Fruit Quality Characterization of New Sweet Cherry Cultivars as a Good Source of Bioactive Phenolic Compounds with Antioxidant and Neuroprotective Potential

Sweet cherries (Prunus avium L.) are highly appreciated fruits for their taste, color, nutritional value, and beneficial health effects. In this work, seven new cultivars of sweet cherry were investigated for their main quality traits and nutraceutical value. The phytochemical profile of three classes of phenolic compounds and the antioxidant activity of the new cultivars were investigated through high-performance liquid chromatography with diode array detection (HPLC-DAD) and spectrophotometric assays, respectively, and compared with those of commonly commercialized cultivars. Cyanidine-3-O-rutinoside was the main anthocyanin in all genotypes, and its levels in some new cultivars were about three-fold higher than in commercial ones. The ORAC-assayed antioxidant capacity was positively correlated with the total anthocyanin index. The nutraceutical value of the new cultivars was investigated in terms of antioxidant/neuroprotective capacity in neuron-like SH-SY5Y cells. Results demonstrated that the new cultivars were more effective in counteracting oxidative stress and were also able to upregulate brain-derived neurotrophic factor (BDNF), a pro-survival neurotrophin, suggesting their potential pleiotropic role in counteracting neurodegenerations.


Introduction
Dietary habits play a crucial role in the prevention of chronic diseases such as cardiovascular diseases, cancer, diabetes, and neurodegenerative disorders [1]. Fruit and vegetables are essential components of the human diet, and their intake is associated with the maintenance of health and the prevention of a wide range of diseases [2].
Sweet cherry (Prunus avium L.) fruits are among the most widely consumed fruits in temperate areas. The annual global production is about 2.2 million tons, the market being led by Turkey, the USA,
All cherry trees were grown under similar agronomic techniques (fertilization, irrigation, and pest control). For each cultivar, samples of randomly picked fruits were collected from four different trees of each cultivar. The collected fruits were used either for the determination of standard quality traits or for the gas chromatographic (GC) determination of simple sugars and acids in the juice. The fruit quality analysis was performed on a sample of 30 fruits/plant. The fruit quality parameters analyzed were fruit color (L*, a*, b* measured with a Konica Minolta color CR-400 Chroma Meter, Tokyo, Japan), fruit diameter, fruit weight, delta absorbance (DA) index (using a Cherry DA-meter, Sinteleia, Bologna, Italy), fruit firmness (using a Güss Fruit Texture Analyzer equipped with a 4-mm tip, Strand, South Africa), soluble solids (using an Atago Pocket Refractometer PAL-1, Tokyo, Japan), fruit pH and acidity (using titration with a Crison Titromatic 1S, Barcelona, Spain), and kernel weight. The color coordinates were used to calculate Chroma and Hue angle.
Finally, aliquots from each sample were also immediately frozen in liquid nitrogen, freeze-dried, and stored at −80 • C until extraction as described below.

Analysis of Fruit Quality Traits and GC Determination of Sugars and Acids in Fruit Juices
Analysis of simple sugar and acid contents was performed using a GC approach following the protocol by Bartolozzi et al. [29]. In detail, 500-µL juice aliquots from each sample were added to 450 µL of imidazole buffer (pH 7.0) and stored at 4 • C until analysis. Aliquots of 500 µL of each diluted juice were dried under a continuous air flow prior to derivatization, and 1 µL of the resulting solution was injected into a VARIAN GC 3900 equipped with electronic flow control (EFC), flame ionization detector (FID), CP-Sil 5 CB capillary column (30 m long), and a CP8410 autosampler (helium as carrier gas and nitrogen as make up gas).

Extraction Procedures
After freeze-drying, pits were discarded and the pulp from cherries picked from the same tree was mixed together and finely ground in a knife mill for 4 × 30 s periods. Pulp powder was then subjected to the "coning and quartering" sampling procedure and two technical replicates were carried out.
Flavonoids and anthocyanins were extracted from samples according to the protocol by Ballistreri et al. [30], with modifications. A 1-g aliquot of freeze-dried powder was thoroughly mixed with 5 mL of methanol/HCl (95.5/0.5, v/v) mixture and homogenized for 1 min (speed 5.5). The suspension was sonicated for 20 min at 35 • C. After centrifugation for 10 min at 1400× g, the supernatant was transferred into another vial and the sediment was extracted twice again with the same procedure. The supernatants were merged, mixed and filtered through Grade 44 (3 µm) ashless filter paper. The final volume of limpid liquid obtained was recorded. The extract was filtered again through a syringe filter (nylon, 0.22 µm pore diameter) from Thermo Fisher Scientific.
Phenolic acids were extracted from samples according to the protocol by Milinović et al. [31], with modifications. A 1-g aliquot of powder was thoroughly mixed with 20 mL of methanol/water (80/20, v/v) mixture and homogenized for 1 min (speed 5.5). After centrifugation for 15 min at 1400× g, the supernatant was filtered through Grade 44 (3 µm) ashless filter paper. The limpid liquid was brought to 25 mL with 85% methanol in a volumetric flask.

