SWEET TASTE ENHANCER

[01] This application claims the benefit of Serial No. 61/598, 121 filed on February 13, 2012, which is incorporated herein by reference in its entirety.

[02] Each reference cited in this disclosure is incorporated herein in its entirety.

TECHNICAL FIELD [03] This disclosure relates generally to sweeteners for reduced-calorie products.

BACKGROUND

[04] Due to its mouth-watering and highly attractive sapid taste profile centering around well balanced sweet and sour taste, as well as a long-lasting mouthfuUness, traditional balsamic vinegars from Modena (TBV) and Reggio Emilia, respectively, are among the key ingredients of top-level cuisine and are highly appreciated by consumers all over the world.

[05] Traditional balsamic vinegars, certified with the protected designation of origin

(PDO) status in the year 2000 (EC Council Regulation No. 813/2000), are produced by following a traditional but rather time consuming manufacturing process (1 ).

Such traditional balsamic vinegars are made from combinations of Trebbiano and Lambrusco grapes. After cooking the grapes in an open vessel, the reduced must undergoes a spontaneous alcoholic and acetic fermentation in a barrel or "badessa." Thereafter, the vinegar is matured in a sequence of at least five casks made from different woods such as, e.g. , oak, acacia, chestnut, cherry, mulberry, and ash, respectively. Once a year an aliquot of the aged vinegar is taken from the smallest barrel ("prelivio") and the respective next bigger barrel is sequential refilled

("travaso"), whereas the biggest barrel is refilled with must from the badessa

("rincalzo"). During the maturation in this so-called "batteria," the vinegar becomes increasingly concentrated by slow evaporation of water. Depending on their age of maturation, two different varieties of the traditional balsamic vinegar of Modena are available, namely "Affmato" and "Extravecchio" matured for at least 12 and 25 years, respectively, whereas of traditional balsamic vinegar from Reggio Emilia three varieties (aragosta (12 years), argento (18 years) and oro (25 years)) are available (1).

[06] In comparison to these premium-quality TBV products, balsamic vinegar of Modena (BV) is produced on an industrial scale from wine vinegar, coloring and flavoring additives, and greatly differ in price and concentration of monosaccharides, organic acids, amino acids, phenolic acids, and furan-2-aldehydes from the traditional, artisanal-type variety (1 -17). In order to differentiate premium quality traditional balsamic vinegar from balsamic vinegars, tremendous research efforts have been targeted towards the identification of marker molecules to analytically monitor quality and validate authenticity and age of traditional balsamic vinegars (13-19). For example, hexose acetates were recently reported in traditional balsamic vinegars and were claimed to be favorably generated with increasing age of maturation ( 18).

[07] Although the volatile, key odorants have been identified in traditional balsamic

vinegar from Modena by means of gas chromatography/olfactometry, quantified by means of stable isotope dilution analysis, and validated by means of aroma reconstitution experiments (20), any systematic studies on the non-volatile sensometabolome coining the highly attractive orosensory profile of a traditional balsamic vinegar are still lacking. Although the key players imparting the typical taste of red wine were recently identified (21 , 22), the taste compounds and/or putative taste modulators in TBV and TB have yet not been systematically investigated.

Moreover, the impact of the artisanal "batteria"-type maturation on taste-active molecules in TBV has not yet been substantiated by means of molecular-sensory studies. Although ellagitannins like vescalagin (1 ) and castalagin (2), Figure 1 , are well-known astringent compounds originating from oak wood (23), their contribution to the taste of matured TBV is unclear.

[08] A systematic study on the entire nonvolatile sensometabolome of traditional balsamic vinegar has not yet been performed. Thus there is a need to identify and quantify taste active and taste modulatory compounds in TBV and TB, to rank them in their sensory impact based on dose-activity considerations, and to validate their sensory relevance by means of taste re-engineering experiments. BRIEF DESCRIPTION OF THE DRAWINGS

[09] FIG. 1. Structures of oak-derived ellagitannins vescalagin (1 ) and castalagin (2),

Acacia-derived (+)-Dihydrorobinetin (3), and hexose mono acetates 6-0-acetyl-a/ - D-glucopyranose (4) and l-O-acetyl-P-D-fructopyranose (5).

[10] FIGS. 2A-F. LC-MS (MRM) analysis. FIG. 2 A, vescalagin ( 1 ) and castalagin (2) in balsamic vinegar (BV); FIG. 2B, (+)-dihydrorobinetin (3) in BV; FIG. 2C, vescalagin (1 ) and castalagin (2) in traditional balsamic vinegar (TBV); FIG. 2D, (+)- dihydrorobinetin (3) in TBV; FIG. 2E, vescalagin (1) and castalagin (2) in balsamic vinegar (BV) spiked with 1 and 2 prior to analysis; FIG. 2F, (+)-dihydrorobinetin (3) in BV spiked with 1 and 2 prior to analysis.

[11] FIG. 3. GAC chromatogram (λ = 220 nm) of the low molecular weight fraction

(TBV-LMW, <5kDa) isolated from TBVM by means of ultrafiltration.

[12] FIG. 4. (A) HILIC-ELSD chromatogram of GAC fraction VII and (B) RP-HPLC- DAD chromatogram of GAC fraction X isolated from the TBV-LMW fraction.

[13] FIG. 5. HMBC NMR-spectrum (500 MHz, MeOD) and chemical structure of 5- acetoxymethyl-2-furaldehyde (6) isolated from GAC fraction X-2; s.s. solvent signal.

[14] FIG. 6. Influence of maturation on the taste profile of intermediary vinegar samples collected from casks A-H of the "batteria" of TBV manufacturing; data are given as the mean of triplicates.

[15] FIGS. 7A-B. Sensomics heatmap calculated from quantitative data of selected

sensometabolites in intermediary vinegar samples A-H collected from the "batteria" after normalizing concentrations based on fresh weight (FIG. 7A) and (B) dry weight (FIG. 7B)

[16] FIG. 8. Reaction scheme showing the formation of sweet taste modulator 6 via the Maillard reaction product 5-hydroxymethyl-2-furaldehyde and acetic acid generated upon fermentation. DETAILED DESCRIPTION

[17] This disclosure provides a sweetness-enhancing component consisting essentially of 5-acetoxymethyl-2-furaldehyde, which can be used to enhance sweetness of reduced calorie products comprising a nutritive sweetener.

[18] Sensory-directed fractionation of traditional balsamic vinegar of Modena (TBV) using ultrafiltration, gel absorption chromatography, RP-HPLC, and HILIC combined with analytical sensory evaluation of fractions in water and a basic taste recombinant, prepared from 54 taste compounds on the basis of quantitative data, led to the identification of the hexose acetates 6-0-acetyl-a/p-D-glucopyranose and O-acetyl-β- D-fructopyranose as sweet-bitter tasting compounds and 5-acetoxymethyl-2- furaldehyde as a previously unknown sweetness modulator. Taste re-engineering experiments confirmed 5-acetoxymethyl-2-furaldehyde to contribute to the typical long-lasting sweet taste quality of TBV. Compared to TBV, balsamic vinegar of Modena (BV) differed significantly by the increased concentration of acetic acid, the significantly lower concentrations of the sweet-modulating 5-acetoxymethyl-2- furaldehyde, the non-volatile organic acids and polyphenols, and the lack of wood- derived ellagitannins. Quantitative monitoring of 37 selected sensometabolites throughout the intermediary steps of an entire "batteria" of eight wooden casks used in TBV manufacturing, followed by data normalization, agglomerative hierarchical cluster analysis, and visualization by use of sensomics heatmapping gave an molecular insight into the alterations occurring during TBV maturation. The sweetness modulating 5-acetoxymethyl-2-furaldehyde was found to be generated with increasing degree of maturation, most likely by esterification of the Maillard reaction product 5-hydroxymethyl-2-furaldehyde with acetic acid.

[19] Driven by the need to discover the key players imparting the typical taste of foods, the research area "sensomics" has made tremendous efforts in recent years in order to map the comprehensive population of sensory active, low-molecular weight compounds, coined sensometabolome, and to catalog, quantify, and evaluate the sensory activity of metabolites which are present in raw materials and/or are produced upon food processing and storage, respectively (22, 24). Aimed at decoding the typical taste signature of food products on a molecular level, the so-called taste dilution analysis (TDA) was developed as an efficient screening tool enabling the sensory-directed identification of taste-active nonvolatiles in foods (25). This allowed the determination of taste active lead molecules in foods such as black tea infusions (26), roasted cocoa (27), double boiled chicken broth (28), dried morel mushrooms (29) as well as red wine (22) or oak matured spirits (23) and, in addition, the identification of various taste modulating compounds such as γ-glutamyl peptides in cheese and beans (30, 31), N-(l-methyl-4-oxoimidazolidin-2-ylidene)-a-amino acids in stewed beef (32), and N2-(l-carboxyethyl)guanosine 5 '-monophosphate in yeast extract (33), respectively.

Addition of 5-acetoxymethyl-2-furaldehyde to Reduced-Calorie Products

[20] A reduced-calorie product is a product wherein the calorie content has been reduced by at least 25% compared to the full-calorie counterpart. The reduced calorie product is produced, for example, by reducing the amount of nutritive sweetener in the product in order to reduce the calorie content. However, reducing the amount of nutritive sweetener also reduces the sweetness of the reduced-calorie product compared to the full-calorie counterpart. It was discovered that 5-acetoxymethyl-2- furaldehyde can be added to reduced-calorie products to enhance the sweet taste of the product thus offsetting the effects of reducing the amount of nutritive sweetener.

Concentrations

[21] 5-acetoxymethyl-2-furaldehyde is included in a reduced calorie product in an amount sufficient to enhance the sweet taste of the product containing a reduced amount of sweetener compared to compositions that are prepared without 5-acetoxymethyl-2- furaldehyde, as judged by human beings or animals or, in the case of formulations testing, as judged by a majority of a panel of, e.g. , eight human taste testers, via procedures commonly known in the field.

[22] The concentration of 5-acetoxymethyl-2-furaldehyde sufficient to enhance sweetness in a reduced calorie product will of course depend on many variables, including the specific type of reduced calorie product and its various other ingredients, as well as natural genetic variability and individual preferences and health conditions of various individuals tasting the reduced calorie product. [23] In some embodiments, the concentration of 5-acetoxymethyl-2-furaldehyde sufficient to enhance sweetness in a reduced calorie beverage product is from about 0.1 to 5.0 mmol/L (e.g. , 0.1 , 0.15, 0.2, 0.25, 0.3, 0.35, 0.4, 0.45, 0.5, 0.75, 1.0, 1.5, 2.0, 3.0, 4.0, or 5.0). For example, in a 4% (=1 16.8 mmol/L) sucrose solution the ratio of nutritional sweetener to 5-acetoxymethyl-2-furaldehyde on a parts basis is 1 ,170: 1 to 23.4: 1. Similar ratios of nutritional sweetener to 5-acetoxymethyl-2-furaldehyde can also be used in reduced calorie foods, such as those described below.

Reduced calorie products

[24] Reduced calorie products containing 5-acetoxymethyl-2-furaldehyde can be

comestible or noncomestible. A "comestible reduced calorie product" is a reduced calorie product that can be consumed as food by humans or animals. A

"noncomestible reduced calorie product" is a reduced calorie product that is intended to be consumed or used by humans or animals not as food. Both comestible and noncomestible reduced calorie products include solids, gels, pastes, foamy materials, semi-solids, liquids, and mixtures thereof. "Animals" include any non-human animal, such as farm animals and pets.

[25] It should be understood that reduced calorie products in accordance with this

disclosure may have any of numerous different specific formulations or constitutions. For example, the formulation of a reduced calorie product in accordance with this disclosure can vary to a certain extent, depending upon such factors as the reduced calorie product's intended market segment, its desired nutritional characteristics, flavor profile and the like. For example, it will generally be an option to add further ingredients to the formulation of some embodiments, including any of the reduced calorie product formulations described below. Additional (i.e. , more and/or other) sweeteners may be added, flavorings, electrolytes, vitamins, fruit juices or other fruit reduced calorie products, tastants, masking agents and the like, flavor enhancers, and/or carbonation typically can be added to any such formulations to vary the taste, mouthfeel, nutritional characteristics, etc. Sweeteners

[26) As discussed above, a nutritive sweetener is present in the reduced calorie products.

In some embodiments of the food and beverage reduced calorie products disclosed herein, additional sweeteners may be included, for example, additional nutritive sweeteners or non-nutritive sweeteners.

[27] Non-limiting examples of nutritive sweeteners include sucrose, liquid sucrose,

fructose, liquid fructose, glucose, liquid glucose, glucose-fructose syrup from natural sources such as apple, chicory, agave, honey, etc., e.g. , high fructose corn syrup, chicory syrup, Agave syrup, invert sugar, medium invert sugar, maple syrup, maple sugar, honey, brown sugar molasses, e.g., cane molasses and sugar beet molasses, sorghum syrup, an mixtures of any of them. Such sweeteners are present in some embodiments in an amount of from about 0.1% to about 20% by weight of the finished food or beverage reduced calorie product, such as from about 6% to about 16% by weight, depending upon the desired level of sweetness for the finished food or beverage reduced calorie product. Nutritive sweeteners may be present in beverage concentrate embodiments up to about 60% by weight of the beverage concentrate.

[28] Non-limiting examples of potent non-nutritive natural sweeteners that may be

included in food and beverage reduced calorie products include rebaudioside A, rebaudioside D, stevioside, other steviol glycosides, Stevia rebaudiana extracts, Lo Han Guo (LHG), e.g. , LHG juice concentrate or LHG powder, thaumatin, monellin, brazzein, monatin, and mixtures of any of them. LHG, if used, may have for example, mogroside V content of from about 2 to about 99%.

[29] Other examples of non-nutritive sweeteners include sorbitol, mannitol, xylitol,

glycyrrhizin, maltitol, maltose, lactose, xylose, arabinose, isomalt, lactitol, trehalulose, ribose, fructo-oligosaccharides, and mixtures of any of them. Optionally, the sweetener component can include erythritol, tagatose, an erythritol and tagatose blend, or polydextrose.

[30] Non-limiting examples of potent non-nutritive artificial sweeteners that may be

included in food and beverage reduced calorie products include peptide based sweeteners, e.g. , aspartame, neotame, and alitame, and non-peptide based sweeteners, e.g., sodium saccharin, calcium saccharin, acesulfame (including but not limited to acesulfame potassium), cyclamate (including but not limited to sodium cyclamate and/or calcium cyclamate), neohesperidin dihydrochalcone, sucralose, and mixtures of any of them. Potent non-nutritive sweeteners typically are employed at a level of milligrams per ounce of food or beverage according to their sweetening power, any applicable regulatory provisions of the country where the food or beverage is to be marketed, the desired level of sweetness of the food or beverage, etc. Mixtures of any of the above nutritive and non-nutritive sweeteners are included within the scope of the disclosed invention. It will be within the ability of those skilled in the art, given the benefit of this disclosure, to select suitable additional or alternative sweeteners for use in various embodiments of the food and beverage reduced calorie products disclosed here.

Reduced calorie beverage products

[31] In some embodiments, beverage concentrates are prepared with an initial volume of water to which the additional ingredients are added. Full strength beverage compositions can be formed from the beverage concentrate by adding further volumes of water to the concentrate. Typically, for example, full strength beverages can be prepared from the concentrates by combining approximately 1 part concentrate with between approximately 3 to approximately 7 parts water. In some embodiments, the full strength beverage is prepared by combining 1 part concentrate with 5 parts water. In some embodiments, the additional water used to form the full strength beverages is carbonated water. In certain other embodiments, a full strength beverage is directly prepared without the formation of a concentrate and subsequent dilution.

