein, and genistein) and β-glycosides (daidzin, glycitin, and genistin) were purchased from LC Laboratories (Woburn, MA, USA). Isoflavone acetylglycosides (acetyldaidzin, acetylglycitin, and acetylgenistin) were obtained from Nacalai tesque (Kyoto, Japan), and malonylglycosides (malonyldaidzin, malonylglycitin, and malonylgenistin) were obtained from GenDEPOT (Katy, TX, USA).

#### *3.2. Soybean Cultivation*

Soybean seeds (*Glycine max* L. cv. *Pungwon*) used in this study were provided by the Pulmuone Food Co. (Chungbuk, Korea). The cultivar 'Pungwon' was registered to the Korea Seed & Variety Service (Gimcheon, Korea) in 2007 and had earlier matu-ration period and high content of isoflavones (more information described in Oh et al. [58]). The soybeans were germinated for 24 h at room temperature in a dark culture room after soaking with distilled water for 4 h. The germinated soybeans were planted in a horticultural soil (Baroker, Seoulbio Co., Eumseong, Korea) in pots (Plastic pot, 24 × 27 × 18 cm) in early June 2020 and then grown in the greenhouse of Kyung Hee University (Yongin, Korea) under natural sunlight. The average of temperature during the soybean growing season was 18–22 ◦C in June; 22–30 ◦C in July and August; 19–26 ◦C in September; 12–23 ◦C in October (based on Korean meteorological administration data). The average solar radiation period was 14 h/day in June to August and 12.5 h/day in September and October (based on Korean meteorological administration data). The average of relative humidity was 45–55% from June to October. The potted soybeans were maintained with several irrigations per week in the early stage of soybean plants and with daily irrigation in the period of seed formation. The soybean seeds were harvested at the immature stage on September 10 when the pods of soybeans contained green seeds that filled the pod cavity and harvested at the mature stage on October 10 when 95% of the pods exhibited the light brown color with dehydrating, as shown in Figure 1 in a previous report [36].

#### *3.3. Physical Characteristics of Immature and Mature Seeds*

The harvested soybean samples were weighed before and after freeze-drying. The dry weight of the immature and mature soybeans was expressed as the weight (g) of 10 raw seeds based on the mean value of ten replicates. The water content (%) was calculated as follows: 100 × [fresh weight (g) − dry weight (g)]/[fresh weight (g)]. Fresh soybeans (0.3 g) were ground with a pestle and a mortar and added 0.6 mL of distilled water to measure the soluble solid content (SSC). After stirring the mixture, the sample was centrifuged at 14,240× *g* for 15 min. The SSC of the supernatant was evaluated using a hand refractometer (Atago Co., Tokyo, Japan) and expressed as degree of Brix (◦Brix).

#### *3.4. Thermal Treatment*

The immature and mature soybean seeds were processed using three thermal processing methods: (1) freeze-drying (FD) at −80 ◦C for 72 h in a vacuum freeze-dryer (IlshinBioBase. Co. Ltd., Dongducheon, Korea) and stored in a −20 ◦C refrigerator; (2) dryheating at 100 ± 3 ◦C for 30, 60, 90, and 120 min with a convective dryer (Koencon Co., Ltd., Hanam, Korea); and (3) wet-heating (steaming) at 100 ± 3 ◦C for 30, 60, 90, and 120 min with a steam cooker. All experiments were carried out in triplicate. The thermally treated samples were freeze-dried and stored in a −20 ◦C refrigerator before isoflavone analysis.

#### *3.5. Extraction of Isoflavones*

All samples were finely ground using a commercial grinder (JL-1000, Hibell, Hwaseong, Korea). The isoflavones extraction was performed by previously described method [59]. Briefly, 20 mg of ground sample mixed with 58% aqueous acetonitrile (1 mL, *v*/*v*) in a shaking incubator for 24 h at 25 ◦C and 120 rpm after sonication for 30 min. The supernatant was obtained after centrifuging at 14,240× *g* for 5 min. Then, two-fold volume of distilled water was added to dilute the supernatant. The diluted supernatant was filtered through a 0.45 µm hydrophilic PTFE membrane syringe filter (Futecs Co., Ltd., Daejeon, Korea) and used for isoflavones analysis.

#### *3.6. Determination of Isoflavones*

Extracts were analyzed using HPLC (Waters 2695 Alliance HPLC; Waters Inc., Milford, MA, USA) with the octadecylsilane column (Prontosil 120–5-C18-SH-EPS 5.0 µm (200 × 4.6 mm; Bischoff, Leonberg, Germany). According to the previously published method [59], the solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile) were used as mobile phase with the flow rate of 0.8 mL/min. The mobile phase B gradient was as follows: 16–25%, 0 to 35 min; 25–50%, 35 to 40 min; 50–65%, 40 to 47 min; 65–16%, 47 to 50 min. The injection volume was 5 µL. The peaks of 12 standard isoflavones were detected at 254 nm (Water 996 photodiode array detector (Waters Inc.)).

#### *3.7. Statistical Analysis*

All experiments were carried out in triplicate with the data expressed as the mean with standard error (*n* = 3). Analysis of variance was performed using SAS software (Enterprise guide 7.1 version, SAS Institute Inc., Cary, NC, USA). Significant differences between experimental treatments were evaluated using Tukey's student range test, with a significance level defined at *p* < 0.05.

#### **4. Conclusions**

This is the first study to compare the patterns of isoflavone changes between soybeans at two maturity levels after thermal processing. Overall, the patterns of the isoflavone changes in the soybeans depended significantly on the soybean maturity and the processing method, which affected the decarboxylation of malonylglycoside or the deesterification of acetylglycoside isoflavones. The decreases in the TI in all the samples were mainly caused by the decrease in the TIMG during thermal processing. The deglycosylation of the three types of malonyl isoflavones was more severe in wet- than in dry-heating regardless of the seed maturity. In the acetylglycoside isoflavones, dry-heating produced a relatively low amount of acetyl isoflavones in the immature seeds compared with that in the mature seeds. The ADZI was relatively less changed in the dry-heated immature seeds but increased significantly in the processed mature seeds. AGLI and AGNI were produced in greater amounts in the dry-heated samples regardless of the seed maturity. The AGNI in wet-heating was the key molecule because its content remained unchanged in the immature soybeans during processing but increased in the mature soybeans, which determined the total amount of acetylglycoside in wet-heating. Wet-heating increased the amount of β-glycoside isoflavones in the mature soybeans more than in dry-heating, while, interestingly, the immature soybeans exhibited the opposite behavior. The aglycone isoflavones were stable under dry-heating, but their contents decreased significantly after wet-heating. The internal moisture content of the soybeans was an important factor affecting the deglycosylation of isoflavones during thermal processing, also confirmed by the verification experiment (Section 2.4). This is the first study to highlight the importance of the internal water content of soybeans on the distribution of isoflavones during thermal processing. The results of the present study will provide basic information on the different uses of immature and mature soybeans after thermal processing.

**Author Contributions:** Conceptualization, S.H.E.; methodology, S.Q., S.J.K. and S.H.E.; software, S.Q.; validation, S.H.E.; formal analysis, S.Q. and S.D.; investigation, S.Q.; resources, S.H.E.; data curation, S.Q. and Y.J.L.; writing—original draft preparation, S.Q.; writing—review and editing, S.J.K. and S.H.E.; visualization, S.Q. and S.D.; supervision, S.H.E.; funding acquisition, S.H.E. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by Radiation Technology R&D program (NRF-2017M2A2A6A050- 18538) through the National Research Foundation of Korea funded by the Ministry of Science and ICT. This work was also supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (NO. NRF-2019R1A2C1009623).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Conflicts of Interest:** The authors declare no conflict of interest.

**Sample Availability:** Samples of the compounds are not available from the authors.

#### **References**


## *Article* **Influence of Freeze-Dried Phenolic-Rich Plant Powders on the Bioactive Compounds Profile, Antioxidant Activity and Aroma of Different Types of Chocolates**

**Dorota Zy˙ zelewicz ˙ 1,\* , Joanna Oracz <sup>1</sup> , Martyna Bilicka <sup>1</sup> , Kamila Kulbat-Warycha <sup>1</sup> and Elzbieta Klewicka ˙ 2**


**Abstract:** In this study, the blueberries (BLUB), raspberries (RASB), blackberries (BLCB), pomegranates pomace (POME) and beetroots (BEET) freeze-dried powders were used as the sources of phenolic compounds to enrich different types of chocolates, substituting a part of the sweetener. It was found that 1% addition of fruit or vegetable powders to chocolates increased the content of total phenolic compounds (flavan-3-ols, phenolic acids and anthocyanins) of enriched dark and milk chocolates compared to the control ones dependent on the powder used. Among the enriched chocolates, the chocolates with the addition of BLUB powder were characterized by the highest total polyphenol content. The highest percentage increase (approximately 80%) in the total polyphenol content was observed in MCH chocolate enriched with BLUB powder. Chocolates incorporated with BLUB, RASB, BLCB and POME powders presented a richer phenolic compound profile than control counterparts. The highest DPPH radical-scavenging capacity was exhibited by the DCH98S chocolate enriched with BEET powder. However, the DCH98ESt chocolates enriched with POME and BEET powders demonstrated the highest FRAP values. An electronic nose analysis confirmed the existence of differences between the profiles of volatile compounds of various types of chocolates enriched with fruit or vegetable powders. Thus, the enrichment of dark and milk chocolates with BLUB, RASB, BLCB, POME and BEET powders seemed to be an interesting approach to enhance bioactivity and to enrich the sensory features of various chocolate types.

**Keywords:** antioxidants; chocolate; free radical scavenging activity; reducing power; functionalization of food; electronic nose analysis

#### **1. Introduction**

In recent years, a rapid development of research on bioactive substances present in plant raw materials and their impact on the human body has been observed. Numerous studies revealed that there was a significant upward trend for the use of the plant-derived natural compounds as antioxidants and functional ingredients. Growing interest in the production and consumption of functional foods with specific pro-health characteristics results from the rationales indicating a close relationship between the consumption of food rich in natural antioxidants and the prevention of degenerative diseases. The main dietary sources of antioxidants are fruits and vegetables, as well as their derived products. The natural antioxidants occurring in plant materials are mainly phenolic compounds, carotenoids, and vitamins [1]. Many of these natural antioxidants, especially phenolic compounds, demonstrate pro-healthy properties. It was confirmed that the presence of these compounds in food plays an important role in the prevention of many civilization diseases, in particular cancer, cardiovascular diseases, as well as diabetes and rheumatoid

**Citation:** Zy˙ zelewicz, D.; Oracz, J.; ˙ Bilicka, M.; Kulbat-Warycha, K.; Klewicka, E. Influence of Freeze-Dried Phenolic-Rich Plant Powders on the Bioactive Compounds Profile, Antioxidant Activity and Aroma of Different Types of Chocolates. *Molecules* **2021**, *26*, 7058. https:// doi.org/10.3390/molecules26227058

Academic Editors: Jan Oszmianski and Sabina Lachowicz

Received: 11 October 2021 Accepted: 19 November 2021 Published: 22 November 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

arthritis [1–3]. However, the biological activity and bioavailability of these compounds highly depend on the molecular weight and chemical structure of these compounds, the matrix of food and their concentration in the consumed products [4].

