The Contribution of Bitter Blockers and Sensory Interactions to Flavour Perception
There is a continued need for the application of flavour modifiers to improve the sensory profile of products within the functional food market. Additionally, psychophysical studies have tended to confine their scope to stimuli that elicit single sensations, and ingredients that are not always of most interest to the food industry. While basic taste-eliciting compounds and odourants have been used in functional food optimisation, modification can also include the addition of bitter-blocking ingredients. This study examines the impact that these modifiers have both alone and in conjunction with each other on the flavour of (+)-catechin containing model functional beverages.
The intensities of sweetness, bitterness, astringency and aroma were rated for (+)-catechin (CAT) aqueous solutions alone and containing a sweetener [sucrose or rebaudioside A (REB)], an odourant (vanilla or black tea), a bitter blocker [ß-cyclodextrin (CD) or homoeriodictyol sodium salt], and all combinations of each.
The use of sweeteners, both alone and in conjunction with bitter blockers, decreased the bitterness of CAT, while odourants had no effect. CD + REB significantly decreased the astringency of CAT. Astringency and bitterness of CAT was not altered by the addition of bitter blockers alone or in combination with odourants. Bitter blockers did not affect intensities of sweetness and aroma.
The use of sweeteners in combination with bitter blockers can lower the bitterness of (+)-catechin. The addition of bitter blockers may be used without a detrimental effect on the flavour profile of model beverages.
Decreasing the bitterness of plant-derived, health-promoting compounds may be achieved through the application of certain sweet eliciting and bitter-blocking compounds, which in turn, may lead to increasing the acceptability of some functional foods for bitter sensitive consumer populations.
KeywordsBitterness Bitter blockers Sensory interactions Flavour Functional foods
The flavour of functional food products is an important driver in both consumer acceptance and purchasing behaviour (Siró et al. 2008). However, due to the addition of bioactive ingredients, flavour can be compromised in such foods. For example, at higher concentrations, the addition of plant-derived compounds can alter bitterness and astringency (Gomez-Carneros and Drewnowski 2000), which may in turn lead to a decrease in the consumer acceptance (Lesschaeve and Noble 2005). Thus, a major challenge facing product developers is creating products that maintain a sensory profile that is acceptable to consumers while still containing an adequate level of ingredients to fulfill bioactivity and/or health claim requirements.
Flavour is a complex perceptual and cognitive phenomenon that involves the combination of various sensations including taste, smell and mouthfeel. Numerous interactions can occur between these sensations, and thus, changes in one component that impacts a sensory modality may affect another modality when part of a complex mixture and modify the overall perceived flavour (Delwiche 2004). Within a food matrix, interactions can occur within a sensory modality (e.g. taste-taste, odour-odour) or across sensory modalities (e.g. taste-odour, taste-mouthfeel). Overall, interactions within the same sensory modality are more effective at altering flavour compared to those across modalities (Gillan 1983). The flavour of food and beverage matrices arises from a combination of physical chemistry, peripheral physiological mechanisms and central cognitive mechanisms. Taste interactions occur both at the periphery (explained below) as well as centrally, where the intensity of one tastant in a mixture is perceived as suppressed or enhanced independent of peripheral mechanisms and physical chemistry within the oral cavity (Keast and Breslin 2002). For example, through split-tongue studies, the addition of sucrose (sweet tasting) to a quinine solution (bitter tasting) results in the mutual suppression of sweetness and bitterness (Kroeze and Bartoshuk 1985). A central cognitive effect can also occur between taste and smell, where the perceived sweetness of sucrose is enhanced by the addition of a congruent odour, such as strawberry (Frank and Byram 1988). However, taste-odour pairings that are typically not associated with each other often do not result in taste enhancement (e.g. sweet taste and peanut butter odour) (Frank and Byram 1988, but see Delwiche and Heffelfinger 2005). Odour-induced enhancement of taste appears to be based on the prior associations made during the previous food and beverage exposure (Delwiche 2004), and thus occur due to neural-regulated, perceptual processes (Small and Prescott 2005).
The sensory interactions may also be based on the peripheral mechanisms (Keast and Breslin 2002). The suppression of bitterness by saltiness is an example of a peripheral physiological effect (Breslin and Beauchamp 1995, 1997). Salt-induced suppression of bitterness may be due to a number of reasons, including alteration of the integrity of the TAS2R receptor cell membrane or interaction with the TAS2R taste transduction pathway (Keast and Breslin 2002). While the sensory interactions have been studied in several conventional foods and beverages, how these interactions impact functional foods and beverages has yet to be explored. Formulations for these products are continually evolving, and often include the addition of compounds that reduce bitterness, or ‘bitter blockers.’ Some bitter blockers lower the perception of bitterness by encapsulating a bitter compound, creating a physical barrier that reduces access of the compound to bind to the taste receptor (Gaudette and Pickering 2013). Other bitter blockers interact with the extracellular polypeptide chain of TAS2Rs (Gaudette and Pickering 2013). While bitter blockers may be effective in simple systems, their addition into a food mixture containing multiple sensory stimuli may alter their bitter inhibiting capacity, and/or impact the interactions that occur between these stimuli, and thus affect the overall perceived flavour. The sensory interactions existent when bitter blockers are used in functional foods and their impact on the overall flavour profile have yet to be examined.