HPLC Determination of Phenolic Compounds
The extracts were injected into a Jasco (Tokyo, Japan) HPLC-DAD system, which consisted of a PU-4180 pump, an MD-4015 PDA detector and an AS-4050 autosampler. The stationary phase was an Agilent (Santa Clara, CA, USA) Zorbax Eclipse Plus C18 reversed-phase column (100 mm × 3 mm I.D., 3.5 µm). The injection volume was 20 µL for all determinations.
The method for flavonoid analysis was adapted from Milinović et al. [31]. Elution was carried out with a mixture of solvent A (water/formic acid 95/5, v/v) and solvent B (acetonitrile), with a composition gradient ranging from 97 to 36% of solvent A and flowing at 0.5 mL/min. A signal at 370 nm was used for quantitative purposes. Recovery values of flavonoids in spiked samples ranged from 79.2 to 91.4% (RSD < 9.4%, n = 6).
The chromatographic method for the analysis of phenolic acids was carried out as previously described [32]. Gradient elution was carried out with a mixture of solvent A (50 mM phosphate buffer, pH 2.5) and solvent B (acetonitrile) flowing at 0.7 mL/min, going from 97 to 50% (v/v) of solvent A. Signals at 254, 280, and 329 nm were used for analyte quantitation. Recovery values of phenolic acids in spiked samples ranged from 80.2 to 91.3% (RSD < 10.7%, n = 6).
The method for anthocyanin analysis was adapted from Ballistreri et al. [30]. Elution was carried out with a mixture of solvent A (water/formic acid/acetonitrile 87/10/3, v/v/v) and solvent B (water/formic acid/acetonitrile 40/10/50, v/v/v), with a composition gradient ranging from 94 to 40% of solvent A and flowing at 1.0 mL/min. A signal at 520 nm was used for quantitative purposes. Recovery values of anthocyanins in spiked samples ranged from 78.6 to 89.3% (RSD < 9.7%, n = 6).

In Vitro Antioxidant Activity Assays
2,2 -diphenyl-1-picrylhydrazide (DPPH) assays were carried out on phenolic acid sample extracts (see above) using a Jasco V-630 double beam spectrophotometer, as described elsewhere [33]. For the setup of calibration curves, 950 µL of 11 µM DPPH in methanol and 50 µL of methanolic Trolox (Tx) solution at different concentrations (0.05-2.00 mM), or 50 µL of methanol (blank solution), were thoroughly mixed in a 2-mL polypropylene vial. For sample analysis, the Trolox solution was replaced with 50 µL of suitably diluted sample. The vial was incubated in the dark at room temperature for 24 h, then the absorbance of the solution was read at 515 nm. Calibration curves were set up plotting the discoloration ratio (i.e., [ABS without TX /ABS with TX ] − 1) as a function of Tx concentration. Trolox equivalents (TE) of the samples were calculated interpolating on the calibration curve.
Oxygen radical absorbance capacity (ORAC) assays were carried out on the same extracts (see above) using a Perkin Elmer (Turku, Finland) Viktor X3 multilabel plate reader, essentially as described by Ou et al. [34]. TE were calculated from the relative area under the curve of the emission intensity vs. time plot.

Cell Culture and Viability
The SH-SY5Y human neuroblastoma cell line was obtained from Merck Italia. Cells were grown in DMEM supplemented with 10% (v/v) of FBS, 2 mM of L-glutamine, 50 U/mL of penicillin, and 50 µg/mL of streptomycin and maintained at 37 • C in a humidified incubator with 5% CO 2 , as previously reported [35]. Cells were used for experiments after inducing their differentiation with RA (10 µM) for 7 days.
To test viability, cells were treated with different concentrations (0.1-100 µg/mL) of five cherry extracts (Burlat, Grace Star, Gabriel, Lorenz, Marysa) for 24 h. Oxidative stress was induced with 700 µM H 2 O 2 for 1 h. Cell viability was evaluated by measuring MTT reduction as previously reported [36]. Briefly, at the end of the experiments, MTT was added to the medium (final concentration 0.5 mg/mL) and incubated for 90 min at 37 • C. After incubation, MTT solutions were removed, DMSO was added, and the absorbance was measured using a Viktor X3 multilabel plate reader at a wavelength of 595 nm.

Intracellular ROS Measurement
The formation of intracellular ROS was evaluated using the fluorescent DCFH-DA probe as previously reported [37]. Briefly, SH-SY5Y cells were treated with 50 µg/mL of each cherry extract for 24 h, then incubated with 10 µM DCFH-DA in DMEM 1% FBS w/o phenol red for 30 min. After removal of DCFH-DA, cells were incubated with 400 µM H 2 O 2 in DMEM 1% FBS w/o phenol red for 15 min. Then, H 2 O 2 was removed and replaced using PBS. Cell fluorescence was measured at 485 nm (excitation) and 535 nm (emission) with a Viktor X3 multilabel plate reader.

Determination of Reduced Glutathione (GSH) Levels
Reduced glutathione (GSH) levels were determined using the monochlorobimane (MCB) fluorometric assay as previously reported [38]. Briefly, SH-SY5Y cells were treated with 50 µg/mL of each cherry extract for 24 h before the induction of oxidative stress with 700 µM H 2 O 2 for 1 h. After treatment, cells were incubated with 50 µM MCB 1% FBS w/o phenol red for 30 min at 37 • C. After incubation, fluorescence was measured at 355 nm (excitation) and 460 nm (emission) with a Viktor X3 multilabel plate reader.