[32] Water is a basic ingredient in the beverage reduced calorie products disclosed here, typically being the vehicle or primary liquid portion in which the remaining ingredients are dissolved, emulsified, suspended or dispersed. Purified water can be used in the manufacture of certain embodiments of the beverages disclosed here, and water of a standard beverage quality can be employed in order not to adversely affect beverage taste, odor, or appearance. The water typically will be clear, colorless, free from objectionable minerals, tastes and odors, free from organic matter, low in alkalinity and of acceptable microbiological quality based on industry and

government standards applicable at the time of producing the beverage. In certain typical embodiments, water is present at a level of from about 80% to about 99.9% by weight of the beverage. In some embodiments, the water used in beverages and concentrates disclosed here is "treated water," which refers to water that has been treated to reduce the total dissolved solids of the water prior to optional

supplementation, e.g., with calcium as disclosed in U.S. patent no. 7,052,725.

Methods of producing treated water are known to those of ordinary skill in the art and include deionization, distillation, filtration and reverse osmosis ("r-o"), among others. The terms "treated water," "purified water," "demineralized water," "distilled water," and "r-o water" are understood to be generally synonymous in this discussion, referring to water from which substantially all mineral content has been removed, typically containing no more than about 500 ppm total dissolved solids, e.g. 250 ppm total dissolved solids.

[33] Juices suitable for use in some embodiments of the beverage reduced calorie products disclosed here include, e.g., fruit, vegetable and berry juices. Juices can be employed in the present invention in the form of a single-strength juice, NFC juice, 100% pure juice, juice concentrate, juice puree, or other suitable forms. The term "juice" as used here includes single-strength fruit, berry, or vegetable juice, as well as concentrates, purees, milks, and other forms. Multiple different fruit, vegetable and/or berry juices can be combined, optionally along with other flavorings, to generate a beverage having the desired flavor.

[34] Examples of suitable juice sources include orange, lemon, lime, tangerine, mandarin orange, tangelo, pomelo, grapefruit, grape, red grape, sweet potato, tomato, celery, beet, lettuce, spinach, cabbage, watercress, rhubarb, carrot, cucumber, raisin, cranberry, pineapple, peach, banana, apple, pear, guava, apricot, watermelon, Saskatoon berry, blueberry, plains berry, prairie berry, mulberry, elderberry, Barbados cherry (acerola cherry), choke cherry, date, coconut, olive, raspberry, strawberry, huckleberry, loganberry, currant, dewberry, boysenberry, kiwi, cherry, blackberry, quince, buckthorn, passion fruit, sloe, rowan, gooseberry, pomegranate, persimmon, mango, rhubarb, papaya, lychee, plum, prune, date, currant, fig, etc. Numerous additional and alternative juices suitable for use in some embodiments will be apparent to those skilled in the art given the benefit of this disclosure.

[35] Non-mineral nutritive compounds such as vitamins can be added to the beverage reduced calorie products. Examples of non-mineral nutritional supplement ingredients are known to those of ordinary skill in the art and include, for example, antioxidants and vitamins, including Vitamins A, D, E (tocopherol), C (ascorbic acid), Bj (thiamine), B ₂ (riboflavin), B ₆ , B 12, and K, niacin, folic acid, biotin, and combinations thereof. The optional non-mineral nutritional supplements are typically present in amounts generally accepted under good manufacturing practices. In some

embodiments, amounts are between about 1% and about 100% RDV, where such RDV are established. In some embodiments, the non-mineral nutritional supplement ingredient(s) are present in an amount of from about 5% to about 20% RDV, where established.

Acid used in beverage reduced calorie products disclosed here can serve any one or more of several functions, including, for example, providing antioxidant activity, lending tartness to the taste of the beverage, enhancing palatability, increasing thirst quenching effect, modifying sweetness and acting as a mild preservative by providing microbiological stability. Any suitable edible acidulant may be used, for example citric acid, malic acid, tartaric acid, phosphoric acid, ascorbic acid, lactic acid, formic acid, fumaric acid, gluconic acid, succinic acid, maleic acid, sodium acid sulfate and/or adipic acid. The acid can be used in solution form, for example, and in an amount sufficient to provide the desired pH of the beverage. Typically, for example, the one or more acids of the acidulant are used in amount, collectively, of from about 0.01% to about 1.0% by weight of the beverage, e.g., from about 0.05% to about 0.5% by weight of the beverage, such as 0.1% to 0.25% by weight of the beverage, depending upon the acidulant used, desired pH, other ingredients used, etc.

As used herein, a "full-calorie" beverage formulation is one fully sweetened with a nutritive sweetener. The term "nutritive sweetener" refers generally to sweeteners which provide significant caloric content in typical usage amounts, e.g., more than about 5 calories per 8 oz. serving of beverage. As used herein, a "potent sweetener" means a sweetener which is at least twice as sweet as sugar, that is, a sweetener which on a weight basis requires no more than half the weight of sugar to achieve an equivalent sweetness. For example, a potent sweetener may require less than one-half the weight of sugar to achieve an equivalent sweetness in a beverage sweetened to a level of 10 degrees Brix with sugar. Potent sweeteners include both nutritive (e.g., Lo Han Guo juice concentrate) and non-nutritive sweeteners (e.g., typically, Lo Han Guo powder). In addition, potent sweeteners include both natural potent sweeteners (e.g., steviol glycosides, Lo Han Guo, etc.) and artificial potent sweeteners (e.g., neotame, etc.). However, for natural beverage reduced calorie products disclosed here, only natural potent sweeteners are employed. Commonly accepted potency figures for certain potent sweeteners include, for example,

Cyclamate 30 times as sweet as sugar

Stevioside 100-250 times as sweet as sugar

Acesulfame-K 200 times as sweet as sugar

Mogroside V 100-300 times as sweet as sugar

Rebaudioside A 150-300 times as sweet as sugar

Aspertame 200 times as sweet as sugar

Saccharine 300 times as sweet as sugar

Neohesperidin dihydrochalcone 300 times as sweet as sugar

Sucralose 600 times as sweet as sugar

Neotame 8,000 times as sweet as sugar

[38] As used herein, a "non-nutritive sweetener" is one which does not provide significant caloric content in typical usage amounts, i.e., is one which imparts less than 5 calories per 8 oz. serving of beverage to achieve the sweetness equivalent of 10 Brix of sugar. As discussed above and used herein, "reduced calorie beverage" means a beverage having at least a 25% reduction in calories per 8 oz. serving of beverage as compared to the full calorie version, typically a previously commercialized full-calorie version. As used herein, a "low-calorie beverage" has fewer than 40 calories per 8 oz. serving of beverage. As used herein, "zero-calorie" or "diet" means having less than 5 calories per serving, e.g., per 8 oz. for beverages.

[39] Natural embodiments of the beverage reduced calorie products disclosed here are natural in that they do not contain anything artificial or synthetic (including any color additives regardless of source) that would not normally be expected to be in the food. As used herein, therefore, a "natural" beverage composition is defined in accordance with the following guidelines: Raw materials for a natural ingredient exists or originates in nature. Biological synthesis involving fermentation and enzymes can be employed, but synthesis with chemical reagents is not utilized. Artificial colors, preservatives, and flavors are not considered natural ingredients. Ingredients may be processed or purified through certain specified techniques including at least: physical processes, fermentation, and enzymolysis. Appropriate processes and purification techniques include at least: absorption, adsorption, agglomeration, centrifugation, chopping, cooking (baking, frying, boiling, roasting), cooling, cutting,

chromatography, coating, crystallization, digestion, drying (spray, freeze drying, vacuum), evaporation, distillation, electrophoresis, emulsification, encapsulation, extraction, extrusion, filtration, fermentation, grinding, infusion, maceration, microbiological (rennet, enzymes), mixing, peeling, percolation,

refrigeration/freezing, squeezing, steeping, washing, heating, mixing, ion exchange, lyophilization, osmose, precipitation, salting out, sublimation, ultrasonic treatment, concentration, flocculation, homogenization, reconstitution, enzymolysis (using enzymes found in nature). Processing aids (currently defined as substances used as manufacturing aids to enhance the appeal or utility of a food component, including clarifying agents, catalysts, flocculants, filter aids, and crystallization inhibitors, etc. See 21 CFR § 170.3(o)(24)) are considered incidental additives and may be used if removed appropriately.

Sweeteners suitable for use in various embodiments of the beverages disclosed herein include nutritive and non-nutritive, natural and artificial or synthetic sweeteners. Suitable sweeteners and combinations of sweeteners are selected for the desired nutritional characteristics, functional characteristics, taste profile for the beverage, mouthfeel and other organoleptic factors. Non-nutritive artificial sweeteners suitable in some embodiments include, for example, peptide based sweeteners, e.g., aspartame, neotame, and alitame, and non-peptide based sweeteners, for example, sodium saccharin, calcium saccharin, acesulfame (including but not limited to acesulfame potassium), cyclamate (including but not limited to sodium cyclamate and/or calcium cyclamate), neohesperidin dihydrochalcone, and sucralose. Alitame may be less desirable for caramel-containing beverages where it has been known to form a precipitate. In some embodiments, the beverage reduced calorie product employs aspartame as the sweetener, either alone or with other sweeteners. In some embodiments, the sweetener comprises aspartame and acesulfame potassium. Other non-nutritive sweeteners suitable in some embodiments include, for example, sorbitol, mannitol, xylitol, glycyrrhizin, neohesperidin dihydrochalcone, D-tagatose, erythritol, meso-erythritol, malitol, maltose, lactose, fructo-oligosaccharides, Lo Han Guo powder, steviol glycosides, e.g., rebaudiosides such as Rebaudioside A, stevioside, etc., xylose, arabinose, isomalt, lactitol, maltitol, trehalulose, and ribose, and protein sweeteners such as monatin, thaumatin, monellin, brazzein, L-alanine and glycine related compounds and mixtures of any of them. Lo Han Guo, steviol glycosides, e.g., rebaudiosides such as Rebaudioside A, stevioside, etc. and related compounds, as discussed further below, are natural non-nutritive potent sweeteners. It will be within the ability of those skilled in the art, given the benefit of this disclosure, to select suitable non-nutritive sweeteners (e.g., one or combination of non-nutritive sweeteners, either alone or together with nutritive sweetener) for a particular embodiment of the beverage reduced calorie products disclosed here.

In some embodiments, the sweetener component can include nutritive, natural crystalline or liquid sweeteners such as sucrose, liquid sucrose, fructose, liquid fructose, glucose, liquid glucose, leucrose, trehalose, glactose, isomaltulose, dextrose, maltodextrin, corn syrup solids, glucooligosaccharides, glucose-fructose syrup from natural sources such as apple, chicory, honey, etc., e.g., high fructose corn syrup, invert sugar, maple syrup, maple sugar, honey, brown sugar molasses, e.g., cane molasses, such as first molasses, second molasses, blackstrap molasses, and sugar beet molasses, sorghum syrup and/or others. Such sweeteners are present in some embodiments in an amount of from about 0.1% to about 20% by weight of the beverage {e.g. , from about 1 % to about 4%, 0.1% to about 3%, 2% to about 10% by weight), depending upon the desired level of sweetness for the beverage. To achieve desired beverage uniformity, texture and taste, In some embodiments, of the natural beverage reduced calorie products disclosed here, standardized liquid sugars as are commonly employed in the beverage industry can be used. Typically such

standardized sweeteners are free of traces of nonsugar solids which could adversely affect the flavor, color or consistency of the beverage. [42] The sweeteners are edible consumables suitable for consumption and for use in beverages. By "edible consumables" is meant a food or beverage or an ingredient of a food or beverage for human or animal consumption. The sweetener or sweetening agent used here and in the claims can be a nutritive or non-nutritive, natural or synthetic beverage ingredient or additive (or mixtures of them) which provides sweetness to the beverage, i.e., which is perceived as sweet by the sense of taste. The perception of flavoring agents and sweetening agents may depend to some extent on the interrelation of elements. Flavor and sweetness may also be perceived separately, i.e., flavor and sweetness perception may be both dependent upon each other and independent of each other. For example, when a large amount of a flavoring agent is used, a small amount of a sweetening agent may be readily perceptible and vice versa. Thus, the oral and olfactory interaction between a flavoring agent and a sweetening agent may involve the interrelationship of elements.

(43) Non-nutritive, high potency sweeteners typically are employed at a level of

milligrams per fluid ounce of beverage, according to their sweetening power, any applicable regulatory provisions of the country where the beverage is to be marketed, the desired level of sweetness of the beverage, etc. It will be within the ability of those skilled in the art, given the benefit of this disclosure, to select suitable additional or alternative sweeteners for use in various embodiments of the beverage reduced calorie products disclosed here.

[44] As mentioned above, some embodiments of the beverages disclosed here employ steviol glycosides, e.g., rebaudiosides such as Rebaudioside A, stevioside, etc. or related compounds or mixtures of any of them for sweetening. These compounds can be obtained by extraction or the like from the stevia plant. Stevia (e.g., Stevia rebaudiana bectoni) is a sweet-tasting plant. The leaves contain a complex mixture of natural sweet diterpene glycosides. Steviol glycosides, e.g., rebaudiosides such as Rebaudioside A, stevioside, etc. are components of Stevia that contribute sweetness. Typically, these compounds are found to include stevioside (4-13% dry weight), steviolbioside (trace), the rebaudiosides, including rebaudioside A (2-4%), rebaudioside B (trace), rebaudioside C ( 1 -2%), rebaudioside D (trace), and

rebaudioside E (trace), and dulcoside A (0.4-0.7%). The following nonsweet constituents also have been identified in the leaves of stevia plants: labdane, diterpene, triterpenes, sterols, flavonoids, volatile oil constituents, pigments, gums and inorganic matter. In at least certain embodiments of the beverage reduced calorie products disclosed herein, non-nutritive sweeteners steviol glycosides, e.g., rebaudiosides such as Rebaudioside A, stevioside, etc. may be included in ready to drink beverage compositions at a weight percent of about 0.1 % to about 10.0%, and preferably between about 0.2% and about 0.75%.

[45] The sweetener Lo Han Guo, which has various different spellings and pronunciations and is abbreviated here in some instances as LHG, can be obtained from fruit of the plant family Cucurbitaceae, tribe Jollifieae, subtribe Thladianthinae, genus Siraitia. LHG often is obtained from the genus/species S. grosvenorii, S. siamensis, S.

silomaradjae, S. sikkimensis, S. africana, S. borneensis, and S. taiwaniana. Suitable fruit includes that of the genus/species S. grosvenorii, which is often called Lo Han Guo fruit. LHG contains triterpene glycosides or mogrosides, which constituents may be used as LHG sweeteners. Lo Han Guo is a potent sweetener which can be provided as a natural nutritive or natural non-nutritive sweetener. For example, Lo Han Guo juice concentrate may be a nutritive sweetener, and Lo Han Guo powder may be a non-nutritive sweetener. Lo Han Guo can be used as the juice or juice concentrate, powder, etc. Preferably LHG juice contains at least about 0.1 %, e.g., from 0.1% to about 15%. mogrosides, preferably mogroside V, mogroside IV, (1 1-oxo-mogroside V), siamenoside and mixtures thereof. LHG can be produced, for example, as discussed in U.S. patent No. 5,41 1 ,755. Sweeteners from other fruits, vegetables or plants also may be used as natural or processed sweeteners or sweetness enhancers in some embodiments of the beverages disclosed here.

