Assessment of Antioxidants in Selected Plant Rootstocks

The service tree (Sorbus domestica) is a wild fruit tree with immense medicinal and industrial value. This study aimed at determining the four major groups of antioxidants (flavonoids, phenolic acids and aldehydes, catechin and procyanidin) in rootstocks of Crataegus laevigata (genotypes O-LE-14 and O-LE-21), Aronia melanocarpa (genotypes O-LE-14 and O-LE-21), Chaenomeles japonica (genotype O-LE-9) and Cydonia oblonga (BA 29) (genotypes O-LE-14 and O-LE-21). Hyperoside (Quercetin 3-D-galactoside) was the most abundant flavonoid compound, since its average content in the rootstocks of Crataegus laevigata (O-LE-21) was 180.68 ± 0.04 μg·g−1. Dihydrokaempherol was the least frequently found flavonoid compound, with an average concentration of 0.43 ± 0.01 μg·g−1 in all the rootstocks of plants considered in this study. Among the phenolic compounds, the most represented one was protocatechuic acid, with 955.92 ± 10.25 μg·g−1 in the rootstocks of Aronia melanocarpa (O-LE-14). On the other hand, the least represented p-Coumaric acid exhibited the average concentration of 0.34 ± 0.01 μg·g−1 in the plant rootstocks. Epicatechin was the most abundant catechin compound, with a content of 3196.37 ± 50.10 μg·g−1 in the rootstocks of Aronia melanocarpa (O-LE-14). The lowest represented catechin compound was epigallocatechin, with the average concentration of 0.95 ± 0.08 μg·g−1 in the screened plant rootstocks. From the procyanidin compounds, the most abundant one was procyanidin b2 in the rootstocks of Crataegus laevigata (O-LE-14), with a concentration of 5550.40 ± 99.56 μg·g−1. On the contrary, procyanidin a2, with an average concentration of 40.35 ± 1.61 μg·g−1, represented the least frequent procyanidin compound in all the plant rootstocks screened herein.


Chemicals
The chemicals used in different procedures during the present study were purchased from Sigma-Aldrich (St. Louis, MO, USA) in ACS purity, unless noted otherwise.

Plants
In this study, seven varieties of rootstocks potentially suitable for S. domestica (as Crataegus laevigata

Preparation of the Plant Samples for Analyzing Flavonoid, Phenolic Acids and Aldehydes, Catechin and Procyanidin Compounds
In this experiment, seven kinds of rootstocks of Sorbus domestica with the control (S. domestica) were used. Methanol (80%, v/v) was used in the extraction of the flavonoid, the phenolic acids and aldehydes and the catechin and procyanidin compounds from the rootstocks of S. domestica. The samples were subjected to lyophilization for 24 h, 0.014 mBar vacuum and −55 • C (Lyophilizer, Labconco, Kansas City, Missouri, USA). An equal weight (20 mg) of samples was taken from each of the seven kinds of rootstocks of S. domestica with the control (Analytical Balance, EP 240A, Precisa, Vienna, Austria). The samples were homogenized in a friction bowl with 1.0 mL of 80% methanol, and 0.05 to 0.1 g of sea sand, until evaporation. The homogenization was repeated twice. Thereafter, the samples were vortexed (Vortex Mixers, VELP Scientifica, Usmate Velate MB, Italy) for 1-2 min and were subsequently centrifuged at 25,000 rpm and 16 • C for 15 min (Centrifuge Z326K, Hermle, Gosheim, Germany). Later, each sample was filtered through a filter (LUT Syringe Filters Nylon, LABICOM s.r.o., Olomouc, Czech Republic). Finally, the samples were pipetted out (400 µL) and analyzed using LC/MS. The results have been recalculated per 1.0 g of plant tissue.

Analysis of the Plant Sample-Extracts Using LC/MS
To determine the selected flavonoid, phenolic acids and aldehydes, and catechin and procyanidin compounds, a high-performance liquid chromatograph (HPLC Agilent 1200 Series) with a triple quadrupole and the mass detector (6460 Triple Quad) LC/MS equipped with ESI ionization were used. For the separation of the flavonoid, the phenolic acids and aldehydes and the catechin and procyanidin compounds, a Zorbax EC 18 column of 50 × 3.0 mm and a particle size of 2.7 µm was used prior to analyzing the compounds of interest. The measured concentration was the average of three measurements (injections) for each sample of triplicate. The acquired data between triplicates varied within RSD 5%.