Generally, cocoa bean and its derived products, such as chocolate and cocoa powder are a good source of natural antioxidants, especially phenolic compounds [5–8]. However, the production of chocolate may result in the loss of up to 80% of phenolic compounds originating from cocoa beans [9]. The largest and the most diverse group of phenolic compounds found in cocoa beans are flavonoids [5,10–12]. The predominant phenolic compounds in raw cocoa beans are proanthocyanidins (58%), monomeric flavan-3-ols (37%) and anthocyanins (4%) [10,11,13]. However, the level of these substances in cocoa beans can vary greatly and depends on a number of factors, mainly the variety, the geographical and environmental conditions during growth, the degree of ripeness, the harvest time, as well as the conditions and duration of storage after harvesting [14–16]. Many studies have shown that the processing of cocoa beans in order to obtain chocolate significantly influences the phenolic compounds content, and thus their activity and bioavailability [10,17–21]. All processing steps, such as fermentation, drying, alkalization and roasting, in addition to beneficial physicochemical, microbiological, and organoleptic changes, cause significant degradation of phenolic compounds [17,18]. During fermentation and drying of the cocoa beans, the phenolic compounds undergo biochemical transformations leading to a reduction in their content. Monomeric flavan-3-ols are enzymatically oxidized to semi-quinones and quinones. Anthocyanins, which are highly unstable and susceptible to various degradation reactions such as enzymatic or non-enzymatic browning, in the presence of the glycosidases are hydrolyzed to anthocyanidins and monosaccharides, mainly arabinose and galactose [13,15,19,22]. Phenolic compound content decreases with thermal processing. Heat exerted during the roasting causes degradation of these substances through non-enzymatic oxidation and polymerization reactions of proteins and protein hydrolysis products, as well as amino acids, polysaccharides and Maillard reaction products leading to the formation of insoluble macromolecular complexes [21,23]. As a result, the concentration of phenolic compounds in raw cocoa beans differs significantly from that in the roasted beans and chocolate [21,24–26].

In recent years, a healthy lifestyle has been promoted. It is mainly manifested by caring for health and taking up physical activity in order to ensure a longer life in a good physical and mental condition. One of the manifestations of caring for health is eating food that is not only of the highest nutritional value, but also has specific health-promoting properties. However, food processing can lead to a reduction in the concentration or even complete loss of many valuable bioactive ingredients positively influencing human health. Hence the efforts of food and nutrition technologists, dieticians and, consequently, many conscious food producers to functionalize (design food products) the aim of which is not only to provide all nutrients, but also to positively influence the body's functions by providing compounds that reduce the risk of developing certain diseases, especially the so-called civilization diseases, or providing compounds that generally improve health and well-being. In the case of chocolate production, the content of bioactive ingredients, such as phenolic compounds, is also reduced. Already at the stage of fermentation of cocoa beans in plantations, their content is reduced by over 50%. Especially anthocyanins and (−)-epicatechin are highly degraded. Drying the beans after fermentation causes a further reduction in the concentration of these compounds. Some authors [5] state that up to 90% of (−)-epicatechin is lost as a result of fermentation and drying of cocoa beans. Further processes leading to obtaining chocolate, especially roasting cocoa beans, only increase the loss of polyphenols. That is why scientists, food technologists and dieticians see the need to enrich food, including chocolate, with an additional portion of bioactive compounds, including antioxidants, and the functional food market is constantly growing and not only small enterprises with niche production but also industrial giants are investing in it [5,16]. The current scientific literature on the subject describes few proposals in this area, starting from the introduction of an additional portion of polyphenols in pure form, in the form

of cocoa liquor from raw beans, cocoa liquor or cocoa powder with increased polyphenol content [27–29] to the addition of spices [30], leaves [31,32] or pomaces [33]. These studies have shown that in this way the polyphenol content and antioxidant activity of chocolates can be increased and their sensory profile can be influenced. There are commercially available chocolates with low or unprocessed fruits, e.g., almonds, nuts, raisins, but their shelf life is shorter than chocolate without such additives, due to, among other things, a higher water content in semi-finished fruit products (raisins) or a high content of unsaturated fatty acids (nuts, almonds) [34]. The introduction of freeze-dried fruit or vegetable powder into chocolate recipes eliminates this inconvenience. To the authors' knowledge, no studies have been carried out in terms of the aromatic profile, polyphenol content and antioxidant activity of different kinds of chocolates enriched with freeze-dried fruit or vegetable powders. For this purpose, we used several types of berries (blueberries, raspberries, blackberries), pomegranate pomace and beetroots and we assumed the aim of the presented research to investigate the effect of phenolic-rich plant powders fortification on the bioactive compounds profile, antioxidant activity and aroma (measured by an objective method using an electronic nose) of different types of chocolates. Berries and pomegranates contain significant amounts of flavonoids, mainly anthocyanins, and possess many beneficial properties for human health [35]. Anthocyanins are considered as strong antioxidants and free radical scavengers that have an antioxidant potential twice as high as that of other antioxidants, such as (+)-catechin and α-tocopherol [36]. Therefore, these fruits could be used as a source of phenolic substances with pronounced antioxidant activity. Powdered dried fruits (including berries) and vegetables are new and valuable ingredients for the chocolate industry. European Union regulations allow the introduction of up to 40% of additional foodstuffs to chocolate recipes [37]. In this context, the enrichment of chocolates with freeze-dried fruit (berries, pomegranate) or vegetable (beetroots) powder, which is a rich source of natural antioxidants, including anthocyanins, can greatly improve functional and health-promoting properties of these products. However, the addition of plant powders rich in anthocyanins or other phenolic compounds may have an effect on the organoleptic properties, including aroma, and in consequence consumer acceptance of these novel chocolate products, depending on the source of the bioactive compounds and the type of product in which they are contained.

#### **2. Results and Discussion**

#### *2.1. Water Content and Activity and Color of Chocolates*

Table 1 shows the water content and activity, as well as the CIE L\*a\*b\* and organoleptic characteristics of the tested chocolates.

Water content is one of the quality parameters of chocolate that affects the properties of this product. The increase in the water content causes the deterioration of the rheological properties of the chocolate. This influences, among other things, an increase in the power consumption of devices during production, problems with pumping of the chocolate mass, coating of cores, forming the product into bars or figures and worse sensory perceptions during its consumption (chocolate sticks to the teeth). The water content of the chocolates should be as low as possible. Bolenz and Glöde (2021) indicate that it should be lower than 0.6% [33]. This is difficult to achieve, especially in the case of white or milk chocolates due to the higher water content in milk powder than in other raw materials used in the production of chocolate. Problems with achieving such a low value of this parameter may also occur in the production of sugar-free chocolates. Sucrose substitutes are often characterized by a higher water content than sugar (sucrose), hence the generally higher moisture content of this type of product. Thus, in practice, it is often aimed at less than 2%. The water content in chocolates also depends on the production technology used, and the type and design of the machines used.

Our results are consistent with the results of other researchers. An example is the study by Godoˇciková et al. (2019) who studied the antioxidant properties and volatile profile of different kinds and types of chocolates produced in Slovakia with various fruit and nut

additions [34]. They found the water content in dark chocolates at 1.1%, in milk chocolates at 1.8% and in white chocolates at 1.6%. Chocolates enriched with plant additives (sea buckthorn, almonds, mulberry, currant, cherry) were characterized by a much higher value of this parameter (8.7–16.6%). In our studies the water content in different types of chocolates varied from 0.93 to 2.19%, with the differences being statistically significant (*p* < 0.05).

Our results indicated that the type of chocolate and functional enrichment affected the water activity of chocolates significantly (*p* < 0.05). The water activity of tested chocolates ranges from 0.378 to 0.490. The present results are within the range reported by other authors [28,34]. Looking at the effect of freeze-dried fruit or vegetable powder on the water activity of chocolates, it was demonstrated that a partial replacement of sweetener in different types of chocolates by 1% freeze-dried phenolic-rich plant powder did not cause an increase of the water activity above the optimal level. Good quality and microbiologically stable chocolates are characterized by low moisture content and water activity of 0.25–0.50 [38]. Interestingly, it can be observed that some chocolate samples enriched with freeze-dried fruit or vegetable powder had lower moisture content and water activity than control ones. The differences in water content and activity could be explained with the changes in the chemical composition of the control and supplemented chocolates.



Blueberries (BLUB), raspberries (RASB), blackberries (BLCB), pomegranates pomace (POME), beetroots (BEET), control chocolate, (CONT). Data are presented as mean ± SD of three replications. The values followed by the same lowercase letter (a–f) within each chocolate type in the same column do not differ significantly according to Tukey's HSD test at *p* < 0.05.

> The color, next to the characteristic chocolate aroma and taste, is one of the basic attributes influencing the quality of chocolate products. These parameters were dependent on the type of chocolate. In our case, it was influenced by such factors as the content of cocoa liquor and related to this sweetener content, presence of freeze-dried plant powder, as well as the presence of powdered milk in milk chocolates. As can be seen in Table 1, the type of chocolate affected the CIE L\*a\*b\* color parameters of the chocolates significantly

(*p* < 0.05). The brightness (L\*) of the surface of tested chocolates ranged from 27.22 to 34.01, while the redness (a\*) and yellowness (b\*) were 3.03–7.99 and 0.93–2.19, respectively. The results suggested that the control chocolates and the supplemented chocolates differed (*p* < 0.05) in terms of values of parameters L\*, a\* and b\*. However, the values of parameter b\* remained more similar. Generally, the addition of freeze-dried phenolic-rich plant powder to DCH, DCH98S, DCH98ESt or MCH significantly (*p* < 0.05) brightened the color on the surface of enriched chocolates compared to the control ones, which could be attributed to some changes in the crystal structure of cocoa butter (polymorphism) [32]. On the other hand, there were only slight differences for the a\* and b\* parameters between the control chocolates and the enriched chocolates of the same type. These results suggested that the addition of 1% freeze-dried phenolic-rich plant powder has no negative effect on the color of the enriched chocolates. Similar results were presented by Muhammad et al. (2018) for milk chocolates supplemented with cinnamon nanoparticles [30].

The organoleptic evaluation of the tested chocolates was in the range of 3.8–4.9 points (Table 1). The DCH chocolates were rated the best, followed by the MCH ones, and the worst were the DCH98Est. The organoleptic quality of the DCH and the MCH chocolates, control and with the addition of BLUB, RASB, BLCB and POME powders was rated as extremely desired. They obtained scores from 4.6 to 4.9 points, while the scores within one group of chocolates, i.e., the DCH or the MCH group, did not differ significantly from each other (*p* ≥ 0.05). For example, in the case of the DCH chocolates, their rating was ranging from 4.7 to 4.9 points. Chocolates with BEET powder were rated the worst, though still in the desired organoleptic quality category. This was due to the earthy taste of the chocolates, which was especially noticeable in the case of milk chocolate (3.8 points). Among dark chocolates, chocolates with 98% cocoa were rated worse than the DCH chocolate. Panelists described them as tart and slightly more acidic than the DCH chocolates.

