Article pubs.acs.org/JAFC
Mechanisms of Degradation of the Natural High-Potency Sweetener (2R,4R)‑Monatin in Mock Beverage Solutions Corin Storkey,†,‡ David I. Pattison,†,‡,§ Dan S. Gaspard,∥ Erik D. Hagestuen,∥ and Michael J. Davies*,†,§ †
The Heart Research Institute, 7 Eliza Street, Newtown, New South Wales 2042, Australia Faculty of Medicine, University of Sydney, Sydney, New South Wales 2006, Australia ∥ Cargill, Inc., 15407 McGinty Road West, Wayzata, Minnesota 55391, United States §
S Supporting Information *
ABSTRACT: The sodium, potassium, or mixed sodium/potassium salt of the naturally occurring high-potency sweetener (2R,4R)-monatin, also known by the common name arruva, degrades over time in model beverage solutions in the presence of light. By use of UHPLC, LC−MS/MS, and peroxide assays, it has been demonstrated that degradation is accelerated by UV/ visible light and the presence of trace metal ions. Data are presented that are consistent with a role for singlet oxygen (1O2), free radicals, and peroxides (both H2O2 and organic peroxides) in monatin oxidation. Separation of degradation products by UHPLC/HPLC or LC−MS/MS provided evidence for the formation of hydroxylated and peroxide species formed on the indole ring (mass increases 16 and 32, respectively) as well as multiple ring and side-chain oxidation and scission products. Model oxidation systems using the photosensitizer Rose Bengal as a source of 1O2 support the proposed photodegradation pathways. KEYWORDS: monatin, tryptophan, photo-oxidation, sweetener, metal ions, radicals, singlet oxygen
■
INTRODUCTION Consumption of sucrose and high fructose corn syrups as sweeteners in beverages has been associated with multiple nutritional and medical problems.1−8 As a result, there has been a shift toward the use of noncaloric, noncariogenic high-potency sweeteners as a substitute for sucrose and high fructose corn syrups. Many currently available potently sweet, low-calorie sucrose substitutes are synthetic materials, with these including acesulfame-K, alitame, aspartame, cyclamate, saccharin, and sucralose.9 Many of these sweeteners have been associated with unpleasant off-flavors or undesirable sweetness profiles. Recent studies have focused on the highly sweet tryptophan derivative monatin (sodium/potassium (2R,4R)-2-amino-4-carboxy-4-hydroxy-5-(3-indolyl) pentanoate) [1], also known as 2-Hydroxy2-(indol-3-ylmethyl)-4-aminoglutaric acid and by the common and usual name arruva (Scheme 1; cf. tryptophan [2]). Monatin is isolated from an African plant, Schlerochiton ilicifolius A. Meeuse (Acanthaceae) and has a sweetness potency ∼1300 times that of sucrose, with this depending on the stereoisomer examined.10
Purified (2R,4R)-monatin has been reported to be over 3000 times sweeter than sucrose at 5% sucrose equivalence, making it one of the most potently sweet naturally occurring substances known.11 Monatin is the only known native amino acid with a highly sweet taste12,13 and has the benefit of containing no sugar component and almost zero calories.14 In addition, monatin has a remarkably clean taste.15,16 These properties make monatin a desirable material for use in tabletop sweeteners, beverages, and other food products. Monatin has two chiral centers leading to four potential stereoisomers; (2R,4R)-monatin, (2S,4S)-monatin, (2R,4S)monatin, and (2S,4R)-monatin, with the first of these exhibiting the most potent sweetness.17 Because of the potential commercial applications of monatin and the low abundance of this material in the native plant, synthetic methods have been devised for production of this compound; these encompass both chemoenzymatic18−23 and chemical24 methods. As a result of the interest in the use of monatin as a highpotency, low calorie sweetener in beverages,15,25,26 extensive studies have been carried out to overcome potential formulation challenges that arise from the chemical stability of monatin under standard beverage conditions. These involve chemical modification of monatin by pH, electromagnetic radiation, and dissolved gases, resulting (in some formulations) in the development of undesirable characteristics, such as discoloration, loss of sweetness, and the formation of unpleasant flavors or aromas.25−28 To prevent or minimize such degradation, mechanistic and stability studies are needed to fully understand