[46] Preservatives may be used in at least certain embodiments of the beverages disclosed here. That is, some embodiments contain an optional dissolved preservative system. Solutions with a pH below 4 and especially those below 3 typically are "microstable," i.e., they resist growth of microorganisms, and so are suitable for longer term storage prior to consumption without the need for further preservatives. However, an additional preservative system can be used if desired. If a preservative system is used, it can be added to the beverage reduced calorie product at any suitable time during reduced calorie production, e.g., in some cases prior to the addition of the sweetener. As used here, the terms "preservation system" or "preservatives" include all suitable preservatives approved for use in food and beverage compositions, including, without limitation, such known chemical preservatives as benzoic acid, benzoates, e.g., sodium, calcium, and potassium benzoate, sorbates, e.g., sodium, calcium, and potassium sorbate, citrates, e.g., sodium citrate and potassium citrate, polyphosphates, e.g., sodium hexametaphosphate (SHMP), dimethyl dicarbonate, and mixtures thereof, and antioxidants such as ascorbic acid, EDTA, BHA, BHT, TBHQ, EMIQ, dehydroacetic acid, ethoxyquin, heptylparaben, and combinations thereof.

[47] Preservatives can be used in amounts not exceeding mandated maximum levels under applicable laws and regulations. The level of preservative used typically is adjusted according to the planned final reduced calorie product pH, as well as an evaluation of the microbiological spoilage potential of the particular beverage formulation. The maximum level employed typically is about 0.05% by weight of the beverage. It will be within the ability of those skilled in the art, given the benefit of this disclosure, to select a suitable preservative or combination of preservatives for beverages according to this disclosure. In certain embodiments of the invention, benzoic acid or its salts (benzoates) may be employed as preservatives in the beverage reduced calorie products.

[48] Other methods of beverage preservation suitable for some embodiments of the

beverage reduced calorie products disclosed here include, e.g., aseptic packaging and/or heat treatment or thermal processing steps, such as hot filling and tunnel pasteurization. Such steps can be used to reduce yeast, mold and microbial growth in the beverage reduced calorie products. For example, U.S. Patent No. 4,830,862 to Braun et al. discloses the use of pasteurization in the reduced calorie production of fruit juice beverages as well as the use of suitable preservatives in carbonated beverages. U.S. Patent No. 4,925,686 to astin discloses a heat-pasteurized freezable fruit juice composition which contains sodium benzoate and potassium sorbate. In general, heat treatment includes hot fill methods typically using high temperatures for a short time, e.g., about 190° F for 10 seconds, tunnel pasteurization methods typically using lower temperatures for a longer time, e.g., about 160° F for 10-15 minutes, and retort methods typically using, e.g., about 250° F for 3-5 minutes at elevated pressure, i.e., at pressure above 1 atmosphere. [49] The beverage reduced calorie products disclosed here optionally contain a flavor composition, for example, natural and synthetic fruit flavors, botanical flavors, other flavors, and mixtures thereof. As used here, the term "fruit flavor" refers generally to those flavors derived from the edible productive part of a seed plant. Included are both those wherein a sweet pulp is associated with the seed, e.g., banana, tomato, cranberry and the like, and those having a small, fleshy berry. The term berry also is used here to include aggregate fruits, i.e., not "true" berries, but that are commonly accepted as a berry. Also included within the term "fruit flavor" are synthetically prepared flavors made to simulate fruit flavors derived from natural sources.

Examples of suitable fruit or berry sources include whole berries or portions thereof, berry juice, berry juice concentrates, berry purees and blends thereof, dried berry powders, dried berry juice powders, and the like.

[50] Examples of fruit flavors include the citrus flavors, e.g., orange, lemon, lime and

grapefruit, and such flavors as apple, grape, cherry, and pineapple flavors and the like, and mixtures thereof. In some embodiments, the beverage concentrates and beverages comprise a fruit flavor component, e.g., a juice concentrate or juice. As used here, the term "botanical flavor" refers to flavors derived from parts of a plant other than the fruit. As such, botanical flavors can include those flavors derived from essential oils and extracts of nuts, bark, roots and leaves. Also included within the term "botanical flavor" are synthetically prepared flavors made to simulate botanical flavors derived from natural sources. Examples of such flavors include cola flavors, tea flavors, and the like, and mixtures thereof. The flavor component can further comprise a blend of various of the above-mentioned flavors. The particular amount of the flavor component useful for imparting flavor characteristics to the beverages of the present invention will depend upon the flavor(s) selected, the flavor impression desired, and the form of the flavor component. Those skilled in the art, given the benefit of this disclosure, will be readily able to determine the amount of any particular flavor component(s) used to achieve the desired flavor impression.

[51] Other flavorings suitable for use in some embodiments of the beverage reduced

calorie products disclosed here include, e.g., spice flavorings, such as cassia, clove, cinnamon, pepper, ginger, vanilla spice flavorings, cardamom, coriander, root beer, sassafras, ginseng, and others. Numerous additional and alternative flavorings suitable for use in some embodiments will be apparent to those skilled in the art given the benefit of this disclosure. Flavorings can be in the form of an extract, oleoresin, juice concentrate, bottler's base, or other forms known in the art. In some embodiments, such spice or other flavors complement that of a juice or juice combination.

[52] The one or more flavorings can be used in the form of an emulsion. A flavoring

emulsion can be prepared by mixing some or all of the flavorings together, optionally together with other ingredients of the beverage, and an emulsifying agent. The emulsifying agent may be added with or after the flavorings mixed together. In some embodiments, the emulsifying agent is water-soluble. Examples of suitable emulsifying agents include gum acacia, modified starch, carboxymethylcellulose, gum tragacanth, gum ghatti and other suitable gums. Additional suitable emulsifying agents will be apparent to those skilled in the art of beverage formulations, given the benefit of this disclosure. The emulsifier in some embodiments comprises greater than about 3% of the mixture of flavorings and emulsifier. In some embodiments, the emulsifier is from about 5% to about 30% of the mixture.

[53] The beverage concentrates and beverages disclosed here may contain additional ingredients, including, generally, any of those typically found in beverage

formulations. These additional ingredients, for example, can typically be added to a stabilized beverage concentrate. Examples of such additional ingredients include, but are not limited to, caffeine, caramel and other coloring agents or dyes, antifoaming agents, gums, emulsifiers, tea solids and cloud components.

Reduced calorie food products

[54] Reduced calorie food products comprise at least one food component, i.e. , any edible material suitable for human or animal consumption, whether or not fully or partially digestible). Non-limiting examples of food components include proteins,

carbohydrates, fats, vitamins, minerals, etc. Food reduced calorie products include, e.g. , oatmeal, cereal, baked goods {e.g. , cookies, crackers, cakes, brownies, breads, etc.); snack foods {e.g. , potato chips, tortilla chips, popcorn, snack bars, rice cakes, etc.) and other grain-based food reduced calorie products; sauces; candies, confections; desserts; coatings, frostings, or glazes; dairy products {e.g. , milk, yoghurt and sour milk drinks, ice cream); spreads (e.g. , jams and preserves, honey, chocolate spreads).

[55] In some embodiments, reduced calorie food products include, e.g. , oatmeal, cereal, baked goods (e.g. , cookies, crackers, cakes, brownies, breads, etc.), snack foods {e.g. , potato chips, tortilla chips, popcorn, snack bars, rice cakes, etc.), and other grain- based food products.

[56] Nothing in this specification should be considered as limiting the scope of this

disclosure. All examples presented are representative and non-limiting. The above- described embodiments can be modified or varied, as appreciated by those skilled in the art in light of the above teachings. It is therefore to be understood that, within the scope of the claims and their equivalents, the embodiments disclosed herein can be practiced otherwise than as specifically described.

EXAMPLE 1

[57] Chemicals. All chemicals used were purchased from Sigma-Aldrich (Taufkirchen,

Germany) and Fluka (Neu-Ulm, Germany), respectively. Isotopic labeled amino acids were purchased from Cambridge Isotope Laboratories (Andover, MA, USA). (+)- Dihydrorobinetin (3) was purchased from Extrasynthese (Genay Cedex, France). For sensory analysis, bottled water (Evian) was adjusted to pH 3.0 with hydrochloric acid (0.1 M). Solvents were of HPLC grade (Mallinckrodt Baker, Griesheim, Germany). Ultrapure water used for chromatography was purified by means of a MilliQ-water Gradient A 10 system (Millipore, Schwalbach, Germany) and deuterated solvents for NMR spectroscopy were supplied by Euriso-Top (Gif-Sur-Yvette, France). Reference samples of vescalagin (1 ) and castalagin (2) were isolated and purified from Quercus alba L. following the procedure reported earlier (34).

[58] Vinegar samples. Samples of traditional balsamic vinegar of Modena (TBV;

"affinato ^' " quality, 12 years) and eight intermediary samples collected from the batteria" were obtained from a local producer in the region of Modena, Italy. The battericT consisted of a sequence of eight casks differing in wood variety and volumes as given in parenthesis: barrel A (acacia, 50 L), B (chestnut, 40 L), C (cherry, 30 L), D (mulberry, 23 L), E (oak, 13 L), F (chestnut, 10 L), G (chestnut, 5 L), and H (chestnut, 5 L). Dry mass (35) and pH value of intermediary vinegar samples A (43.91%, pH 2.65), B (55.23%, pH 2.62), C (63.75%, pH 2.58), D

(70.75%, PH 2.52), E (70.95%, pH 2.52), F (72.76%, pH 2.48), G (73.98%, pH 2.49), and H (74.45%, pH 2.49) were determined as given in parenthesis. In addition, balsamic vinegar of Modena (BV) was obtained from another local producer in the region of Modena, Italy. All samples were kept at 4°C (40°F) in the dark until used for analysis.

Ultrafiltration. An aliquot (10 mL) of the TBV was diluted with deionized water (200 mL) and placed into a Vivacell 250 static gas pressure filtration system (Vivascience, Hannover, Germany), which was equipped with a 5 kDa molecular weight cutoff 5000 MWCO PES membrane (Vivascience, Hannover, Germany) preconditioned by rinsing the membrane with deionized water (300 mL) twice. After sealing and pressurizing (4 bar) the Vivacell 250 cell using an air pressure controller, the cell was placed on a 3005-type laboratory shaker (GFL, Burgwedel, Germany) operating at 200 rpm at 20°C. After the first filtration, the retentate was suspended with deionized water (200 mL) and filtered again under pressure (4 bar). This procedure was repeated twice, the retentate was then taken up in deionized water (50 mL), and filtrate and retentate were lyophilized to afford the low (< 5kDa; TBV-LMW) and the high molecular weight fraction (> 5kDa; TBV-HMW) in yields of 656.3 and 8.7 g/L, respectively. The isolates obtained for both fractions in various separations were pooled accordingly and, then, stored at -20 °C (-4°F) in the dark until further analysis.

Gel Adsorption Chromatography (GAC). A defined amount (1.0 g) of the lyophilized TBV-LMW fraction was dissolved in MeOH/water (20/80, v/v; 10 mL) and transferred onto the top of a XK 50/100 glass column (Amersham Pharmacia Biotech, Uppsala, Sweden) filled with a slurry of Sephadex LH 20 (GE Healthcare, Munich, Germany), which was conditioned with MeOH/water (20/80, v/v; adjusted to pH 4.0 with a 1% aqueous formic acid). Chromatography was carried out at a flow rate of 1.3 mL/min by eluting the column sequentially with aliquots (400 mL, each) of

MeOH/water containing 20, 40, or 80% methanol, respectively, followed by pure methanol (400 mL). Monitoring the effluent at 220 nm by means of an UV-2575-type UV-Vis detector (Jasco, GroB-Umstadt, Germany), the individual GAC fractions were collected every 10 min by means of a LKB Bromma 7000 Ultrorac fraction collector and combined to give a total of ten fractions, namely I-X. The fractions obtained were separated from solvent in vacuum, freeze-dried and stored at -20 °C (-4°F) until used for further analysis.

[61] Identification of Taste-Active Compounds in GAC Fraction VII. An aliquot (100 mg) of the GAC fraction VII was dissolved in acetonitrile/water (50/50, v/v; 5 mL) and, after membrane filtration, was analyzed by semi-preparative hydrophilic interaction liquid chromatography (HILIC) on a 300 x 21.5 mm i.d., 10 μηι, TSKgel Amide-80 column (Tosoh Bioscience, Stuttgart, Germany) equipped with a 75 x 21.5 mm i.d., 10 μιη, guard column of the same type (Tosoh Bioscience). Using a flow rate of 6.0 mL/min, chromatography was performed using 1 % aqueous acetic acid as solvent A, and acetonitrile containing 1% acetic acid as solvent B. Starting with 5% A and increasing A to 100% within 20 min, the effluent was monitored by means of an evaporative light scattering detector (ELSD) and fractionated into five subfractions, namely VII- 1 to VII-5, which separated from solvent under vacuum and then lyophilized twice. Comparison of chromatographic (HILIC) and mass spectrometry data, followed as well as co-chromatography with the corresponding reference compounds led to the identification of 1 -O-acetyl-B-D-fructopyranose in fraction VII- 2, 6-O-acetyl- - and 6-O-acetyl-B-D-glucose in VII-3 and VII-4, and D-glucose in fraction VII-5, respectively.

[62] Identification of Taste Modulating Compounds in GAC Fraction X. An aliquot (50 mg) of GAC fraction X (Figure 3) was dissolved in MeOH/water (20/80, v/v; 1 mL) and, after membrane filtration, an aliquot (200 μί) was injected into the preparative HPLC system connected to a 250 x 21.2 mm i.d., 5 μηι, Microsorb RP18 column (Varian, Darmstadt, Germany). Using a flow rate of 18 mL/min, chromatography was performed using 0.1 % aqueous formic acid (solvent A), and 0.1% formic acid in methanol (solvent B). Starting with 0% solvent A and increasing A to 100% within 20 min, the effluent was monitored at 280 nm and separated into two subfractions, namely X-l and X-2, which were freed from solvent under vacuum and lyophilized. LC-MS/MS and 1 D/2D-NMR studies led to the identification of 5-hydroxymethyl-2- furaldehyde and 5-acetoxymethyl-2-furaldehyde (6). Comparison of chromatographic (RP-18) and spectroscopic data (UV-Vis, LC-MS/MS, and NMR) with those of the corresponding reference compound confirmed the identity of these molecules. [63] 5-Hydroxymethyl-2-furaldehyde: UV-Vis (MeOH): Imax = 280 nm; LC-MS (ESI+): m z 127 (70%, [M+H]+), 148 (100%, [M+Na]+). 1 H and 13C NMR data were identical with those measured for the commercially available reference compound.

[64] 5-Acetoxymethyl-2-furaldehyde (6): UV-Vis (MeOH), X ax = 280 nm; LC-MS

(ESI+): m/z 169 (70%, [M+H]+), 191 (100%, [M+Na]+); 1 H-NMR (400 MHz, MeOD, COSY): δ 2.20 [s, 3H, H-C(l)], 5.17 [s, 2H, H-C(3)], 6.71 [d, 1 H, J = 4.0 Hz, H-C(5)], 7.39 [d, 1H, J = 4.0 Hz, H-C(6)], 9.59 [s, 1H, H-C(8)]; 13C-NMR (100 MHz, MeOD, HSQC, HMBC, DEPT): δ 19.1 [C(l)], 57.4 [C(3)], 1 12.2 [C(5)], 122.5 [C(6)], 153.0 [C(7)], 156.0 [C(4)], 170.5 [C(2)], 178.2 [C(8)].

[65] Synthesis of 6-O-Acetyl-a/B-D-glucopyranose (4), l-O-Acetyl-B-D-fructopyranose (5), and 6-0-Acetyl-a/B-[ 13C6]-D-glucopyranose ([13C6J-1 ). D-Glucose or D- fructose (3 mmol), respectively, was dissolved in dry pyridine/THF (4/1 , v/v; 5 mL) and, after addition of acetic anhydride (1.5 mmol), the solution was stirred at 0°C. After 30 min, deionized water (1 mL) was added, pyridine was removed in vacuum, the residue was diluted with water (50 mL), freeze-dried, and, then, separated by means of preparative HPLC on a 250 x 21.2 mm i.d., 5 μπι, Microsorb RP18 column (Varian, Darmstadt, Germany). Using 1 % aqueous formic acid as the eluent, separation was performed isocratically, monitoring the effluent by means of an ELSD. The effluents containing the reaction reduced calorie products were collected and freeze-dried twice to afford the title compounds 6-O-acetyl-a/B-D-glucopyranose (0.9 mmol) and 1 -O-acetyl-B-D-fructopyranose (0.8 mmol) as white solids with purities of >98%. For synthesis of 6-0-aCetyl-a/B-[13C6]-D-glucopyranose, D-glucose-13C6 (1 mmol) and acetic anhydride (0.3 mmol) were reacted in dry pyridine/THF (4/1 , v/v; 2 mL) and purified as detailed above for the natural 13C abundant analogue.

