Extruded corn gruels containing linden flowers: quantitation of phenolic compounds and selected quality characteristics

Abstract Extrusion-cooking of plant materials may enhance antioxidant activity and improve health benefits. Selected antioxidant polyphenols in extruded corn gruels enriched with different amounts of linden flowers were determined by LC-ESI-MS/MS and quality characteristics were determined. Phenolic content increased with Tiliae inflorescentia addition and was not decreased by high-temperature extrusion. Linden flower incorporation into instant gruels should be limited to 10% to retain acceptable sensory properties. Graphical Abstract


Introduction
Extrusion-cooking has become a popular high temperature, short time processing method for vegetable raw materials in snacks, pasta, flat bread, instant gruels, baby foods, confectionery, modified starches and pet food [1]. Grits, gruels and instant cereals are basic carbohydrate sources in baby and infant diets. These products should be processed to improve their dispersibility and digestibility, since a 3-4 month-old baby's pancreas has a limited ability to digest starch. Extrusion-cooking is much more efficient than traditional methods of infant cereals processing, e.g. drum or roll drying [2,3].
Extrusion-cooking temperature, screw speed, moisture content, feed rate and residence time distribution are crucial for extrudate nutritional characteristics and antioxidant activity. The process changes texture, gelatinizes starch, cross-links proteins, and creates flavors [4]. It may inactivate antinutritional factors, denature undesirable enzymes or reduce lipid oxidation [5].
Extrusion-cooking may enhance food antioxidant activity and form resistant starch or insoluble fiber [6][7][8]. It retains more nutrition and active compounds than other thermal treatments, especially slow cooking or deep frying, but it is comparable to blanching, freezing or fermentation [9]. Analysis of extruded product extract showed higher antioxidant activity than in the raw materials; phenolics may be released [10][11][12][13][14].
Khanal et al. [15] reported the effects of extrusioncooking on procyanidin monomers and dimers in grape seed and pomace. It appears to increase the level of low molecular weight bioactive compounds (procyanidins etc.) and biologically important monomers and dimers from polymer chains [16]. Extruded product enrichment with bioactive compounds is limited by secondary plant metabolites' low thermal stability and their susceptibility to changes during processing [5,17].
Fruits and vegetables are the most popular components added to cereal extrudates, but some waste products or functional compound isolates may be used [8,13,[18][19][20]. Linden flower (Tiliae inflorescentia) is one of the herbal supplements most often used. Its main constituents include flavonoids (quercetin glycosides, kaempferol glycosides, tiliroside), phenolic acids, essential oils, phytosterols, organic acids, tannins, mucilage, minerals, niacin, and vitamin C [21][22][23]. Its pharmacological activity is in feverish colds, infectious diseases, and bronchitis, but it also has sedative and diuretic actions [24,25]. It is used as extracts or infusions (teas), but it may also be a pro-health component in infant or baby food.
Our aim was to determine selected polyphenols in extruded corn gruels enriched with different amounts of linden flower by LC-ESI-MS/MS and to evaluate their quality characteristics.

Extrusion-cooking procedure
Blends of corn grits and ground linden flowers were prepared by mixing dry components in weight ratios of 100:0, 99:1, 97:3, 95:5, 90:10 and 80:20. The blended samples were conditioned to 15% moisture by spraying with water and mixing continuously for 10 minutes.
Blends were processed in a TS-45 single screw extrusioncooker (ZMCh Metalchem, Gliwice, Poland) with barrel length / screw diameter = 12, and screw compression ratio 3:1. The barrel temperatures were 120-135-140°C in the first zone, second zone, and forming die. The screw speed was 125 rpm and a 3 mm circular open die shaped the extrudate [29,30]. After cooling (extrudate moisture content was 6.0%) extrudates were ground in an iG5A laboratory grinder (TestChem, Poland) to less than 1 mm for instant gruels [3,29]. Samples were stored in polyethylene bags at room temperature.

Extraction procedure
Before extraction dry plant material was milled and sieved. Three 2 g portions were each extracted 3 times with 40 mL 80% aqueous ethanol in a Sonorex RK 100H ultrasonic bath (Bandelin Electronic, Germany) (20 kHz, 100 W) for 30 min at 60°C [26]. Extracts were filtered, combined and evaporated to dryness. Residues were dissolved in 10 mL of methanol.

