Identification and Quantification of Naphthoquinones and Other Phenolic Compounds in Leaves, Petioles, Bark, Roots, and Buds of Juglans regia L., Using HPLC-MS/MS

The present study was designed to identify and quantify the major phenolic compounds in different Juglans regia L. (common walnut) tissues (leaves, petioles, bark, roots, buds), to define the compositions and contents of phenolic compounds between these tissues. A total of 91 individual phenolic compounds were identified and quantified, which comprised 8 hydroxycinnamic acids, 28 hydroxybenzoic acids, 11 flavanols, 20 flavonols, 22 napthoquinones, and 2 coumarins. Naphthoquinones were the major phenolic group in leaves, petioles, bark, and buds, as >60% of those identified, while hydroxybenzoic acids were the major phenolic group in side roots, as ~50% of those identified. The highest content of phenolic compounds was in the J. regia main root, followed by side roots and buds, leaves, and 1-year-old bark; the lowest content was in petioles and 2-year-old bark. Leaves, roots, and buds of J. regia represent a valuable source of these agro-residues.


Introduction
The Persian, English, or common walnut (Juglans regia L.) is a valuable tree nut and a well-known member of the Juglandaceae family. Walnuts are the third most consumed nut in the world, and they are known for their high content of phenolic compounds [1,2]. Over the last two decades, much attention has been paid to characterizing the contents of phenolic compounds in various plant materials, as these can have beneficial effects on human health. For example, phenolic compounds can reduce the risk of cardiovascular and degenerative diseases by preventing oxidative stress and oxidation of biological macromolecules [2]. Numerous studies have also demonstrated human health benefits of such bioactive compounds, in terms of potential protection against cancers, diabetes, and cardiovascular diseases, as well as showing anti-allergen, antimicrobial, anti-inflammatory, and antioxidant activities, among others [2,3]. Phenolic compounds can also be used effectively as functional ingredients in foods, as they prevent lipid oxidation, and mold and bacterial growth [4].
Juglans regia is recognized as a rich source of phenolic compounds. The kernel, fresh green fruit, husks, shell skins, leaves, bark, and roots have been comprehensively studied for use in the food, cosmetic, and pharmaceutical industries [2,5,6]. Leaves of J. regia are known to contain considerable amounts of phenolic compounds, which are mainly attributed with the excellent pharmacological and therapeutic properties associated with these leaves [2,3,7]. Leaves and petioles are easily available in large quantities, while the other parts of the tree, such as bark, roots, and buds, are not abundant, and whole plants would have to be cut down to obtain them.
Walnut leaves have historically served as a source of health-promoting compounds, and have been used extensively in conventional medicine due to their anthelmintic, purga-tive, antidiarrheal, astringent properties, and for the treatment of hemorrhoidal symptoms and venous insufficiency [2,8]. Extracts of walnut leaves are also reported to have antiscrofulous, hypotensive, antifungal, keratolytic, hypoglycemic, and sedative activities [9][10][11][12]. While the leaves have been extensively studied and the contents of their phenolic compounds quantified, there have been no studies that have identified or quantified the phenolic compounds in the petioles. Like the leaves, the petioles are easily abundant, and might serve as a good source of phenolic compounds.
Leaves and petioles are the most easily available of the plant tissues, although bark, roots, and buds might also be good sources of phenolic compounds [2]. However, these cannot be harvested from the trees in the same way as leaves and petioles. Alternatively, these plant tissues can be considered as agro-residues when the trees are cut for timber or when an orchard is too old to be economically sustainable. The efficient use of these walnut agro-residues would be a strategy to simultaneously help to increase the economic return for farmers and companies while protecting the environment, as the efficient use and recovery of such secondary metabolites might be used to generate functional ingredients to substitute for synthetic chemicals, thus also adding more value to the walnut industry [13,14]. To effectively recover and use the phenolic compounds in walnut leaves, petioles, bark, roots, and buds, the chemical profiles of each of these agro-residues need to be defined, especially with respect to the individual phenolic compounds that they contain. Nonedible tissues of J. regia are indeed considered as good sources of naphthoquinones and flavonoids. Naphthoquinones have significant toxicity due to their nonspecific mechanisms of action, which can be observed for juglone and its allopathic effects. Due to these properties, many studies have explored the biological and toxicological activities of naphthoquinones, to potentially discover and develop new drugs [15].
The aim of the present study was to determine the phytochemical compositions of walnut leaves, petioles, 1-year-old and 2-year-old bark, side roots and main roots, and buds, and to thus extend the discussion on the possible uses of these bioactive molecules from J. regia. As only leaves, buds, and bark have been studied in particular [5], the present study also provides interesting insights into the biochemical compositions of walnut roots and petioles, for which scientific information on their chemical constituents is scarce. Quantification of phenolic compounds across these different plant parts will also provide valuable data on their contents, and will demonstrate where the extraction of individual phenolic compounds might be meaningful. This study thus defines the many phenols that can be identified in the agro-residues of these different walnut tree tissues, and proposes a new direction for future studies for the agro-food, cosmetic, and pharmaceutical industries.