Two-way ANOVA revealed a significant effect of chocolate type (*p* < 0.001), phenolicrich plant powder (*p* < 0.05) and their interaction (*p* < 0.01) for the water content and activity, the CIE L\*a\*b\* color parameters and organoleptic characteristic of the tested chocolates (Table S1, Supplementary Materials).

#### *2.2. Profile and Concentrations of Bioactive Compounds*

The results show that there was considerable diversity between the profile and the levels of bioactive compounds in the control and the enriched chocolates. It is worthwhile noting that raw cocoa beans are known to be a good source of phenolic compounds, mainly flavonoids, but most of them are lost during the production of chocolate and thus even dark chocolate, which contains more cocoa parts than milk chocolate, has a significantly lower level of phenolic compounds. As a result of fermentation and drying, the concentration of phenolic compounds, especially flavonoids, decreases by more than 80% [26]. In the case of unstable anthocyanins, the losses may reach even up to 90% [10]. Heat treatment may also cause the transformation of the remaining anthocyanins into colorless chalcones, which spontaneously degrade to suitable phenolic acids or polymerize and condense with other phenolic compounds to form brown polymeric pigments. Therefore, chocolates obtained in the conventional processing of cocoa beans contain significantly less flavan-3-ols and do not contain anthocyanins. To minimize these losses, various plant extracts (e.g., raspberry leaf and green tea extracts) or dried fruits (e.g., cherry, black mulberry, currant, sea buckthorn) rich in phenolic compounds have been added to the white, milk or dark chocolates for enhancing their functional properties [27,30,31]. The application of freeze-dried fruit (blueberries, raspberries, blackberry, pomegranates pomace) and beetroot powder was exploited to enrich different kinds of chocolates, such as dark (DCH, DCH98S and DCH98ESt) and milk (MCH) chocolates, enhancing their health-promoting properties.

The addition of tested freeze-dried fruit or vegetable powders to different types of chocolates led to a significant (*p* < 0.05) enrichment of these products with phenolic compounds, mainly anthocyanins and phenolic acids (Table 2). It was found that a 1% addition of these powders to chocolates increased the content of total phenolic compounds

(flavan-3-ols, phenolic acids and anthocyanins) of enriched dark and milk chocolates compared to the control ones dependent on the powder used. The highest percentage increase (approximately 80%) in the total polyphenol content was observed in the MCH chocolate enriched with BLUB powder (Table 3). The greatest increase in the total content of polyphenols in all obtained chocolates was caused by the addition of BLUB powder followed by the addition of BEET powder. The BLUB powder, in all obtained chocolates, enriched the polyphenol composition by approximately 32.8 mg/100 g DM of anthocyanins, which were absent in the control chocolates. The greatest amount of flavan-3-ols and phenolic acids was introduced into the chocolates also via BLUB powder. Therefore, the profile and concentrations of phenolic compounds in chocolates containing tested freezedried plant powders are influenced by the phytochemical composition of cocoa liquor and plant powder used as a functional additive. The contents of individual phenolic compounds detected in the control chocolates and in the chocolates made with 1% addition of the fruit or vegetable powder in the product recipe are presented in Table 3 (powders were added instead of part of sweetener).

**Table 2.** Phenolic compounds content and antioxidant properties of studied freeze-dried phenolic-rich plant powders.


Blueberries (BLUB), raspberries (RASB), blackberries (BLCB), pomegranates pomace (POME), beetroots (BEET), control chocolate, (CONT). Data are presented as mean ± SD of three replications. The values followed by the same lowercase letter (a–e) in the same row do not differ significantly according to Tukey's HSD test at *p* < 0.05. nd—not detected.




**Table 3.** *Cont.*

nd—not detected. Blueberries (BLUB), raspberries (RASB), blackberries (BLCB), pomegranates pomace (POME), beetroots (BEET), control chocolate, (CONT). Data are presented as mean ± SD of three replications. The values followed by the same lowercase letter (a–f) within each chocolate type in the same row do not differ significantly according to Tukey's HSD test at *p* < 0.05.

> The addition of plant powders made the chocolates differ in the total phenolics content and the qualitative and quantitative composition of polyphenols (Table 3). The feature that distinguished them is primarily the content and concentration of anthocyanins. The control and chocolates with BEET powder did not contain anthocyanins. Chocolates with the addition of BLUB powder included in their composition four anthocyanins: cyanidin-3,5-*O*-diglucoside (Cy-3,5-diGlu), cyanidin-3-*O*-xyloside (Cy-3-Xyl), delphinidin-3,5-*O*diglucoside (Del-3,5-diGlu) and pelargonidin-3,5-*O*-diglucoside (Pel-3,5-diGlu), whose total concentration was approximately 33 mg/100 g DM. Cy-3,5-di Glu and Pel-3,5-diGlu were present in the highest amounts, each at concentrations of above 10 mg/100 g DM. Chocolates with the addition of RASB powder included only one anthocyanin—cyanidin-3-*O*-glucoside (Cy-3-Glu) in a concentration of approximately 3.7 mg/100 g DM. There were four anthocyanins in the chocolates with BLCB powder, i.e., Cy-3-Glu, cyanidin-3- *O*-rutinoside (Cy-3-Rut), Cy-3-Xyl and cyanidin-3-(600-malonyl)-glucoside (Cy-3-(600-Mal-Glu)), the total concentration of which was approximately 5 mg/100 g DM, with Cy-3-Glu being the highest amount of 4.4 mg/100 g DM. On the other hand, in chocolates with POME powder there were three anthocyanins, i.e., Cy-3,5-diGlu, Del-3,5-diGlu and delphinidin-3- *O*-glucoside (Del-3-Glu), with a total concentration of 0.14 mg/100 g DM, with Cy-3,5-diGlu being the highest amount of 0.1 mg/100 g DM.

> Seven phenolic compounds were identified in all types of control chocolates and chocolates enriched with BEET powder, including four flavan-3-ols (catechin—Cat, epicatechin— Ecat, procyanidin B2—PC B2 and procyanidin C1—PC C1) and three phenolic acids (gallic acid—GA, protocatechuic acid—PA and *p*-hydroxybenzoic acid—*p*-HBA). The presence

of these compounds in chocolates has already been described by other authors [29,39,40]. Chocolates enriched with RASB powder showed eight phenolic compounds, seven of which were present in the control chocolates and in chocolates made with vegetable powder in addition to Cy-3-Glu. The dark and milk chocolates containing BLCB powder showed the presence of eleven phenolic compounds and eight of these phenolics were identified in raspberry-enriched chocolates in addition to Cy-3-Rut, Cy-3-Xyl and Cy-3-(600 - Mal-Glu). The samples of DCH, DCH98S, DCH98Est and MCH made with BLUB powder also showed eleven phenolic compounds, eight of which were present in chocolates with BLCB powder in addition to Cy-3,5-diGlu, Del-3,5-diGlu and Pel-3,5-diGlu. The different types of chocolates containing POME powder showed the presence of ten phenolic compounds and nine of these phenolics were identified in BLUB-enriched chocolates in addition to Del-3-Glu. The predominant phenolic compound in all tested chocolates was Ecat. The second most abundant compound was PC B2, followed by Cat, PC C1, GA and *p*-HBA. Moreover, Cy-3,5-diGlu, Pel-3,5-diGlu, Cy-3-Glu and Cy-3-Xyl were found in significant quantities but only in chocolates made with berry powders. Depending on the evaluated chocolate types and functional enrichment, the concentrations of individual phenolic compounds varied significantly (*p* < 0.05) within tested chocolates.

As can be seen in Table 3, the sum of phenolic compounds differed significantly (*p* < 0.05) between control and enriched chocolates of different types, with the highest total phenolics found for DCH98S chocolate enriched with BLUB powder and the lowest for control MCH (the results ranged from 89.24 to 358.92 mg/100 g DM).

The major phenolic compounds in both control and chocolates enriched with fruit or vegetable powders were flavan-3-ols, which represented approximately 72–92% of total phenolics levels. The total amount of the investigated flavan-3-ols within the different chocolate types ranged from 79.66 mg/100 g DM in control milk chocolate to 294.82 mg/100 g DM in DCH98S chocolate enriched with BEET powder. The content of these compounds was similar to the amount reported by other authors [29,39,40] in different kinds of chocolates. Considerable amounts of phenolic acids were also found in tested chocolates. The concentrations of these phenolics represented approximately 8–11% of the average amounts of total phenolics. Among investigated samples, the highest phenolic acids levels were found in DCH98S chocolate enriched with either BLUB or POME powders. In turn, significant quantities of anthocyanins were found in chocolates enriched with fruit powders but mainly in those made with berry powders. Depending on the evaluated chocolate types and functional enrichment, the concentrations of anthocyanins varied significantly (*p* < 0.05) within chocolates enriched with fruit powders and ranged from 0.14 to 32.99 mg/100 g DM. In this study, no anthocyanins were found in the control chocolates. Moreover, chocolates with the addition of POME powder were characterized by very small amounts of these compounds.

The results indicated that the addition of all fruit and vegetable powders to different types of chocolates led to substantial changes in the levels of phenolic compounds of enriched chocolates (Table 3). A two-way ANOVA revealed that the content of all phenolic compounds, apart from anthocyanins, varied significantly with the type of chocolate (*p* < 0.001), phenolic-rich plant powder (*p* < 0.001) and their interaction (*p* < 0.001) (Table S2, Supplementary Materials). In addition, the total anthocyanins, Cy-3-Glu, Cy-3-Rut, Cy-3,5-diGlu, Cy-3-Xyl, Cy-3-(600-Mal-Glu), Del-3,5-diGlu and Del-3-Glu concentrations were significantly affected by phenolic-rich plant powder, but there was not a statistically significant interaction between the effects of the type of chocolate and the phenolic-rich plant powder. The addition of all tested fruit and vegetable powders increased considerably the total content of phenolic compounds of all chocolate types compared to the control ones. Irrespective of the chocolate type, there was a significant increase in the level of flavan-3-ols and phenolic acids of the enriched chocolates. It was also observed that the addition of berry powders also caused a significant (*p* < 0.05) increase in the total anthocyanins content in both dark and milk chocolates with respect to control chocolates, while the amount of these pigments in chocolates made with POME powder only slightly increased compared

to control ones. Overall, the greatest increment in the total contents of three classes of phenolic compounds of all types of chocolates was caused by the supplementation of BLUB powder. The increase in phenolic contents in fruit- or vegetable-supplemented chocolates reflects the addition of specific functional enrichment. Godoˇciková et al. (2017) reported also that specific types of dried fruits, for example black mulberry, rich in anthocyanins, was also suitable to enhance the concentration of bioactive substances of chocolate even with a lower cocoa solids content [41]. Recently, Martini, Conte and Tagliazucchi (2018) demonstrated that the enrichment of dark chocolates with Sakura green tea leaves or turmeric powder is an effective technique to improve the health-enhancement of the final product [42]. classes of phenolic compounds of all types of chocolates was caused by the supplementation of BLUB powder. The increase in phenolic contents in fruit- or vegetable-supplemented chocolates reflects the addition of specific functional enrichment. Godočiková et al. (2017) reported also that specific types of dried fruits, for example black mulberry, rich in anthocyanins, was also suitable to enhance the concentration of bioactive substances of chocolate even with a lower cocoa solids content [41]. Recently, Martini, Conte and Tagliazucchi (2018) demonstrated that the enrichment of dark chocolates with Sakura green tea leaves or turmeric powder is an effective technique to improve the health-enhancement of the final product [42]. Our data supported the possible application of berries and POME powders to the

ols and phenolic acids of the enriched chocolates. It was also observed that the addition of berry powders also caused a significant (*p* < 0.05) increase in the total anthocyanins content in both dark and milk chocolates with respect to control chocolates, while the amount of these pigments in chocolates made with POME powder only slightly increased compared to control ones. Overall, the greatest increment in the total contents of three

*Molecules* **2021**, *26*, x 10 of 24

Our data supported the possible application of berries and POME powders to the formulation of both dark and milk chocolates with increased phenolic compounds, mainly anthocyanins, which are never found in a given type of cocoa bean or are lost during their processing. formulation of both dark and milk chocolates with increased phenolic compounds, mainly anthocyanins, which are never found in a given type of cocoa bean or are lost during their processing.