Analysis of mRNA Expression
SH-SY5Y cells were treated with 50 µg/mL of each cherry extract for 24 h and, after treatment, total RNA was extracted using an RNeasy Mini Kit, following the manufacturer's protocol. The yield and purity of RNA were measured using a NanoVue spectrophotometer (GE Healthcare, Milan, Italy). mRNA was reverse-transcribed into cDNA starting from 1 µg of total RNA using an iScript cDNA Synthesis Kit (BIO-RAD, Hercules, CA, USA), following the manufacturer's protocol. The subsequent PCR was performed in a total volume of 10 µL containing 2.5 µL (12.5 ng) of cDNA, 5 µL SsoAdvanced Universal SYBR Green Supermix (BIO-RAD), and 0.5 µL (250 nM) of each primer. Primers used were: BDNF, 5 CAAAAGTGGAGAACATTTGC3 (forward) 5 AACTCCAGTCAATAGGTCAG3 (reverse); glutathione reductase (GR), 5 GACCTATTCAACGAGCTTTAC3 (forward) 5 CAACCACCTTTTCTTCCTTG3 (reverse); NAD(P)H quinone oxidoreductase (NQO1), 5 AGTATCCACAATAGCTGACG3 (forward) Antioxidants 2020, 9, 677 6 of 21 5 TTTGTGGGTCTGTAGAAATG3 (reverse); and ribosomal protein S18 (RPS18), 5 CAGAAGGATGT AAAGGATGG3 (forward) 5 TATTTCTTCTTGGACACACC3 (reverse) from Merck Italia. RPS18 was used as a reference gene. cDNA amplification was started by activating the polymerase for 30 s at 95 • C, followed by 40 cycles of 5 s at 95 • C and 30 s at 60 • C. A melt curve was run to ensure quality control and generation of a single product. Normalized expression levels were calculated relative to control cells according to the 2 −∆∆CT method [39].

Statistical Analysis
Each experiment was performed at least three times, and all values are represented as mean ± standard error median (SEM). One-way analysis of variance (ANOVA) was used to compare differences among groups, followed by Tukey's test or Dunnett's test (Prism 6; GraphPad Prism Software, San Diego, CA, USA). Values of p < 0.05 were considered statistically significant. Correlation analysis was carried out with the same software.
A canonical discriminant analysis of some collected data (ORAC values, anthocyanin, phenolic acid and flavonol contents) was performed with the library 'candisc' of the R software (3.6.1 version). Correlations among different fruit traits and biochemical variables were calculated using Statistica 12 software.

Physical-Chemical Characteristics of Sweet Cherry Cultivars
The selected set of cultivars uniformly covers the sweet cherry harvesting season in the Emilia-Romagna region, starting from Burlat (May 23rd) and ending with Stephany (June 17th). The complete fruit quality description of each cultivar is reported in the Supporting Information materials. In general, significant differences among cultivars were identified for all the measured quality traits. The new varieties from the 'Sweet' series produce large fruits, with a caliper ranging from about 28 mm (Aryana) to more than 32 mm (Stephany and Valina; Figure S1). As expected, early varieties (Burlat, Aryana, Marysa, and Lorenz) bear fruits slightly smaller than those of late cultivars (Stephany and Valina), whose fruits reached weights of more than 17 g each. As expected, there is a high correlation (0.78-0.80; Table S1) between fruit and kernel sizes, and Aryana and Marysa produced the fruits with the smallest kernels.
Regarding the red color at maturity, fruits of Lapins, Marysa, and Valina showed higher values of a*, b*, and chroma (bright red) compared to those of Gabriel, Lorenz, and Stephany (dark red; Figure S2). The DA harvesting index results were strongly negatively correlated with some fruit color parameters like a* and chroma (−0.90; Table S1); therefore, the highest DA indexes were recorded on the dark red fruiting cultivars, having low values of either a* or chroma.
As concerns soluble solids, the lowest values were recorded in Lapins, Marysa, and Burlat, while the highest ones in Lorenz, Stephany and Gabriel ( Figure S3). Regarding titratable acidity, Burlat, Lapins, and Aryana were the least acidic fruits, while Gabriel and Stephany were the most acidic ones. The acidity of the last cultivar was very high (12.22 mg/L) in comparison with all the others, but the global taste was well balanced, because the soluble solid content was also the highest (19.47 • Brix).
The GC analysis of the fruit juices revealed that malic acid represented more than 93% of total acid content in Gabriel, Grace Star, Valina, Stephany, and Lapins ( Figure S4). Burlat, Aryana, and Lorenz contained less than 90% of malic acid and around 7-8% of succinic acid. Quinic acid was the third acid, in order of importance, in cherry fruits, but its content only ranged between 1.9 and 3.6% of the total. A high correlation (0.86) was found between titratable acidity and malic acid content using GC (Table S2).
Regarding monosaccharides and analogues, their content is variable among cultivars; glucose and fructose represented the most important sugars of cherry juices (46.5 and 41.0%, respectively), followed by sorbitol, while others (sucrose, inositol or xylose) were present only in trace amounts ( Figure S5). Glucose content ranged from 44% (Saretta and Stephany) to 47-48% (Marysa, Burlat and Aryana). The highest fructose percentage was found in Burlat and Gabriel (43-44%) and the lowest in Saretta and Stephany (about 39%). A high correlation (0.98) was observed between fructose and glucose contents, while correlation with total soluble solid content was much lower (0.41; Table S3).