[66] 6-O-Acetyl-a/B-D-glucopyranose, 4, Figure 1 : LC-MS (ESI+): m z 240 (100,

[M+NH4]+), 223 (20, [M+H]+); 1H-NMR (400 MHz, D20, COSY): δ 2.03 [s, 6H, H-C(8a/P)], 3.14 [dd, 1H, J = 4.0, 8.0 Hz, Η-0(2β)], 3.31-3.61 [m, 6H, H-C(4a), H- C(4P), H-C(3P), H-C(2a), Η-0(5β), H-C(3a)], 3.92 [ddd, 1H, J - 4.0, 8.0, 12.0 Hz, H- C(5a)], 4.14 [dd, 1 H, J = 4.0, 12.0 Hz, H-C(6Pa)], 4.22 [dd, 2H, J = 4.0, 8.0 Hz, H- C(6o0], 4.29 [dd, 1 H, J = 4.0, 12.0 Hz, H-C(6pb)], 4.54 [d, 1 H, J = 8.0 Hz, H-C(l p)], 5.10 [d, 1H, J = 4.0 Hz, H-C(la)]; 13C-NMR (100 MHz, D20, HSQC, HMBC): δ 20.1 [C(8a/P)], 20.2 [C(8p/a)], 63.4 [C(6a/p)], 63.6 [C(6p/a)], 69.1 [C(5a>], 69.5 [C(4ct/P)], 69.6 [0(4β/α)] _} 71.4 [C(2a)], 72.6 [C(5p>], 73.3 [C(3a)], 74.0 [C(2P)], 75.5 [C(3P)], 92.0 [C(la)], 96.0 [C(lp)], 174.09 [C(7a/p)], 174.1 1 [C(7p/a)].

[67) l-O-Acetyl-P-D-fructopyranose, 5, Figure 1 : LC-MS (ESI+): m/z 240 (100,

[M+NH4J+), 223 (20, [M+H]+); 1H-NMR (400 MHz, D20, COSY): δ 2.08 [s, 3H, H-C(8)], 3.63 [dd, 1 H, J = 4.0, 12.0 Hz, H-C(6a)], 3.71 [d, 1 H, J = 8.0 Hz H-C(4)], 3.84 [dd, 1 H, J = 4.0, 12.0 Hz, H-C(5)], 3.93-3.99 [m, 2H, H-C(6b), H-C(3)], 4.12 [d, 2H, J = 4.0 Hz, H-C(l)]; 13C-NMR (100 MHz, D20, HSQC, HMBC): δ 20.1 [C(8)], 63.4 [C(6)], 65.6 [C(l)], 67.8 [C(4)], 68.8 [C(3)], 69.2 [C(5)], 96.8 [C(2)], 173.6 [C(7)].

[68] 6-0-Acetyl-a/B-[ 13C6]-D-glucopyranose, [ 13C6J-1 : LC-MS (ESI+): m/z 246 (100, [M+NH4]+), 229 (20, [M+H]+); 1H-NMR (500 MHz, D20, COSY): δ 2.06 [s, 6H, H-C(8a/p)], 3.04 [m, 0.5H, J = 4.0, 12.0 Hz, H-C(2P)], 3.22-3.82 [m, 7.5H, H-C(2a), H-C(2P), H-C(3 ), H-C(3P), H-C(4a), H-C(4P), H-C(5a), H-C(5p)], 4.01-4.43 [m, 4H, H-C(6a), H-C(6P)], 4.49 [d, 0.5H, J = 8.0 Hz, H-C(i p)], 4.75 [d, 0.5H, J = 8.0 Hz, H-C(l p)], 4.98 [d, 0.5H, J = 4.0 Hz, H-C(la)], 5.32 [d, 0.5H, J = 4.0 Hz, H- C(1 P)]; 13C-NMR (125 MHz, D20, HSQC, HMBC): δ 20.1 [C(8a/p)], 20.2

[C(8p/a)], 63.5 [d, J = 44.0 Hz, C(6a), C(6P)], 68.7-69.8 [m, C(5a), C(4a), C(4P)], 71.0-74.3 [m, C(2a), C(5P), C(3a), C(2P)], 75.5 [m, C(3P)], 92.1 [d, J = 46.2 Hz, C(la)], 96.0 [dt, J = 3.75, 5.0, 46.2 Hz, C(i p)], 174.12/174.13 [C(7a/p)].

[69] Synthesis of 5-[13C2]-Acetoxymethyl-2-furaldehyde ([13C2]-6). [13C2]-

Acetylchloride (3 mmol) was added dropwise to a solution of 5-hydroxymethyl-2- furaldehyde (4 mmol) in THF/triethylamine (5/1 , v/v; 6 mL). After stirring for 18 h at 20°C, deionized water (1 mL) was added and the title compound was isolated by means of preparative HPLC on a 250 x 21.2 mm i.d., 5 μηι, Microsorb RP18 column (Varian, Darmstadt, Germany) using water/methanol (60/40, v/v) as the isocratic eluent. 5-[13C2]-Acetoxymethyl-2-furaldehyde (0.5 mmol) was obtained as a white powder and its structure verified by means of LC-MS/MS and NMR spectroscopy.

[70] 5-[13C2]-Acetoxymethyl-2-furaldehyde, [13C2]-6: UV-Vis (MeOH): λπΐ3χ = 284 nm; LC-MS (ESI+): m/z 271 (70, [M+H]+), 291 ( 100, [M+Na]+); 1 H-NMR (400 MHz, MeOD, COSY): δ 1.95 [d, 1.5H, J = 8.0 Hz H-C(l)], 2.21 [d, 1.5H, J = 8.0 Hz, H-C(l )], 5.51 [d, 2H, J = 4.0 Hz, H-C(3)], 6.70 [d, 1H, J = 4.0 Hz, H-C(5)], 7.37 [d, 1 H, J = 4.0 Hz, H-C(6)], 9.57 [s, 1 H, H-C(8)]; 13C-NMR (100 MHz, MeOD, HSQC, HMBC, DEPT): δ 19.1 [d, J = 200 Hz, C(l)], 57.3 [C(3)], 1 12.1 [C(5)], 122.6 [C(6)], 153.0 [C(7)], 156.0 [C(4)], 170.5 [d, Jl ,2 = 200 Hz, C(2)], 178.2 [C(8)].

EXAMPLE 2

Identification and Quantitative Analysis of Candidate Taste-Active Compounds.

[71] Soluble Carbohydrates, Alditols, Organic acids, and Minerals. Balsamic vinegar as well as intermediate samples A-H were diluted with deionized water for

quantification of carbohydrates (1/10000, v/v), alditols (1/1000, v/v), organic acids (1/200, v/v), anions and cations (1/50, v/v), respectively. After membrane filtration (0.45 μπι), aliquots (5-25 μί) of the diluted samples were analyzed by means of high- performance ion chromatography using an ICS 2500 ion chromatography system (Dionex, Idstein, Germany) following the protocol reported recently (28). Gluconic acid was quantitatively determined in vinegar samples by means of an enzyme kit (R- Biopharm, Darmstadt, Germany) according to the manufacturer's instructions.

[72] Amino Acids. Free amino acids were quantified by stable isotope dilution analysis by means of HILIC-MS/MS and following a modified literature protocol (32). The balsamic vinegar samples were diluted with water (1/50; v/v), membrane-filtrated (0.45 μπι), and an aliquot (990 μί) of the sample was then spiked with an aliquot (10 μί) of the internal standard solution containing all isotope labelled amino acids mentioned below (1 mg/L final concentration, each). Aliquots (10 μί) were injected into HPLC-MS/MS system 1 equipped with a 150 x 2.0 mm i.d., 5 μιη; TS gel Amide-80 column (Tosoh Bioscience, Stuttgart, Germany). Using acetonitrile containing 5% of an aqueous solution of ammonium acetate (5 mmol/L, pH 3.0) as eluent A and an aqueous ammonium acetate solution (5 mmol/L, pH 3.0) as eluent B, chromatography was carried out at a flow rate of 0.2 mL/min, starting with a mixture of 85% A and 15% B for 3 min, then increasing the content of B within 7 min to 25% then within 5 min to 50% and, finally to 100% within another 3 min. The following amino acids and their corresponding isotope labelled standards were analyzed in the positive electrospray ionization mode (ESI+) using the mass transitions and declustering potential (DP, in V), entrance potential (EP, in V), collision energy (CE, in V), and cell exit potential (CXP, in V) given in parentheses: glycine (m/z 76.1→76.0; +31/+10/+5/+6), glycine- 13C2-15N (m/z 79.0→79.0; +41/+10/+5/+5), L-alanine (m/z 90.1→90.0; +26/+10/+5/+6), L-alanine-13C3 (m/z 93.0→93.0;

+41/+10/+5/+5), L-serine (m/z 106.1→60.0; +26/+10/+17/+4), L-serine-13C3 (m/z 109.0→62.0; +38/+10/+16/+5), L-proline (m/z 1 16.1→70.0; +21/+10/+21/+4), L- proline-13C5-15N (m/z 122.0→75.0; +73/+10/+25/+5), L-valine (m/z 1 18.1→72.1 ; +21/+10/+15/+6), L-valine- 13C5 (m/z 124.0→77.0; +64/+ 10/+ 14/+ 10), L-threonine (m/z 120.1→73.9; +36/10/+17/+6), L-threonine- 13C4-15N (m/z 125.0→78.0;

+32/+10/+14/+5), L-leucine/L-isoleucine (m/z 132.1→86.0; +41 +10/+15/+6), L- leucine-13C2 (m/z 134.1→87.9; +46/+10/+15/+6), L-isoleucine-13C6 (m/z

139.1→92.0; +39/+ 10/+ 14/+ 10), L-asparagine (m/z 132.9→73.9; +46/+10/+19/+6), L-asparagine-15N2 (m/z 135.0→75.0; +39/+10/+20/+5), L-aspartic acid (m/z 134.1→87.9; +46/+10/+15/+6), L-aspartic acid-13C4-15N (m/z 139.1→92.0;

+39/+ 10/+ 14/+ 10), L-glutamine (m/z 147.0→84.0; +46/+10/+23/+6), L-glutamine- 13C5 (m/z 152.0→88.0; +43/+10/+23/+10), L-glutamic acid (m/z 148.1→84.0;

+31/+10/+23/+6), L-glutamic acid-13C5-15N (m/z 154.0→89.0; +38/+10/+23/+10), L-lysine (m/z 147.0→84.0; +46/+10/+23/+6), L-lysine-13C6-15N2 (m/z

155.0→90.0; +44/+10/+23/+10), L-methionine (m/z 150.1 -→104.0; +31/+10/+15/+8), L-methionine-d3 (m/z 153.1→107.0; +50/+ 10/+ 14/+ 10), L-histidine (m/z

156.1→1 10.0; +41/+10/+21/+8), L-histidine- 13C6 (m/z 162.0→1 15.0;

+46/+10/+21/+10), L-phenylalanine (m/z 166.0→120.0; +51/+10/+19/+10), L- phenylalanine-d5 (m/z 171.0→125.0; +48/+ 10/+ 19/+ 10), L-arginine (m/z

175.1→70.1 ; +36/+10/+33/+4), L-arginine- 13C6 (m/z 181.0→74.0;

+78/+10/+36/+5), L-tyrosine (m/z 182.1→136.0; +26/+ 10/+ 19/+ 10), L-tyrosine-d4 (m/z 186.0-→140.0; +38/+ 10/+ 19/+ 10), L-tryptophan (m/z 205.1→146.0;

+41/+10/+25/+12) and L-tryptophan-d5 (m/z 210.0→150.0; +40/+10/+26/+10). The isotope labeled standards and the analytes were mixed in six molar ratios from 0.04 to 8.0 keeping a constant concentration of the internal standard. After triplicate LC- MS/MS analysis, calibration curves were calculated by plotting peak area ratios of each analyte to the respective internal standard against concentration ratios of each analyte to the internal standard using linear regression (correlation coefficients > 0.99 for each compound).

Phenols and 5-Hydroxymethyl-2-furaldehyde. Following a literature procedure with some modifications (22), an aliquot (1 mL) of the vinegar samples was applied onto a Strata C I 8-E Giga Tube, 55μπι, 7θΑ, RP- 18 cartridge ( 10g/60 mL, Phenomenex, Aschaffenburg, Germany), which was equilibrated with methanol, followed by water ( 100 mL, each). After flushing the cartridge with water ( 100 mL), the target analytes were eluted with methanol ( 100 mL) and, after removing the solvent in vacuum, the residue was taken up in acetonitrile/water (20/80, v/v; 500 μί) and an aliquot (15 μί) was injected into the LC-MS/MS-system 2 equipped with a 150 x 2 mm i.d., 5 μπι, Synergy Fusion RP18 column (Phenomenex, Aschaffenburg, Germany). Using acetonitrile containing 1 % formic acid as solvent A and 1 % aqueous formic acid as solvent B, chromatography was performed with a flow rate of 0.25 mL/min starting with 0% of solvent A for 1 min, then increasing solvent A to 50% within 20 min and to 100% within 1 min and, finally, keeping solvent A at 100% for 2 min. Phenolic acids and esters were analyzed in the positive electrospray ionization mode (ESI+), using the mass transitions and declustering potential (DP, in V), entrance potential (EP, in V), cell entrance potential (CEP, in V), collision energy (CE, in V), and cell exit potential (CXP, in V) given in parentheses: caffeic acid (m/z 181 .1—+89.0;

+21/+6/+ 16/+41/+4), gallic acid (m/z 171.1→108.9; +31 /+ 12/+24/+25/+4), ferulic acid (m/z 195.1→145.1 ; +21/+6/+18/+21/+4), gentisinic acid (m/z 155.1— » 125.0; +46/+12/+14/+27/+4), quinic acid (m/z 193.1→ 147.0; +26/+9/+16/+ 13/+4), vanilline (m z 167.1→92.9; +26/+8/+12/+ 19/+4), p-coumaric acid (m/z 165.1→1 19.2;

+21/+6/+ 14/+25/+4), protocatechuic acid (m/z 155. l -→ 123.2; +41/+12/+ 12/+ 15/+4), syringic acid (m/z 199.2- 140.2; +31/+9/+ 18/+21 /+4), vanillic acid (m z

169.1→93.0; +21/+1 1/+14/+ 19/+4), syringaldehyde (m/z 183.2→ 123.0;

+26/+4/+16/+ 17/+4), 5-hydroxymethyl-2-furaldehyde (m/z 126.9→ 109.0;

+26/+9/+10/+15/+4), gallic acid ethyl ester (m z 199.2→ 127.1 ; +31 /+6/+16/+17/+4), vanillic acid ethyl ester (m z 167.2→1 1 1.0; +26/+10/+ 16/+17/+4). Quantitative analysis was performed by external calibration in triplicate by comparing the peak areas obtained for the corresponding mass traces with those of defined standard solutions (0.05 - 5.0 mg/L) of each reference compound in 20% aqueous methanol.

