Authentication of Citrus spp. Cold-Pressed Essential Oils by Their Oxygenated Heterocyclic Components

Citrus essential oils are routinely adulterated because of the lack of regulations or reliable authentication methods. Unfortunately, the relatively simple chemical makeup and the tremendous price variations among Citrus varieties encouraged the interspecies adulteration of citrus oils. In this study, a sensitive UPLC-MS/MS method for the quantitation of 14 coumarins and furanocoumarins is developed and validated. This method was applied to screen the essential oils of 12 different Citrus species. This study, to our knowledge, represents the most comprehensive investigation of coumarin and furanocoumarin profiles across commercial-scale Citrus oils to date. Results show that the lowest amount was detected in calamansi oil. Expressed oil of Italian bergamot showed the highest furanocoumarin content and the highest level of any individual furanocoumarin (bergamottin). Notable differences were observed in the coumarin and furanocoumarin levels among oils of different crop varieties and origins within the same species. Potential correlations were observed between bergapten and xanthotoxin which matches with known biosynthetic pathways. We found patterns in furanocoumarin profiles that line up with known variations among the Citrus ancestral taxa. However, contrary to the literature, we also detected xanthotoxin in sweet orange and members of the mandarin taxon. Using multivariate analysis, we were able to divide the Citrus oils into 5 main groups and correlate them to the coumarin compositions.


Introduction
Citrus essential oils (EOs) have several applications in cosmetics, the food industry, and the flavor and fragrance industry. They are also utilized as natural preservatives because of their wide range of biological activities, which include antioxidant and antimicrobial actions [1]. These strong biological activities are attributed to the presence of terpenes, flavonoids, carotenes, and coumarins [2]. Several studies have investigated the volatile makeup of various parts of Citrus species due to their significant economic importance. All cold-pressed Citrus oils contain a portion of non-volatiles fundamentally made of simple coumarins, psoralens, and methoxy-flavones [3]. Coumarins (1,2-benzopyrones) are a huge family of naturally occurring secondary metabolites. Psoralens, also known as furanocoumarins (FCs), are a large family of compounds commonly found in Rutaceae, Apiaceae, and Fabaceae, with Rutaceae containing the highest concentrations [4,5]. FCs contain a furan ring fused to a coumarin core [6]. The fusion helps separate the FCs into linear or angular structural forms. FCs have shown a potential to elicit variable degrees of phototoxic skin reactions. In comparison to angular FCs, linear FCs have often been proven to cause phototoxic responses at lower doses [7].
While studying the volatile composition of Citrus oils, the nonvolatile fractions are hard to detect under standard gas chromatography conditions because of their limited volatilities, relatively polar or heat-liable nature. These nonvolatile ingredients may hold the secret to constructing a perfect analytical strategy for interspecies adulteration detection. the secret to constructing a perfect analytical strategy for interspecies adulteration detection. This essential fraction of the cold-pressed oil can be used to identify speciesspecific patterns and establish Citrus species fingerprinting. For instance, creating synthetic bergamot oils or adulterating bergamot oils with similar Citrus oils like bitter orange are simple strategies to boost profits. Both strategies make it essentially impossible for consumers to detect the difference. The non-volatile fraction contributes very little to the Citrus oils' aroma, but because of its high complexity, commercial unavailability, or extremely high cost in comparison to the Citrus oils themselves, it is more difficult to manipulate. Previous studies report the separation and identification of coumarins and FCs in Citrus peel extracts and oils using gas chromatography-mass spectroscopy (GC-MS) after derivatization [8], high-performance liquid chromatography (HPLC) [8], enzyme-linked immunosorbent assay (ELISA) [9], reversed-phase (RP)-HPLC [10,11], HPLC-diode array detector (DAD) [12], HPLC-nuclear magnetic resonance (NMR) [13], ultra-performance liquid chromatography coupled with mass spectrometry (UPLC-MS) [14,15], LC-MS [16], and HPLC-UV-MS [17].