Phenolic Compound Profile
The characterization of new cultivars for their phenolic profile was carried out and compared to that of the three reference cultivars Burlat, Grace Star, and Lapins.
The main classes of phenolic compounds in the pulp of sweet cherry are anthocyanins, phenolic acids (hydroxycinnamic acid derivatives), and flavonols; thus, HPLC-DAD analyses were carried out to identify and quantify compounds belonging to these three classes of secondary metabolites.
Among new cultivars, Gabriel, Lorenz, and Aryana were the richest in cyanidin-3-O-rutinoside ( Figure 1A), and in particular, Gabriel had levels about twice those of the three reference cultivars. All other new accessions showed cyanidin-3-O-rutinoside levels rather similar among them, and significantly higher than those of Burlat.
As concerns cyanidin-3-O-glucoside ( Figure 1B), the highest levels were found in Burlat, which showed a concentration about seven times higher than the other reference genotypes. Among new accessions, Aryana showed the highest level, while all the others did not differ significantly among each other and compared to Grace Star and Lapins.
As concerns peonidin-3-O-rutinoside levels ( Figure 1C), cultivars can be divided into three groups: the richest one was Stephany, with more than 20 mg/100 g FW, followed by Lorenz, Gabriel, Saretta, Aryana, and Lapins, with levels ranging from 10 to 20 mg/100 g FW; the group with the lowest levels (less than 10 mg/100 g FW) was represented by Valina, Marysa, Grace Star, and Burlat.
Phenolic acid profiles are shown in Figure 2. In all accessions, neochlorogenic acid was the main hydroxycinnamic acid derivative, followed by coumaroylquinic, chlorogenic, and caffeic acids, respectively.
Neochlorogenic acid levels ranged from 15 to 65 mg/100 g FW ( Figure 2A). Within this range, three groups could be clearly defined: the high-level group, comprising Marysa and Lapins, with levels higher than 50 mg/100 g FW; the intermediate group, comprising Grace Star, Saretta, and Stephany, with levels ranging from 30 to 50 mg/100 g FW, and the low-level group, comprising Burlat, Aryana, Valina, Gabriel, and Lorenz, with levels below 30 mg/100 g FW.
Coumaroylquinic acid content ranged from 1.5 to 17 mg/100 g FW. Lorenz had the highest and Marysa the lowest coumaroylquinic acid levels ( Figure 2B). Chlorogenic and caffeic acid content ranged from 1 to 4 mg/100 g FW in all cultivars ( Figure 2C,D), and slight fluctuations were observed among them. For caffeic acid, all new cultivars showed levels significantly higher than Burlat, and among them, Aryana, Stephany, Marysa, Gabriel, and Lorenz were the richest ones ( Figure 2D).
Within the class of flavonols, quercetin-3-O-rutinoside (rutin) was the most abundant compound. Quercetin-3-O-glucoside and kaempferol-3-O-rutinoside were also detected in all cultivars, but at very low levels, in the 0.02-1.0 mg/100 g FW range for the former and 0.08-1.2 mg/100 g FW for the latter (data not shown). As concerns cyanidin-3-O-glucoside ( Figure 1B), the highest levels were found in Burlat, which showed a concentration about seven times higher than the other reference genotypes. Among new accessions, Aryana showed the highest level, while all the others did not differ significantly among each other and compared to Grace Star and Lapins. Coumaroylquinic acid content ranged from 1.5 to 17 mg/100 g FW. Lorenz had the highest and Marysa the lowest coumaroylquinic acid levels ( Figure 2B). Chlorogenic and caffeic acid content ranged from 1 to 4 mg/100 g FW in all cultivars ( Figure 2C,D), and slight fluctuations were observed among them. For caffeic acid, all new cultivars showed levels significantly higher than Burlat, and among them, Aryana, Stephany, Marysa, Gabriel, and Lorenz were the richest ones ( Figure 2D). Rutin patterns in all genotypes are shown in Figure 3. Lorenz and Gabriel had the highest rutin levels among all genotypes, significantly higher compared to all reference cultivars.