#### *2.3. Antioxidant Activity 2.3. Antioxidant Activity*

In order to evaluate the freeze-dried phenolic-rich plant powders' contribution to the antioxidant properties of the different types of chocolates, the DPPH radical scavenging activity and the ferric-reducing ability in control and enriched chocolates were determined and the results have been presented in Figure 1I and II, respectively. In order to evaluate the freeze-dried phenolic-rich plant powders' contribution to the antioxidant properties of the different types of chocolates, the DPPH radical scavenging activity and the ferric-reducing ability in control and enriched chocolates were determined and the results have been presented in Figure 1 (I) and (II), respectively.

**Figure 1.** (**I**) DPPH radical scavenging activity of different types of chocolates enriched with various freeze-dried phenolic-rich plant powders, expressed as IC50 values. (**II**) Ferric reducing antioxidant **Figure 1.** (**I**) DPPH radical scavenging activity of different types of chocolates enriched with various freeze-dried phenolic-rich plant powders, expressed as IC50 values. (**II**) Ferric reducing antioxidant power (FRAP value) of different types of chocolates enriched with various freeze-dried phenolic-rich plant powders. Data are expressed as the mean of triplicate ± SD. Bars with the same lowercase letter (a–f) within each type of chocolate do not differ significantly according to Tukey's HSD test at *p* < 0.05. Blueberries (BLUB), raspberries (RASB), blackberries (BLCB), pomegranates pomace (POME), beetroots (BEET), control (CONT).

The DPPH• scavenging capacity was expressed as IC<sup>50</sup> values, the concentration at which 50% inhibition of free radical scavenging activity is observed. The range of IC<sup>50</sup> values of the analyzed chocolates was 3.05–28.12 mg/mg DPPH (Figure 1I). Two-way ANOVA revealed the significant effect of chocolate type (*p* < 0.001), freeze-dried phenolicrich plant powders (*p* < 0.001) and their interaction for the DPPH radical scavenging activity (Table S2).

Overall, dark chocolates revealed better antioxidant properties than milk chocolates, which agrees with the higher content of phenolic compounds found in the dark chocolates. The antioxidant activity of chocolates is usually attributed to the presence of monomeric flavan-3-ols, polymeric procyanidins and Maillard reaction products (e.g., melanoidins) that are well known to possess effective scavenging activity of free radicals [43].

The results indicated that the enrichment of chocolates with various fruit or vegetable powders can improve the antioxidant activity of chocolates, depending on the chocolate type and plant powder used as a functional additive. As compared to control chocolates, the supplementation with almost all freeze-dried fruit and vegetable powders caused significant (*p* < 0.05) decreases in the DPPH free radical scavenging activity in many types of obtained chocolates. Interestingly, among all studied chocolates, the highest DPPH radical scavenging capacity was exhibited by the DCH98S chocolate enriched with BEET powder. Nevertheless, the results indicated that the addition of BEET powder to DCH and DCH98ESt chocolates reduced DPPH radical scavenging activity significantly (*p* < 0.05) compared to the control. The observed differences in the free radical scavenging capacity of chocolates of different types with the same additive and different sweeteners may be due to different reaction mechanisms occurring during the preparation of the chocolates, including the Maillard reaction, the degradation of phenolic compounds of higher molecular weight to smaller phenolics and/or various transformations of flavan-3-ols and phenolic acids. In addition, other authors reported that supplementation of a sour cherry puree with sucrose or erythritol significantly declines its free radical scavenging activity [44]. They showed that the addition of natural sweeteners to a sour cherry puree resulted in a significant reduction in phenolic compounds, mainly flavan-3-ols. These phenomena may be due to the intermolecular interactions between the hydroxyl group from phenolic compounds in chocolate and phenolic-rich plant powders and a hydroxyl group in sucrose or sugar alcohol molecules [44,45].

It was observed that dark chocolates (98%) sweetened with erythritol and stevia instead of sucrose supplemented with RASB, BLCB and POME powders exhibited a higher DPPH radical-scavenging activity as compared to control DCH98ESt samples. Regarding the DCH chocolates, the enrichment with POME powder significantly increased (*p* < 0.05) the DPPH antioxidant capacity compared to the control sample. The obtained data highlighted, moreover, that the addition of freeze-dried berries, POME and BEET powders to MCH samples caused a significant (*p* < 0.05) increase in DPPH radical scavenging activity. As expected, the lowest free radical scavenging abilities was exhibited by the control milk chocolate. Our results also show that the partial sweetener substitution by BLCB powder caused the most pronounced reduction of the DPPH radical-scavenging potential of all dark chocolates. This result may be attributed to the synergistic and antagonistic interaction that results from the coexistence of many antioxidant compounds in enriched chocolates [41]. Some reports revelated that interactions between flavan-3-ols and anthocyanins might accelerate the degradation of anthocyanin pigments that further react giving rise to polymeric brown pigments [44]. It should be understood that the antioxidant activity of a mix is not the sum of the antioxidant activities of each of the components, due to interactions between the components. Therefore, it is difficult to predict in advance the result of food functionalization into components with antioxidant properties, e.g., the antioxidant potential of the product, inhibition of the growth of cancer cells or other biological properties. The effects can be surprising. Godoˇciková et al. (2017) observed that dark chocolates enriched with mulberry and sea buckthorn exhibited higher DPPH scavenging activity than control ones. Antioxidant capacities significantly increased with the addition of capsules of bioactive compounds [41].

The addition of freeze-dried phenolic-rich plant powders led to significant (*p* < 0.001) differences in the reducing capacity of all types of chocolates (Table S2). The ferric-reducing antioxidant power of the tested chocolate samples varied from 140.15 to 354.38 µmol TE/g DM (Figure 1II).

Among investigated samples, the DCH98ESt chocolates enriched with POME and BEET powders demonstrated the highest FRAP values. The obtained results indicated that the reducing capacity of dark chocolates made with sucrose noticeably decreases after supplementation of BLCB, POME and BEET powders. While in the case of dark chocolate sweetened with erythritol and stevia instead of sucrose, the addition of all functional powders led to a considerable increase in the reducing power compared to the control ones. This phenomenon may be ascribed to the interaction of phenolic compounds with sucrose. For example, Shalaby et al. (2016) showed that the introduction of sucrose to green tea significantly decreased its antioxidant potential [45]. In the present study, irrespective of the functional enrichment type, there was a significant increase in the ferric-reducing ability of all supplemented milk chocolates. Other authors have also observed that the addition of different sweeteners to fruit puree and green or black tea affects their antioxidant properties in different ways [44,45]. Therefore, we can conclude that the observed differences in DPPH and FRAP values of dark chocolates (98%) with the same powders and different sweeteners may be due to different reaction mechanisms occurring between phenolic compounds and sucrose and erythritol with stevia glycosides. Interestingly, unlike dark chocolates, the addition of BLCB powder caused the greatest increase in the reducing capacity of milk chocolate. Our results also showed that the addition of anthocyanin-rich BLUB powder caused an increase in reducing power regardless of the type of chocolate. This increase may be attributed to the fact that anthocyanins can act as reducing agents mainly through the electron-transfer mechanism [36].

Other studies have also reported that the addition of fruit or other phenolic-rich plants to chocolates either increase or decrease antioxidant activity evaluated by FRAP and DPPH assays [32,33,41,46]. These distinct differences may be attributed to the increased interaction between phenolics and other compounds, including carbohydrates, sweeteners, and proteins, present in chocolate and in fruits or vegetables. It is well known that both non-oxidized and oxidized phenolic compounds have a strong affinity to proteins, polysaccharides, alkaloids and Maillard reaction products, and may form insoluble complexes [46]. The results provide strong evidence that the interactions of mixtures of antioxidant compounds might generate synergic or inhibitor effects and can enhance or inhibit the antioxidant activity or even modify their reaction mechanisms [41].

#### *2.4. Electronic Nose Analysis of Chocolates*

Chocolate aroma depends on the combination of many volatile compounds (VCs) derived from cocoa beans and other ingredients, such as sucrose (sweetener), milk, and flavors, formed or modified during the roasting, alkalization and conching stages. It is well known that the typical chocolate flavor is mainly formed due to the Maillard reactions and the Strecker degradation of flavor precursors, such as free amino acids, short-chain peptides, and reducing sugars (e.g., glucose) during roasting [47,48]. However, it has to be noticed that functional additives rich in phenolic compounds having a positive effect on biological activity of enriched chocolates, can also affect the sensory properties of the final product. For example, phenolic compounds are responsible for the specific astringent and bitter taste of the raw beans and influence the stability and digestibility of the products obtained from them as a result of the formation of complexes mainly with polysaccharides, proteins, methylxanthines and Maillard reaction products. Therefore, these compounds are playing an important role in shaping the sensory characteristics of chocolates, fruits, and vegetables, as well as products obtained from them [48].

In our study, volatile compounds in chocolates were determined by using an electronic nose. A total of twenty-six VCs were identified, including alcohols, phenols, aldehydes, ketones, esters, acids, pyrazines, furfural, lactone, and sulfide compound in the all-enriched chocolates (Table 4). The results revealed that the addition of berries, POME and BEET powders significantly (*p* < 0.05) influenced the sensory attributes of the resultant chocolates. A two-way ANOVA revealed that the content of all VCs, apart from vanillin, varied significantly with chocolate type (*p* < 0.001), freeze-dried phenolic-rich plant powders (*p* < 0.001) and their interaction (*p* < 0.001) (Table S3, Supplementary Materials). The most important compounds of control and supplemented chocolates were acetic acid, benzaldehyde, 2-methylpropanal, 3-methylbutanal, 2-furfural, 2,5-dimethylpyrazine, pentanal and phenylethylacetate. The compound found in the highest concentration in all samples was acetic acid, which is associated with sour, pungent, and unpleasant notes. This compound is the highest odor-active compound in unroasted cocoa beans. Despite the fact that during further processing of cocoa beans acetic acid concentration decreases by over 70%, it is still the highest odor-active compound in roasted cocoa beans, cocoa mass and chocolates obtained from them [47–50]. The second compound found in the highest concentration in all tested chocolates is benzaldehyde. This compound has a pleasant fruity-type odor and a fruity-type flavor.