Plant Materials
Samples of walnut leaves, petioles, bark, roots, and buds were obtained from 2-year-old plants of J. regia (n = 10). The plants were grown in Slovenia from mixed seeds of known and unknown cultivars, as commonly used for rootstock for seedling production of J. regia. As older plants cannot be dug up whole without damaging the main and lateral roots, 2-year-old plants were used. This also provided more accurate results on the basis of the phenolic compounds in the whole of each plant tissue, rather than just for a part of the tissue. All of the plants were grown and collected from the Experimental Field of the Biotechnical Faculty of Ljubljana University (Slovenia; 46 • 2'54 N; 14 • 28 22 E; 295 m a.s.l.). The samples were obtained from a total of 10 plants, with two plants used as one replicate for analysis, defining a total of five replicates per tissue. Two plants were used for each replicate because there would have been insufficient material for the bud and petiole analyses if a single plant was used. The plants were transported to the laboratory of the Department of Agronomy in the Biotechnical Faculty, where the tissues were carefully separated. The roots and bark were further subdivided: the bark according to 1 year or 2 years of plant stem growth, and the roots according to the main root and side roots, as shown in Figure 1.
of the Department of Agronomy in the Biotechnical Faculty, where the tissues were carefully separated. The roots and bark were further subdivided: the bark according to 1 year or 2 years of plant stem growth, and the roots according to the main root and side roots, as shown in Figure 1. After separation of the tissues, they were immediately frozen in liquid nitrogen, lyophilized, and stored at −20 °C prior to further analysis, to avoid oxidation of the compounds that they contained.

HPLC-Mass Spectrometry Analysis of the Individual Phenolic Compounds
Analysis of the phenolic compounds was carried out on an UHPLC system (Vanquish; Thermo Scientific, Waltham, MA, USA), with a diode array detector at 350 nm to detect flavonols, and at 280 nm to detect hydroxycinnamic acids, hydroxybenzoic acids, flavanols, napthoquinones, and coumarins. A C18 column (Gemini; 150 × 4.60 mm, 3 μm; Phenomenex, Torrance, CA, USA) was used to separate the phenolic compounds. The After separation of the tissues, they were immediately frozen in liquid nitrogen, lyophilized, and stored at −20 • C prior to further analysis, to avoid oxidation of the compounds that they contained.

HPLC-Mass Spectrometry Analysis of the Individual Phenolic Compounds
Analysis of the phenolic compounds was carried out on an UHPLC system (Vanquish; Thermo Scientific, Waltham, MA, USA), with a diode array detector at 350 nm to detect flavonols, and at 280 nm to detect hydroxycinnamic acids, hydroxybenzoic acids, flavanols, napthoquinones, and coumarins. A C18 column (Gemini; 150 × 4.60 mm, 3 µm; Phenomenex, Torrance, CA, USA) was used to separate the phenolic compounds. The spectra were recorded from 200 nm to 600 nm, and the other parameters were as described by Medic et al. [6].
Tandem mass spectrometry (MS/MS; LTQ XL; Thermo Scientific, Waltham, MA, USA) with heated electrospray ionization operating in negative ion mode was used for identification of the phenolic compounds. The parameters were as described by Medic et al. [6]. The data were processed using the Xcalibur 2.2 software (Thermo Fisher Scientific Institute, Waltham, MA, USA). For identification and quantification of known compounds, external standards were used. For identification of unknown compounds, MS fragmentation and literature data were used, with quantification using the most relevant similar standards. As the contents of juglone, hydrojuglone, and 1,4-naphthoquinone are usually very low and other compounds can interfere with their quantification on HPLC, more accurate content quantification was obtained using MS/MS (as above). Hydrojuglone-β-D-glucopyranoside was quantified by both UHPLC and MS/MS to compare the accuracy of the UHPLC and MS quantification for the compounds that were present at higher levels. The rest of the compounds were quantified using the UHPLC system. The contents of the individual phenolic compounds are given as grams per kilogram dry weight.
Acetonitrile and formic acid for the mobile phases were of HPLC-MS grade (Fluka Chemie GmbH, Buchs, Switzerland). The water used for all sample preparation, solutions, and analyses was bi-distilled and purified using a Milli-Q water purification system (Millipore, Bedford, MA, USA).