Total Anthocyanin Index, Total Phenolic Acid Index, and In Vitro Antioxidant Activity
Total anthocyanin index (TAI) and total phenolic acid index (TPAI), which were calculated as the sum of all individual concentrations, are shown in Figure S6. The pattern was very similar to that of the major components of each class, namely cyanidin-3-O-rutinoside and neochlorogenic acid,

Total Anthocyanin Index, Total Phenolic Acid Index, and In Vitro Antioxidant Activity
Total anthocyanin index (TAI) and total phenolic acid index (TPAI), which were calculated as the sum of all individual concentrations, are shown in Figure S6. The pattern was very similar to that of the major components of each class, namely cyanidin-3-O-rutinoside and neochlorogenic acid, respectively.
The antioxidant activity (AA) of fruit methanolic extracts was measured using the DPPH and ORAC assays (Figure 4). The pattern of AA, according to the DPPH assay, showed that Gabriel was the most powerful antioxidant accession, reaching levels of about 10 mmol TE/100 g FW ( Figure 4A). As concerns AA results measured using the ORAC assay, differences among genotypes were much less pronounced in comparison to those of DPPH, and differences observed between new and reference cultivars were not significant ( Figure 4B). Nevertheless, a good correlation (0.53) was found between ORAC-assayed AA and TAI (Table S4).   The results of the canonical discriminant analysis of the AA (ORAC), and anthocyanin, phenolic acid, and flavonol contents is shown in Figure 5. The first two components explain around 73% (46.2 and 26.2) of the observed variability. The first component is significantly related to the content of neochlorogenic acid (+0.83), while the second one is related to peonidin-3-O-rutinoside (+0.80) and cyanidin-3-O-rutinoside (+0.66) contents. Based on this explorative analysis, two reference cultivars (Burlat and Grace Star) and three new cultivars (Gabriel, Lorenz and Marysa) resulted to be very diverse in respect to the contents of phenolic compounds and were selected for the subsequent evaluation of the protective effects of cherry extracts.

Neuroprotective Effect of Cherry Extracts against Oxidative Stress
The cytotoxicity of the different cherry extracts was evaluated using the MTT assay, treating differentiated SH-SY5Y cells with the extracts (0.1-100 µg/mL) for 24 h ( Figure S7). No cytotoxicity was observed, as cell viability was not influenced by the exposure to extracts up to 100 µg/mL. In order to assess the potential protective role of the tested extracts, cells were pretreated with 0.1-100 µg/mL of each extract for 24 h before exposing them to H2O2 for 1 h (Figure 6) to induce oxidative stress as previously reported [35]. Gabriel, Lorenz, Grace Star, and Marysa cultivars were able to significantly increase viability in cells exposed to H2O2. In particular, Gabriel was the most effective one, with significant protection observed at all concentrations higher than 1 µg/mL, while Lorenz produced significant protection at two concentrations (10 and 50 µg/mL). Burlat did not show any protection against oxidative stress.
As 50 µg/mL was the only concentration that led to significant protection in most extracts, this concentration was used for subsequent experiments. To better characterize the protection mechanisms, the ability of the cherry extracts to counteract H2O2-induced intracellular ROS production was investigated using the DCFH-DA assay. As reported in Figure 7A, in agreement with viability data, Gabriel, Grace Star, Marysa, and Lorenz led to a significant reduction in intracellular ROS levels in cells exposed to H2O2, while Burlat did not produce any significant effect.

Neuroprotective Effect of Cherry Extracts against Oxidative Stress
The cytotoxicity of the different cherry extracts was evaluated using the MTT assay, treating differentiated SH-SY5Y cells with the extracts (0.1-100 µg/mL) for 24 h ( Figure S7). No cytotoxicity was observed, as cell viability was not influenced by the exposure to extracts up to 100 µg/mL. In order to assess the potential protective role of the tested extracts, cells were pretreated with 0.1-100 µg/mL of each extract for 24 h before exposing them to H 2 O 2 for 1 h (Figure 6) to induce oxidative stress as previously reported [35]. Gabriel, Lorenz, Grace Star, and Marysa cultivars were able to significantly increase viability in cells exposed to H 2 O 2 . In particular, Gabriel was the most effective one, with significant protection observed at all concentrations higher than 1 µg/mL, while Lorenz produced significant protection at two concentrations (10 and 50 µg/mL). Burlat did not show any protection against oxidative stress. Antioxidants 2020, 9, x FOR PEER REVIEW 13 of 23 Since GSH is the most abundant intracellular antioxidant, the effect of the cherry extracts on intracellular GSH levels was evaluated using the MCB assay. SH-SY5Y cells were treated with 50 µg/mL of each extract for 24 h and then exposed to H2O2 ( Figure 7B). As expected, H2O2 treatment As 50 µg/mL was the only concentration that led to significant protection in most extracts, this concentration was used for subsequent experiments. To better characterize the protection mechanisms, the ability of the cherry extracts to counteract H 2 O 2 -induced intracellular ROS production was investigated using the DCFH-DA assay. As reported in Figure 7A, in agreement with viability data, Gabriel, Grace Star, Marysa, and Lorenz led to a significant reduction in intracellular ROS levels in cells exposed to H 2 O 2 , while Burlat did not produce any significant effect. Antioxidants 2020, 9, x FOR PEER REVIEW 14 of 23 led to significant reduction of GSH levels in respect to control cells. Grace Star and Gabriel extracts significantly increased GSH levels when compared to H2O2-exposed cells, up to values comparable to those of control cells. Marysa, Lorenz, and Burlat did not increase GSH levels compared to peroxide.

