Freeze-dried grape skin powder enhances the antioxidant status of fortified processed cheese spread more than the oven-dried grape skin powder does.

Grapes of the white variety ‘Müller Thurgau’ were harvested from the Prostřední Hory (Bzenec, Czech Republic) vineyard track during September 2017. After pressing the grape berries to obtain a liquid for wine manufacturing, a portion of grape pomace was immediately stored at -20 °C in an evacuated plastic package. Before processing, grape pomace was thawed and the needles and seeds were removed using an analytical sieve (mesh size 0.5 × 0.5 cm). Grape skins (GS) were dried at the following conditions: oven-drying (OD) was performed in laboratory air-forced oven (HS62A, Chirana, Brno, Czech Republic) at 46 °C for 24 h. Freeze-dried (FD) samples were prepared at -40 °C (12 Pa) for 48 h (CoolSafe 100-4, Trigon Plus, Čestlice, Czech Republic). Dried grape skins were milled at 5000 rpm for 10 s with a Grindomix GM 200 (Retsch GmbH, Haan, Germany) and sieved to obtain particles <800 μm. Both GS-OD and GS-FD powders were stored in a tightly sealed plastic pack at -20 °C until use. The contents of crude protein (Method 960.52), fat (Method 920.39), moisture (Method 934.01), ash (Method 930.05) and total dietary fibre (Method 985.29) were determined according to AOAC (Horwitz, 2000) procedure in duplicate. The total content of saccharides was calculated from differences.

The composition of raw materials, including Eidam cheese (dry matter ≈50 g.100 g^-1 and fat in dry matter ≈30 g.100 g^-1, 8-week maturity), butter (dry matter ≈84 g.100 g^-1 and fat in dry matter ≈98 g.100 g^-1) and water, was adjusted to obtain processed cheese with dry matter ≈37 g.100 g^-1 and fat in dry matter ≈50 g.100 g^-1. A ternary mixture of monosodium dihydrogenphosphate (19%), disodium hydrogen phosphate (37%), tetrasodium diphosphate (22%) and the sodium salt of polyphosphate (22%) was used in a total concentration of 2.8 g.100 g^-1. For the preparation of functional processed cheese spread (PCS), OD-GS and FD-GS powders were added to produce PCSOD-GS and PCSFD-GS samples. Both grape skin powders were added at 1.0 and 2.0% (w/w) levels. Processed cheese without grape skin powder served as a control. Model processed cheese was manufactured in Stephan UMC-5 (Stephan Machinery GmbH, Halmen, Germany) equipment with indirect heating as is described in the flow chart below (Figure 1). Eidam block cheese and butter were cut into small pieces, put into the kettle and minced for 30 s. Then water, emulsifying salts and GS powder were added into the blend. The mixture was heated at 90 °C for 13 min at a constant agitation of 1500 rpm. Samples were poured into 80 g polystyrene doses with caps. The packages were cooled down to 6 °C and stored at -20 °C to avoid deterioration. Dry matter (DM), fat and protein contents were evaluated according to Method 969.19, Method 2001.14 and Method 960.52 as described in Horwitz (2000), respectively, in three repetitions.

A glass test tube with 1.0 g of dried GS sample and 10.0 mL of 50% methanol solution was put in ultrasonic bath Sonorex TK52 (Bandelin Electronic, Berlin, Germany) for 30 min. A clear supernatant after centrifugation at 4100 rpm for 5 min (Universal 320, Hettich, Tuttlingen, Germany) was obtained.

Extract of PCS samples was obtained according to the procedure described in Přikryl et al. (2018), i.e. 1.0 g of frozen PC sample was extracted into 10.0 mL of 50% methanol solution in an ultrasonic bath for 30 min. Subsequently, elimination of proteins and salts was performed using the procedure of Khalifa, Omar and Mohamed (2017) with a slight modification; the pH of extract was adjusted to 4.0 using HCl (2 M) and precipitated proteins were removed by centrifugation at 6000 rpm for 10 min. Then, the pH of clear supernatant was adjusted to 7.0 using NaOH (1 M) followed by centrifugation at 6000 rpm for 10 min to remove remaining proteins and salts. Supernatants were kept refrigerated and used for antioxidant assays and HPLC analysis. Two extracts per sample were prepared, and each extract was measured in duplicate, resulting in a sample size N = 4.