Castalagin, Vescalagin, and (+)-Dihydrorobinetin. Following a literature protocol for ellagitannin analysis (23) with slight modifications, aliquots ( 1 mL) of vinegar samples were applied on a Strata C I 8-E Giga Tube, 55μπι, 7θΑ, RP- 18 cartridge ( 10g/60 mL, Phenomenex, Aschaffenburg, Germany), which was equilibrated with methanol, followed by water ( 100 mL, each). After flushing the cartridge with water (100 mL), the target analytes were eluted with methanol (100 mL) and, after removing the solvent in vacuum, the residue was taken up in acetonitrile/water (20/80, v/v; 500 μί) and an aliquot (15 μί) was injected into the LC-MS/MS-system 1 equipped with a 150 x 2 mm i.d., 5 μηι, Luna Phenylhexyl column (Phenomenex, Aschaffenburg, Germany). Using methanol containing 1% formic acid as solvent A and aqueous 1% formic acid as solvent B, chromatography was performed with a flow rate of 0.25 mL/min starting with a mixture of 10% solvent A and 90% solvent B for 1 min, then increasing the content of solvent A within 12 min to 100%, which was then kept isocratic for additional 5 min. Vescalagin (1), castalagin (2), and (+)- dihydrorobinetin (3) were analyzed in the in the MRM mode by use of negative electrospray ionization using the mass transitions and declustering potential (DP, in V), entrance potential (EP, in V), collision energy (CE, in V), and cell exit potential (CXP, in V) given in parentheses: vescalagin/castalagin (m/z 933.0→301.0; -150/- 10/-70/-7; m/z 933.0→631.0; -150/-10/-40/-29); (+)-dihydrorobinetin (m/z

303.0→ 174.9; -75/-10/-26/-1 1 ; m/z 303.0→ 108.9; -75/-10/-44/-5). For quantitation of the wood-derived compounds 1-3, six standard solutions of vescalagin (1) castalagin (2), and (+)-dihydrorobinetin (3) were prepared in balsamic vinegar of Modena (BV). To achieve this, aliquots of the BV (1 mL) were spiked with different amounts of stock solutions of compounds 1-3 (final concentration: 0.5 - 5.0 mg/L, each), followed by sample-work up as detailed above. After triplicate LC-MS/MS analysis, calibration curves were calculated by plotting the peak area of each analyte against the concentration using linear regression, showing linear responses with correlation coefficients of > 0.99 each. Quantitative analysis was performed in triplicate by comparing the peak areas obtained for the corresponding mass traces in the samples with those of the standard solutions of both reference compounds.

5-Acetoxymethyl-2-furaldehyde (6). Vinegar samples were diluted 1 :200 (v/v) with de-ionized water, membrane-filtrated (0.45 μιη), and, after spiking a sample aliquot (990 μί) with the internal standard solution (10 μί) containing 5-[13C2]- acetoxymethyl-2-furaldehyde (200 μg/L), aliquots (10 μί) were injected into the LC- MS/MS system 2 connected to a 150 x 2 mm i.s., 5 μιη, Luna PFP column

(Phenomenex, Aschaffenburg, Germany). Using acetonitrile containing 1% formic acid as solvent A and aqueous 1% formic acid as solvent B, chromatography was performed with a flow rate of 0.2 mL/min starting with 0% of solvent A for 2 min, then increasing solvent A to 100% within 15 min, followed by an isocratic elution for additional 2 min. 5-acetoxymethyl-2-furaldehyde and 5-[13C2]-acetoxymethyl-2- furaldehyde were analyzed in the positive electrospray ionization mode (ESI+), using the mass transitions and declustering potential (DP, in V), entrance potential (EP, in V), cell entrance potential (CEP, in V) collision energy (CE, in V), and cell exit potential (CXP, in V) given in parentheses: 5-acetoxymethyl-2-furaldehyde (m/z 168.9→109.0; +16/+7/+14/+17/+4), 5-[13C2]-acetoxymethyl-2-furaldehyde (m/z 171.0→109.0; +1 1/+8/+ 14/+17/+4). The isotope labeled standard and the analyte were mixed in eight molar ratios from 0.1 to 8.0 keeping a constant concentration of the internal standard. After triplicate LC-MS/MS analysis calibration curves were prepared by plotting peak area ratios of each analyte to the respective internal standard against concentration ratios of each analyte to the internal standard using linear regression (correlation coefficient > 0.99).

6-O-Acetyl-a/B-D-gIucopyranose (4) and l-O-Acetyl-B-D-fructopyranose (5). The balsamic vinegar samples were diluted with water (1/50; v/v), membrane-filtrated (0.45 μηι), and an aliquot (990 μί) of the sample was then spiked with an aliquot of an internal standard solution (10 μί) containing 6-0-acetyl-a/B-[13C6]-D- glucopyranose (50 mg/L). Aliquots (10 μί) were injected into LC-MS/MS-system 2 connected to a 150 x 2.0 mm i.d., 3 μη , TSKgel NH2-100 column (Tosoh Bioscience, Stuttgart, Germany). Using acetonitrile containing 5% of an aqueous ammonium acetate solution (5 mmol/L) and 1% formic acid as eluent A and an aqueous ammonium acetate solution (5 mmol/L) containing 1% formic acid as eluent B, chromatography was performed isocratically with 89% solvent A at a flow rate of 0.2 mL/min for 10 min. 6-0-Acetyl-a/B-D-glucopyranose, l-O-Acetyl-B-D- fructopyranose and the corresponding isotope labelled standard were analyzed in the positive electrospray ionization mode (ESI+) using the mass transitions and declustering potential (DP, in V), entrance potential (EP, in V), cell entrance potential (CEP, in V), collision energy (CE, in V), and cell exit potential (CXP, in V) given in parentheses: 6-O-acetyl-a/B-D-glucopyranose and l-O-acetyl-B-D-fructopyranose (m/z 240.1→205.1 ; +1 1/+7/+20/+13/+4; 240.1→187.0; DP: +1 1/+7/+20/+17/+4); 6- 0-acetyl-a/B-[13C6]-D-glucopyranose (m/z 246.1→21 1.1 ; +1 1/+3/+28/+15/+4). For recording of calibration curves the isotope labeled standard and the two analytes were mixed in six molar ratios from 0.4 to 8.0 keeping a constant concentration of the internal standard. After triplicate LC-MS/MS analysis calibration curves were prepared by plotting peak area ratios of each analyte to the respective internal standard against concentration ratios of each analyte to the internal standard using linear regression, showing linear responses with correlation coefficients of > 0.99, each.

EXAMPLE 3

Preparation of Taste Recombinants.

[77] A basic taste recombinant (rTBVI-VI) was prepared by dissolving the tastants of TBV summarized in groups I- VI (Table 2) in their "natural" concentrations in bottled water and adjusting the pH-value of this solution to 3.0 with trace amounts of hydrochloric acid (0.1 M). In addition, an extended taste recombinant (rTBVI-VII) with the taste compounds given in groups I-VI (Table 2) and the TBV-HMW fraction (group VII in Table 2), each in its "natural" concentration, and a total taste recombinant (rTBVtotal) containing all tastants from groups I-VIII of TBV were prepared. In addition, a total taste recombinant (rBVtotal) was prepared containing the individual taste compounds of BV, each in its "natural" concentration (Table 2). After equilibration for 12 h, the overall taste profiles of rTBVI-VI, rTBVI-VII, rTBVtotal, and rBVtotal were evaluated by means of taste profile analysis.

EXAMPLE 4

Analytical Sensory Experiments.

[781 General Conditions, Panel Training. In order to familiarize the subjects with the taste language used by our sensory group and to get them trained in recognizing and distinguishing different qualities of oral sensations in analytical sensory experiments, 12 assessors (eight women and four men, age 26-39 years), who gave the informed consent to participate the sensory tests of the present investigation and had no history of known taste disorders, participated for at least two years in sensory training sessions with purified reference compounds by using the sip-and-spit method as reported earlier (23, 26, 29), but performing the experiments at pH 3.0, adjusted with hydrochloric acid (0.1 mol/L). For intensity scaling, test solutions, containing a tastant in defined concentrations, were used to calibrate the panel forjudging the intensities 0, 2.5, and 5.0. Prior to sensory analysis, the fractions or compounds isolated were analytically confirmed to be essentially free of solvents and buffer compounds.

[79] Taste Recognition Threshold Concentrations. Threshold concentrations of sour sweet, bitter, salty, or umami tasting compounds were determined in bottled water adjusted to pH 3.0 with trace amounts of hydrochloric acid (0.1 M), using triangle tests with ascending concentrations of the stimulus following the procedure reported in the literature (23). To overcome carry-over effects, astringent tasting compounds were evaluated by use of the already reported half tongue test (36). Values between individuals and separate sessions did not differ more than plus or minus one dilution step; as a result, a threshold value of e.g. 900 μιτιοΙ/L for gluconic acid represents a range of 450- 1800 μιηοΙ/L.

[80] Taste Profile Analysis. Aqueous 1 +2 dilutions of vinegar samples TBV and BV as well as the taste recombinant solutions rTBV and rBV, respectively, were presented to the trained sensory panelists, who wore nose clips in order to prevent cross-modal interactions, and were asked to evaluate the taste qualities bitter, sour, sweet, salty, umami, astringent and mouthfulness/viscosity on an intensity scale from 0 (not detectable) to 5 (strongly detectable). For taste profile analysis of vinegar fractions, the freeze-dried samples were taken up in bottled water in their "natural"

concentrations and the pH value of the solutions was adjusted to 3.0 by adding trace amounts of hydrochloric acid (0.1 mol/L). Aqueous 1+2 dilutions (v/v) of this stock solutions were then presented to the sensory panelists who were asked to rate the intensity of the individual taste qualities as given above.

[81] Taste Dilution Analysis (TDA). Aliquots of the lyophilized GAC fractions were taken up in "natural" ratios in water (10.0 mL), adjusted to pH 3.0 with trace amounts of hydrochloric acid (0.1 mol/L), diluted stepwise 1+1 (v/v) with acidified water (pH 3.0), and, then, used for the determination of the taste dilution (TD) factor as detailed in the literature (25, 36).

[82] Comparative Taste Profile Analysis. Lyophilized GAC fractions (I-X) were taken up in their "natural" concentrations either in water (10 mL), or in basic taste recombinant solution (rTBVI-VI; 10 mL), and the pH-value was adjusted to 3.0 using trace amounts of hydrochloric acid (0.1 mol/L). These solutions were then presented to the trained sensory panel, which was asked to rate the intensity of the taste qualities sweet, sour, astringent, mouthfulness, bitter, salty, and umami taste on a scale from 0 (no taste impression detectable) to 5 (strong taste impression) in comparison to the non-spiked recombinant rTBVI-VI as control.

[83] Determination of Sweetness Modulatory Activity. The sweetness modulating activity of 5-hydroxymethyl-2-furaldehyde and 5-acetoxymethyl-2-furaldehyde (6) were determined in 4% sucrose solution (pH 3.0) containing 1% ethanol using a three alternative forced-choice test (3-AFC) (37), meaning that the panelists had to choose the differing sample out of three samples, one containing 4% sucrose and the stimuli and the other two just containing 4% sucrose, whereby the panelists were forced to choose one of the samples, even if they didn't taste a difference in sweetness.

Concentrations of the stimuli ranged from 0.2 to 10.0 mmol/L for 5-hydroxymethyl-2- furaldehyde and from 0.2 to 2.0 mmol/L for 5-acetoxymethyl-2-furaldehyde (6).

[84] High Performance Liquid Chromatography (HPLC). The HPLC system (Jasco, GroB- Umstadt, Germany) consisted of two PU-2087 Plus pumps, a DG-2080-53 degasser, a LG-2080-02 gradient unit, an AS-2055 Plus autosampler with a 100 loop, a Rh 7725i injection valve with a 1000 loop (RJ eodyne, Bensheim, Germany), a MD- 2010 Plus multiwavelength detector, and a 85 Sedex LT-ELSD (Sedere, Alfortville, France).

[85] Liquid Chromatography/Mass Spectrometry (LC-MS/MS). LC-MS/MS

measurements were acquired on two different systems. LC-MS/MS-system 1 : a API 4000 Q-Trap LC-MS/MS system (Applied Biosystems Sciex Instruments, Darmstadt, Germany) connected to a 1200 HPLC-system (Agilent, Waldbronn, Germany); LC- MS/MS-system 2: a API 3200 LC-MS/MS system (Applied Biosystems) connected to a 1 100 HPLC-system (Agilent, Waldbronn, Germany). Ion spray voltage was set at 5500 V in the ESI+ mode and at -4500 V in the ESI- mode, source temperature was set at 425 °C, nitrogen was served as curtain gas (20 psi) and declustering potential was set at +/-25 V, respectively. Both mass spectrometers were operated in the full scan mode for monitoring of positive or negative ions. Fragmentation of [M+H]+ or [M-H]- pseudo molecular ions into specific reduced calorie product ions was induced by collision with nitrogen (4x10-5 Torr) and a collision energy of +/-25 V.

Declustering potential (DP), entrance potential (EP), collision cell entrance potential (CEP), collision energy (CE), and cell exit potential (CXP) were tuned for each individual compound by flow injection (10 μΐνππη) via syringe pump injection, detecting the fragmentation of the [M+H]+ or [M-H]- pseudomolecular ions into specific reduced calorie product ions after collision with nitrogen (4.5x10-5 Torr).

[86) Nuclear Magnetic Resonance Spectroscopy (NMR). 1 H, 13C, COSY, HSQC and HMBC experiments were performed on a Bruker DRX-400 or an Avance-III-500 spectrometer, the latter of which was equipped with a Cryo-CTCI probe (Bruker, Rheinstetten, Germany). MeOD or D20 were used as solvents and

trimethylsilylpropionic acid-d4 (TMSP) as the internal standard. Data processing was performed by using Topspin software (version 2.1 ; Bruker) as well as Mestre-C software (version 4.8.6; Mestrelab Research, Santiago de Compostella, Spain).

[87] Multivariate Analysis. Data analysis was performed within the programming and visualization environment R (version 2.10.0) (38). The sensomics heatmap (Figure 6) was calculated using the heatmap.2 function of R based on the raw concentration data (see supporting information) normalized to the TBVM content of each

sensometabolite. The dendrogram was constructed by means of an agglomerative average linkage algorithm (39), whereas the distance between two clusters is defined as the average of distances between all pairs of objects and each pair is made up of one object from each group.

EXAMPLE 5

[88] To identify the compounds responsible for the typical taste of traditional balsamic vinegar of Modena (TBV), a 1 +2 dilution of the vinegar and, in comparison of a balsamic vinegar of Modena (BV), was presented to 12 trained panelists who were asked to rate the intensities of the taste descriptors sweet, sour, bitter, astringent, umami, salty, as well as the impression of mouthfulness on a linear scale from 0 (not detectable) to 5 (strongly detectable). Sourness and sweetness of the TBV sample were rated with the highest intensities of 3.6 and 2.3, respectively, followed by astringency (2.2) and mouthfulness (1.7), whereas bitterness was perceived with a lower score of 0.6 only (Table 1). In comparison, the BV sample exhibited a significantly more sour (4.3) and less sweet ( 1.5) taste profile with also lower scores judged for mouthfulness (1.0). In both samples, umami and salty taste were hardly perceived with an average intensity of 0.2. The mean values obtained for each orosensory impression in triplicate analysis was used to calibrate the sensory panel for the precise sensory evaluation of TBV and the fractions isolated therefrom in the following.