Separation of Flavonoid Compounds
The column was held at 60 • C. The mobile A phase consisted of 100% methanol, whereas the mobile B phase was 0.2% acetic acid. The flow rate of the mobile phase was kept at 0.7 mL·min −1 . The compounds were eluted with a linear upward gradient: 0 min (90% B), 2 min (40% B), 4 min (0% B) and 6 min (90% B). The triple quadrupole mass detector was operated in the negative mode. Gas (nitrogen) temperature was kept at 350 • C, the gas flow rate was set to 13 L·min −1 , the pressure nebulizer had a value of 50 psi, the temperature of the focusing gas was 400 • C, the flow rate of the focusing gas was set at 12 L·min −1 and the voltage on the capillary tube amounted to 4000 V (Table 1

Separation of Phenolic Acids and Aldehydes
The column was held at 45 • C. The mobile A phase consisted of 100% methanol, and the mobile B phase was 0.2% acetic acid. The flow rate of the mobile phase was kept at 0.6 mL·min −1 . The compounds were eluted with a linear upward gradient: 0.00 min (82% B), 0.17 min (82% B), 0.51 min (70% B), 1.70 min (45% B), 4.00 min (45% B) and 6.00 min (82% B). The triple quadrupole mass detector was operated in negative mode. The gas (nitrogen) temperature was kept at 300 • C, the gas flow rate was set to 12 L·min −1 , the pressure nebulizer had a value of 60 psi, the temperature of the focusing gas was 300 • C, the flow rate of the focusing gas was set at 11 L·min −1 and the voltage on the capillary tube amounted to 3500 V (Table 2).

Separation of Catechin and Procyanidin Compounds
The column was held at 45 • C. The mobile A phase consisted of 100% methanol, and the mobile B phase was 0.2% acetic acid. The flow rate of the mobile phase was kept at 0.6-0.7 mL·min −1 . The compounds were eluted with a linear upward gradient: 0.00 min (85% B), 0.17 min (85% B), 0.51 min (75% B), 1.70 min (70% B), 4.00 min (70% B) and 6.00 min (85% B). The triple quadrupole mass detector was operated in the negative mode. The gas (nitrogen) temperature was 300 • C, the gas flow rate was set to 12 L·min −1 , the pressure nebulizer had a value of 45 psi, the temperature of the focusing gas was 250 • C, the flow rate of the focusing gas was set at 11 L·min −1 and the voltage on the capillary tube amounted to 3500 V (Table 3).

Statistics Methodology
The data were processed using MICROSOFT EXCEL ® (Microsoft, Redmond, WA, USA) and STATISTICA CZ Version 12.0 (StatSoft CR s.r.o., Prague, Czech Republic). The data are expressed as mean ± standard deviation (S.D.), unless otherwise noted (EXCEL). The statistical significance of the measured data was determined using STATISTICA CZ. The Anderson-Darling test was used to test the normality of the data. Differences with p < 0.05 were considered significant and were determined by using a one-way ANOVA test and a post-hoc Dunnett's test, which was applied as a means of comparison to the control group. Moreover, for the exploratory data analysis (EDA) cluster analysis, the principle component analysis (PCA) and the correlation were done.

LC/MS-Based Profile of the Test Plant Extracts
A LC/MS analysis was performed to determine different flavonoid, phenolic, catechin and procyanidin compounds in selected rootstocks of different plants. The determination of the occurrence and content of flavonoid, phenolic acids and aldehydes, catechin, and procyanidin compounds was done using high-performance liquid chromatography with mass detection.

Contents of Phenolic Acids and Aldehydes
Protocatechuic acid was the most frequent phenolic compound, followed by 3,4-Dihydroxybenzaldehyde, syringic acid, vanilic acid and vanillin in selected rootstocks of plants ( Figure 2A).