Effect of Cherry Extracts on Antioxidant Enzyme and BDNF Expression
The cherry extracts could act through the modulation of the endogenous antioxidant system, involved in the formation of GSH and hydroquinones in the cells. To evaluate this possibility, the expression of two endogenous antioxidant enzymes, namely GR and NQO1, was analyzed using RT-PCR ( Figure 8). Cells were treated with 50 µg/mL of the five extracts before measuring changes in mRNA expression. Treatment with extracts from Grace Star, Gabriel, Lorenz, and Marysa cultivars was able to significantly up-regulate NQO1 and GR, while Burlat did not influence the expression of these enzymes. Gabriel was the most effective one in up-regulating GR. Interestingly, a positive correlation was found between GR expression and TAI (0.89, data not shown). Since GSH is the most abundant intracellular antioxidant, the effect of the cherry extracts on intracellular GSH levels was evaluated using the MCB assay. SH-SY5Y cells were treated with 50 µg/mL of each extract for 24 h and then exposed to H 2 O 2 ( Figure 7B). As expected, H 2 O 2 treatment led to significant reduction of GSH levels in respect to control cells. Grace Star and Gabriel extracts significantly increased GSH levels when compared to H 2 O 2 -exposed cells, up to values comparable to those of control cells. Marysa, Lorenz, and Burlat did not increase GSH levels compared to peroxide.

Effect of Cherry Extracts on Antioxidant Enzyme and BDNF Expression
The cherry extracts could act through the modulation of the endogenous antioxidant system, involved in the formation of GSH and hydroquinones in the cells. To evaluate this possibility, the expression of two endogenous antioxidant enzymes, namely GR and NQO1, was analyzed using RT-PCR (Figure 8). Cells were treated with 50 µg/mL of the five extracts before measuring changes in mRNA expression. Treatment with extracts from Grace Star, Gabriel, Lorenz, and Marysa cultivars was able to significantly up-regulate NQO1 and GR, while Burlat did not influence the expression of these enzymes. Gabriel was the most effective one in up-regulating GR. Interestingly, a positive correlation was found between GR expression and TAI (0.89, data not shown).
The levels of brain-derived neurotrophic factor (BDNF) are strongly reduced in different neurodegenerative diseases [27]. For this reason, the effect of the different extracts on BDNF expression levels was investigated. Cells were treated with 50 µg/mL of the extracts before measuring changes in BDNF mRNA expression. Interestingly, all the extracts were able to significantly up-regulate BDNF in comparison to control cells (Figure 9), suggesting a protective role of cherries in neurodegeneration, possibly through a mechanism other than the antioxidant activity. To better understand the relationship between the global effect of the different extracts on these proteins and neuroprotection, a canonical discriminant analysis was conducted ( Figure 10). Data evidenced that Gabriel is the cultivar with the highest neuroprotective activity, and this seems to be related to the up-regulation of GR; in fact, the first component explains 91.2% of the variability, and this is mainly due to GR. On the other hand, Burlat falls in the opposite part of the graph underlining its lower impact on these proteins and its lower neuroprotective activity. The levels of brain-derived neurotrophic factor (BDNF) are strongly reduced in different neurodegenerative diseases [27]. For this reason, the effect of the different extracts on BDNF expression levels was investigated. Cells were treated with 50 µg/mL of the extracts before measuring changes in BDNF mRNA expression. Interestingly, all the extracts were able to significantly upregulate BDNF in comparison to control cells (Figure 9), suggesting a protective role of cherries in neurodegeneration, possibly through a mechanism other than the antioxidant activity. To better understand the relationship between the global effect of the different extracts on these proteins and neuroprotection, a canonical discriminant analysis was conducted ( Figure 10). Data evidenced that Gabriel is the cultivar with the highest neuroprotective activity, and this seems to be related to the up-regulation of GR; in fact, the first component explains 91.2% of the variability, and this is mainly due to GR. On the other hand, Burlat falls in the opposite part of the graph underlining its lower impact on these proteins and its lower neuroprotective activity.

Discussion
Results here obtained show that the cherry cultivars clearly differ from each other in fruit quality parameters and in health-promoting compound content. Indeed, the pattern of polyphenolic compounds known to characterize Prunus avium fruits showed significant quantitative differences among cultivars.
Fruit quality traits, as well as health-promoting compounds, are known to be strongly influenced by the environment, orchard management [40], and harvesting date [5], but a rather wide genetic variability among the selected cultivars has been observed for the compounds analyzed in this study.
The main quality traits of the new cultivars confirmed their overall excellent organoleptic attributes compared to reference cultivars, such as for weight, firmness, color, and sugar/acidity ratio [5,41]. One of the main peculiarities of these new cultivars is their different ripening times, which collectively cover a time period from the middle of May to the end of June, thus allowing uniform coverage of the very short harvest season for sweet cherries.
Sugar and acid composition, as well as flesh firmness, are known to be key factors in determining the fruit quality and taste of sweet cherries. In general, sugar and acid composition is in line with those reviewed by Chockchaisawasdee et al. [42]. The correlation between total sugars and soluble solids is significant but not very high (0.44, data not shown), but it is known that the Brix value is affected also by other compounds such as acids and amino acids [42], as confirmed by the observed significant positive correlation (0.52, data not shown) between soluble solids and fruit acidity.