First, GS were screened for the presence of the following phenolic compounds: quercetin, rutin, (+)-catechin, (-)- epicatechin, resveratrol, caffeic acid, p-cumaric acid and ellagic acid. Secondly, phenolic substances identified in GS extracts were determined in PCSOD-GS and PCSFDGS extracts. Before injection, each extract was filtered through a syringe filter (0.45 μm, Labicom, Olomouc, Czech Republic). Samples were injected into an Agilent 1290 Infinity (Agilent Technologies, Santa Clara, CA, USA) equipped with a degasser, an autosampler, a binary pump, a thermostated column compartment and DAD detector. A Zorbax Elipse Plus 1.8 μm C18 (50 × 2.1 mm; Agilent Technologies) column thermostated at 40 °C was used. For the analysis, 2 μL of the sample were injected. A mixture of 0.01 M ammonium acetate adjusted to pH 3.1 using formic acid (solution A) and acetonitrile (solution B) was used as mobile phase with a flow rate 0.6 mL.min-1. The gradient for solution B was 0-3 min at 3%; 10 min at 20% and 20 min at 80%. The signal was detected at 280 nm. The identification of each peak in chromatograms of the extracts was carried out by comparing retention time and absorption spectrum against a pure standard. Quantitative determinations were done using calibration plots of a selected external standard.

Total phenolic content (TPC) was determined by measuring the complex of antioxidants with Folin- Ciocalteau reagent at 765 nm (DU 530, Beckman Coulter Inc., Brea, USA) using the procedure described in Přikryl et al. (2018). The results were expressed as gallic acid (GAE) equivalents (mg.g^-1 of dry matter (DM)). Total flavonoid content (TFC) was determined using aluminium chloride assay (Denni and Mammen, 2012). The increase of absorbance at 415 nm was proportional to the increase in the content of flavonoids. Results were reported as quercetin (QUE) equivalents (mg.g^-1 DM).

ABTS (2,2´-azino-bis-3-ethylbenzthiazoline-6-sulphonic acid) radical scavenging activity assay was adopted from our previous study (Červenka, et al., 2018). The reaction between ABTS•⁺ and antioxidants was monitored at 734 nm, and the results were reported as Trolox equivalents antioxidant capacity (TEAC) in mg.g^-1 DM.

Reducing power (RP) of extracts was determined via the formation of Prussian blue at 700 nm (ferric-ferrous complex) according to the procedure of Pavithra and Vadivukkarasi (2015).

Order statistic methods for small sample size were used throughout this study. The mean and its deviation were calculated according to Horn’s procedure (Horn, 1983) and were expressed as pivot half sum (P_L) and pivot range (R), respectively. Nonparametric statistical methods were used in this study. The sign test was applied for pair-wise comparisons between means for grape powder samples. As far as processed cheese is concerned, the pair-wise comparison procedure was performed using Tukey’s method. A two-factor Kruskal-Wallis analysis of variance (ANOVA) was applied to determine whether the drying method of GS powder (factor A) and the amount added to PCS sample (factor B) influenced the antioxidant properties in fortified PCS samples. To determine associations among variables, Spearman rank-order correlation coefficients (r) were assessed. All the statistical treatments were done at the probability P = 95%) (Statistica CZ, 12.0, StatSoft CR s.r.o., Prague).