**Table 4.** The content of volatile compounds in different types of chocolates.



**Table 4.** *Cont.*


**Table 4.** *Cont.*

nd—not detected. Blueberries (BLUB), raspberries (RASB), blackberries (BLCB), pomegranates pomace (POME), beetroots (BEET), control (CONT). Data are expressed as the relative peak area (in percentage) of each compound and presented as mean ± SD of three replications. The values followed by the same lowercase letter (a–f) within each chocolate type in the same row do not differ significantly according to Tukey's HSD test at *p* < 0.05.

> The presence of pyrazines, aldehydes and furfural was attributed to Maillard reactions. Among the aldehydes characteristic of the Strecker degradation, which is one of the main stages of the Maillard reaction, 3-methylbutanal, 2-methylpropanal and dimethyl disulfide derived from the decomposition of leucine, valine, and methionine, respectively, were determined. 3-Methylbutanal and 2-methylpropanal are very important compounds that have a positive effect on the development of the characteristic chocolate aroma of cocoa products [47–50].

> The principal component analysis (PCA) showed that dark and milk chocolates made with the addition of different functional additives are markedly different in terms of their VCs and thus clustered separately (Figure 2).

**Figure 2.** Principal component analysis (PCA) scores plot of aroma signals of different types of chocolate enriched with different freeze-dried phenolic-rich plant powders: DCH (**square**), DCH98S (**triangle**), DCH98ESt (**diamond**) and MCH (**dot**). Blueberries (BLUB—blue), raspberries (RASB—pink), blackberries (BLCB—dark blue), pomegranates pomace (POME—violet), beetroots (BEET—dark red), control (CONT—gray).

It is clear that each type of chocolate enriched with various functional additives was clearly distinguished by PC1 into two clusters, which suggested that the substitution of sweetener by 1% of fruit or vegetable powder resulted in the majority of the variance in the VCs composition compared to the corresponding control chocolate. It was demonstrated that almost all enriched chocolates could be distinguished from the control chocolates due to the abundance of some aldehydes, ketones, alcohols and acetic acid contents and the emergence of γ-nonalactone, which was not present in the control chocolates. In all chocolates enriched with berries, pomegranates pomace and beetroot powders the same classes of VCs were observed which were identified in control chocolates, in addition to γ-nonalactone. From the detected VCs, mainly 3-methylbutanal, phenylethylacetate, 2-phenylethanol, 2,5-dimethylpyrazine, 2,3-butanediol were positively correlated with chocolate aroma. However, the presence of benzaldehyde and pentanal with bitter and pungent notes origin from lipid oxidation was negatively correlated with chocolate flavor quality [47–49].

3-Methylbutanal produce key cocoa aromas such as malty and chocolate notes. Phenylethylacetate and 2-phenylethanol confer pleasant flowery and honey flavor notes enhancing flavor impression. 2,3-Butanediol, with the natural odor of cocoa butter, has been considered an important compound that could alter the overall aroma of chocolate [47–49].

Generally, the concentration of 2,3-pentanedione, γ-nonalactone and dimethyl trisulfide benzaldehyde, pentanal was significantly increased by the addition of fruit or vegetable powder. As demonstrated in Table 4, significant differences (*p* < 0.05) in the content of acetic acid were found between chocolates with different functional additives. The results showed that, depending on the functional ingredient type, a substantial change in the content of acetic acid was observed in all samples. Interestingly, all chocolates supplemented with RASB and BEET powder contained significantly lower amounts of acetic acid while those enriched with BLUB and POME powder had higher amounts of acetic acid than control chocolates. Independent of chocolate type, chocolates supplemented with BLUB and POME powders showed the highest content of acetic acid, while those enriched with

RASB and BEET powders exhibited the highest content of benzaldehyde. On the other hand, the concentrations of 3-methylbutanal were higher in the control than in almost all enriched chocolates, mainly those made with BLUB or POME powder. Nevertheless, almost all of the enriched chocolates showed higher levels of alcohols, such as 2-phenylethanol and 2,3-butanediol, which are desirable to obtain cocoa products with flowery and honey aromas [49]. Thus, chocolates made with the addition of berries, pomegranates pomace and beetroot powders may have a good consumer acceptability when compared to control dark and milk chocolates.

#### **3. Materials and Methods**

#### *3.1. Materials*

The research materials were chocolates supplemented with lyophilized fruits and vegetables rich in flavonoids, including anthocyanins and phenolic acids. Chocolates were obtained from the following raw materials: Cocoa liquor (with 55% *w*/*w* of fat) and butter were purchased from Barry Callebaut (Łód´z, Poland). Sugar, alkalized cocoa powder (with 10% *w*/*w* of fat), skimmed milk powder (with 1.5% *w*/*w* of fat), soy lecithin, polyglycerol polyricinoleate—PGPR emulsifier and ethyl vanillin were obtained from WIEPOL Zakład Pracy Chronionej Ireneusz Wielimborek (Sierpc, Poland). Erythritol with the addition of stevia (99% erythritol and 1% stevia) was purchased from Domos Polska Sp. z o.o. (Czosnów, Polska). Plants, such as blueberries (BLUB), raspberries (RASB), blackberries (BLCB), pomegranates (POME), and beetroots (BEET) were bought on the local market.

#### *3.2. Chemicals and Reagents*

Standards of catechin (Cat), epicatechin (ECat), procyanidin B2 (PC B2), procyanidin C1 (PC C1), gallic acid (GA), protocatechuic acid (PA), *p*-hydroxybenzoic acid (*p*-HBA), cyanidin-3-*O*-glucoside (Cy-3-Glu), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox), 2,20 -azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS), 2,2-diphenyl-1-(2,4,6-trinitrophenyl) hydrazyl (DPPH), 2,4,6-tri(2-pyridyl)-s-triazine (TPTZ), sodium acetate, ferric chloride hexahydrate, ferrozine, and ammonium acetate were purchased from Sigma-Aldrich (St. Louis, MO, USA). HPLC grade methyl tert-butyl ether (MTBE) and methanol were purchased from J.T. Baker (Deventer, The Netherlands). All other reagents were of analytical grade and were purchased from Chempur (Piekary Sl ˛askie, Poland). Chromacol PTFE syringe filters (0.2 ´ µm pore size) were purchased from Shim-Pol (Izabelin, Poland). Ultrapure water (resistivity 18.2 MΩ cm), obtained from a Milli-Q purification system (Millipore, Bedford, MA, USA), was used for all analyses.

#### *3.3. Lyophilization of Fruits and Vegetables*

Plant materials, i.e., pomegranates pomace, whole berries and beetroots cut into cubes, were frozen at −80 ◦C for 48 h. Then they were freeze-dried in a BETTA2-8LSC plus Christ freeze drier (Osterode am Harz, Germany) for 24 h. The initial parameters of the process were—pressure: 1 millibar, shelf temperature: 5 ◦C, while final parameters—pressure: 0.1 millibar, shelf temperature: 5 ◦C. The obtained freeze-dried products were then ground to a fine powder in an XB-9103 MPM PRODUCT knife mill (Milanówek, Poland) and stored in glass containers.

#### *3.4. Preparation of Chocolates*

Four types of chocolate with 1% addition of fruit or vegetable powders were prepared during the study: dark chocolate 53% (~53% cocoa) with the total fat content of 35% (*w*/*w*) and sweetened with sucrose (DCH), dark chocolate 98% (~98% cocoa) with the total fat content of 51% (*w*/*w*) sweetened with sucrose (DCH98S), dark chocolate 98% (~98% cocoa) with the total fat content of 51% (*w*/*w*) sweetened with erythritol with stevia (DCH98ESt), milk chocolate (~40% cocoa, 20% skimmed milk powder) with the total fat content of 36% (*w*/*w*) and sweetened with sucrose (MCH). Recipes of chocolates are given in Table 5.


**Table 5.** Recipes of examined chocolates.

DCH—dark chocolate with 53% cocoa content sweetened with sucrose, DCH98S—dark chocolate with 98% cocoa content sweetened with sucrose, DCH98ESt—dark chocolate with 98% cocoa content sweetened with erythritol with stevia, MCH—milk chocolate with 40% cocoa content sweetened with sucrose.

> Furthermore, control chocolates were obtained in which an additional portion of the sweetener was added to the chocolates instead of the freeze-dried fruits or vegetables. For example, in the DCH, the sucrose percentage was then 45.59% instead of 44.59%.

> The preparation process of chocolates consisted of the following basic stages: grinding in the ball mill type M-5 (Promet, Łód´z, Poland; loading up to 6 kg), conching in the K-5 type conch (Promet, Łód´z, Poland; loading up to 6 kg), tempering in the temperer model T8 type Temperatrice containing a closed cooling system with an additional AC current heat exchanger (Pomati Group Srl, Zona Ind. Mirandolina, Italy; loading up to 8 kg), moulding, cooling and wrapping.

> The cocoa butter, liquor and eventually cocoa powder were liquefied in the preheated to 70 ◦C ball mill. Next, sugar, milk powder (in the case of MCH) and 50% of the amount of lecithin were dosed into the ball mill. The grinding was carried out at 70 ◦C for 50 min (DCH), at 85 ◦C for 90 min (DCH98S and DCH98ESt) or at 55 ◦C for 60 min (MCH) with a rotational velocity of 75 rpm to obtain a particle size in the range of 20–25 µm, which was measured using the micrometric screw NSK Digitrix-MARK II ELECTRONIC MICROMETER with the electronic readout of the results (Japan Micrometer MFG. Co., Ltd., Tokyo, Japan). After this time, the obtained cocoa mass was transferred to preheated to 50 ◦C conch for further homogenization and emulsification. Then, freeze-dried phenolicrich plant powders were introduced into the conch. After 45 min of conching, ethyl vanillin, PGPR emulsifier and the remaining part of the lecithin were added to the mass, and then conching was continued for 15 min. Next, the mass was subjected to tempering. For this purpose, the temperature of the mass was lowered in the temperer from 50 to 32 ◦C in the case of dark chocolates and to 28 ◦C in the case of milk chocolate. These temperatures were maintained for 15 min. Finally, chocolate masses were poured to preheated to 30 ◦C (dark chocolates) or 27 ◦C (milk chocolate) forms, cooled to 18 ◦C in a cooling tunnel (Promet, Łód´z, Poland), and removed from the forms. Chocolate bars were wrapped in aluminum foil and subjected to analysis. All chocolates were obtained in triplicate.

#### *3.5. Water Content and Water Activity Determination*

Water content was determined by drying the ground chocolate samples mixed with sand at 102–105 ◦C to constant weight [28].

The water activity of the chocolates was determined by using HYGROPALM AW1 meter (Rotronic, Helvetia, Switzerland) equipped with a digital probe AW-DIO [28].

#### *3.6. Color Determination*

The color was determined using a trichromatic reflection colorimeter Konica Minolta CR-400 with Spectra Magic NX 1.3 software (Konica Minolta Sensing INC., Osaka, Japan). The results were expressed in accordance with CIE L\*a\*b\* system (D65 illuminant and 10◦ viewing angle) [51].

