J. Agric. Food Chem. XXXX, xxx, 000 A Synephrine Content of Juice from Satsuma Mandarins (Citrus unshiu Marcovitch) KLAUS DRAGULL,ANDREW P. BREKSA III,* AND BRIAN CAIN Western Regional Research Center, Agricultural Research Service, U.S. Department of Agriculture, 800 Buchanan Street, Albany, California 94710 Synephrine, the main protoalkaloid in Citrus species, is commonly analyzed as the active component in citrus peel-containing herbal supplements, but the edible parts of mandarins have been largely ignored. The synephrine concentration has been determined in the juices of Citrus unshiu mandarins harvested from 10 different groves located in a major growing region in California. For comparison, the physicochemical properties of the juices, including pH, conductivity, soluble solids content, and titratable acidity, were also measured. The synephrine values among 10 groves ranged from 73.3 to 158.1 mg L-1. Repeat sampling of fruit from the 10 locations showed that the intragrove variability in synephrine concentrations ranged from 1.0 to 27.7% CV and was grove dependent. Among the physicochemical properties, titratable acidity weakly correlated with synephrine, and for one sample a low maturity index was linked to high synephrine content. The overall mean synephrine concentration of 92.8 mg L-1 is up to 6-fold higher than values previously determined for orange juices and suggests that mandarin juice could constitute a significant dietary source of synephrine. Furthermore, the results suggest that grove location and maturity affect synephrine content. KEYWORDS: Citrus; juice; synephrine; Citrus unshiu; Citrus reticulata; mandarin; tangerine; Satsuma; variability INTRODUCTION ucts, that the presence of synephrine in citrus fruits and products - Synephrine (Figure 1) is a phenolic amine that was initially became of interest (6 13). With the resurgence in the interest isolated as synthetic product and used pharmacologically as a in synephrine, multiple analytical methods for the quantitation vascoconstrictor and bronchiectatic agent (1). The presence of of synephrine have been reported, and a review of these methods synephrine in citrus leaves was first reported in 1964 by Stewart was recently completed by Pellati et al. (14). A direct com- et al. (2), who in the course of evaluating the micronutrient parison of the synephrine concentrations reported among all contents of several hundred varieties determined that the greatest these papers is difficult due to the ambiguous sample descrip- synephrine concentrations were found in mandarin and orange tions found in some and differences in sample preparation. leaves. No synephrine was detected in the leaves from grapefruit, However, as a whole, they support the original observations pummelo, or shaddock trees; neither was it detected in roots of made by Stewart et al. (2) and have additionally contributed by any of the citrus varieties tested or in leaf samples taken from identifying sour orange as a citrus with synephrine concentra- trees suffering from Mn deficiencies (2). Results obtained in a tions comparable to those found in mandarins (Citrus reticulata follow-up study examining the juice concentrations of syneph- Blanco, Citrus unshiu Marcovitch). rine corresponded well with those observed for leaves, but were As indicated above, the literature contains a number of at concentrations 10-20 times less than those found in papers detailing the development and demonstration of leaves (3, 4). A subsequent analysis of the phenolic amine analytical methods for determining synephrine concentrations. content of Citrus and another 187 plant species demonstrated The emphasis on method development has resulted in a wide that Citrus possessed the greatest concentrations of synephrine breadth of Citrus samples analyzed and within those samples and that the highest concentrations in citrus were at least 10- the identification of those cultivars with the highest syneph- fold higher than those in the other plants evaluated (5). Citrus rine concentrations. However, the variability in synephrine spp. are the only known plants having synephrine in the edible concentrations within the cultivars tested is still largely portions. unknown because in most cases only single representative It was not until recently, when sour orange (Citrus aurantium samples were used. The main objective of this study was to L.) derived botanicals and extracts were introduced to consumers close this gap. An understanding of the variability in as alternatives to the FDA-banned ephedrine-containing prod- concentration, and eventually the effects of environment and * Corresponding author [telephone (510) 559-5898; fax (510) 559- agricultural practices, will be essential if citrus growers and 5849; e-mail [email protected]. processors are to capitalize on the economic potential of 10.1021/jf801225n This article not subject to U.S. Copyright. Published XXXX by the American Chemical Society Published on Web 09/05/2008 B J. Agric. Food Chem., Vol. xxx, No. xx, XXXX Dragull et al. Table 1. Summary of