Contents of Phenolic Acids and Aldehydes
Protocatechuic acid was the most frequent phenolic compound, followed by 3,4-Dihydroxybenzalde hyde, syringic acid, vanilic acid and vanillin in selected rootstocks of plants (Figure 2A). observed in all the rootstocks of plants in the present study.
(C) Syringic acid, vanilic acid and vanilin Interestingly, syringic acid, vanilic acid and vanilin occurred in higher concentrations in the extract of Crataegus laevigata (O-LE-14) rootstocks. In the case of syringic acid, the control exhibited 40.65 ± 1.21 μg·g −1 ), whereas 243.66 ± 7.93 μg·g −1 of syringic acid occurred in Crataegus laevigata (O-LE-14) rootstocks. The concentration of vanilic acid was very similar to that of syringic acid (238.90 ± 9.68 μg·g −1 ), and the control showed 47.94 ± 3.41 μg·g −1 of syringic acid. However, the concentration of vanilin, when compared to the control, was surprising. In the extract of rootstocks of Crataegus laevigata (O-LE-14), 242.78 ± 13.21 μg·g −1 of vanillin was observed, whereas its concentration in the control was only 2.73 ± 0.01 μg·g −1 . The other flavonoid compounds revealed concentrations not exceeding 189 μg·g −1 . Figure 2B shows the major steps followed during the preparation of samples for the LC/MS analysis. The experimental details can be found in Section 2.    Figure 2B shows the major steps followed during the preparation of samples for the LC/MS analysis. The experimental details can be found in Section 2.

Contents of Selected Catechin and Procyanidin Compounds
Catechin and procyanidin compounds were also determined in selected rootstocks of plants. Catechin, epicatechin and epigallocatechin were selected among the catechins for analysis ( Figure 3A).

Statistical Analysis
The Anderson-Darling test revealed that all data had a normal distribution. The statistical significance of the differences between the control sample and the other samples, including the majority of anti-oxidative compounds, was tested at p < 0.001, employing a one-way ANOVA test and a post-hoc Dunnett's test (Table S1). In two cases, the significance level was lower (p < 0.050), but the control group was still significantly different. Finally, no significant differences were found for quercetin between the control sample and BA 29 (O-LE-14) (p < 0.322); the same is true for dihydrokaempferol between the control sample and BA 29 (O-LE-21) (p < 1.000), Crataegus laevigata (O-LE-14) (p < 0.216) and Aronia melanocarpa (O-LE-14) (p < 0.819).
Regarding the samples, the exploratory data analysis (EDA) revealed the most similar sample to the control sample, based on the cluster analysis ( Figure 4A Figure 4A). Regarding the anti-oxidative compounds, the PCA analysis ( Figure 4B) revealed the major four main groups of compounds with good in-group correlations. The closer the vectors of the compounds were in the projection (Figure 4B), the more significant was the correlation between the compounds. The first group consisted of p-coumaric acid, caffeic acid, eriodictyol, pentahydroxychalcone, p-hydroxybenzaldehyde, rutin (Q-3-rutinoside), gallic acid and p-hydroxybenzoic acid. The second group consisted of naringenin chalcone, chlorogenic acid, cryptochlorogenic acid, 3,4-Dihydroxybenzaldehyde, catechin, epicatechin, homoeriodictyol, quercetin and salicylic acid. The third group consisted of vitexin, isovitexin and procyanidin a2. The fourth group consisted of