Method Validation
The LC-MS/MS chromatogram of 14 coumarins using the MRM acquisition mode is shown in Figure 2. Specificity, precision, accuracy, linearity, intermediate precision, and LOQ results are summarized in Table 1. The method proved specific to the target compounds since no interferences were found in any of the processed blanks. All com- The LC-MS/MS chromatogram of 14 coumarins using the MRM acquisition mode is shown in Figure 2. Specificity, precision, accuracy, linearity, intermediate precision, and LOQ results are summarized in Table 1. The method proved specific to the target compounds since no interferences were found in any of the processed blanks. All compounds met the acceptance criterion of RSD% ≤ 10 based on the precision and intermediate precision results. Compound recovery percentages ranged from 94.07 to 114.53% of the expected value. The linearity of the calibration curve of the 14 compounds was well correlated (r ≥ 0.98) within a range of 0.00010.1 ppm. The LOQ values ranged from 0.0001 to 0.005 ppm. These findings demonstrate that the developed method is suitable for analyzing the 14 targeted compounds in EOs.     (Table 3). The non-volatile components are far more species-specific than the volatile components, which have comparable patterns in different Citrus oils. We found patterns in FC profiles that correspond with published differences among the Citrus ancestral taxa [15,18]. Our findings are in line with previous reports that found a mixture of FCs from the bergapten, xanthotoxin, and isopimpinellin clusters in EOs derived from the citron (C. medica) and papeda (C. micrantha) ancestral taxa [15]. In this study, EOs derived from fruits of the mandarin taxa (mandarin, clementine, and tangerine) showed low total coumarin (2.44-149.45 ppm) and FC levels (2.44-149.45 ppm), not aligning with a previous report that this taxon is nearly devoid of FCs [14,15]. Interestingly, trioxsalen and toncarine were not detected in any of the Citrus EOs. Epoxybegamottin was absent from calamansi, clementine, mandarin, kaffir lime, and petitgrain oils. The content of psoralen was almost negligible in most of the Citrus EOs but was relatively high in white grapefruit EO (82.65 ± 0.76 ppm). Furthermore, large amounts of xanthotoxin were detected in bergamot and lime EOs. Previous reports indicate that xanthotoxin is absent from sweet orange (C. sinensis, pummelo taxon) and the mandarin taxa [14,15] while we found 4.85 ± 0.32 ppm in sweet orange EO and 0.33-15.24 ppm xanthotoxin in the mandarin taxa EOs. Bergamottin and 5-geranyloxy-7-methoxycoumarin were reported in mandarin, lemon, and lime oils but not in orange oil [17]. The differences between our findings and previous studies could be due to genetic and/or environmental impacts on FC biosynthesis [19]. Our LOQ, however, may be lower than that of other reports because it was based on the weight of EO rather than the weight of fresh fruit peel. Alternative explanations for the differences in our findings include genetic admixture in Citrus varieties or contamination during processing and handling.

Multivariate Analysis
In order to examine the similarities and relationships between the coumarin compositions and the Citrus essential oils, AHC and PCA were carried out based on a data matrix comprised of 28 Citrus "types" and 12 coumarin components. Based on > 25% similarity, the AHC shows five groups (Figure 3): Group 1 (bergamot from Italy and bergamot from Brazil), Group 2 (lime and lemon from Germany), Group 3 (yuzu, red and white grapefruit), Group 4 (a large group composed of oranges, tangerines, clementines, mandarins calamansi, and petitgrains), and Group 5 (lemons). The PCA analysis (Figure 4) of the Citrus essential oils indicates that F1 and F2 explain 78.43% of the variation in coumarin compositions among the Citrus types. The bergamot group (Group 1) is positively correlated with bergamottin, bergapten, and xanthotoxin; the lemon group (Group 5) positively correlates with biacangelicol and oxypeucedanin as well as citropten and 5-geranyloxy-7-methoxycoumarin. The grapefruit and yuzu group (Group 3) correlate with 6 ,7 -epoxybergamottin and psoralen. Group 4 (oranges, mandarins, clementines, etc.) are characterized as having relatively low levels of coumarins. A positive correlation was found between bergapten and xanthotoxin (2 structures related by a common precursor in biosynthesis [15]). characterized as having relatively low levels of coumarins. A positive correlation was found between bergapten and xanthotoxin (2 structures related by a common precursor in biosynthesis [15]).

Essential Oil Samples
Citrus volatile oils from trusted suppliers were obtained from the collection of the Aromatic Plant Research Center (APRC, Lehi, UT, USA). A total of 374 cold-pressed Citrus oil samples from the APRC collection are listed in Table 4. A simple dilute and shoot technique (1 µL oil in 999 µL of methanol) was used for sample preparation. Further dilution was performed whenever needed.