#### *3.7. UHPLC-DAD-ESI–MS/MS Analysis of Phenolic Compounds*

The phenolic compounds were extracted according to the method described by Zy˙ zelewicz et al. (2018), with some modifications [ ˙ 28]. Briefly, the accurately weighed defatted chocolate samples were extracted 3 times in an orbital shaker at room temperature for 30 min at 150× *g* with a mixture of acetone/water/acetic acid (70/29.5/0.5, *v*/*v*/*v*). The mixture was centrifuged at 4000× *g* for 10 min and the supernatant from each extraction were combined and evaporated under a stream of nitrogen. The residues were dissolved in methanol and filtered through a 0.20 µm pore size PTFE syringe filters. Finally, the samples were analyzed for the content of phenolic compounds using a UHPLC+ Dionex UltiMate 3000 system equipped with a UV–Vis diode array detector (Thermo Fisher Scientific Inc., Waltham, MA, USA), and a Transcend™ TLX-2 multiplexed LC system equipped with Q-Exactive Orbitrap mass spectrometer (Thermo Scientific, Hudson, NH, USA) using a heated electrospray ionization (ESI) interface (HESI–II). Samples (10 µL) were injected on an Accucore™ C18 column (150 mm × 2.1 mm i.d., 2.6 µm; Thermo Fisher Scientific Inc., Waltham, MA, USA). The column temperature was set at 30 ◦C. The mobile phase and gradient program were used as previously described by Oracz, Nebesny and Zy˙ zelewicz (2019), with some modifications [ ˙ 52]. The 2-phase solvent system used for phenolic compounds separation was composed of 0.1% formic acid in water as solvent A and 0.1% formic acid in acetonitrile as solvent B. The flow rate was 0.35 mL/min and the gradient was as follows: 0–8 min, 1–5% B; 8–15 min, 5–8% B; 15–20 min, 8–10% B; 20–25 min, 10–15% B; 25–35 min, 15–20% B; 35–40 min, 20–25% B; 40–50 min, 25–90% B; 50–53 min, 90% B; 53–58 min, 90–1% B. Finally, the initial conditions were held for 7 min for column re-equilibration and for 5 min as a re-equilibration step. UV–Vis detection was performed at 280 nm for flavan-3-ols and phenolic acids and at 520 nm for anthocyanins. Instrument control, data acquisition, and evaluation were conducted with Chromeleon 6.8 Chromatography Data System, Qexactive Tune 2.1, Aria 1.3.6, and Thermo Xcalibur 2.2 software, respectively. Phenolic compounds were identified by comparing their retention times, UV–Vis absorbance spectra, full scan mass spectra, and MS/MS fragmentation patterns with their corresponding standards analyzed under identical conditions and previous literature reports [53,54]. Quantification was carried out using an external standard method. The concentration of individual flavan-3-ols and phenolic acids was determined based on peak area and calibration curves derived from corresponding reference compounds. For the quantification of anthocyanins, the calibration curves of Cy-3-Glu were used. All measurements were conducted in triplicate and results were expressed as mg phenolic compound per 100 g chocolate dry mass (mg/100 g DM).

#### *3.8. Free Radical Scavenging Assay*

The free radical scavenging activity was determined using the DPPH assay [28]. The analytical samples were prepared using serial dilutions of chocolate extracts in methanol. For each sample, experiments were conducted in triplicate. Finally, the mean concentration of the test chocolate extracts at which the concentration of the DPPH free radicals was reduced by 50% (IC50) was calculated.

#### *3.9. Ferric Reducing Antioxidant Power Assay*

The ferric reducing ability (FRAP) was evaluated using the method of Oracz and Zy˙ zelewicz (2019) [ ˙ 55]. For each sample, experiments were conducted in triplicate. The results were expressed as µmol Trolox equivalents per gram of chocolate DM (µmol TE/g DM).

#### *3.10. Electronic Nose Analysis of Tested Chocolates*

Electronic nose (E-nose) analysis of volatile flavor compounds was carried out according to the method of Rottiers et al. (2019), with some modifications [50]. The E-nose analyses were performed using a commercial Heracles II electronic nose (Alpha MOS, Toulouse, France), equipped with an HS-100 autosampler, a sensor array unit, and 2 columns working in parallel mode: a non-polar column (MXT5: 5% diphenyl, 95% methylpolysiloxane, 10 m length and 180 lm diameter) and a slightly polar column (MXT1701: 14% cyanopropylphenyl, 86% methylpolysiloxane, 10 m length and 180 lm diameter). An accurately weighed 1.0 g chocolate sample was put into 20-mL screw vials sealed with a magnetic cap with polytetrafluorethylene-silicone septa and placed in the auto-sampler. The vials were incubated in a shaker oven for 20 min at 50 ◦C and shaken at 500 rpm. Next, a syringe sampled 1000 µL of the headspace and then injected it into the gas chromatograph with 2 flame ionization detectors. The thermal program started at 50 ◦C (held for 2 s) and increased up to 250 ◦C at 3 ◦C/s and held for 21 s. The total separation time was 100 s. The calibration of the apparatus was carried out using a solution of alkanes (from *n*-hexane to *n*-hexadecane). The retention times of *n*-alkanes were used to determine the Kovats indices and identify the volatile compounds using AromaChemBase software (Alpha MOS, Toulouse, France). Each sample was measured in triplicate. Instrument control, data acquisition, and evaluation were conducted with Alphasoft 14.2 and AroChembase (Alpha MOS, Toulouse, France) softwares. The principal component analysis (PCA) was performed using AlphaSoft software (Alpha MOS, Toulouse, France) to determine the dissimilarities among the same types of chocolates in terms of volatile components.

#### *3.11. Organoleptic Evaluation of Chocolates*

The organoleptic evaluation of chocolates was carried out according to Zy˙ zelewicz et al. ˙ (2018) [28] in our specialist sensory analysis laboratory. The evaluation was made by ten panelists using a 5-point scale with the relevant significance coefficients, in which 5 points corresponded to the best quality and 1 point to the worst. The sensory attributes of the chocolates, i.e., appearance in the packet, shape, color, consistency (hardness, smoothness), conchoidal fracture, aroma, taste, and upper and lower surface glossiness were evaluated. Final assessments were presented on a 5-point scale, according to which 5 meant extremely desirable quality, 4 was desirable quality, 3 was tolerable quality, 2 represented dislike, and 1 was for a defective product.

#### *3.12. Statistical Analysis*

The results are presented as mean ± standard deviations of 3 replicates. The one-way analysis of variance (ANOVA) was used to determine if there were significant differences between the physicochemical properties, phenolic compounds and antioxidant activity observed in the control and enriched chocolate samples. Where effects of supplementation of anthocyanin-rich were significant, the means were compared with Tukey's HSD (Honestly Significant Difference) at *p* < 0.05, using the Statistica 13.0 software (StatSoft, Inc., Tulsa, OK, USA). The effects of the types of chocolates and different freeze-dried phenolic-rich plant powders and their interaction on phenolic content, antioxidant activity and volatile compounds content in chocolates were tested by means of two-way ANOVA.

#### **4. Conclusions**

The results of the present study revealed that the enrichment of dark and milk chocolates with berries, pomegranates pomace and beetroot powders caused an increase in the amount of phenolic compounds, including flavan-3-ols, anthocyanins and phenolic acids.

Chocolates enriched with BLUB powder were characterized by the highest total polyphenol content. Our findings were further supported by the enhanced free radical scavenging activity and reducing capacity of different types of chocolates supplemented with fruit or vegetable powders. All chocolates with the addition of BLUB and RASB powders obtained an extremely desirable assessment in the organoleptic evaluation. The analysis of aroma compounds (volatile compounds) with the use of an electronic nose showed that berries, pomegranates pomace and beetroot powders have been successfully added to produce both dark and milk chocolates with acceptable sensory quality.

The physicochemical and sensory analysis results indicated that up to 1% of tested freeze-dried phenolic-rich plant powders can be successfully added to produce milk and dark chocolates with increased contents of polyphenols and good sensory properties.

**Supplementary Materials:** The following are available online, Table S1: Two-way ANOVA analysis of physicochemical characteristic and organoleptic assessment of different types of chocolates enriched with various freeze-dried phenolic-rich plant powders; Table S2: Two-way ANOVA analysis of the content of individual phenolic compounds and antioxidant properties of different types of chocolates enriched with various freeze-dried phenolic-rich plant powders; Table S3: Two-way ANOVA analysis of the content of volatile compounds in different types of chocolates enriched with various freeze-dried phenolic-rich plant powders.

**Author Contributions:** Conceptualization, D.Z. and J.O.; methodology, D. ˙ Z. and J.O.; validation, ˙ D.Z. and J.O.; formal analysis, D. ˙ Z., J.O. and M.B.; investigation, D. ˙ Z., J.O. and M.B.; resources, D. ˙ Z., ˙ J.O. and E.K.; data curation, D.Z. and J.O.; writing—original draft preparation, D. ˙ Z., J.O. and K.K.-W.; ˙ writing—review and editing, D.Z. and J.O.; supervision, D. ˙ Z.; project administration, D. ˙ Z.; funding ˙ acquisition, D.Z. All authors have read and agreed to the published version of the manuscript. ˙

**Funding:** This research was funded by Lodz University of Technology, Institute of Food Technology and Analysis, grant number 501/5-54-1-2.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** All the data are included in the present study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations**



#### **References**


## *Article* **Effect of Inoculated Lactic Acid Fermentation on the Fermentable Saccharides and Polyols, Polyphenols and Antioxidant Activity Changes in Wheat Sourdough**

**Ewa Pejcz 1,\* , Sabina Lachowicz-Wi´sniewska <sup>1</sup> , Paulina Nowicka <sup>2</sup> , Agata Wojciechowicz-Budzisz <sup>1</sup> , Radosław Spychaj <sup>1</sup> and Zygmunt Gil <sup>1</sup>**


**Abstract:** Inoculation of sourdough allows the fermentation medium to be dominated by desired microorganisms, which enables determining the kinetics of the conversion of chemical compounds by individual microorganisms. This knowledge may allow the design of functional food products with health features dedicated to consumers with special needs. The aim of the study was to assess the dynamics of transformations of fermentable oligosaccharide, disaccharide, monosaccharide and polyol (FODMAP) compounds from wheat flour as well as their antioxidant activity during inoculated and spontaneous sourdough fermentation. The FODMAP content in grain products was determined by the fructan content with negligible amounts of sugars and polyols. To produce a low-FODMAP cereal product, the fermentation time is essential. The 72 h fermentation time of *L. plantarum*inoculated sourdough reduced the FODMAP content by 91%. The sourdough fermentation time of at least 72 h also positively influenced the content of polyphenols and antioxidant activity, regardless of the type of fermentation. The inoculation of both *L. plantarum* and *L. casei* contributed to a similar degree to the reduction in FODMAP in sourdough compared to spontaneous fermentation.