Solid phase extraction (SPE)
Crude extracts were purified by SPE. 5 mL was passed through a previously conditioned Bakerbond C18 SPE column. Polyphenols were eluted with 5 mL 60% aqueous methanol followed by 10 mL of 30% aqueous methanol. The combined extracts were evaporated to dryness and dissolved in 10 mL of methanol in a volumetric flask. The procedure was repeated three times.

LC-ESI-MS/MS analysis
Analysis was performed using reversed-phase highperformance liquid chromatography with electrospray ionization mass spectrometry (LC-ESI-MS/MS). An Agilent 1200 HPLC system (Agilent Technologies, USA) equipped with a binary gradient solvent pump, degasser, autosampler and column oven connected to a 3200 QTRAP Mass spectrometer (AB Sciex, USA) was used.
For each compound the optimum multiple reaction mode (MRM) conditions were determined in the infusion mode. The data was acquired and processed using Analyst 1.5 software (AB Sciex, USA). Triplicate injections were made for each standard and sample. Analytes were identified by comparing retention time and m/z values obtained by MS and MS 2 with those of standards obtained under the same conditions (Tables 1 and 2). The calibration curves obtained in MRM mode were used for quantitation; peak areas were compared with calibration curves generated by three repeated injections of known standards at seven concentrations (0.005-50 ng μL −1 ) [27]. Linearity ranges for calibration curves were determined. The limits of detection (LOD) and quantitation (LOQ) for phenolic compounds were determined at signal-to-noise ratios of 3:1 and 10:1 by injecting a series of dilute solutions with known concentrations. Calibration parameters, LOD and LOQ values are in Table 3.

Physical properties
Extrudate expansion, which varies with processing and composition, is an important quality factor and should be as great as possible. The radial expansion index is the ratio of extrudate to die diameter [30]. Ten replicate measurements were done for each recipe. The water absorption index (WAI) is the ratio of gel to dry sample. It was determined as described previously [31] with some modification. Ground extrudates (0.7 g) were suspended in 7 mL room temperature water in plastic tubes and mixed. After gentle stirring for 10 min the tubes were closed and centrifuged for 10 min at 12500 x g in a T24 centrifuge (VEB MLW MEDIZINETECHNIK, Leipzig, Germany) at ambient temperature (21°C). WAI was the ratio of gel remaining after supernatant removal to the weight of original dry solids. The supernatant was dried in an air oven at 105°C to constant weight. The water solubility index (WSI) is the percentage of dry matter recovered after the supernatant is evaporated from the WAI determination. WAI and WSI determinations were replicated in triplicate.
For sensory tests gruels were prepared by mixing with 45°C water (20:80 gruel:water) and served. Sensory characteristics include the taste, color, flavor, consistency, and overall quality. A 15-member semi-trained panel judged gruels on a 5-point scale (1 = weak, 5 = very good). Acceptability was evaluated on a 9-point hedonic scale, where: 1=dislike extremely, 9=like extremely. Gruels were deemed acceptable if their mean acceptability scores were above 5 [34].

Determination of polyphenols content
Our objective was the qualitative and quantitative analysis (LC-ESI-MS/MS) of phenolic extracts of corn gruels containing linden flower. Ultrasound assisted extraction [26] followed by SPE was precise and accurate (Tables 2  and 3). Ethanol, 80% aqueous ethanol, methanol, and 80% aqueous methanol were extractants; 80% aqueous ethanol gave higher yields of all the polyphenols because of their relatively polar nature.
Precision was evaluated by intra-day and inter-day tests. Intra-day experiments were performed by replicate analysis of six aliquots of the same sample within one day. Inter-day tests were carried out on three consecutive working days in the same way as intra-day experiments. Three peak area measurements for each component were carried out. The standard deviations are in good agreement with the requirements for a developed method ( Recovery studies were performed to assess accuracy. Crude extracts were spiked with standard solution (three concentration levels) and SPE and analysis were carried out as for real samples. The experiment was repeated three times. Recoveries ranged from 92.9% (tiliroside) to 102.3% (gentisic acid), demonstrating the method's accuracy. a SD -standard deviation (n=3), b BQLpeak detected, concentration lower than the LOQ but higher than the LOD Phenolic content increased with the addition of linden flower (Table 4). In gruels without linden addition only ten phenolic compounds were identified. These were 4-OH-benzoic, p-coumaric, salicylic, ferulic, and synapic acids as well as rutin, isoquercetin, kaempferol-3rutinoside, astragalin, and apigenin-7-glucoside. In Tiliae inflorescentia extract twenty polyphenols were found. The additional ones were gallic, protocatechuic, vanillic, caffeic, syringic, gentisic, rosmarinic acids along with hyperoside, quercetrin, and tiliroside. In gruel with 5% and 10% linden addition eighteen phenolic compounds were identified, and nineteen were found in gruel with 20% Tilia. Most polyphenols increased proportionally to the linden addition, but some phenolic acids (e.g. p-cumaric and ferulic acids) did not increase significantly with additional flower. This may be due to limited solubility of these acids in the extractant or acid content variation in the linden.
Extrusion process conditions preserve the antioxidant compounds; high-temperature extrusion-cooking does not deactivate polyphenols in raw Tilia inflorescentia. This agrees with Özer and coworkers [35] who found that extrusion-cooking had no effect on total phenolic content in extrudates containing chickpea, corn, oat, carrot and hazelnut.