As can be seen from Table 1, there were no statistical differences between the contents of crude protein, fat, total dietary fibre and ash for oven-dried and freeze-dried GS powders. On the other hand, the type of drying process affected the TPC and antioxidant properties of samples. Freeze-dried GS powder has a higher level of TPC (19.97 ±1.60 mg GAE.g^-1 DM) and more than twice as high TEAC_ABTS value (127.10 ±27.28 mg Trolox.g^-1 DM) in comparison with the oven-dried GS samples. It has been previously concluded that the freeze-drying process is superior for the preservation of antioxidant compounds, and probably higher efficiency of extraction due to pronounced disruption of plant cells via the formation of ice crystals (Kamiloglu, et al., 2016, Kamiloglu and Capanolgu, 2014). However, such a release of flavonoid compounds from plant cells may cause their pronounced exposure to oxygen. As was described in a study of Nunes et al. (2016), oven drying of guava powder released more soluble flavonoids than the freeze-drying technique. In the chromatogram (Figure 2) of dried grape skin powder extract, fourteen well-resolved peaks were observed. However, from eight selected polyphenolic compounds, only the presences of (+)-catechin, (-)-epicatechin and rutin were confirmed in grape skin powder samples using retention time and absorption spectra in the current study. There was no difference in the number of peaks/compounds in oven-dried and freeze-dried GS powder extracts. The chromatograms differed in the heights of the peak showing that oven-dried GS powder extracts contained more polyphenolic substances.

As shown in Table 1, significantly higher contents of (+)- catechin, (-)-epicatechin and rutin were observed in ovendried GS powders, which corresponded to TFC values.

Table 2 shows the differences in the main composition of fortified processed cheese. The addition of grape skin powders significantly increased the protein content in processed cheese samples when added in 2% (w/w) levels, i.e. from 112.5 ±5.0 g.kg^-1 (in control) to 128.4 ±5.7 and 129.7 ±2.3 g.kg^-1 for processed cheese enriched with freeze-dried and oven-dried GS powders, respectively (p <0.05). Since we used the same amount of ingredients (cheese, butter, polyphosphate salts) for the preparation of all of the processed cheese samples, the variation in protein content was due to the addition of grape skin powder. Although the values of dry matter content increased linearly with the level of addition of grape skin powder, the effect of fortification was recognized as insignificant (p >0.05). On the contrary, Khan et al. (2018) found that fortification of Gouda cheese with the mango kernel did not significantly change the protein content.

Pair-wise comparison tests revealed that only processed cheese with 2% (w/w) of oven-dried GS powder had significantly higher dry matter content than that found in control (p <0.01). Fat and ash contents remained similar for all the processed cheese samples (p >0.05).

The content of phenolic individuals is depicted in Figure 3B and show their higher content in processed cheese fortified with freeze-dried grape skin powder at both levels. Total phenolic content of processed cheese samples was determined using Folin-Ciocalteau’s assay (Figure 3A). Significant increase of TPC values was obtained for processed cheese supplemented with GS powder at 2% (w/w) level from 0.13 ±0.02 mg GAE.g^-1 DM (control sample) to 0.47 ±0.07 and 0.54 ±0.07 mg GAE.g^-1 DM for oved-dried and freeze-dried GS powder, respectively. The addition of OD-GS powder to processed cheese spread resulted in the increase in TFC values from 0.25 ±0.10 mg QET.g^-1 DM (control sample) to 0.37 – 0.43 mg QET.g^-1 DM without respect to GS level. Freeze-dried GS powder enhanced processed cheese samples with flavonoids at a higher level (2%, w/w). It is interesting to note that even the control sample exhibited TFC value that then slightly increased with the addition GS powder. It has been reviewed that antioxidant properties of milk and milk products are due to the presence of sulphur-containing amino acids, vitamins, enzymes, peptides and oligosaccharides (Khan, et al., 2019; Usta and Yilmaz- Ersan, 2013; Atmaca, 2004; Egger and Ménard, 2017; Alenisan, et al., 2017). Although oven-dried GS powder exhibited higher levels of (+)-catechin, (-)-epicatechin and rutin, their content in PCS samples with OD-GS powder decreased in comparison with PCS fortified with FD-GC powder. TPC values showed very high correlation with the content of all phenolic constituents (0.746 < r < 0.826, p <0.01) while low correlation coefficients have been observed for TFC values showed values where (r = 0.610, p <0.05).