**Keywords:** sourdough fermentation; inoculation; lactic acid bacteria; FODMAP; fructans; antioxidant activity

#### **1. Introduction**

Sourdough is traditionally prepared by mixing flour with water, and subjecting this mixture to a multi-stage spontaneous fermentation, which is carried out by exogenous flour microflora, including mainly 104–10<sup>7</sup> CFU/g of bacteria and yeast [1]. In order to shorten the technological process and increase its repeatability, it is an increasingly common practice to add starter cultures to sourdough. A group of lactic acid bacteria (LAB) plays a key role in these processes and has a long and safe history of use and consumption in fermented foods and beverages [2]. Another solution is to inoculate fermented products, including bakery sourdoughs, with pure cultures of bacteria or yeast proliferated to a desired number of colony-forming units [3,4]. Sourdough fermentation allows the fermentation medium to be dominated by desired microorganisms, which enables determining the kinetics of the conversion of chemical compounds of flour by individual microorganisms, and their targeted selection [5].

Cereal products make up a significant proportion of food consumed by the worldwide population. Wheat bread is considered a rich source of fermentable oligosaccharides, disaccharides, monosaccharides and polyols (FODMAPs) due to a high content of fructans, formed by the aggregation of fructose molecules. FODMAPs are easily fermentable, highly

**Citation:** Pejcz, E.;

Lachowicz-Wi´sniewska, S.; Nowicka, P.; Wojciechowicz-Budzisz, A.; Spychaj, R.; Gil, Z. Effect of Inoculated Lactic Acid Fermentation on the Fermentable Saccharides and Polyols, Polyphenols and Antioxidant Activity Changes in Wheat Sourdough. *Molecules* **2021**, *26*, 4193. https://doi.org/10.3390/ molecules26144193

Academic Editor: Daniel Cozzolino

Received: 16 June 2021 Accepted: 8 July 2021 Published: 10 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

osmotic carbohydrates, including fructooligosaccharides (FOSs), galactooligosaccharides (GOSs), lactose, fructose and polyols (notably sorbitol and mannitol) [6–8].

The effect of FODMAPs on human health is determined by the amount of sugar delivered to the body within food. The appropriate intake of FODMAPs has a positive impact on human health because certain FODMAP sugars exhibit prebiotic effects [9,10]. The excess intake of FODMAP-rich products (above 20 g/day) can lead to sugar accumulation in the intestines, which in turn may induce various gastric ailments, which are acute in people suffering from irritable bowel syndrome [6,8].

Irritable bowel syndrome (IBS) is a gastrointestinal disorder that can appear in persons of various ages, genders and ethnical origins. It affects 4–20% of the population. Its typical symptoms usually appear after the intake of FODMAP-containing food products and include abdominal discomfort and stomachache, accompanied by flatulence, constipation or diarrhea [11,12]. Simple sugars and polyols exhibit a stronger osmotic effect, whereas saccharides such as fructans, FOSs and GOSs are more susceptible to fermentation by the intestinal microbiome [8,13].

Research has shown that FODMAP components trigger clinical symptoms in IBS patients [14–16]. One of the diets most often recommended by dietitians to help combat IBS symptoms is the low-FODMAP diet. Its principle is to reduce the intake of food products containing short-chain carbohydrates, which are rapidly absorbable in the human gastrointestinal tract [17]. In addtion, dietitians advise paying attention to the fructose:glucose ratio in consumed food products and recommend that their levels are similar or a higher glucose content. This can help improve the intestine's capability to absorb fructose [6]. Food products rich in these compounds include cereal products rich in fructans [11].

Fructans are not digested nor absorbed in the human digestive tract [6]. When ingested in small amounts, fructans have some health benefits but their excess can cause various ailments of the digestive tract [9]. The low-FODMAP dietary guidelines recommend substituting traditional bread with gluten-free products [11,14,18]. Wheat bakery products have significantly higher contents of protein, dietary fiber, minerals and vitamins than the gluten-free ones. Therefore, the exclusive consumption of gluten-free products can lead to deficiencies of these compounds in the body [19].

The FODMAP content in bread depends on both flour type and bread-making method [20]. The content in bread can be reduced in many ways, one of which is to use sourdough in the bread-making process. Another means is to appropriately select microorganisms responsible for the fermentation and degradation of sugars that trigger the gastrointestinal disorders [8]. In wheat bread, the above goal can also be achieved by extending fermentation time, which not only improves the flavor values of bread but also effectively decreases FODMAP content [20]. Fructans present in high quantities in cereal kernels can be degraded during sourdough fermentation. The consumption of sourdough bread has been proved to have a beneficial effect on mitigating irritable bowel syndrome symptoms [8]. In order to produce a low-FODMAP bread, LAB should also be added to the sourdough as they enhance the metabolic activity of fermenting flora. Apart from their capability to metabolize fructans, LAB can also convert free fructose to mannitol. In addition, they produce α-galactosidase, i.e., an enzyme responsible for breaking the bonds between the molecules of sugars constituting GOSs [8]. In turn, the enzymes capable of mannitol conversion are secreted by, e.g., *Lactobacillus delbrueckii*, *Lactobacillus casei*, *Lactobacillus plantarum* and *Lactobacillus salivarius* [7,21]. Bread produced with sourdough requires longer fermentation, which entails multiple changes in the carbohydrate composition. Microbial invertase rapidly degrades flour saccharose into glucose and fructose. Afterward, glucose is consumed as a source of energy, whereas fructose can be reduced by heterofermentative LAB to mannitol. All fermentable carbohydrates are rapidly depleted in the first hours of fermentation, whereas the carbohydrates featuring a high degree of polymerization (like fructans) are consumed later [8,22].

Sourdough fermentation used in bread making improves the nutritional value and antioxidative properties of bread, as well as its taste, aroma, texture and stability, and finally the bioaccessibility of its elements [23]. The antioxidant activity of the components of sourdough depends on the type of inoculum used for fermentation [24] and sourdough fermentation time [25]. The aim of the study was to assess the dynamics of transformations of FODMAP compounds from wheat flour as well as the antioxidant activity of nutrients of flour during inoculated and spontaneous sourdough fermentation.

#### **2. Results and Discussion**

#### *2.1. Dynamics of pH Changes during Fermentation*

Table 1 shows the results of the pH measurement of spontaneously fermented and lactobacilli-inoculated wheat sourdoughs. In each type of sourdough, the greatest decrease in pH was observed after the first 24 h of fermentation. During fermentation, LAB produce lactic acid, which results in a lower pH level [26]. In spontaneously fermenting and *L. casei*inoculated sourdough after the first day of fermentation, pH remained at a similar level. A further slight decrease in pH was observed in *L. plantarum*-inoculated sourdough when the fermentation time was extended to 72 h. The study by Menezes et al. [8] also showed the greatest decrease in the pH level in the first hours of wheat dough fermentation, until relatively stable values were achieved after several stages. Fluctuations in the pH level affect the action of amylases. A study by Struyf et al. [27], showed that lowering the pH level has an effect on maltose release but has no effect on other saccharides.

**Table 1.** pH of wheat sourdough during fermentation.


Values represent the means of four replicates. Mean values in columns with different letters are significantly different according to Duncan test at *p* ≤ 0.05.

#### *2.2. Dynamics of FODMAP Content Change during Fermentation*

Changes in the FODMAP content in the sourdoughs during their fermentation are presented in Table 2. Fructans constituted the majority of these compounds in the tested samples. The content of fructans in the sourdough was influenced by the fermentation time and the type of LAB used. Each extension of the fermentation time resulted in a significant decrease in the content of fructans in the sourdough compared to the control, which was non-fermented sourdough. For each of the sourdough types, the content of fructans decreased with the fermentation time and reached the lowest values after 72 h of fermentation. A similar relationship between the extension of the fermentation time and the decrease in the content of fructans was observed by Struyf et al. [28], where after 1 h of fermentation, more than half of the fructans were degraded in the dough compared to the content of fructans present in the flour. In the study by Gélinas et al. [29], it was found that 20% of fructans were degraded after the dough-mixing process. Then, by fermenting the dough with yeast for 180 min, the fructan content was reduced by 82% compared to the amount of fructans present after mixing the dough. For fermentation lasting 24 h, the sourdough fermented with *L. plantarum* achieved the lowest content of fructans among the analyzed sourdoughs. However, in the case of 48 h and 72 h fermentation, the lowest fructan content was observed in sourdoughs inoculated with *L. casei*. Fraberger et al. [30] tested 13 strains of microorganisms for their ability to reduce fructans and found that the metabolism of microflora contributed to a significant reduction in the content of fructans in the dough compared to the control sample. Sourdough fermented with *L. casei* bacteria reached a lower content of fructans faster compared to sourdough fermented with *L. plantarum* and this could be due to the higher activity of *L. casei* enzymes than *L. plantarum* [7].


**Table 2.** The content of FODMAP components (g/100 g d.m.) in wheat sourdough.

Nd: not detected. Values represent the means of two replicates. Mean values in columns with different letters are significantly different according to Duncan test at *p* ≤ 0.05.

> The non-fermented sourdough control sample did not contain free glucose (Table 2). After 24 h of spontaneous fermentation, the glucose content was 0.06 g/100 g d.m., then after 48 h its value increased to 0.08 g/100 g d.m., and after 72 h it dropped back to 0. In the case of sourdough inoculated with *L. plantarum*, after 24 h and after 48 h of fermentation, the glucose content was 0.05 g/100 g, and after 72 h, its content in the sourdough decreased to 0. In the case of sourdough fermented with *L. casei*, the content of glucose increased to 0.2 g/100 g d.m. after 24 h of fermentation, and after both 48 and 72, its value dropped to 0. The glucose level in sourdough is determined by the content of damaged starch and the activity of β-amylase and amyloglucosidase [22]. It was also found that it is a factor blocking the transformations of, among others, sucrose, raffinose and mannitol. A fermentation time of 72 h led to a complete reduction of glucose in the sourdough. Further changes in glucose may result in the formation of CO2, lactate, acetate and ethanol [7,21,27]. No fructose content was observed in any of the analyzed sourdough. It is consumed quickly and can also be converted into mannitol by lactobacilli [7].

> The presence of mannitol was not found in any of the analyzed sourdough during the first 24 h of fermentation, because mannitol is formed from the degradation of fructose, which is transformed in the later stages of fermentation [31]. No mannitol was detected in the spontaneously fermented sourdough for 24 as well as 48 h, and after 72 h its value increased to 0.007 g/100 g d.m. In the sourdough with the addition of *L. casei* bacteria, after 48 h, the mannitol content was found at the level of 0.006 g/100 g of dry matter, and after 72 h, the content decreased to 0.002 g/100 g d.m. In sourdough fermented with *L. plantarum*, the level of mannitol remained at 0 during 72 h of fermentation. Gänzle [21] claims that the degradation of mannitol requires the enzymes of lactobacilli found, among others, in *L. casei* bacteria. In the spontaneously fermented sourdough, mannitol was present only after 72 h of fermentation, which results from the metabolism of fructose. It is converted into mannitol by lactobacilli, therefore in pure bacterial cultures fructose was degraded to mannitol faster than in the case of spontaneously fermenting sourdough. Mannitol metabolism, however, may be inhibited by the presence of glucose [7,21].