Physical properties of enriched snacks
Radial expansion indices are in Table 5. Increased linden decreased expansion (R 2 = 0.954). The difference between reference corn and extrudates with the highest linden addition was 30.8%, which may reflect increased fiber in the products.
WAI measures granule or starch water absorption after swelling in excess water. WSI is the free polysaccharide or polysaccharide released after water addition [8,31]. The WAI decreased significantly as additive increased (R 2 = 0.9715), attributed to decreasing starch content with increasing linden and less water absorption by starch. WAI (Table 5) ranged from 5.51 g g -1 for corn gruels to 4.65 g g -1 for extrudates containing the most linden. Replacement of starchy raw materials by vegetables, fruits, or highfiber additives reduces the starch undergoing swelling and gelatinization during processing so WAI is usually much lower with more additives and can be temperaturedependent [36]. Testing barley-tomato pomace extrudates, Altan and coworkers [37] reported that WAI decreases from 7.03 to 6.10 g g -1 with increasing temperature and pomace level.
WSI is often used as an indicator of starch degradation. It depends mostly on starch granule disruption, amylose and amylopectin depolymerization, starch gelatinization and the consequent starch solubilization generated in the extruder by pressure, heating, shearing and residence time [1,8,38]. WSI results (Table 5) showed decreased solubility with linden enrichment (from 15.32% for corn extrudates to 7.86% for the highest additive level). Increasing additives increases fiber content and decreases starch, reducing solubility.
Linden addition decreased L* values (R 2 = 0.953, Table 5). Green-red balance varied from -3.94 for a sample without additives to -2.66 for 10% and 0.28 for 20% linden; the product became more red with increasing additive. The nature of the linden used may be key to the redness  Table 1. See Experimental section for details.  Table 2. See Experimental section for details. and yellowness improvements with increasing additive (R 2 = 0.9133 and 0.9765). b* increased from 24.43 for control gruels to 30.87 for extrudates with 20% linden. Yellowness was observed in all samples because carotenoids are present in both corn grits and linden.
Color, flavor, taste and consistency characteristics are presented in Table 6. The best colors were for corn snacks and gruels with 1% additive; increasing linden lowered the color results (R 2 = 0.9607). The panelists noted that linden flower addition darkened and gave a red tint to gruels, confirmed by instrumental measurements (Table 5). Flavor and taste notes also decreased with increased linden content (R 2 = 0.9722 and 0.9896), because of the distinctive linden odor and taste. Good sensory properties were found with linden addition limited to 5%, with overall quality notes 3.86−3.12. More additive significantly lowered the sensory notes to 2.45 and 2.01 for extrudates with 10 and 20% linden. These high proportions also lowered gruel acceptability (R 2 = 0.9716) ( Table 6) due to the darker color, intense flavor and distinctive linden taste. Acceptability results were below 6.0 for those with 10% and 20% linden .

Conclusions
High-temperature extrusion-cooking did not decrease the polyphenolics present in both raw materials. Linden herb enriched instant gruels demonstrate the incorporation of nutritionally functional components into food products by extrusion-cooking. 10 and 20% linden incorporation lowered extrudate expansion, reduced WAI and WSI values and considerably lowered instant gruel acceptance due to intense herbal taste and flavor. Linden in instant gruels should be below 10% to improve nutritional characteristics and retain acceptable sensory properties.
Acknowledgment: This work used equipment purchased by the project "The equipment of innovative laboratories doing research on new medicines used in the therapy