Antioxidant capacity of processed cheese samples in terms of TEAC_ABTS significantly increased with the increase of GS level (p <0.05). Surprisingly, TEAC_ABTS values were approximately tenfold higher (0.31 – 0.72 mg Trolox.g^-1 DM) for processed cheese with 1% (w/w) of GS in comparison with that of the control, followed by an additional tenfold increase (5.22 – 6.69 mg Trolox.g^-1 DM) when 2% (w/w) of GS powders were added (Figure 4). Such an increase may be attributed to the higher temperature used for melting the processed cheese in this study (90 °C for 13 min). It was previously published that pasteurization enhanced antioxidant properties of grape juice by releasing some polyphenolic compounds that were previously bound to other molecules (Genova, Tosetti and Tonutti, 2016; Fuleki and Ricardo-da-Silva, 2003). For instant, general increase in ABTS scavenging was detected in fresh and technologically harvested grape juices followed by pasteriuzation at 78 °C for 30 min (Genova, Tosetti and Tonutti, 2016).

In our previous work, the antioxidant properties of processed cheese fortified by rutin or quercetin increased with the increase of melting temperature as measured by ABTS•⁺ assay (Přikryl, et al., 2018). In addition, polyphenolics contributed differently to antioxidant capacity measured by ABTS•⁺, as was observed in a study of Lingua et al. (2016). They found that (-)-epicatechin, peonidin-3-glucoside and peonidin-3-acetylglucoside positively correlated with the assay, while pigment A and siringetin-3-glucoside had a negative effect. Similarly, strong positive correlation was observed for TEAC_ABTS values and the content of (+)-catechin, (-)-epicactechin and rutin (r = 0.955, p <0.001; r = 0.739, p <0.01 and r = 0.951, p <0.001, respectively). Considering that high correlation coefficient reflects the strong association between variables, we may conclude that TEAC_ABTS was mainly influenced by (+)-catechin and rutin contents followed by TPC values (r = 0.877, p <0.01).

The addition of GS powder had influence on reducing the power of processed cheese samples (Figure 4), where an increase was observed with the increase of GS powder level. Even though oven-dried and freeze-dried GS powders had similar RP values (see Table 1), processed cheese samples enriched with freeze-dried GS powder showed significantly higher RP values than in the case of incorporation of oven-dried GS powder, particularly at the 2% level (w/w) (p <0.05). Those discrepancies can be explained by the formation of new compounds, which may have enhanced the antioxidant capacity of the samples. Although the masking of antioxidant properties of various plant-based extracts by the addition of milk or whey proteins is common in literature, an interaction of protein and polyphenolic compounds may also increase antioxidant capacity. For instance, the mixing of α-casein or β-casein with epigallocatechin gallate led to the increase of inhibition of ABTS•⁺ during storage (Almajano, Delgado and Gordon, 2007). In a study by Sęczyk, Świeca and Gawlik-Dziki (2017), the addition of green coffee extract into soymilk significantly elevated ABTS radical scavenging ability (3.5-fold) and reducing power (13.8-fold). The study of interactions between polyphenolic compounds and β-conglycinin revealed that antioxidant capacity increased after the formation of the protein-phenolic complex (Zhao, et al., 2018; Murray, 2002). The authors also demonstrated the increase of antioxidant capacity of the protein-phenolic mixture after heating at 90 °C for 30 min. Thus, we suppose a formation of new products during the preparation of enriched process cheese with enhanced antioxidant activity towards ABTS•⁺. Both TPC and TFC values had the greatest influence on reducing power. Strong positive associations have been found for RP vs. TPC (r = 0.97, p <0.001) and RP vs. TFC content (r = 0.944, p <0.001).

Rutin, (+)-catechin and (–)-epicatechin also showed strong positive correlations with reducing powder (p <0.001) of processed cheese enriched with GS powder, i.e. r = 0.859; 0.842 and 0.917, respectively.

Two-factor Kruskall-Wallis ANOVA was applied to study the effect of grape skin powder levels and type of drying used for their preparation (oven-dried vs. freezedried). As can be seen from Table 3, the drying technique has a significant effect on the TPC values, all the phenolic individuals, and reducing power (p <0.01) in processed cheese spreads where the addition of freeze-dried GS powder assured their higher values. Total flavonoids and TEAC_ABTS values were not influenced by the drying technique. The effect of GS powder level was found to be significant for all parameters except for (-)-epicatechin content. As expected, the addition of GS powder at a higher level was reflected in higher values of TPC and TFC values, rutin and (+)-catechin contents, and antioxidant capacities in terms of TEAC_ABTS and reducing power.