[89] In order to separate the taste-active compounds based on molecular weight

differences, TBV was separated by means of ultrafiltration using a polyethersulfone membrane with a 5 kDa cut-off to obtain the low molecular weight (TBV-LMW, < 5kDa) and the high molecular weight fraction (TBV-HMW, > 5kDa) after freeze- drying as amorphous powders. To evaluate their sensory impact, both fractions were taken up in bottled water in their "natural" concentrations, 1+2 diluted with table water, and, after adjusting the pH to 3.0, were analyzed by means of a comparative taste profile analysis using the 1+2 diluted TBVM as reference (Table 1 ). The intensities of the orosensory descriptors sweet, bitter, salty, and umami judged for the TBV-LMW fraction were rather close to those of the TBV sample, thus

demonstrating the key molecules imparting these sensations do exhibit molecular weights below 5 kDa. In comparison to TBV, sourness, astringency, and

mouthfulness were judged with lower scores in the TBV-LMW fraction. The comparatively low impact of sourness in the TBV-LMW fraction might be explained by the loss of the volatile acetic acid upon sample lyophilization. The low astringency score of 1.4 reported for the TBV-LMW fraction indicated that the overall astringency perception of TBV (2.2) is only partially due to low molecular weight compounds and is complemented by astringent macromolecules in the TBV-HMW fraction judged with an intensity of 0.9. This is well in line with recent findings on the equally important contribution of low (<5kDa) and high molecular weight compounds (>5kDa) to the overall astringency of red wine (22).

[90] Identification and Quantitative Analysis of Basic Taste Molecules in TBV. In order to evaluate the sensory impact of basic taste active compounds to the taste profile of TBV, 2 monosaccharides, 8 alditols, 8 organic acids, 4 cations and 2 inorganic anions were identified and quantitatively analyzed by means of high-performance ion chromatography (Table 2). In addition, gluconic acid was quantified by means of an enzymatic assay. Moreover, 8 phenolic acids, 2 phenolic acid esters, vanilline, and 5- hydroxymethyl-2-furaldehyde were identified and quantified by means of RP-HPLC- MS/MS and a total of 15 free amino acids by means of HILIC-MS/MS.

[91] As TBV is aged in wooden barrels, the vinegar sample was analyzed for the presence of vescalagin (1 ; Figure 1) and castalagin (2), both ellagitannins have been reported as astringent molecules migrating from oak wood into wine and whiskey upon barrel maturation (23). As sensory evaluation revealed an intense astringent impression above the recognition threshold concentration of 23 μιηοΙ/L (Table 2), the

dihydroflavonol (+)-dihydrorobinetin (3) was analyzed in TBV as it is reported as a marker molecule for storage of vinegars in Acacia barrels (40). After SPE cartridge clean-up, compounds 1-3 were analyzed by means of RP-HPLC-MS/MS in TBV and for comparison also in the BV sample, the latter of which is not matured in wooden barrels. As expected, recording the characteristic mass transitions of vescalagin (1), castalagin (2) and (+)-dihydrorobinetin (3) in the BV sample did not show any signal (A/B; Figure 2). In contrast, the analysis of the TBV sample clearly demonstrated the presence of the wood-derived polyphenols 1 -3 (C/D; Figure 2). Finally, matrix calibration by spiking the BV sample with the reference compounds, followed by HPLC-MS/MS analysis confirmed the absence of 1 -3 in BV and, for the first time, led to the successful identification of these polyphenols in TBV (E/F; Figure 2). Matrix- calibrated HPLC-MS/MS quantitation revealed concentrations of 51 and 38 μπιοΙ/L for the ellagitannins vescalagin (1 ) and castalagin (2) and 1 μηιοΙ/L for (+)- dihydrorobinetin (3).

[92] After quantitative analysis, the taste recognition threshold concentrations of the

compounds were determined and a dose-over-threshold (DoT)-factor was calculated for each compound from the ratio of the concentration and the threshold concentration (26). As we aimed to elucidate the key metabolites for each individual taste quality, the single taste compounds identified in TBV were grouped into classes differing in their taste qualities (Table 2).

[93] Among the sweet tasting molecules (group I) in TBV, fructose and glucose were

evaluated with the highest DoT-factors of 157.0 and 101.0, followed by glycerol with a value of 1.7, and L-proline, inositol, sorbitol, erythritol, and xylitol evaluated with DoT-factors between 0.1 and 0.4 (Table 2). Tartaric acid, gluconic acid, glycolic acid, malic acid, acetic acid, citric acid, and succinic acid exceeded their taste threshold concentrations with tastant group II comprising the sour tasting molecules. Group III consisted of bitter tasting amino acids and minerals, among which only calcium and magnesium ions exceeded their thresholds by a factor of 4.0 and 3.1 , respectively (Table 2). All astringent molecules of TBV were summarized in group IV, but only the ellagitannins castalagin (2), vescalagin (1), as well as 5-hydroxymethyl-2- furaldehyde showed high DoT-Factors of 46.4, 34.5, and 2.9, respectively. (E)- Caffeic acid, gentisinic acid, p-coumaric acid, and gallic acid were evaluated with DoT-factors between 0.1 and 0.4, whereas all the other polyphenols were more than 10-fold below their taste threshold concentrations (Table 2). Among the group of salty tasting components (group V), the cations sodium and potassium and the anions chloride and phosphate were judged with DoT-factors between 1.5 and 3.1 in TBV, whereas none of the amino acids in group VI exceeded their recognition threshold for umami taste.

Re-Engineering the Taste of TBV. To confirm the results of the instrumental analysis and to check whether the compounds identified can already create the typical taste of TBV, an aqueous taste recombinant containing 54 taste-active compounds, each in its "natural" concentration (Table 2), was prepared and, after pH adjustment (pH 3.0), the taste profile of this basic taste recombinant (rTBVI-VI) was compared to that of the TBV (Table 3). The intensities detected for sweetness, sourness, bitterness, and astringency matched rather well those determined for TBV, whereas the mouthfulness was judged significantly less intense in the tastant cocktail. In addition, the sensory panel reported the sweetness of TBV to be more long lasting when compared to rTBVI-VI. In order to investigate the impact of the high-molecular weight components of vinegar on the mouthfulness perception, an additional taste

recombinant (rTBVI-VII) was made by spiking rTBVI-VI with the TBV-HMW fraction in its "natural" concentration. Comparative taste profile analysis of rTBVI- VII revealed an increase in mouthfulness from 1.1 to 1.6, thus demonstrating the macromolecular components to play an important role in mouthfulness perception. Despite the presence of the TBV-HMW fraction, the recombinant rTBVI-VII induced a less long-lasting sweetness perception when compared to the authentic TBV or the TBV-LMF fraction, respectively, thus indicating that the basic taste recombinant is lacking compounds modulating the sweetness perception. [95] Sensory-Directed Fractionation of the TBV-LMW Fraction. In order to locate the molecules responsible for the long-lasting sweet taste of TBV, the TBV-LMW fraction was fractioned by means of gel absorption chromatography (GAC) on Sephadex LH-20 material using a methanol/water gradient. Monitoring the effluent by means of UV/Vis detection, the TBV-LMW fraction was separated into ten fractions, namely I-X (Figure 3), which were individually collected and freeze-dried twice.

[96] In a first set of experiments, an aliquot of each GAC fraction was dissolved in bottled water in its "natural" concentration, which means in the amounts obtained from the GAC column, and evaluated by means of a taste dilution analysis (TDA) to evaluate their intrinsic taste impact (Table 4). The highest TD-factors of 256 and 128 were found for sweetness in fraction VII and sourness in fraction IX, respectively, followed by astringency in fraction IX and sourness in fraction III, both judged with a TD- factor of 64 (Table 4). Besides sour taste (16), bitterness (4), and astringency (2), the sensory panel reported on a sweet taste in fraction VI but with a low TD factor of 1. Except for the sensory inactive fraction I, the remaining fractions revealed some sour, bitter, and astringent taste impressions or combinations thereof with TD-factors ranging from 1 to 32 (Table 4).

[97] A second set of experiments were aimed at the discovery of taste-modulating

molecules in the individual GAC-fractions. Therefore, a solution of the basic taste recombinant rTBVI-VI was used as the matrix solution for the localization of candidate sweetness modulators. To achieve this, aliquots of the individual GAC- fractions were dissolved in the 1 +2-diluted rTBVI-VI solution in their "natural" concentrations and were then evaluated by means of a comparative taste profile analysis using the likewise diluted, blank rTBVI-VI as control. Out of the fractions I- X, spiking the rTBVI-VI solution with fractions VII and X induced a increased sweetness intensity and a more long-lasting sweetness perception, respectively, when compared to rTBVI-VI alone (Table 4). Therefore, further experiments were targeted towards the sensory active compounds imparting the enhanced sweetness or more long-lasting sweetness in fractions VII and X, respectively.

[98] Identification of Taste and Taste Modulatory Compounds in TBV. Fraction VII was separated by means of hydrophilic liquid interaction chromatography (HILIC) to afford five subfractions, namely VII- 1 to VII-5 (Figure 4 A), all of which imparted sweet and/or bitter taste. MS- and HPIC analysis revealed fructose and glucose as the main components in subfraction VII-5. In addition, LC-MS (ESI+) analysis revealed m/z 223 ([M+H]+) and 240 ([M+NH4J+) as the pseudomolecular ions of the compounds eluting in four fractions VII- 1 to VII-4, thus indicating the presence of isomers. As the molecular weight of 220 Da was well in agreement with those of monosaccharide acetates reported in literature (40), reference compounds of 6-0- acetyl-a/B-D-glucopyranose (4, Figure 1) and 1 -O-acetyl-P-D-fructopyranose (5, Figure 1) were synthesized by acetylation of D-glucose and D-fructose, respectively, with acetic anhydride in pyridine and dry THF, followed by structure verification by means of LC-MS/MS and 1 D/2D-NMR-experiments. Comparison of MS data and retention times (RP-HPLC, HILIC), followed by co-chromatography with the corresponding reference compound revealed 6-O-acetyl-a-D-glucopyranose and 6-0- acetyl-B-D-glucopyranose (4, Figure 1 ) as bitter-sweet compounds eluting in fractions VII-3 and VII-4, respectively, and 1 -O-acetyl- -D-fructopyranose (5, Figure 1 ) as the bitter-sweet compound eluting in fraction VII-2. Sensory evaluation of the mixture of 6-O-acetyl-a/p-glucopyranose showed a taste threshold concentration of 12.3 mmol/L for sweet and bitter taste, whereas 1 -O-acetyl-P-D-fructopyranose revealed a threshold of 16.9 mmol/L for sweet and 21.2 μη οΙ/L for bitter taste.

Separation of fraction X by means of RP-HPLC and monitoring the effluent at 280 nm resulted in two peaks, namely X-l and X-2 (Figure 4B). Comparison of MS data and retention time with those of the reference compound, followed by co- chromatography led to the identification of peak X-l as 5-hydroxymethyl-2- furaldehyde. LC-MS/MS analysis of peak X-2 showing sweet modulating activity in rTBVI-VI in a degustation experiment (data not shown) revealed a pseudomolecular ion ([M+H]+) of m z 168 as well as the fragment ions m/z 109 and m/z 81 for the compound eluting in fraction X-2, thus suggesting the presence of a 5- hydroxymethyl-2-furaldehyde moiety. The 1H and 13C NMR spectra of that unknown compound revealed five proton resonance signals integrating for eight protons and a total of eight carbon signals (Figure 5). The proton signals H-C(3), H- C(5), H-C(6), and H-C(8) resonating at 5.17, 6.71 , 7.39, and 9.59 ppm were assigned as the protons of the 5-hydroxymethyl-2-furaldehyde moiety. In addition, the singlet signal resonating at 2.02 ppm and integrating for three protons indicated the presence of a methyl group. The 13C NMR spectrum exhibited the resonance signal at 178.2 ppm as expected for the aldehyde carbonyl atom C8 and, in addition, another carbon signal C(2) at 170.5 ppm, thus suggesting the presence of an acetyl ester moiety. Heteronuclear couplings (HMBC) were observed between the carbonyl carbon C(2) and the methyl residue H-C(l) as well as the methylene protons H-C(3) (Figure 5), thus leading to the identification of that compound as 5-acetyoxymethyl-2- furaldehyde (6).

[100] 5-Acetoxymethyl-2-furaldehyde (6) and, in comparison, 5-hydroxymethyl- furaldehyde were sensorially evaluated for their intrinsic taste in water (pH 3.0, 1% ethanol) and in a 4% aqueous sucrose solution (pH 3.0, 1% ethanol) for sweet taste modulating properties. Whereas neither 5, nor 6 showed any intrinsic taste up to a concentration of 2 mmol/L (water), the presence of 5-acetoxymethyl-2-furaldehyde (6) induced a significant change in sweet taste quality as well as a more long-lasting sensation in the 4% sucrose solution with a recognition threshold of 1.5 mmol/L (a- level: 0.05). In comparison, the non-acetylated compound 5 did not show any sweetness modulating activity. Although this compounds which is reported as an aging marker of TBV (10, 14), the sensory impact of 5-acetoxymethyl-2-furaldehyde (6) has not been reported until now.

[101] Quantitative Analysis of 6-0-Acetyl-a/B-D-glucopyranose (4), l -O-Acetyl-P-D- fructopyranose (5), and 5-Acetoxymethyl-2-furaldehyde (6) in TBV. To determine the concentrations of compounds 4-6 in TBV, 6-0-acetyl-a/B-[13C6]-D-glucopyranose and 5-[13C2]-acetoxymethyl-2-furaldehyde were synthesized and used as internal standards for the development of stable isotope dilution analyses

[102] After spiking the vinegar samples with defined amounts of the 6-O-acetyl-a/B-

[13C6]-D-glucopyranose and 5-[13C2]-acetoxymethyl-2-furaldehyde, followed by sample clean-up, the natural [13C]-abundant and [13C6]-6-0-acetyl-a/B-D- glucopyranose (4) and 1 -O-acetyl-p-D-fructopyranose (5) were analyzed by means of HILIC-MS/MS and natural [13C] -abundant and [13C2]-5-acetoxymethyl-2- furaldehyde (6) was analyzed by means of RP-HPLC-MS/MS. The TBV sample contained 13823 and 8329 μι οΐνΐν 1 -O-acetyl-P-D-fructopyranose (5) and 6-0- acetyl-a/p-D-glucopyranose (4) as well as 315 μι οΐνί. of 5-acetoxymethyl-2- furaldehyde (6) (Table 2). Calculation of DoT-factors revealed values of 0.7 and 0.8 for the intrinsic sweetness of the hexose acetates 5 and 4, respectively, and a value of 0.2 for 5-acetoxymethyl-2-furaldehyde based on its threshold concentration in a 4% sucrose solution (Table 2).

[103] In order to answer the question as to whether the differences in sweet taste quality of the taste recombinants (rTBVI-VI, rTBVI-VII) and the authentic TBVM is due to compounds 4-6, a total taste recombinant (rTBVtotal) was prepared by spiking the taste recombinant rTBVI-VII with l-O-acetyl-P-D-fructopyranose, 6-O-acetyl-a/p-D- glucopyranose, and 5-acetoxymethyl-2-furaldehyde, each in its "natural"

concentration. By means of a three-alternative forced choice test (p<0.1), rTBVI-VII and rTBVtotal could be significantly differentiated by the sensory panel. Comparative taste profile analysis did not reveal any significant increase in sweet taste intensity from rTBVI-VII to rTBVtotal, but the panelists reported on a change in sweet taste quality as well as an increase in the duration of the sweetness perception of rTBVtotal, the taste profile of which closely matched that of the authentic TBV (Table 3). Omission of the hexose acetates 4 and 5 from rTBVtotal could not be significantly differentiated from rTBVtotal containing compounds 4-6 by means of a three-alternative forced choice test (data not shown), thus demonstrating 5- acetoxymethyl-2-furaldehyde (6) to be a natural sweet taste modulator in the TBV matrix.

