Nine organic apple varieties, namely `Remo`, `Rewena`, `Relinda`, `Rebela`, `Freedom`, `Pinova`, `Florina`, `Topaz` and `Dalienette`, were purchased from a certified organic producer from the north-western region of Romania, Bihor County (autumn harvest 2020 and 2021) and their bioactive components were analysed. The fruits were processed into juice using a Bosch MES4000 juicer (München, Germany) and the remaining by-products were analysed (peel, pulp pomace and whole fruit pomace). Briefly, the peel of the fruits was obtained by peeling the apples before juice extraction. The peeled apples were then processed into juice and pulp pomace sample was obtained. Whole fruits were also processed into juice, obtaining this way the whole fruit pomace. Some fruits were processed after harvest (raw material) and some after 9 months of storage in a cold chamber (Besseling Group B.V., Blokker, the Netherlands) at (1.0±0.5) °C and 95 % RH (after storage). Immediately after processing, the obtained by-products were stored at -80 °C. Prior to analysis, the samples were lyophilised using a CHRIST LyoCube Alpha 2-4 freeze-dryer (Martin Christ, Osterode am Harz, Germany). The obtained results were expressed on dry mass basis.

Standard phenolic compounds and reagents: 37 % hydrochloric acid and gallic acid were purchased from Carl Roth GmbH (Karlsruhe, Germany). Ascorbic, formic, acetic and o-phosphoric acid, acetonitrile, methanol, Folin & Ciocalteu’s phenol reagent and 1,1-diphenyl-2-picrylhydrazyl (DPPH) were purchased from Sigma-Aldrich Chemie GmbH, Merck (Steinheim, Germany), 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) was purchased from Acros Organics, Fisher Scientific (Geel, Belgium) and isoquercitrin (quercetin-3-glucoside), hyperoside (quercetin-3-galactoside), quercitrin (quercetin-3-rhamnoside), (+)-catechin, (-)-epicatechin, procyanidine B2, phloretin, cyanidin-3-O-galactoside and phloridzin were purchased from Extrasynthese (Genay, France).

The total dry matter was determined by weighing 0.5 g of ground sample material, then drying it at 105 °C on the MAC 500 PARTNER thermo-balance (Radwag, Radom, Poland). The results were expressed in percentage.

To 1 g of raw material, 2 mL of 2 % o-phosphoric acid were added and triturated for 1 min at room temperature. The mixture was quantitatively passed into a 15-mL centrifuge tube and then brought to a final volume of 10 mL with 2 % o-phosphoric acid. After extraction, all samples were centrifuged (centrifuge 5430R; Eppendorf, Hamburg, Germany), filtered and immediately analysed by HPLC technique using the method adapted from Chanforan et al. (10). An Agilent 1200 chromatograph (Agilent Technologies, Santa Clara, CA, USA) equipped with a diode array detector (UV-DAD) was used for HPLC analysis. All data were recorded and processed with the Agilent ChemStation B.04.03 software (Agilent Technologies).

To 1 g of ground sample, 10 mL of 70 % aqueous methanol were added and incubated overnight in the dark at room temperature. After that, the extracts were shaken at 500 rpm for 1 h and then centrifuged at 2935×g and 4 °C for 10 min using a centrifuge 5430R (Eppendorf). The supernatant was recovered in centrifuge tubes and the residue was re-extracted two more times with 10 mL of 70 % aqueous methanol. All three supernatants were combined and then the volume of each extract was adjusted to 30 mL with the extraction solvent. The total phenolic content of the extract solutions was determined by the Folin-Ciocalteu spectrophotometric method described by Georgé et al. (11). The absorbance was measured at 760 nm using a UV–Vis spectrophotometer Specord 210 Plus (Analytik Jena, Jena, Germany). The results were expressed in mg of gallic acid equivalents (GAE) per gram of dry mass.