UPLC-MS/MS Analyses
Coumarins were quantified using a NEXERA UPLC system (Shimadzu Corp., Kyoto, Japan) equipped with a mass spectrometer (Triple quadrupole, LCMS8060, Shimadzu, Kyoto, Japan). Target compounds were chromatographed on a Shimadzu Nexcol C 18 column (1.8 µm, 50 × 2.1 mm) with a C 18 guard column (Restek, Bellefonte, PA, USA) at 40 • C. The mobile phase consisted of 0.1% formic acid in water (A) and 0.1% formic acid in methanol (B). The compounds were eluted using the following gradient: %10 B at 0 min, %20 B at 0.74 min, %60 B at 5.88 min, %90 B at 10 min, held at %100 B for 4 min, and %10 for 4 min before the next injection. The flow rate was maintained at 0.2 mL/min, and the injection volume was 1 µL. The UPLC system was connected to the MS by electrospray ionization (ESI) operating in positive ion mode. The interface, desolvation line, and heating block temperatures were 350, 250, and 400 • C, respectively. The capillary voltage was 4.5 kV, and CID gas was set at 350 kPa. Nebulizing gas flow was set at 3.0 L/min, and heating and drying gas were set at 10.0 L/min. The detection was completed in multiple reaction monitoring mode (MRM) ( Table 5). Samples were run in triplicates with external standards in between. Each run contained a quality control (QC) standard, and at least one QC standard was run at the beginning and the end of the run. The acquired chromatographic results were processed in LabSolutions Insight software version 3.2 (Shimadzu). For each compound, calibration curves (0.005, 0.001, 0.0025, 0.005, 0.01, 0.025, 0.05, and 0.1 ppm) were drawn by linking its peak area and its concentration.

Method Validation
Method validation was executed according to the USP<1225> Validation of compendial procedures [31] and ICH harmonized tripartite guideline validation of analytical procedures: text and methodology Q2(R1) [32]. Specificity, precision, accuracy, linearity, intermediate precision, and limit of quantification (LOQ) were determined using standard solutions. Distilled yuzu essential oil was used as a matrix (total coumarins < 0.001 ppm). To prove the specificity of the method, standard solution mixtures and at least three blanks were processed to demonstrate the absence of interferences with the elution of the analytes. Precision and repeatability were determined by injecting six sample preparations spiked to a final concentration of 0.04 ppm and then calculating the RSD% between injections which may reach 10% for each. For the intermediate precision, the repeatability experiment was repeated on a second day and performed by a second analyst with the acceptance criterion of RSD ≤ 10 for each compound and each analyst. To determine the recoveries (accuracy) of the target compounds, three individually prepared samples of yuzu oil were spiked with three concentrations of the standard (LOQ, 0.04, and 0.05 ppm in triplicates). Recoveries were calculated by comparing the absolute peak areas with a reference measurement which must be within 80-120% of the expected value. Five concentrations from 0.001 to 0.1 ppm were used to determine linearity and a coefficient of determination (r) higher than 0.98 was needed. The data obtained during the linearity, precision, and accuracy studies were used to assess the range of the method for the target compounds. The acceptable range was defined as the concentration interval over which linearity, precision, and accuracy are acceptable. To estimate the LOQ, standard mixtures at low concentrations (0.0005 to 0.01 ppm) were analyzed. The calculated LOQ was determined using the signal-to-noise (S/N) ratio (10:1) and then injected 6 times. The acceptance criterion for the LOQ was RSD ≤ 15%. A calibration curve based on the linear range was prepared and injected to estimate the quantity of coumarins in the oil samples. Additionally, QC standards at low (0.05 ppm) and high (0.1 ppm) concentrations were used.

Multivariate Analysis
The average coumarin concentrations (12 compounds) in the Citrus samples were used as variables in the multivariate analysis. First, the data matrix was standardized by subtracting the mean for each compound concentration and dividing it by the standard deviation. For the agglomerative hierarchical cluster (AHC) analysis, the 24 Citrus samples were treated as operational taxonomic units (OTUs). Pearson correlation was selected as a measure of similarity, and the unweighted pair group method with arithmetic average (UPGMA) was used for cluster definition. Principal component analysis (PCA) was performed for the visual comparison of the coumarin compositions of the different Citrus groups using the 12 coumarin components as variables, with a Pearson correlation matrix. The AHC and PCA analyses were performed using XLSTAT v. 2018.1.1.62926 (Addinsoft, Paris, France).

Conclusions
In this study, we developed and validated a simple and sensitive UPLC-MS/MS method for the detection and quantification of 14 selected oxygen heterocyclic compounds (coumarins and furanocoumarins). Targeted screening using this method was successfully completed for the essential oils of 12 different Citrus species. To our knowledge, this is the most comprehensive investigation of coumarin and furanocoumarin profiles across commercial-scale Citrus oils to date. The lowest amount was detected in calamansi oil. Expressed oil of Italian bergamot showed the highest furanocoumarin content and the highest level of any individual furanocoumarin (bergamottin). Remarkable differences were observed in the coumarin and furanocoumarin levels among oils of different crop varieties and origins within the same species. We found potential correlations between bergapten and xanthotoxin which matches with known biosynthetic pathways. Patterns in furanocoumarin profiles lined up with known variations among the Citrus ancestral taxa. Using multivariate analysis, we were able to divide the Citrus oils into 5 main groups (bergamots; lime and German lemon; yuzu and grapefruit; oranges, tangerines, clementines, mandarins, calamansi, and petitgrains; and lemons) and correlate them to the coumarin compositions.

Conflicts of Interest:
The authors declare no conflict of interest.