Physicochemical Properties and Synephrine Concentrations (n) 30) soluble solids total titratable maturity conductivity, content (SSC), acidity (TTA), index synephine, ° -1 Figure 1. Chemical structure of synephrine. pH mS Brix % (w/v) (SSC/TTA) mg L low 3.12 1.89 10.8 0.84 8.95 54.5 delivering dietary levels of synephrine through their fresh high 3.59 4.57 15.6 1.74 14.93 160.2 fruits and products. Tangerine or mandarin oranges are ideally average 3.43 2.73 12.7 1.02 12.65 98.1 median 3.44 2.70 12.5 0.96 12.48 92.8 suited to be utilized in this way because they already possess % CV 3.21 19.84 8.66 19.60 12.51 28.10 many properties important to consumers (e.g., easy peeling, seedless, etc.) and have some of the highest reported -1 synephrine concentrations (i.e., Cleopatra, 280 mg L )(3). was dissolved in deionized water, diluted to volume with the same, For this study, we chose to characterize the synephrine and stored at 4 °C until analysis. concentrations found in peeled fruit from commercial groves Samples and standard (10 µL) were analyzed under isocratic located in Placer County (California) of the Owari selection conditions (9)at25°C on a HPLC system consisting of a Waters 2695 of Satsuma mandarin (C. unshiu Marcovitch). Placer County (Milford, MA) coupled to a SpectraSystem UV6000LP photodiode array is a major mandarin-producing region in the state. To gain detector (Thermo Separation Products, San Jose, CA) set to scan from better insight into the variations in synephrine concentration 190 to 450 nm. The mobile phase, ACN/H2O (90:10 v/v) with found within and across the groves, mandarins were obtained ammonium acetate at 10 mM concentration, was filtered through a 0.45 µm Teflon filter prior to use. The flow was 1 mL min-1, and the from 10 different groves. Physicochemical properties of the wavelength used for quantification was 225 nm. Injections were harvested fruits were measured in addition to the synephrine performed in triplicate. Four-point calibration curves were established concentrations to facilitate comparison of the samples. with the synephrine standard covering 22.4-240 µgmL-1, and each level was injected before and after each sample set as standard brackets. MATERIALS AND METHODS R2 values were typically 0.997 or greater. After 12 sample injections, -1 Materials and Chemicals. ACN (HPLC grade), ammonium acetate the column was washed with 90% ACN (aq) for 15 min at 1 mL min , -1 (enzyme grade), and water (HPLC grade) were purchased from Fisher followed by 100% ACN for 15 min at 1 mL min . Scientific (Fair Lawn, NJ). Alternatively, water was deionized to g18.1 Samples were typically analyzed within 24 h. In some instances, a MΩ/cm resistance using a Barnstead NANOpure Deionization System white precipitate formed, presumably pectins, in samples that were (Dubuque, IA) and filtered through a 0.45 µm type HA membrane filter stored for several hours under refrigeration prior to analysis. These (Millipore, Billerica, MA) prior to use. The (()-synephrine standard samples were reclarified by centrifugation to protect the HPLC column. was purchased from ChromaDex (Santa Ana, CA). Initial tests had shown that neither storage (up to 1 week) nor Plant Materials. Fruits [Owari selection of Satsuma mandarin (C. recentrifugation had any measurable effect on quantification results. unshiu Marcovitch)] were harvested from each of 10 different groves located in Placer County, CA. Harvesters were instructed to pick mature RESULTS AND DISCUSSION fruits of random sizes from throughout the entire grove. Fruits were harvested in mid-November 2007 and stored for <1 week at 5 °C until From among the available methods for quantifying synephrine sampling. we selected Pellati et al.’s recently reported method based upon Sample Preparation and Measurement of Physicochemical Prop- a pentafluorophenyl stationary phase (9) because of its analysis erties. For each location, healthy and undamaged fruits were randomly time and direct detection of synephrine by UV. This method, divided into three groups. For each group, 10 fruits were used for the and a version modified for detection by mass spectrometry put analysis. Because the peels of mandarin fruits are known to contain synephrine (9), the use of a juice reamer was ruled out to avoid forth by Nelson et al. (10), were developed for the analysis of potentially contaminating the juice with synephrine derived from the extracts derived from dried fruit materials, dietary supplements, peel. Fruits were therefore peeled manually (the peelers wore nitrile and standard reference materials. In contrast, we chose clarified gloves) and then blended in an Osterizer Classic mixer (now Jarden juice as our intended target to reduce the number of sample Consumer Solutions, Boca Raton, FL) until no chunks of pulp remained preparation steps. However, in applying clarified mandarin juices (10-20 s). The resulting juice was clarified by centrifugation at 4700 to the pentafluorophenyl stationary phase, we observed the rpm for 4 min using a Marathon 8K from Fisher Scientific Ltd.