The DPPH test was adapted from a method described by Bujor et al. (12) with some modifications. The extract used for this DPPH analysis was the same as that used for the determination of the total phenolic content. Briefly, 0.2 mL of the sample extract was added to 2 mL of 0.2 mM DPPH solution in methanol (prepared daily, protected from the light and kept on ice). The mixtures were shaken for 30 min in the dark at 500 rpm using a KS 260 homogeniser (IKA, Staufen, Germany). The absorbance was then measured at 515 nm. Methanol was used as a blank reference. The results were expressed in mg of Trolox equivalents (TE) per g of dry mass.

Approximately 0.3 g of sample was ground in a mortar and 5 mL of methanol acidified with 1 % HCl were added. The extracts were shaken at 500 rpm for 15 min and then centrifuged at 2935×g and 4 °C for 5 min using a centrifuge 5430R (Eppendorf). The supernatant was recovered in centrifuge tube and the residue was re-extracted two more times with 5 mL of methanol with 1 % HCl. All three supernatants were combined and then the volume of each extract was adjusted to 15 mL with the extraction solvent. After extraction, the samples were subjected to solid phase extraction (C18 Bond Elute cartridge 200 mg, 3 mL; Agilent Technologies) based on Bujor et al. (12) method. UPLC-PDA analyses were performed using a Waters ACQUITY UPLC chromatograph (Waters, Milford, MA, USA). Individual compounds were quantified as mass fraction of apple powder with external standards measured at 520 nm. For anthocyanin quantification, cyanidin-3-O-galactoside was used as standard. The results were expressed as ideain on dry mass basis.

For the extraction of phenolic compounds, 100 mg of freeze-dried sample was dispersed in 1200 µL of dried methanol acidified with 1 % acetic acid in a 1.5-mL Eppendorf vial. The extraction was carried out in an ultrasonic bath for 15 min. After filtration (PTFE, 0.45 µm), the reaction medium was injected (20 µL) directly into the HPLC system. HPLC/DAD analyses were carried out in independent triplicates using an Agilent 1200 chromatograph (Agilent Technologies) equipped with a DAD. All the data were recorded and processed with the Agilent ChemStation B.04.03 software (Agilent Technologies). Separations were achieved as described by Bujor et al. (12) using a LiChroCART (LiChrospher PR-18 5 µm) column (250 mm×4 mm i.d.; Merck, Darmstadt, Germany) with a guard column (Licrospher RP-18 5 µm column; Merck) operated at 30 °C. Phenolic compounds were identified by comparison of their retention time and UV-visible spectra with those of standards and literature (13-15). Individual compounds were quantified as mass fractions in mg/g.

All data are expressed as mean value±standard deviation (S.D.). Results were submitted to JMP Statistical Discovery™ from SAS (16). Two-way analysis of variance (ANOVA) was carried out to examine the statistical differences among groups for all analysed parameters (variety/stage) using the standard least squares method. To investigate the overall effect of storage time measured at (1.0±0.5) °C, the model effects used were variety, stage (raw material/after storage) and their interaction (variety×stage). Values were considered significant at p<0.05.

The main effects of variety, stage and their interaction on the dry matter of apple waste after cold storage of the fruit were analysed (Fig. 1). In contrast to the fruit pulp (p>0.05), the percentage of dry mass in apple peel and fruit pulp from 2020 varied significantly among varieties (p<0.0001) over time (p<0.0001). As shown by the interaction effect (p<0.0001), we observed significant trends in both peel and fruit pulp with a significant decrease in Florina, Topaz and Dalinette varieties, while Pinova showed an increase in peel. In contrast, all apple waste evaluated in 2021 showed large differences among varieties during storage (p<0.0001). Overall, the most notable differences were found in the pulp pomace, with the highest decrease in the percentage of dry mass in Pinova, Florina and Topaz varieties.