[104] Comparison of Taste Compounds in TBV and BV. In order to answer the question as to whether the differences in the taste profiles of TBV and BV is reflected by the differences in the concentrations of individual taste compounds, the total of 59 taste compounds were quantitatively determined in the TBV sample and DoT-factors were calculated (Table 2). First, an aqueous taste recombinant (rTBtotal) containing all taste-active molecules, each in its "natural" concentration (Table 2), was prepared and its taste profile compared to that of the authentic TB sample in order to functionally confirm the results of the instrumental analysis (Table 3). As the taste profile of rTBtotal matched rather well that of TBV, the taste molecules summarized in Table 2 were considered the key molecules imparting the taste profile of BV.

[105] Comparing the concentrations of the individual taste molecules in TBV and BV

revealed 2-fold higher values for the sour tasting acetic acid, but 2-3 times lower values for the sweet tasting monosaccharides and glycerol (group I, Table 2) and the bitter-sweet tasting hexose acetates 6-O-acetyl-a/B-D-glucopyranose (4) and l-O- acetyl-P-D-fructopyranose (5), as well as 4 times lower levels of the sweet modulating 5-acetoxymethyl-2-furaldehyde (6), thus being well in line with the observed differences in sweet/sour balance between samples TBV and BV. On the other hand, the concentration of the major non-volatile acids gluconic acid, glycolic acid, malic acid, and tartaric acid were 14, 7, 6 and 3 times lower when compared to TBV. Within the group of bitter and astringent compounds, BV showed significantly lower DoT- factors for the minerals as well as the polyphenolic acids, and the wood-derived ellagitannins 1 and 2 as well as (+)-dihydrorobinetin (3) were not detectable at all in BV, thus indicating the lack of any wood maturation in industrial BV manufacturing.

[106] Sensory Evaluation and Quantitative Sensomics Profiling of Storage Levels during TBV Ageing. In order to gain more detailed insight into the taste development in TBVM reduced calorie production, each intermediary sample collected from the "batteria" of barrel A (acacia), B (chestnut), C (cherry), D (mulberry), E (oak), F (chestnut), G (chestnut), and H (chestnut) throughout a full-scale TBV manufacturing process was analyzed by means of comparative taste profile analysis using the final TBV sample (Table 1) as the reference. With increasing degree of maturation, a slight decrease in sourness and an increase of sweetness and mouthfulness was observed (Figure 6). In comparison, perceived astringency did not seem to be significantly influenced by maturing.

[107] Aimed at correlating the sensory data with the presence of the individual tastants, a total of 37 selected sweet, sour, or astringent sensometabolites were quantitatively determined in the barrel samples A-H (Figure 7). In order to determine the

multivariate distances between the respective sensometabolites throughout the maturation process, the concentrations determined for each compound in samples A- H were normalized and a hierarchical cluster analysis was performed on the basis of the normalized data, respectively. The results were visualized in a sensomics heatmap that was combined with hierarchical agglomerative clustering of the 37

sensometabolites (Figure 7). In order to consider concentration effects by the evaporation of water throughout the "batteria" cascade, the concentrations used for these calculations are based on fresh weight (A; Figure 7) and dry matter (B), respectively. The cluster analysis quantifies the degree of similarity between the sensometabolites by calculating the distance between all possible pairs of molecules. The two most similar sensometabolites were then grouped together and the distance measure recalculated. This iterative process was continued until all sensometabolites were members of a single cluster. This resulting hierarchical clustering is visually displayed as a dendrogram (Figure 7A, B). The closer the sensometabolites are to each other in the dendrogram, the smaller the differences in their concentration patterns throughout the entire TBVM manufacturing process.

[108] Based on fresh weight calculation, the hierarchical analysis arranged the

sensometabolites into the two large clusters 1 and 2 subdivided into the smaller clusters l a and l b, as well as 2a-2c (Figure 7A).

[109] Cluster l a (Figure 7A) consisted of vescalagin ( 1 ), glycolic acid, quinic acid, and gluconic acid, following a mixed trend throughout the maturation procedure. For example, gluconic acid went through a maximum in barrel E fitting well to decreasing microbial activity with increasing age (/, 6). In comparison, highest levels of vescalagin were observed in barrel H made from fresh chestnut collaborating well with its release from chestnut (41).

[110] Cluster l b (Figure 7A) comprised the sensometabolites increasing in concentration upon maturation, among them the sweet tasting fructose, glucose, 1 -O-acetyl-P-D- fructopyranose, 6-0-acetyl-a/p-D-glucopyranose, glycerol, sorbitol, xylitol, erythritol, mannitol, and the sweetness modulating 5-acetoxymethyl-2-furaldehyde (6), being well in line with the increased sweetness impact developing from sample A to H. In addition, the sour tasting organic acids malic acid, citric acid, lactic acid, tartaric acid, and the astringent compounds gallic acid and 5-hydroxymethylfurfural increased with maturation age.

[I l l] Whereas ribitol clustered separated from all the other compounds (cluster 2a), the large cluster 2b (Figure 7A) consisted of sensometabolites decreasing in concentration during maturation of the vinegar. Among these, the sour tasting acetic acid, the sweet tasting arabitol, as well as the astringent phenolic compounds gentisic acid, protocatechuic acid, gallic acid methyl ester, gallic acid ethyl ester, p-coumaric acid, syringic acid, ferulic acid, caffeic acid, syringaldehyde, and vanilline, respectively.

[112] Cluster 2c (Figure 7A) grouped the astringent tasting (+)-dihydrorobinetin (3),

castalagin (2), vanillic acid, and p-hydroxybenzoic acid, all of which were present in highest concentrations in sample A taken from the Acacia barrel, thereafter decreasing strongly throughout the maturation cascade.

[113] In order to remove concentration effects by the evaporation of water throughout the "battaria" cascade, another hierarchical analysis was performed using the quantitative data based on dry matter of barrel samples A-H. The sensometabolites were arranged into three main clusters ( l a, l b, and 2 in Figure 7B). Cluster l a (Figure 7B) comprises the sensometabolites which are partially degraded with increasing degree of maturation, namely the astringent phenolic compounds, the sour tasting acetic acid, gluconic acid, malic acid, and lactic acid, as well as mannitol. Cluster l b contained sensometabolites which were detected in highest concentrations in barrel sample A and dropped drastically already in barrel B such as, e.g. (+)-dihydrorobinetin (3).

[114] In contrast, cluster 2 (Figure 7B) summarized the sensometabolites which are

increasing in concentration. Among these substances are xylitol, sorbitol, erythritol, citric acid, the monosaccharides glucose und fructose as well as the hexose acetates 4 and 5, as well as the Maillard reaction reduced calorie product 5-hydroxymethyl-2- furaldehyde and its acetylation reduced calorie product 5-acetoxymethyl-2- furaldehyde (6). Whereas the data on the increasing concentrations of the

sensometabolites in cluster l b of the wet weight calculation (Figure 7A) do also contain concentration effects due to water evaporation, the dry-matter data in cluster 2 (Figure 7B) clearly demonstrate that these compounds are generated with increasing maturation time.

[115] In conclusion, sensory-guided fractionation of traditional balsamic vinegar from

Modena (TBV), followed by quantitative analysis and taste recombination

experiments led to the identification of the key sensometabolites, amongst which 5- acetoxymethyl-2-furaldehyde was discovered for the first time as a natural sweet taste modulator. Compared to TBV, balsamic vinegar of Modena (BV) differed

significantly by the increased concentration of acetic acid, the significantly lower concentrations of the sweet-modulating 5-acetoxymethyl-2-furaldehyde (6), the nonvolatile organic acids and polyphenols, and the lack of wood-derived ellagitannins

[116] Moreover, quantitative profiling of 37 sensometabolites contributing to sweetness, sourness, and astringency of balsamic vinegar revealed a comprehensive insight into the process-induced evolution of sensometabolites throughout a full-scale TBV manufacturing process including a "batteria" of eight casks, e.g. the sweet modulating 5-acetoxymethyl-2-furaldehyde (6) is proposed to be generated by the esterification of the Maillard reaction reduced calorie product 5-hydroxymethyl-2-furaldehyde with the fermentation reduced calorie product acetic acid upon maturation in the "batteria" (Figure 8). The data obtained by means of hierarchical analysis and sensomics heatmapping offers the scientific basis for a knowledge-based optimization of the taste profile of TBV by technological means.

EXAMPLE 6

Taste Experiments

[117] Three beverage samples were prepared:

1. 1.5 mM 5-Acetymethyl-2-Furanaldehyde in CSD I base w/o carbonation

2. 1.5 mM 5-Acetymethyl-2-Furanaldehyde + 4% Sucrose in CSD I base w/o carbonation

3. 4% Sucrose in CSD I base w/o carbonation

[118] The samples were tasted by seven panelists. The results are shown in the table below..

LITERATURE CITED

(1) Giudici, P.; Gullo, M; Solieri, L.; Falcone, P.M. Technical and Microbial Aspects of Traditional Balsamic Vinegar and Their Influence on Quality and Sensorial Properties. Adv. Food Nutr. Res. 2009, 58, 137-182.

(2) Cocchi, M.; Durante, C; Grandi, M.; Lambertini, P.; Manzini, D.; Marchetti, A.

Simultaneous Determination of Sugars and Organic Acids in Aged Vinegars and Chemometric Data Analysis. Talanta 2006, 69, 1 166-1 175.

(3) Plessi, M.; Monzani, A.; Coppini, D. Determination of the Monosaccharide and Alcohol Content of Balsamic and other Vinegars by Enzymatic Methods. Agric. Biol. Chem. 1988, 52, 25-30.

(4) Cocchi, M.; Ferrari, G.; Manzini, D.; Marchetti, A.; Sighinolfi, S. Study of the

Monosaccharides and Furfurals Evolution during the Preparation of Cooked Grape Musts for Aceto Balsamico Tradizionale Production. J. Food Eng. 2007, 79, 1438- 1444.

(5) Plessi, M.; Monzani, A.; Coppini, D. Quantitative Determination of Acids and

Derivatives in Balsamic and other Vinegars. Sciences des Aliments 1989, 9, 179-183.

(6) Sanarico, D.; Motta, S.; Bertolini, L.; Antonelli, A. HPLC Determination of Organic Acids in Traditional Balsamic Vinegar of Reggio Emilia. J. Liq. Chromatogr. Relat. Technol. 2003, 75, 2177-2187.

(7) Cocchi, M.; Lambertini, P.; Manzini, D.; Marchetti, A.; Ulrici, A. Determination of Carboxylic Acids in Vinegars and in Aceto Balsamico Tradizionale Di Modena by HPLC and GC Methods. J. Agric. Food Chem. 2002, 50, 5255-5261 .

(8) Plessi, M.; Bertelli, D.; Miglietta, F. Extraction and Identification by GC-MS of

Phenolic Acids in Traditional Balsamic Vinegar from Modena. J. Food Comp. Anal. 2006, 19, 49-54.

(9) Verzelloni, E.; Tagliazucchi, D.; Conte, A. Relationship between the antioxidant

properties and the phenolic and flavanoid content in traditional balsamic vinegar. Food Chem. 2007, 105, 564-571.

(10) Chinnici, F.; Masino, F.; Antonelli, A. Determination of Furanic Compounds in

Traditional Balsamic Vinegars by Ion-Exclusion Liquid Chromatography and Diode- Array Detection. J. Chromatogr. Sci. 2003, 41, 305-310. (1 1) Theobald, A.; Miiller, A.; Anklam, E. Determination of 5-Hydroxymethylfurfural in Vinegar Samples by HPLC. J. Agric. Food Chem. 1998, 46, 1850-1854.

(12) Blasi, F.; Riccieri, R.; Brutti, M.; Del Signore, A.; Simonetti, M.S.; Cossignani, L.;

Damiani, P. HRGC-MS of free enantiomeric amino amino acids in aged balsamic vinegars from Modena. J. Commodity Sci. 2002, 41, 334-346.

(13) Masino, F.; Chinnici, F.; Bendini, A.; Montevecchi, G.; Antonelli, A. A Study on

Relationships among Chemical, Physical, and Qualitative Assessment in Traditional Balsamic Vinegar Food Chem. 2008, 106, 90-95.

(14) Del Signore, A. Chemometric Analysis and Chemical and Physical Parameters of

Traditional Balsamic Vinegars from Modena J. Commodity Sci. 2001, 40, 33-48.

(15) Falcone, P.M.; Chillo, S.; Giudici, P.; Del Nobile, M.A. Measuring rheological

properties for applications in quality assessment of Traditional Balsamic Vinegar:

Description and preliminary evaluation of a model. J. Food Eng. 2007, 80, 234-240.

(16) Masino, F.; Chinnici, F.; Franchini, G.C.; Ulrici, A.; Antonelli, A. A Study of the

Relationships among Acidity, Sugar and Furanic Compound Concentrations in Set of Casks for Aceto Balsamico Tradizionale of Reggio Emilia by Multivariate Techniques. Food Chem. 2005, 92, 673-679.

(17) Caligiani, A.; Acquotti, D.; Palla, G.; Bocchi, V. Identification and quantification of the main organic compounds of vinegars by high resolution Ή-NMR spectroscopy. Anal. Chim. Acta 2007, 585, 1 10-1 19.

(18) Cirlini, M.; Caligiani, A.; Palla, G. Formation of glucose and fructose acetates during maturation and ageing of balsamic vinegars. Food Chem. 2009, 112, 51 -56.

(19) Cocchi, M.; Bro, R.; Durante, C; Manzini, D.; Marchetti, A.; Saccani, F.; Sighinolfi, S.;

Ulrici, A. Analysis of sensory data of Aceto Balsamico Tradizionale di Modena

(ABTM) of different ageing by application of PARAFAC models. Food Oiial. Prefer. 2006, 77, 419-428.

(20) Grohnke, G. Decoding the aroma of Balsamic Vinegar and Parmigiano Reggiano

Cheese by concepts of molecular sensory science. PhD thesis, Technische Universitat Munchen, 2010.

(21) Hufnagel, J.C.; Hofmann, T. Orosensory-Directed Identification of Astringent

Mouthfeel and Bitter- Tasting Compounds in Red Wine. J. Agric. Food Chem. 2008, 56, 1376-1386.

(22) Hufnagel, J.C.; Hofmann, T. Quantitative reconstruction of the nonvolatile

sensometabolome of a red wine. J. Agric. Food Chem. 2008, 56, 9190-9199. (23) Glabasnia, A.; Hofmann, T. Sensory-Directed Identification of Taste-Active

Ellagitannins in American (Quercus alba L.) and European Oak Wood (Quercus robur L.) and Quantitative Analysis in Bourbon Whiskey and Oak-Matured Red Wines. J. Agric. Food Chem. 2006, 54, 3380-3390.

(24) Toelstede, S.; Hofmann, T. Quantitative Studies and Taste Re-Engeneering

Experiments toward the Decoding of the Nonvolatile Sensometabolome of Gouda Cheese. J. Agric. Food Chem. 2008, 56, 5299-5307.

(25) Frank, O.; Ottinger, H.; Hofmann, T. Characterization of an intense bitter-tasting

1 H,4H-quinolizinium-7-olate by application of the taste dilution analysis, a novel bioassay for the screening and identification of taste-active compounds in foods. J. Agric. Food Chem. 2001, 49, 231-238.

(26) Scharbert, S.; Hofmann, T. Molecular definition of black tea taste by means of

quantitative studies, taste reconstitution, and omission experiments. J Agric. Food Chem. 2005, 55, 5377-5384.

(27) Stark, T.; Bareuther, S.; Hofmann, T. Sensory-guided decomposition of roasted cocoa nibs (Theobroma cacao) and structure determination of taste-active polyphenols. J. Agric. Food Chem. 2005, 53, 5407-5418.

(28) Dunkel, A.; Hofmann, T. Sensory-directed identification of β-alanyl dipeptides as contributors to the thick-sour and white-meaty orosensation induced by chicken broth. J. Agric. Food Chem. 2009, 57, 9867-9877.