Statistical Analysis
The Anderson-Darling test revealed that all data had a normal distribution. The statistical significance of the differences between the control sample and the other samples, including the majority of anti-oxidative compounds, was tested at p < 0.001, employing a one-way ANOVA test and a post-hoc Dunnett's test (Table S1). In two cases, the significance level was lower (p < 0.050), but the control group was still significantly different. Finally, no significant differences were found for quercetin between the control sample and BA 29 (O-LE-14) (p < 0.322); the same is true for dihydrokaempferol between the control sample and BA 29 (O-LE-21) (p < 1.000), Crataegus laevigata (O-LE-14) (p < 0.216) and Aronia melanocarpa (O-LE-14) (p < 0.819).
Regarding the samples, the exploratory data analysis (EDA) revealed the most similar sample to the control sample, based on the cluster analysis ( Figure 4A Figure 4A). Regarding the anti-oxidative compounds, the PCA analysis ( Figure 4B) revealed the major four main groups of compounds with good in-group correlations. The closer the vectors of the compounds were in the projection (Figure 4B), the more significant was the correlation between the compounds. The first group consisted of p-coumaric acid, caffeic acid, eriodictyol, pentahydroxychalcone, p-hydroxybenzaldehyde, rutin (Q-3-rutinoside), gallic acid and p-hydroxybenzoic acid. The second group consisted of naringenin chalcone, chlorogenic acid, cryptochlorogenic acid, 3,4-Dihydroxybenzaldehyde, catechin, epicatechin, homoeriodictyol, quercetin and salicylic acid. The third group consisted of vitexin, isovitexin and procyanidin a2. The fourth group consisted of vanilin, syringic acid and vanillic acid. One compound, protocatechuic acid, was found to be distant from the other compounds and groups of compounds, and therefore it did not exhibit a correlation with the other compounds. This was also confirmed in the correlation matrix (Table S2).
Antioxidants 2020, 9, x FOR PEER REVIEW 10 of 14 vanilin, syringic acid and vanillic acid. One compound, protocatechuic acid, was found to be distant from the other compounds and groups of compounds, and therefore it did not exhibit a correlation with the other compounds. This was also confirmed in the correlation matrix (Table S2).  (Tables S1 and S2).

Discussion
Generally, antioxidants have not yet been deeply investigated in connection with the affinity in grafted woods. Hudina et al. reported arbutin as the most abundant phenolic compound in the phloem above and below the graft union, followed by procyanidin B1 and chlorogenic acid [27]. Assuncao et al. considered gallic and sinapic acids as the markers of graft/scion compatibility [42]. The authors identified high concentrations of gallic and sinapic acids together with catechin as the cause of decreased affinity. A lower abundance in gallic acid, sinapic acid and catechin in the more compatible combination could be related to a lesser oxidative stress environment of the grafts, consequently promoting a better development of the graft union. Generally, the concentration of flavanols (particularly epicatechin) decreases at the graft interface compared to the surrounding woody tissues. Presumably, the wood has a high concentration of flavanols, which gets diluted as the callus cells develop [43].
The research carried out by Canas et al. [44] on grapevine, among other things, showed that catechin, epicatechin, ferulic acid and caffeic acid seem to have an important involvement in incompatibility, owing to the different content between graft partners, with higher accumulation above the graft union. Other authors highlighted that a quantitative difference in the phenolic compounds produced by heterospecific grafts may result in metabolic dysfunctions between the cells of the scion-rootstock in the graft union [28].

Discussion
Generally, antioxidants have not yet been deeply investigated in connection with the affinity in grafted woods. Hudina et al. reported arbutin as the most abundant phenolic compound in the phloem above and below the graft union, followed by procyanidin B 1 and chlorogenic acid [27]. Assuncao et al. considered gallic and sinapic acids as the markers of graft/scion compatibility [42]. The authors identified high concentrations of gallic and sinapic acids together with catechin as the cause of decreased affinity. A lower abundance in gallic acid, sinapic acid and catechin in the more compatible combination could be related to a lesser oxidative stress environment of the grafts, consequently promoting a better development of the graft union. Generally, the concentration of flavanols (particularly epicatechin) decreases at the graft interface compared to the surrounding woody tissues. Presumably, the wood has a high concentration of flavanols, which gets diluted as the callus cells develop [43].
The research carried out by Canas et al. [44] on grapevine, among other things, showed that catechin, epicatechin, ferulic acid and caffeic acid seem to have an important involvement in incompatibility, owing to the different content between graft partners, with higher accumulation above the graft union. Other authors highlighted that a quantitative difference in the phenolic compounds produced by heterospecific grafts may result in metabolic dysfunctions between the cells of the scion-rootstock in the graft union [28].