Statistical Analysis
The data were collated using Microsoft Excel 2016, and analyzed using R commander. For each methodology (tissue), five repetitions were performed. The data are expressed as means ± standard error (SE), and one-way analysis of variance (ANOVA) with Tukey's test was used to determine significant differences between the data. Statistical means at a 95% confidence level were calculated, to determine the significance of the differences.

Identification of Individual Phenolic Compounds in Walnut
A total of 91 phenolic compounds were tentatively identified in these J. regia plant tissues, based on the literature and the use of standard compounds. Of these 91 phenolic compounds, 21 were identified using standards. Fragmentation of both the standards and addition of external standards to the samples were used to confirm the identities. The remaining 70 phenolic compounds were tentatively identified according to their specific fragmentation patterns and pseudo molecular ions [MH] − . The identified phenolic compounds are shown in Table 1. HPLC-MS full scans, along with the compounds identified, are included in the Supplementary Materials, as Figures S1-S7. Most of the phenolic compounds identified in these J. regia tissues were in the roots (41), followed by petioles, bark, and buds (38), with the least identified in the leaves (37). The majority of hydroxycinnamic acids were detected in leaves (6), hydroxybenzoic acids in roots (25), flavanols in petioles (7), and flavonols in leaves and petioles (13). Napthoquinones were similar in leaves, petioles, bark, and buds (13), and in roots (12). The only two coumarins identified were in the roots. Interestingly, of the 28 hydroxybenzoic acids, none were identified in leaves and petioles; conversely, of the 20 flavonols, none were identified in roots.
For the eight hydroxycinnamic acids: neochlorogenic acid (3-caffeoylquinic acid), cryptochlorogenic acid (4-caffeoylquinic acid), and chlorogenic acid (trans-5-caffeoylquinic acid) were identified through their fragmentation in addition to an external standard; 3-p-coumaroylquinic acid was identified through its fragmentation pattern of MS m/z 337 and MS 2 m/z 163, 191, and 173, and by the retention time previously reported by Medic et al. [5] in J. regia; p-coumaroylquinic acid was identified through its fragmentation pattern; caffeic acid derivatives were identified through their fragmentation patterns of MS n m/z 179; and p-coumaric acid derivatives were identified through their fragmentation patterns of MS n m/z 163 and 119, as reported previously by Medic et al. [5].
The hydroxybenzoic acids included the identification of 28 phenolic compounds. Here, bis-HHDP-glucose 1 and 2, tellimagrandin isomers (digalloyl-HHDP-glucose) 1 and 2, and strictin/isostrictin isomer (galloyl-HHDP-glucose) were previously identified and quantified in J. regia pellicle [1], and are here reported for the first time in the other tissues of J. regia. Ellagic acid and its derivatives were identified through their fragmentation patterns of MS n m/z 301 and MS n + 1 m/z 257, 229, and 185, and gallic acid derivatives through their fragmentation patterns of MS n m/z 169 and 125, as reported by Medic et al. [1]. Many of the ellagic and gallic acid derivatives had also been previously identified in J. regia pellicle [1], and bark and buds [5]. 3-O-Methylellagic acid-4-O-β-D-arabinopyranoside was identified in bark and roots through its fragmentation pattern of MS m/z 447 and MS 2 m/z 315 and 300, as previously reported for Caesalpinia ferrea bark by Wyrepkowski et al. [16], and here for the first time in J. regia. 1-O-(4-Hydroxy-3,5-dimetoxybenzoyl)-D-glucopyranoside was also identified in bark and roots, through its fragmentation pattern of MS m/z 359, MS 2 m/z 299, 239, and 197 and MS 3 m/z 153 and 181, as previously reported by Huo et al. [17] for Juglans mandshurica, and here for the first time in J. regia.
Of the 11 flavanols, (+)catchin and (−)epicatechin were identified through their fragmentation in addition to external standards. (epi)Catechin derivatives were identified through their fragmentation patterns of MS n m/z 245, 205, and 179, and procyanidin derivatives through their fragmentation patterns of MS n m/z 577 and MS n + 1 m/z 425, 407, and 289, as previously reported by Medic et al. [5]. Epigallocatechin was identified through its fragmentation pattern of MS n m/z 179, 221, and 125, and MS n + 1 m/z 165, 151, 137, and 109, as previously reported by Ambigaipalan et al. [18]. Epigallocatechin was reported here for the first time in J. regia.
There were 20 flavonols identified here, with many previously reported by Medic et al. [5,6]. The flavonols included the identification of three groups of compounds: (i) quercetin glycosides, through their fragmentation patterns of MS n m/z 301 and MS n + 1 m/z 179 and 151; (ii) kaempferol glycosides, through their fragmentation patterns of MS n m/z 284 and 285, and MS n + 1 m/z 255 and 227; and (iii) myricetin glycosides, through their fragmentation patterns of MS n m/z 316 and 317, and MS n + 1 m/z 179 and 151, as reported by Viera et al. [19], Santos et al. [20], and Medic et al. [5].
The two coumarins identified in the roots were isofraxidin and isofraxidin derivative, and these were identified through their fragmentation patterns of MS n m/z 221, MS n + 1 m/z 206 and 191, and MS n + 2 m/z 177, 163, and 135, according to Tsugawa et al. [21]. These compounds are reported here for the first time in J. regia, or any other Juglans species.