(29) Rotzoll N., Dunkel A., Hofmann T. Quantitative studies, taste reconstitution, and

omission experiments on the key taste compounds in morel mushrooms (Morchella deliciosa Fr.). J. Agric. Food Chem. 2006, 54, 2705-271 1.

(30) Toelstede, S.; Hofmann, T. A Series of Kokumi Peptides Impart the Long-Lasting Mouthfulness of Matured Gouda Cheese. J. Agric. Food Chem. 2009 57, 1440-1448.

(31) Dunkel, A.; Koster, J.; Hofmann, T.; Molecular and Sensory Characterization of γ- Glutamylpeptides as Key Contributors to the Kokumi Taste of Edible Beans (Phaseolus vulgaris L.). J. Agric. Food Chem. 2007, 55, 6712-6719.

(32) Sonntag, T.; Kunert, C; Dunkel, A.; Hofmann, T. Sensory-guided identification of N- (l-methyl-4-oxoimidazolidin-2-ylidene)-a-amino acids as contributors to the thick-sour taste and mouth-drying orosensation of stewed beef juice. J. Agric. Food Chem. 2010, 58, 6341-6350. (33) Festring, D.; Hofmann, T. Discovery of N2-(l-carboxyethyl)guanosine 5'- monophosphate as an umami-enhancing Maillard-modified nucleotide in yeast extracts. J. Agric. Food Chem. 2010, 58, 10614-10622.

(34) Glabasnia, A.; Hofmann, T. Identification and Sensory Evaluation of Dehydro- and Deoxy-ellagitannins Formed upon Toasting of Oak Wood (Quercus alba L.). J. Agric. Food Chem. 2007, 55, 4109^11 18.

(35) Analysis of foods - Determination of the dry mass of sugar varieties (vacuum drying) (reference method). Official collection of analysis methods § 64 LFGB (in German) 2002 L 39.00-2.

(36) Schabert, S.; Holzmann, N.; Hofmann, T. Identification of the astringent taste

compounds in black tea by combining instrumental analysis and human bioresponse. J Agric. Food Chem. 2004, 52, 3498-3508.

(37) Analysis of foods - Sensorial Test Procedures; Triangle Test (according to DIN EN ISO 4120). Official collection of analysis methods § 64 LFGB (in German) 2007 LOO.90-7.

(38) R Development Core Team. R: A language and environment for statistical computing;

R Foundation for Statistical Computing: Vienna, Austria, 2008; URL R-project.org

(39) Sokal, R. R.; Michener, C. D. A statistical method for evaluating systematic

relationships. Sci. Bull. 1958, 38, 236-244.

(40) Cerezo, A.B.; Espartero, J.L.; Winterhalter, P.; Garcia-Parrilla, M.C.; Troncoso, A.M.

(+)-Dihydrorobinetin: a Marker of Vinegar Aging in Acacia (Robinia pseudiacacia) Wood. J. Agric. Food Chem. 2009, 57, 9551-9554.

(41) Wieser, H.; Jugel, H.; Belitz, H. D. Relationships between structure and sweet taste of amino acids. Z Lebensm.-Unters.-Forsch. 1977, 164, 277-282.

(42) Warendorf, T. Taste-active compounds in bouillon (in German). PhD thesis, Technische Universitat Miinchen, 1991.

(43) Warmke, R. Identification of taste compounds in Emmental cheese and determination of concentration changes during ripening (in German). PhD thesis, Technische

Universitat Miinchen, 1997. Table 1. Taste Profile Analysis of 1+2 Dilutions of Traditional Balsamic Vinegar of Modena (TBV), Aqueous Solutions of the Low Molecular Weight (TBV-LMW, <5kDa) and the High Molecular Weight Fraction (TBV-HMW, >5 kDa) obtained by

Ultrafiltration, and of Balsamic Vinegar of Modena (BV).

Intensity for individual taste quality in ^a taste quality TBV TBV-LMW TBV-HMW BV sweet 2.3 2.4 (±0.3) 0.2 (±0.2) 1.5 (±0.2) sour 3.6 1.8 (±0.3) 0.3 (±0.3) 4.3 (±0.3) bitter 0.6 0.5(±0.1) 0.0(±0.0) 0.8(±0.2) astringent 2.2 1.4(±0.3) 0.9(±0.3) 2.5(±0.3) mouthfulness 1.7 1.1(±0.3) 0.2(±0.2) 1.0(±0.3) umami 0.2 0.2(±0.1 ) 0.0(±0.0) 0.2(±0.1 ) salty 0.2 0.2(±0.1 ) 0.2(±0.2) 0.2(±0.1 ) tensities were judged on a linear scale from 0 (no taste impression) to 5 (strong taste impression) by twelve trained panelists. The data is given as mean of triplicates; std. dev. is given in parenthesis.

Table 2. Taste Qualities, Taste Recognition Thresholds,

Concentrations, and Dose-over-Threshold (DoT) Factors of

Nonvolatile Sensometabolites in Traditional Balsamic Vinegar of

Modena (TBV) and Balsamic Vinegar of Modena (BV). taste compound TC ^b [μπιοΙ/L] cone. [μιηοΙ/L] (RSD in %) DoT ^d

TBV BV TBV BV

group I: sweet tasting compounds

fructose 10200 ^e 1601 1 1 1 (±0.6) 642583 (±0.0) 157.0 69.1 glucose 18000 ^s 1818333 (±0.3) 704760 (±0.0) 101 .0 35.7 glycerol 81 100 ^e 134377 (±0.6) 55191 (±1.2) 1.7 0.7

L-proline 25000 ^f 10830 (±6.2) 12458 (±0.5) 0.4 <0.1 inositol 17700 ^e 5372 (±0.4) 1298 (±3.8) 0.3 0.1 sorbitol 33800 ^e 10534 (±3.2) 490 (±9.7) 0.3 <0.1 erythritol 36300 ^e 6612 (±9.7) 1638 (±9.6) 0.2 <0.1 xylitol 12500 ^e 1766 (±5.7) 358 (±6.3) 0.1 <0.1 mannitol 40000 2900 (±3.8) 23 16 (±4.0) <0.1 0.1 arabitol 43 100 2200 (±3.4) 536 (±3.5) <0.1 <0.1 ribitol 45300 1400 (±4.9) 456 (±8.9) <0.1 <0.1

L-methionine 5000 ^g 17 (±3.9) 54 (±3.0) <0.1 <0.1

L-alanine 12000 ^f 563 (±8.9) 1042 (±7.8) <0.1 <0.1

L-serine 25000 ^f 575 (±5.5) 4722 (±5.4) <0.1 <0.1 glycine 25000 ^f 268 (± 10.3) 837 (± 1.7) <0.1 <0.1

L-threonine 35000 ^f 19 (±6.9) 61 (±4.8) <0. l <0. 1 group II: sour tasting compounds

tartaric acid 292 ^e 31450 (±1.2) 10466 (±0.3) 107.7 35.8 gluconic acid 900 60302 (±3.4) 4342 (±5.7) 67.0 4.8 glycolic acid 600 23946 (±4.3) 3325 (±9.2) 39.9 5.5 malic acid 3690 ^e 1073 15 (±0.4) 19366 (±3.5) 29.0 5.2 acetic acid 19900 ^e 371 161 (± 1.0) 649564 (±0.8) 18.7 32.6 citric acid 2600 ^e 1 1929 (± 1.3) 2282 (±8.0) 4.6 0.9 succinic acid 900 ^e 2420 (±8.7) n.d. 2.7 n.d. lactic acid 15480 ^e 7302 (±3.9) 9786 (±4.4) 0.5 0.6 group III: bitter tasting compounds

calcium 6200 ^{, k} 24578 (±3.8) 10101 (±0.2) 4.0 1.6 magnesium 6400 ^ilk 20024 (±1.5) 8335 (±2.6) 3.1 1.3

L-arginine 75000 7691 (±0.7) 14322 (±5.4) 0.1 0.8

L-leucine 1 1000 ⁶ 143 (±7.3) 531 (±2.3) <0.1 <0. 1

L-tyrosine 4000 ^g 77 (±6.0) 281 (±5.2) <0.1 <0.1

L-isoleucine 10000 ^s 193 (±2.7) 453 (±8.4) <0.1 <0.1

L-valine 30000 ¹ 226 (±6.0) 410 (±4.9) <0.1 <0.1

L-phenylalanine 45000 ^s 88 (±4.2) 336 (± 1.4) <0.1 <0.1

L-histidine 45000 ^g 29 (±2.9) 175 (± 1.2) <0.1 <0.1 group IV astringent compounds

castalagin (2) 1.1 ^m 51 (± 1.1 ) n.d. 46.4 n.d. vescalagin (1) 1.1 ^m 38 (±4.4) n.d. 34.5 n.d.

5-hydroxymethylfurfural 10000 ^m 29419 (± 1.8) 8681 (±5.1) 2.9 0.9 trans-caffeic acid 72 ^m 29 (±2.7) 23 (±0.5) 0.4 0.3 gentisic acid 122 ^m 38 (± 1.9) 10 (± 1.4) 0.3 <0.1 coumaric acid 139 ^m 35 (±2.5) 23 (±0.9) 0.2 0.2 gallic acid 292 ^m 32 (± 10.6) 17 (±5.7) 0.1 <0.1 p-hyroxybenzoic acid 665 ^m 17 (± 1.8) 9.1(±0.0) <0.1 <0.1 quinic acid 579 15 (±2.5) 1.3 (± 1.5) <0.1 <0.1 protocatechuic acid 206 ^m 10 (±3.8) 0.7 (±3.5) <0.1 <0.1 vanillic acid 315 ^m 20 (±8.2) 6.3 (±4.9) <0.1 <0.1 ferulic acid 67 ^m 2.4 (±9.4) 1.8 (±2.4) <0.1 <0. 1 vanilline 829 ^m 7.2 (± 1 .7) 0.7 (±9.8) <0.1 <0.1 gallic acid methyl ester 232 ^m 3.3 (±3.5) 1.6 (± 1.1 ) <0.1 <0.1 gallic acid ethyl ester 185 ^m 2.9 (±5.8) 46 (±9.3) <0.1 <0.1 syringaldehyde 330 ^m 8.7 (±3.4) 0.4 (± 1.5) <0.1 <0.1

(+)-dihydrorobinetin (3) 23 1.1 (±1.8) n.d. <0.1 n.d. group V: saltv compounds

potassium 13000 ^{i k} 58025 (±3.6) 7131 1 (±4.6) 3.1 5.5 phosphate 5000 ^h '' 12230 (±2.9) 7366 (±2.1 ) 2.4 1.5 sodium 3900 ^{i k} 7608 (±9.8) 14667 (±2.1 ) 2.0 3.8 chloride 3900 '·' 6025 (±2.0) 3640 (±1. 1) 1.5 0.7 group VI: umami-like compounds

L-aspartic acid 600 ' 528 (± 12.8) 1776 (±9.5) 0.9 3.0

L-glutamic acid 1 100 ' 179 (±8.3) 591 (±0.6) 0.1 0.5 group VII: astringent polymers

HMW-fraction (>5kDa) n.d. 8.7 g/L n.d. n.d. n.d. group VIII: acetvlated compounds

12300 ° 0.7 0.2

4 8329 (±7.4) 2468 (±8.7)

12300 ^p 0.7 0.2

21200 ° 0.7 0.2

5 13823 (±4.8) 5643 (±7.5)

16900 ^p 0.8 0.3

6 1500 " 315 (±3.1) 76 (±6.9) 0.2 <0. 1 a Taste-active compounds were determined in TBV and

BV, if not stated otherwise; ^b Taste threshold

concentrations (TC) are given as the mean of triplicates

in bottled water and were determined by means of a three

alternative forced choice test, or were taken from

literature; ^c Concentration (μιηοΙ/L) in TBVM; ^d Dose- over-threshold (DoT) factor is calculated as the ratio of

concentration and taste threshold; ^e Value taken from

(26); ^f Value taken from (41); ^g Value taken from (42); ^h

Value taken from (43); ' Value taken from (24); ^k TC

determined for the corresponding chloride salt; ¹ TC

determined for the corresponding sodium salt; ^m Value

taken from (20); " TC for sweet taste enhancement

determined in the 4% sucrose solution; ⁰ TC for bitter

taste; ^p TC for sweet taste. Table 3. Sensory Evaluation of Traditional Balsamic Vinegar of Modena (TBV, 1 +2 diluted) and Balsamic Vinegar of Modena (BV, 1+2 diluted) and the Corresponding Taste Recombinants rTBV and rBV (each 1+2 diluted intensity for individual taste quality ³ in taste quality TBV rTBV,.vi ^b rTBV,.v„ ^b rTBV _total ^b BV ^r TB _t0 ia| ^b sweet 2.3 2.0(±0.3) 2.1 (±0.2) 2.4(±0.2) 1.5 1.3(±0.2) sour 3.6 3.8(±0.3) 3.4(±0.2) 3.6(±0.2) 4.3 4.2(±0.3) bitter 0.6 0.5(±0.1 ) 0.5(±0.1 ) 0.5(±0.1 ) 0.8 0.8(±0.2) astringent 2.2 2.0(±0.1) 2.1 (±0.1 ) 2.2(±0.2) 2.5 2.2(±0.3) mouth fulness 1.7 1.1 (±0.2) 1.6(±0.2) 1.6(±0.3) 1.0 1.0(±0.2) umami 0.2 0.2(±0.1 ) 0.2(±0.1 ) 0.2(±0.1 ) 0.2 0.2(±0.1) salty 0.2 0.2(±0.1 ) 0.2(±0.1 ) 0.2(±0.1) 0.2 0.2(±0.1 ) ^a Intensities were judged on a scale from 0 (not detectable) to 5 (strongly detectable) by twelve trained panelists. The data is given as the mean of triplicates. ^b rTBV[.vi contained the tastants in groups I-VI in concentrations given in Table 2 in water (pH 3.0); rTBVj.vi was prepared by dissolving the HMW fraction (8.7 g/L) in rTBV _I-V i; rTBV, ₀ , _a i was prepared by spiking rTBV|. _V u w tastant group VIII (compounds 4-6) in concentrations given in Table 2; rBV, _ota i contained the tastants in groups I-VIII in concentrations given in Table 2 in water (pH 3.0).

Table 4. Taste Dilution Analysis (TDA) of GAC-Fractions I-X Dissolved in Bottled Water and Comparative Taste Profile Analysis (cTPA) of GAC Fractions I-X Dissolved in Basic Taste Recombinant (rTBVi.vi).

Fraction TDA in water ^b cTPA in TTBVLVI ^C

No ^a TD factor taste quality change in taste quality

I <1 n.d. n.d.

II 16 sour n.d.

16 astringent

1 bitter

III 64 sour n.d.

2 bitter

rv 32 astringent n.d.

16 sour

16 bitter

V 32 astringent n.d.

16 sour

16 bitter

VI 16 sour n.d.

4 bitter

2 astringent

1 sweet

VII 256 sweet increased sweetness ^d

16 sour

VIII 16 sour increased sourness ^d

8 astringent

2 bitter

1 sweet

IX 128 sour increased sourness and

64 astringent astringency ^d

X 16 sour more long-lasting sweetness ^d

8 astringent

4 bitter

a Numbering of GAC fractions corresponds to Figure 3. ^b Taste dilution analysis (TDA) was carried out after dissolving the individual GAC fractions in bottled water (pH 3.0) in their "natural" concentration ratios. ^c The individual GAC fractions were dissolved in a 1 -2 dilution of the basic taste recombinant solution rTBV|. _V i (pH 3.0) containing all tastant groups I-VI given in Table 2. The descriptors given by each panellist were collected and those given by at least nine out of the twelve panellists are given. The rTBVi.vi solution lacking any GAC fractions was used as control, n.d. no difference detectable. ^d p<0.05