The effect of fruit storage on vitamin C content was analysed in apple peel waste over a 9-month period (Fig. 2). Our results show significant differences among the different varieties (p<0.0001), with Topaz and Dalinette having the highest vitamin C mass fraction in both the 2020 and 2021 harvests. The vitamin C mass fraction increased significantly after storage (stage, p<0.0001) in the 2021 harvest for Rebela, Freedom, Pinova and Topaz varieties, in contrast to Rewena and Dalinette, as shown by the interaction effect (p<0.0001). In 2020, we observed different patterns over time (p<0.0001) in the development of vitamin C content, with only the Florina variety having a higher mass fraction, which was measured in the final stage. When looking at the pulp and fruit pomace, the vitamin C content was not detected (<LOD) regardless of the harvest year. This is consistent with the distribution patterns within the fruit, which mainly show an accumulation in the peel.

In a study conducted by Ma et al. (17), apple peel contained between 0.79 and 1.30 mg/g ascorbic acid, depending on the used drying method. Freeze-dried samples had 0.99 mg/g ascorbic acid, hot air-dried samples had values of 0.79-1.03 mg/g and heat pump-dried samples had 1.30 mg/g ascorbic acid. However, Kschonsek et al. (18) also found ascorbic acid in whole fruits of the apple varieties Golden Delicious (124.3 mg/100 g), Granny Smith (99.2 mg/100 g), Jonagold (189.1 mg/100 g) and Jonathan (210.2 mg/100 g).

Fig. 3 shows the values of total phenolic content (TPC), expressed as gallic acid equivalents (GAE) (mg/g), during storage in the apple waste obtained from the fruit harvested in 2020 and 2021. Apart from the apparently similar trends observed overall, there were clear differences in TPC values among the varieties (p<0.0001*). For example, Rewena and Relinda had the highest TPC in the fruit pulp during storage in 2020, followed closely by Topaz and Dalinette, while only Relinda and Dalinette (2021) showed significant values. The phenolic content measured in the peel showed the highest values compared to the fruit pulp and pulp pomace in both years. With seemingly closely related distribution patterns among the tested varieties (2020, p=0.0092; 2021, p<0.0001), several contrasting situations occurred. Specifically, Rewena and Pinova varieties differ in an opposite TPC development in 2020 compared to 2021, with the latter showing a decrease in the final stage.

The pulp pomace from both 2020 and 2021 showed similar results, with higher TPC overall at the end of storage (p<0.0001). Moreover, significant differences can be observed between varieties, with Relinda having the highest TPC at the end of storage in 2020 (p<0.0001), while Relinda and Dalinette had similar contents during time in 2021 (p<0.0001).

Similar results were also obtained by Nile et al. (5), with a value of (3.5±0.1) mg/g of the total phenolic content of powder. In their study, Suárez et al. (19) also determined a value of TPC of (3.6±0.2) g/kg on dry mass when analysing apple pomace. Kschonsek et al. (18) determined values, expressed as GAE, of 521.9 (Golden Delicious), 581.0 (Granny Smith), 1224.8 (Jonagold) and 1212.8 mg/100 g (Jonathan) in the fruit peel, while in the pulp pomace the obtained values were 276.4 (Golden delicious), 163.3 (Granny Smith), 177.5 (Jonagold) and 361.7 mg/100 g (Jonathan).