An Occurrence and Contents of Selected Flavonoid Compounds
According to Hudina et al. [27], a higher concentration of hyperoside (Q-3-galactoside), isoquercitin (Q-3-glucoside), quercetine (Q-3-rhamnoside) and rutine (Q-3-rutinoside) in pear rootstock tissues bellow the graft union may indicate incompatibility between the graft and the scion. Although the samples of the current study were taken directly from the graft union, higher concentrations of hyperoside (Q-3-galactoside) and isoquercitine (Q-3-glucoside) were measured at both C. leavigata variants (O-LE-14, O-LE-21) when compared to the control (good affinity). In both C. leavigata variants, the visible incompatibility was not observed, but future disaffinity cannot be ruled out. On the other hand, in variants with visible disaffinity (Chaenomeles japonica (O-LE-9)) and both Aronia melanocarpa (O-LE-14 and O-LE-21), the concentrations of these flavonoids did not exceed those of the control. In quercitine (Q-3-rhamnoside) and rutin (Q-3-rutinoside), none of the variants exceeded the concentration of the control, suggesting that they are not signaling an incompatibility between the service tree and the tested rootstocks.
For the rest of the flavonoids, to the best of our knowledge, no information is available in connection with the disaffinity of incompatible scions and rootstocks. However, among these, the concentrations of naringenin chalocone, quercetin and homoeriodictyol, which are widespread in other species [45][46][47], were increased in both variants of Aronia melanocarpa, the incompatible rootstocks.

An Occurrence and Contents of Selected Phenolic Acids and Aldehydes
The variants with incompatible rootstocks (Aronia melanocarpa (O-LE-14, O-LE-21) and Chaenomeles japonica (O-LE-9)) had the highest concentrations of 3,4-Dihydroxybenzaldehyde, salicylic acid, chlorogenic acid and cryptochlorogenic acid compared to the control. High chlorogenic acid concentration was previously reported as the signal of disaffinity in pear trees [27], which is in agreement with our results. P-coumaric acid was analyzed in rootstock affinity tests in the work of Usenik et al. [48], where high amounts of this acid were accumulated in apricot scions when disaffinity occurred. In the present study, almost no p-coumaric acid was measured in all variants, which in turn suggested that p-coumaric acid does not play a role in the disaffinity of Sorbus domestica.
Based on the results, the highest concentrations at all rootstocks were measured for protocatechuic acid, for which higher concentrations than the control were measured in incompatible Aronia melanocarpa (O-LE-14, O-LE-21) and Chaenomeles japonica (O-LE-9), but also in one out of two compatible variants: BA29 (O-LE-21) and Crataegus laevigata (O-LE-21). It seems that protocatechuic acid does not affect or signal the compatibility of the graft and the rootstock. The same could be stated for vanilin, vanilic and syringic acid, where the highest concentrations were observed at compatible rootstocks ( Figure 2).

An Occurrence and Contents of Selected Catechin and Procyanidin Compounds
Epicatechin and catechine are well known flavonoids, which increase when disaffinity occurs [27,28,41,45]. Our results are in accordance with this information, as incompatible rootstocks of Aronia melanocarpa of both variants and Chaenomeles japonica had the highest concentration of these substances. Although higher concentrations of catechine and epicatechine were measured in some of the compatible rootstock variants (Figure 3a) when compared to the control, they were not as high as those of incompatible rootstocks. This effect, together with a higher concentration of epicatechin than catechin in tissues, which was observed in the present study, was described by Musacchi [28]. On the other hand, procyanidin b1 and b2 are potentially involved in a graft incompatibility in pear trees [27]. In our study, the highest concentrations of both procyanidins were measured in incompatible Aronia melanocarpa variants, Chaenomeles japonica ( Figure 3B), which proves their incompatibility. However, high concentrations -higher than those of the control were also measured in compatible Crataegus laevigata variants ( Figure 3B). Apart from the results of isoquercitine (Q-3-glucoside) and hyperoside (Q-3-galactoside) discussed above, there is a suspicion that Crateagus leavigata will show the incompatibility symptoms in next years. Procyanidin a2 and c1 did not show any pattern of signaling incompatibility.

Conclusions
This study has presented the results of the pilot analysis of the major flavonoids, phenolic acids and aldehydes, catechin and procyanidin compounds in the selected rootstocks of different plants. Thirteen flavonoid and phenolic compounds, 3 catechin compounds and 4 procyanidin compounds were determined and thoroughly analyzed in this study. The study outcomes related with the amounts of antioxidants and other important substances in grafted plants (not only woods) may provoke future studies on the subject prior to an elucidation of the other compounds. Additionally, novel biochemical studies aimed at elucidating the biochemical mechanisms of affinities during grafting may also be done based on the clues revealed here in the present study.