Discussion
Results here obtained show that the cherry cultivars clearly differ from each other in fruit quality parameters and in health-promoting compound content. Indeed, the pattern of polyphenolic compounds known to characterize Prunus avium fruits showed significant quantitative differences among cultivars.
Fruit quality traits, as well as health-promoting compounds, are known to be strongly influenced by the environment, orchard management [40], and harvesting date [5], but a rather wide genetic variability among the selected cultivars has been observed for the compounds analyzed in this study.
The main quality traits of the new cultivars confirmed their overall excellent organoleptic attributes compared to reference cultivars, such as for weight, firmness, color, and sugar/acidity ratio [5,41]. One of the main peculiarities of these new cultivars is their different ripening times, which collectively cover a time period from the middle of May to the end of June, thus allowing uniform coverage of the very short harvest season for sweet cherries.
Sugar and acid composition, as well as flesh firmness, are known to be key factors in determining the fruit quality and taste of sweet cherries. In general, sugar and acid composition is in line with those reviewed by Chockchaisawasdee et al. [42]. The correlation between total sugars and soluble solids is significant but not very high (0.44, data not shown), but it is known that the Brix value is affected also by other compounds such as acids and amino acids [42], as confirmed by the observed significant positive correlation (0.52, data not shown) between soluble solids and fruit acidity.
Glucose and fructose represent the main sugars in sweet cherries, as reported elsewhere [43]. These two sugars are reported to account for about 90% of the total sugars in cherry fruits [42]. The total amount of these two sugars and their relative amounts are essential in determining the overall sweetness of the fruits, since the sweetness power of fructose is known to be higher than that of glucose [44].
The qualitative profile of hydroxycinnamic acid derivatives in all cultivars analyzed here reflects that reported by Picariello et al. for sweet cherries [45], with neochlorogenic, coumaroylquinic, and chlorogenic acids as main compounds. According to these authors, the relative content of hydroxycinnamates greatly varies depending on genotypes. In all cultivars here analyzed, neochlorogenic acid represents the predominant compound, and this agrees with results reported by other authors [43,46,47], although a different ratio between neochlorogenic and coumaroylquinic or chlorogenic acid has also been observed in some cases [30,48,49]. The levels of these colorless polyphenols found in the new cultivars were very similar to those reported by some authors in other cultivars [46]; however, higher or lower concentrations have been reported elsewhere as well [43,49,50]. These quantitative fluctuations are expected, since they are related to several factors, including genotype, growing conditions, ripeness stage, post-harvest storage, sample preparation, extraction procedures, and analytical techniques.
Results concerning anthocyanin profiles are consistent with most of the results reported in the literature, which identify cyanidin-3-O-rutinoside, cyanidin-3-O-glucoside, and peonidin-3-O-rutinoside as the major anthocyanins and peonidin-3-O-glucoside as a minor component [11,46,51]. The concentrations of cyanidin-3-O-rutinoside in all cultivars here analyzed ranged between 92.7 and 291.3 mg/100 g FW. These values are similar to those found by Nawirska-Olszaňska et al. [51] in old cherry cultivars grown in the Czech Republic, but were significantly higher than those reported by other authors; Ballistreri et al. [30] found levels of cyanidin-3-O-rutinoside ranging from 5.69 to 68.28 mg/100 g FW in twenty-four cultivars grown in Sicily (Italy), and Mozetič et al. [52] reported levels between 28.2 and 62.1 mg/100 g FW in five Slovenian cultivars. Even lower levels (between 1.09 and 14.8 mg/100 g FW) were reported by Usenik et al. [43] in different accessions grown in Slovenia. Besides the numerous factors known to influence the phenolic compound composition and concentrations in fruits, these differences can also be due to the fact that fresh fruits were utilized as the starting material in most reported studies, while lyophilized fruits were used here, and this could provide, e.g., higher extraction yields. Anthocyanin content was shown to be correlated with color parameters scored on fresh fruits. The higher the content of anthocyanins, mainly cyanidin-3-O-rutinoside and -glucoside, the darker the fruit color was, as expressed by lower L* and a* values (negative correlation of −0.53 and −0.47 against TAI, Table S4). The TAI of dark red fruits (Lorenz, Gabriel, and Stephany) was twice that of bright red fruits (Burlat).
The AAs of cherry extracts, measured using two widely used assays, were not univocal and different trends were observed between DPPH and ORAC. This is not surprising, since these assays are very different in their underlying chemical reaction and measurement procedures, and discrepancies in the AA trend measured using different assays have been reported. According to Ou et al. [53], who carried out one of the most comprehensive AA studies to date on 927 vegetables, the ORAC method was considered more chemically relevant to chain-breaking AA, while the stable DPPH radical bears no similarity to the highly reactive, transient radicals formed in the cells during lipid peroxidation. Our results indicate that a rather good correlation exists between the ORAC-assayed AA and TAI (Table S4), suggesting that this class of phenolic compounds gives a good contribution to the antioxidant capacity.