Similar to TPC, the antioxidant activity (DPPH) showed similar patterns among varieties within each type of apple waste (Fig. 4). Higher values for the DPPH radical scavenging activity expressed as Trolox equivalents (TE) indicate higher antioxidant activity. In 2020, the fruit pulp showed significant differences depending on the variety (p<0.0001), with Rewena and Relinda showing slightly higher antioxidant activity in the final stage. However, despite the insignificant main effect of storage (stage, p<0.0001), the significant decrease in Florina during this period corresponds to the significant effect exhibited by the variety×stage interaction (p<0.0001). A similar behaviour as for TPC of fruit pulp in 2021 was also shown for the DPPH values. In Relinda and Dalinette, the values increased over time, while in Rebela and Freedom there was a significant decrease in antioxidant activity, which contributed to the significant effect of the interaction variety×stage (p<0.0001). In apple peels, we observed similar DPPH trends to those for TPC to some extent. There is a wide distribution across varieties (p<0.0001) during storage in 2020 (p<0.0001), with particularly low antioxidant activity at the end of storage (p<0.0001). In 2021, peel showed significant differences between varieties over time (p=0.0008), but without the main effect of storage (p=0.5627) on DPPH. Pinova variety had higher values at the initial stage, followed by a decrease over time. In 2020, pulp pomace showed a slightly decreased DPPH values over time (p=0.0004), to some extent especially in the varieties Florina, Topaz and Dalinette. In contrast, in 2021 the antioxidant activity of pulp pomace did not change significantly over time, as shown by the main effect of stage (p=0.9542) and the interaction (p=0.0857). However, notable differences were observed among varieties (p=0.0013) during storage.

Similarly, Nile et al. (5) obtained the antioxidant activity of apple pomace of (72.6±1.6) %. In addition, a study conducted by Kschonsek et al. (18) showed values of antioxidant activity, expressed as TE, between 2.4 and 9.9 mmol/100 g in apple peel and between 0.8 and 2.3 mmol/100 g in pulp pomace.

The anthocyanin mass fraction was evaluated in the peel and fruit pomace of apples from the 2020 and 2021 harvests during storage (Fig. 5). In contrast to the fruit pulp from 2021, where no differences were observed between varieties over time, the fruit pulp from 2020 showed similar patterns in anthocyanin mass fraction among varieties, with the exception of the Rewena variety, measured at the beginning and end of the trial. It was characterised by a higher decrease in anthocyanins during storage, further highlighting the significant main effects and their interaction (variety, p=0.0001; stage, p=0.0001; p=0.0003). In contrast, the apple peel showed remarkable variations between the two stages in 2020, with lower anthocyanin mass fraction at the end of storage. The highest mass fraction was found in the Rewena variety (variety, p<0.0001), which decreased abruptly over time, followed by Relinda, where no obvious change in anthocyanins occurred. The significant main effect of storage (stage, p<0.0001*) and the interaction (p<0.0001) are further emphasised by the decrease in anthocyanins, which was different for Pinova, Florina, Topaz and Dalinette varieties. When considering the peel of the 2021 samples, the anthocyanin mass fraction showed a steady trend over time. However, given the decrease in the Rewena and Dalinette varieties (variety, p<0.0001), the interaction effect during storage is evident (p<0.0001).

Table S1 and Table S2 show the distribution of phenolic compounds in the apple peel, pulp pomace and fruit pomace of 9 apple cultivars during storage in two harvest years (2020 and 2021). The content of phenolic compounds in the analysed apple by-products has been reported previously by many authors. In this case, for the 2020 harvest year, it was observed that the identified polyphenolic compounds were found in higher mass fractions in the fruit peel both at the initial analysis and at the end of storage. Similarly, for the 2021 harvest year, higher mass fractions of polyphenolic compounds were initially determined in the fruit peel; however, at the end of storage, the highest mass fractions were mainly found in the whole fruit pomace. Briefly, of the measured phenolics, higher mass fractions of phloridzin, followed by cholorogenic acid were found in pulp pomace and fruit pomace, while higher mass fractions of procyanidin B2, quercitrin and chlorogenic acid were found in the peel. The distribution of phenolic compounds is generally maintained during storage and over the two analysed harvest years. Łata et al. (20) studied the variation of phenolic composition in apple peels and whole fruits of 19 apple cultivars. Of the identified compounds, the highest mass fractions on dry mass basis of chlorogenic acid (1.13 mg/g), followed by flavan-3-ols, rutin and phloridzin, were found in whole fruits. However, except for chlorogenic acid, higher mass fractions of compounds were found in the apple peel than in the whole fruit. Similar trends were observed in our study for the samples harvested in 2020, while the values for the 2021 harvest year were slightly lower.