Scheme 1. Structures of the Amino Acid Derivative Monatin (2R,4R)-2-amino-4-carboxy-4-hydroxy-5-(3-indolyl) pentanoate or 2-Hydroxy-2-(indol-3-ylmethyl)-4aminoglutaric Acid) [1] and the Amino Acid Tryptophan (Trp) [2]
Received: Revised: Accepted: Published: © 2014 American Chemical Society
3476
September 18, 2013 February 6, 2014 March 16, 2014 March 17, 2014 dx.doi.org/10.1021/jf404198w | J. Agric. Food Chem. 2014, 62, 3476−3487
Journal of Agricultural and Food Chemistry
Article
South Rydalmere, NSW, Australia) following 2 μL injections. Monatin eluted at 5.07 min, and the corresponding monatin lactone (lactone) at 4.33 min. Peak areas were determined using Lab Solutions 5.32 SP1 (Shimadzu). Accurate mass and MS/MS analyses of monatin products (20 μL sample injections) were carried out using an Agilent 1290 UHPLC system (Agilent Technologies, Santa Clara, CA, USA) coupled to an Agilent 6530 QTOF instrument (Agilent Technologies) operating in both positive and negative ionization modes with an Agilent Jet Stream electrospray source. The QTOF spectrometer was operated under the following conditions: source gas temperature of 350 °C, drying gas flow of 10 L/min, nebulizer gas pressure of 20 psig, sheath gas temperature of 385 °C, sheath gas flow of 12 L/min, VCap of 3000 V, and a fragmentor voltage of 125 V. Ultrahigh purity nitrogen was used as the collision gas for MS/MS experiments. Analysis of Monatin Oxidation Products by LC−MS/MS. A Thermo LCQ Deca XP MAX Plus (Thermo Fisher, North Ryde, NSW, Australia) ion-trap instrument coupled to a Finnigan Surveyor HPLC (Thermo Fisher) system was used with electrospray ionization in positive ion mode. Samples were held at 10 °C prior to injection (20 μL), before separation at 40 °C at a flow rate of 100 μL min−1. Compounds were eluted using the following gradient: 100% solvent A for 10 min, then a linear increase to 40% B over 20 min, followed by a rapid increase to 100% B over 5 min, a 5 min wash with 100% B, before returning to 100% A over 2 min and reequilibration with solvent A for 18 min, to give a total run time of 70 min. The electrospray needle was held at 4.5 kV. Helium was used as the collision gas, and nitrogen was used as the sheath and sweep gas set to 70 and 30 arbitrary units, respectively. The temperature of the heated capillary was 250 °C. Selective ion monitoring (SIM) detection was carried out with an isolation width of ±1 mass unit from the specified m/z values as required, with collision energies of 28 or 35%. Quantification of Peroxides by the FOX Assay. Peroxides were quantified using a modified FOX assay, with data reported as H2O2 equivalents as the reaction stoichiometry for monatin-derived peroxides with the FOX reagent is unknown.34 The FOX reagent was prepared from xylenol orange (3,3-bis[N,N-bis(carboxylmethyl)amino-methyl]o-cresolsulfonephthalein, Na+ salt), sulfuric acid, and iron(II) sulfate (5 mM, 0.5 M, and 5 mM, respectively). Organic peroxide levels were determined after treatment (15 min) of samples with catalase (50 μL of a 1 mg mL−1 solution) to remove H2O2 before addition of the FOX reagent. Previous studies have verified that organic peroxides are not affected by catalase.33,35 Sample Concentration for UHPLC and LC−MS/MS Analysis. Photolyzed monatin standard beverage samples were concentrated as necessary using the following procedure. Ten milliliters of solution was treated overnight with washed Chelex 100 resin (BioRad, Gladesville, NSW, Australia) to remove trace metal ions. The supernatant was removed, and 1 mL aliquots were dried by vacuum centrifugation (18 h). Dried samples were reconstituted in 100 μL of Nanopure water (giving a 10-fold concentration of the sample) before filtering using 0.2 μm centrifugal filters. Samples were kept at 40 μM), with lower, but significant, levels (≤4 μM) of monatin peroxides. The peroxide levels achieved in D2O buffer were higher (up to 60 μM) than in samples made up using H2O or Chelex-treated buffer (57 and 46 μM, respectively) after 42 days of illumination (Figure 3). The corresponding peroxide levels in samples kept in the dark were significantly lower (