> The total FODMAP content before fermentation was 1.153% d.m. and was determined by the fructan content of the flour. The FODMAP content of wheat is influenced by its variety. Ziegler et al. [20] studied the content of compounds from the FODMAP group in two wheat flour varieties and showed that it is from 1.24 ± 0.38 to 2.01 ± 0.42 g/100 g d.m. The fermentation of the flour always resulted in a significant decrease in the FODMAP content, but with a different effect depending on the type of sourdough used and its duration. In the spontaneously fermenting sourdough, the FODMAP content decreased with the extension of the fermentation time, and it reached the lowest value after 72 h. The FODMAP content in the spontaneously fermenting sourdough in the study by Menezes et al. [8] was

0.553 g/100 g d.m. and 0.603 g/100 g d.m. depending on various parameters of sourdough fermentation. A similar effect was observed in *L. casei*-inoculated sourdough, but with slight difference in FODMAP content after 48 and 72 h of fermentation. Sourdough fermentation with the addition of *L. plantarum* resulted in the lowest FODMAP content after 72 h and was constant after 24 and 48 h. In the study of Menezes et al. [8], it is claimed that the sourdough biotechnology requires a longer fermentation time than is usually used in bread making (0.5–3 h). Carbohydrates such as sucrose, maltose, glucose and fructose are depleted quickly during the first hours of fermentation, while higher-polymerized carbohydrates such as fructans are used later, so longer fermentation of sourdough will degrade all FODMAP components more efficiently. Comparing sourdoughs after 24 h of fermentation, the one with the addition of *L. plantarum* had the lowest content of FODMAP, while after 48 and 72 h of fermentation, the lowest FODMAP concentration was in sourdough with *L. casei*. Finally, after 72 h of fermentation with the addition of *L. casei*, the lowest FODMAP level of 0.076 g/100 g d.m. was achieved, which is a reduction of their content by 93%. It is important to select the microorganisms responsible for the fermentation of the sourdough. Appropriate LAB have enzymes that degrade FODMAP components, and they also have the ability to lower the pH of the environment, thanks to which the activity of the enzymes increases, which leads to a reduction in the FODMAP content. By lowering the FODMAP content in wheat bread, it is possible to reduce the symptoms of irritable bowel syndrome [7,8,30].

#### *2.3. Dynamics of Polyphenolic Compounds and Antioxidant Activity Changes during Fermentation*

The total content of polyphenols and the antioxidant activity of sourdoughs are presented in Table 3. The content of polyphenols in the sourdough was higher after each type of fermentation than before. However, the content of polyphenols in the analyzed material did not totally change. The matrix of the components of flour and sourdough was loosened during fermentation and water-extractable polyphenols were released. The fermentation process may increase the antioxidant activity by increasing the amount of easily extractable phenolic compounds [24]. Spontaneously fermenting sourdough reached the highest content of polyphenols after 24 h of fermentation, after which their amount remained on a similar level. *L. casei*-inoculated sourdough contained the highest amounts of polyphenols after 48 and 72 h of fermentation. The content of polyphenols in *L. plantarum*inoculated sourdough increased significantly after 24 h of fermentation and then again after 72 h. Chis, et al. [32] observed an increase in the content of polyphenols with the fermentation time with the addition of *L. plantarum*, which is explained by their proteolytic activity's influence on the polyphenol profile. LAB can affect polyphenols, improving their solubility [33].

The antioxidant activity measured by both ABTS and FRAP methods of the spontaneously fermenting sourdough increased significantly after 24 h of fermentation and then after 72 h. In the study of Banu et al., 2010 [24], the addition of starter cultures containing *Lactobacillus rhamnosus* to the dough increased the antioxidant activity compared to spontaneous fermentation. In this study, the sourdough inoculated with *L. casei* showed a higher antioxidant activity against the ABTS radical after 24 h fermentation than before, and the highest value was achieved after 72 h of fermentation. A significant increase in the ability to reduce iron ions of this sourdough took place only after 72 h of fermentation. The antioxidant activity of *L. plantarum*-inoculated sourdough increased significantly after 72 h of fermentation. In the study of Banu et al. [24], antioxidant activity (measured with ABTS and DPPH methods) of 20 h spontaneously fermented dough was almost two times higher than before fermentation. Colosimo et al. [25] observed a significant increase in polyphenols and antioxidant activity with the fermentation time of the sourdough, which should last 72 h and preferably 96 h. In a study by Rodríguez et al. [34], *L. plantarum* was able to increase the antioxidant activity and improve the aroma profile of the product by degrading certain phenolic components through the metabolic activity of the LAB. The metabolic activity of LAB influences the levels of bioactive ingredients, which allows for an

increase in antioxidant activity. During fermentation with their participation, antioxidant peptides are released, which increases the amount of phenols and antioxidant activity by acidification and hydrolysis of more complex and glycosylated forms [24,35]. Extending fermentation to 72 h resulted in an increase in the antioxidant activity of sourdoughs by 83 to 98% compared to the samples before fermentation, regardless of the type of sourdough fermentation. Sourdough fermentation can remove peptides associated with human intolerance to grain products. It can also lead to the production of bioactive peptides with antioxidant potential, which may affect the bioavailability of nutrients [25].


**Table 3.** The content of polyphenolic compounds and antioxidant activity of wheat sourdough.

Values represent the means of three replicates. Mean values in columns with different letters are significantly different according to Duncan test at *p* ≤ 0.05.

#### **3. Materials and Methods**

*3.1. Material*

Wheat flour type 650 was supplied from GoodMills (Stradunia, Poland). The flour particle size was 93 ± 0.3 µm, it had falling number of 390.5 ± 1.0 and contained 14.72 ± 0.02% protein (data not shown). Lyophilizates of two safe and well-described species of lactic acid bacteria: *Lactobacillus casei*, catalogue number 20,011 and *Lactobacillus plantarum*, catalogue number 20,174, were purchased from DSMZ—German Collection of Microorganisms and Cell Cultures (Leibniz, Germany).

Lactobacilli were grown in Man, Rogosa and Sharp medium (MRS) (Sigma-Aldrich, Hamburg, Germany) and incubated under aerobic conditions at 37 ◦C until the late exponential growth phase was reached (about 24 h). Cells were harvested by centrifugation at 10,000 rpm for 10 min at 4 ◦C. Dilutions were made in saline solution plated on MRS 273 agar, resulting in a concentration of about 10<sup>9</sup> CFU/mL.

The next multiplication of microorganisms took place by preparing a mixture of 100 g of flour, 300 mL of water and 20 mL of liquid microorganism culture (*L. casei* and *L. plantarum*). A mixture without the addition of bacteria was prepared based on the spontaneous fermentation of microorganisms found naturally in the flour. The fermentation lasted three days at 28 ◦C.

Sourdoughs were made from a combination of flour (500 g), water (500 mL) and the appropriate liquid sourdough prepared in the previous step (50 mL). The fermentation of sourdoughs was carried out for 24, 48 and 72 h at a temperature of 28 ◦C.

#### *3.2. Methods*

3.2.1. Dynamic of Fermentation

The pH of the sourdoughs was determined in four replicates after 24, 48 and 72 h of fermentation using the potentiometric method. The pH of the non-fermented sourdough was used as a control. The samples were frozen, freeze dried, ground and vacuum packed for further determinations.

#### 3.2.2. Determination of Fructans

The content of fructans in the freeze-dried sourdough samples was determined using the fructan determination kit based on AOAC Method 999.03 [36], which is based on the determination of the fructose content in the samples resulting from the enzymatic breakdown of fructans. Using a spectrophotometer, the fructose content was measured at a wavelength of λ = 410 nm. The determination was performed in duplicate.

#### 3.2.3. Determination of Sugar and Polyol Content by HPLC-ELSD

Preparation of samples for the determination of sugar and polyol content consisted of adding 10 g of the analyzed sample into a volumetric flask, filling the volumetric flask to 50 mL and boiling and shaking the samples in a boiling water bath for 20 min. Then, 100 mL of cooled samples was made up with distilled water, 10 mL of the extract was centrifuged (10,000 rpm, 10 min) and the samples were filtered on a Sep-Pak C-18.

The content of sugars and polyols was determined by the HPLC method coupled with a light scattering detector. A 40 µL sample was injected by an autosampler (L-7200) onto a Unison UK-Amino 3 µL (3 mm × 250 mm) column (Imtakt, Kyoto, Japan). Detection was performed using an evaporative light scattering detector (PL-ELS 1000) with the following input parameters: evaporator temperature −80 ◦C; nebulizer temperature −80 ◦C; nitrogen flow −1.2 SLM. The elution was performed at 30 ◦C in an isocratic flow using 85% acetonitrile solution at a flow rate of 0.7 mL/min. FODMAP content was identified by comparing with standard HPLC area measurements. The measurements were performed in duplicate and the results were expressed in grams/100 g dry weight of the product. The sum of the FODMAPs was calculated from the fructan content and those identified in the samples: fructose, mannitol and glucose.

#### 3.2.4. Determination of Polyphenolic Compounds and Antioxidant Activity

The extraction for the antioxidant capacity was conducted following a protocol described by Lachowicz et al. [37]. The total polyphenolic content of the sourdough samples was determined using the Folin–Ciocalteu spectrophotometric method [38]. The absorbance at 765 nm was measured after 1 h, using the UV-2401 PC spectrophotometer (Shimadzu, Kyoto, Japan). The results were expressed as mg of gallic acid equivalents (GAE) per 100 g of dry sourdough. Data were expressed as the mean value for three measurements. The ABTS and FRAP methods were carried out with the methods described by Re et al. [39] and Benzie and Strain [40]. The absorbance was measured at 734 nm and 593 nm using the UV-2401 PC spectrophotometer (Shimadzu, Kyoto, Japan). The results of antiradical capacity were expressed as Trolox equivalents in mmol per 100 g of dry sample. Data were expressed as the mean value for three measurements.

#### *3.3. Statistic Analysis*

The results were statistically analyzed with the Statistica 13.3 software package (Stat-Soft, Tulsa, OK, USA). One-way ANOVA at *p* ≤ 0.05 was calculated and homogeneous groups according to the Duncan test were estimated.

#### **4. Conclusions**

The FODMAP content in grain products turned out to be determined by the fructan content with negligible amounts of sugars and polyols. To produce a low-FODMAP cereal product, the fermentation time is essential, and its extension to 72 h or more allows for a strong reduction in the content of these compounds. A sourdough fermentation time of at least 72 h also positively influences the content of polyphenols and antioxidant activity, regardless of the type of fermentation. The inoculation of both *L. plantarum* and *L. casei* contributed to a similar degree to the reduction of FODMAPs in sourdough compared to spontaneous fermentation. Knowledge of the processes that take place during the fermentation of inoculated sourdoughs may allow the production of food products designed according to the needs of consumers.

**Author Contributions:** Conceptualization, E.P. and Z.G.; methodology, E.P., S.L.-W. and P.N.; investigation, E.P., S.L.-W., P.N.; writing—original draft preparation, E.P.; writing—review and editing, R.S. and A.W.-B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Wrocław University of Environmental and Life Sciences; grant name "Innovative scientist", grant number N060/0033/20.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Results will be available from the corresponding author.

**Acknowledgments:** The authors thank Patrycja Jurczyk for contributing to the research.

**Conflicts of Interest:** The authors declare no conflict of interest.

**Sample Availability:** Sourdough lyophilisates are available from the authors upon reasonable request.

#### **References**