Quantification of Individual Phenolic Compounds in Walnut
The highest contents of the phenolic compounds were in the J. regia main roots, followed by the side roots and buds, then leaves and 1-year-old bark, with the lowest in the petioles and 2-year-old bark, as shown in Figure 2A.
The reason why the roots showed the highest content of phenolic compounds compared to other tissues was mainly because of their higher content of hydroxybenzoic acids and naphthoquinones, which are known for their defense mechanisms against pathogens [22,23]. As the soil contains more pathogens then are present above ground [23], higher phenolic content, especially content of hydroxybenzoic acids and naphthoquinones, was expected to be found in underground tissues as observed. Therefore, the roots represent a particularly good source of hydroxybenzoic acids, while, the leaves, roots, and buds all contained high levels of naphthoquinones. On the other hand, the contents of flavanols and flavonols, which are known to have health-promoting effects [6], were the highest in leaves and buds. Both flavanols and flavonols are typically found in leaves and are considered to have a defensive role against viral and bacterial infections that usually affect leaves [6], therefore the highest content was expected to be present in the leaves as observed. Both petioles and 2-year-old bark were less suitable sources of phenolic compounds compared to the other tissues. Previously, buds have been suggested as the best source of phenolic compounds [5]; however, as shown here, the roots contained almost twice the levels of the buds. Roots are also more abundant, and thus the extraction of phenolic compounds would be more meaningful for roots, rather than buds, especially when old walnut orchards are dug up. The results of these total analyzed phenolic compounds in bark and buds of juvenile plants was in agreement with the content of total analyzed phenolic compounds in bark and buds of 24-year-old J. regia plants. The compositions were, however, slightly different, as juvenile plants contained higher levels of naphthoquinones and hydroxybenzoic acids and lower levels of flavanols, compared to the 24-year-old plants reported by Medic et al. [5]. The reason why the roots showed the highest content of phenolic compounds compared to other tissues was mainly because of their higher content of hydroxybenzoic acids and naphthoquinones, which are known for their defense mechanisms against pathogens [22,23]. As the soil contains more pathogens then are present above ground [23], higher The highest relative contents of naphthoquinones were seen for leaves, flavanols for petioles, hydroxycinnamic acids for bark, hydroxybenzoic acids and coumarins for roots, and flavonols for buds ( Figure 2B). The total naphthoquinones were the major phenolic group in leaves, petioles, bark, and buds, where they represented >60% of the phenolic compounds identified; conversely, hydroxybenzoic acids were the major phenolic group in side roots, at~50% of the phenolic compounds identified. The total phenolic compounds in roots was higher than any previously reported for J. regia kernel [1], husk [5], leaves [22], shoots [24], bark [5], or buds [5], which further justifies the use of roots as a valuable source of phenolic compounds; instead, the use of petioles cannot be justified. The contents of the total and individual phenolic compounds in each of the tissues analyzed here are given in Table 2.