Taken together, the results presented here clearly indicate that the profile of bioactive polyphenol compounds is strictly dependent on genotype, and as concerns new cultivars, Marysa was the richest cultivar in colorless polyphenols, Gabriel in anthocyanins, and Lorenz in rutin. Compared to reference cultivars, Gabriel had a TAI value two-three times higher, and Lorenz had a rutin content almost twice higher, while in the case of phenolic acids the results were not so univocal. This indicates that, generally, sweet cherry accessions resulting from the breeding program can be considered as rich in health-promoting compounds as the reference cultivars, or even more so.
The classes of compounds identified in the new cherry cultivars have been reported to exhibit neuroprotective activity [54,55]. Thus, we evaluated the protective effect of the five most diverse cultivars (as assessed using PCA analysis) against oxidative stress in neuron-like differentiated SH-SY5Y cells. These cells are widely used as an in vitro model of neuronal function [56] and their exposure to H 2 O 2 triggers metabolic modifications leading to cell death and mimicking the events that might occur during neurodegeneration [57]. Interestingly, treatment with the new cherry cultivar extracts protected cells against oxidative stress induced by H 2 O 2 . This effect was associated to a significant reduction in intracellular ROS and a significant increase in GSH levels. In particular, Gabriel, which is characterized by the highest content of anthocyanins, was the most effective one in this regard. It has been indicated that anthocyanins can reach the brain in vivo, as absorbed anthocyanins and their metabolites are able to cross the blood-brain barrier [58,59]. Our data agree with the results reported by Leong et al. [60], who observed that cherry cultivars richest in anthocyanins were more effective in protecting Caco-2 cells from H 2 O 2 -induced damage compared to other cultivars. On the contrary, another study compared three different extracts of Saco sweet cherry, namely the non-colored fraction, colored fraction, and total extract [61]. The colored fraction was the least effective in protecting Caco-2 cells against oxidative stress induced by tert-butylhydroperoxide. The antioxidant activity of cherry extracts was also observed in neuronal cells. In particular, Matias et al. [62] demonstrated that a phenolic-rich extract from the Saco cherry cultivar, mainly composed of anthocyanins, counteracts H 2 O 2 -induced damage in SK-N-MC cells.
The cultivars that performed better in terms of neuroprotective potential (Gabriel and Lorenz) were especially rich in cyanidin-3-O-rutinoside, peonidin-3-O-rutinoside, rutin, and some minor phenolic acids (i.e., caffeic acid), while in the Burlat fruits, the same compounds are in much lower amounts. From our data, among anthocyanins, rutinoside forms seem to be more effective than glucoside ones. The observed neuroprotective activity of caffeic acid is in line with recent studies that demonstrated profound effects of this compound in the brain, including protection from toxicity induced by a variety of agents and/or experimental models of Alzheimer's disease [63].
To better clarify the intracellular antioxidant mechanisms of these extracts we evaluated their ability to modulate two antioxidant enzymes: GR and NQO1. GR is an important part of the cellular antioxidant defense as it maintains the intracellular ratio of reduced/oxidized glutathione (GSH/GSSG) [64]. The important antioxidant activity of NQO1 is related to its catalytic mechanism; in fact, it catalyzes exclusively the two-electron reduction of quinones to their corresponding hydroquinones, thus avoiding the generation of highly reactive semiquinones and ROS [65]. Interestingly, the new cherry cultivars triggered a significant up-regulation of these two enzymes, and Gabriel, the most effective one in counteracting oxidative stress, significantly increased GR expression in respect to the other cultivars, suggesting that the observed neuroprotective effect could mainly be ascribed to the up-regulation of this enzyme. Other authors demonstrated that anthocyanins are able to up-regulate GR and NQO1 [66], but to our knowledge, this is the first time that this up-regulation has been observed after cherry extract treatments. In the brain, BDNF has a key role in survival, maintenance, and regeneration of specific neuronal populations [67]. Depletion of this neurotrophic factor has been associated with different neurodegenerative disorders such as Parkinson's and Alzheimer's diseases [26,27] and strategies able to increase its levels are considered of great therapeutic interest. All the tested extracts significantly increased BDNF levels and this seems not to be related to the anthocyanin content as also cultivars with low TAI had a strong effect on BDNF expression. Williams et al. [68] observed that a blueberry extract increased BDNF levels in mouse hippocampus and ascribed this effect not only to anthocyanins but also to flavonols.

Conclusions
In conclusion, our data suggest that these cherry extracts possess neuroprotective activity in vitro, but in vivo studies should be carried out to assess real bioavailability. In fact, cherries are overwhelmingly consumed as fresh fruit, so the study was focused on their most common use, i.e., food use. This is the reason why FW-based results have been shown throughout. In our opinion, the "take home message" is that some of the studied cherry cultivars possess an added value in terms of bioactive compounds with respect to other cultivars, and we hope that in the future the quality of cherries could be evaluated not only in terms of their organoleptic properties, but also for their nutraceutical potential. Moreover, the precise role of the studied compounds has to be further investigated to understand if their positive effect on neuroprotection is due to the total amounts or to their balance. Nevertheless, the variability observed for these compounds in the analyzed cultivars, especially in the newly bred cultivars, could be a key step in selecting the best parents to be used in future breeding programs and pave the way to further improve the nutraceutical value of future sweet cherry cultivars.