There are two main objectives of this study: (1) to investigate the impact of within- (taste-taste) and cross- (taste-odour) modal sensory interactions on the overall flavour profile of (+)-catechin-containing aqueous solutions and (2) to determine the impact of bitter blockers on the overall flavour profile of (+)-catechin-containing aqueous solutions.
(+)-Catechin represents a useful ‘model’ molecule for the investigation of bitter compounds with functional properties as it is relatively inexpensive, is currently used in functional food/beverage formulations (Gaudette and Pickering 2013), and elicits both bitterness and astringency at relatively low concentrations (Peleg et al. 1999; Gaudette and Pickering 2012). We hypothesize that a within modality interaction, specifically taste-taste (bitter-sweet), will be more effective than cross-modality interactions, specifically taste-odour (bitter-vanillin), at decreasing the bitterness of (+)-catechin. In binary taste-odour matrices, we expect sweet-associated odours to enhance sweetness and suppress bitterness. For bitter associated odours, we expect the opposite effect. In ternary matrices, two different tastants will be present (bitter and sweet) in addition to an odourant. We anticipate that vanilla, a ‘sweet’ aroma, will enhance sweetness, and as a result, decreases the perceived bitterness of (+)-catechin. We hypothesize that black tea, an aroma associated with bitterness, will lead to an odour-induced enhancement of bitterness and suppression of sweetness.
We further hypothesize that the addition of bitter blockers to aqueous solutions of (+)-catechin containing sweeteners or odourous compounds will alter the perceived sweetness and aroma intensity of these solutions. While the addition of bitter blockers that target TAS2Rs should decrease the bitterness of (+)-catechin, which in turn, may increase the intensity of sweetness and aroma, bitter blockers that encapsulate bitterants may also bind other chemicals, such as sweeteners and odourants, resulting in a decrease in perceived sweetness and odour intensity. Thus, the addition of bitter blockers to complex matrices can have an impact on the overall flavour profile that is difficult to predict. This investigation will serve to guide the use of these flavour modifiers within the functional food industry.
Material and Methods
Selection of Stimuli
Stimuli administered during data collection
CAT + sweetener OR bitter blocker OR odourant
CAT + sweetener + bitter blocker
CAT + sweetener + odourant
CAT + bitter blocker + odourant
CAT + sweetener + bitter blocker + odourant
Sucrose (SUC; Sigma-Aldrich) and rebaudioside A (REB; PureCircle USA, Inc.) were selected as sweeteners, representing a traditional sweetener and a plant-based, high-potency sweetener, respectively. Concentrations were derived from Schiffman et al. (1995), and the isosweetness of these levels was confirmed via bench top tasting (n = 7, data not shown) (Table 1). Vanillin (V; Sigma-Aldrich) and black tea flavour (T; Firmenich, Inc., NJ, USA) were chosen to represent aromas cognitively associated with sweetness and bitterness/astringency, respectively. The concentration of black tea flavour was initially based on the manufacturer’s recommendations, and adjusted after bench tasting to be isointense with vanillin, which was solubilized in water at ambient temperature.
Samples and Preparation
Treatments administered during data collection sessions consisted of aqueous solutions of (+)-catechin alone (control), or as part of binary, ternary or quaternary mixtures with sweeteners, bitter-blockers and/or odourants (Table 1). All solutions were prepared using fresh deionized water (Millipore RiOs 16 Reverse Osmosis System, MA, USA). A one milliliter of 99 % pure food grade ethanol (Liquor Control Board of Ontario, St. Catharines, Ontario, Canada) was used to fully solubilize (+)-catechin prior to its addition to solutions made in 100 mL volumetric flasks. This concentration of ethanol is below the perception threshold in water (Thorngate and Noble 1995). Samples were prepared in foil-wrapped volumetric flasks to protect (+)-catechin from possible light-induced polymerization (Peleg et al. 1999). The samples were then transferred to airtight, 120 mL amber coloured glass bottles (Fisher Scientific, Rockford, IL, USA). The headspace was filled with nitrogen gas, and the samples were stored in darkness at 4 °C and were replaced every 5 days. Based on our previous work (Gaudette and Pickering 2012), (+)-catechin is stable and remains in monomeric form during these storage conditions for at least 5 days.
Participants, Screening and Training
The participants were staff or students of the university (n = 15, 34 ± 10 years, seven males) between the ages of 22 and 52 who volunteered their time. All training and data collection sessions were held in the university’s dedicated sensory evaluation lab. Screening and training of participants involved correctly identifying and rating the intensity of eight aqueous solutions, each containing a taste (sweet, sour, salty, bitter and umami), astringent, or in-mouth aroma (tea, vanilla) stimulus. Compounds and concentrations were based on the previous literature (Keast 2003; Green and George 2004; Pickering et al. 2006) and modified as needed through bench testing; sucrose (sweetness, 3.0 × 10−1 M), citric acid (sourness, 3.0 × 10−3 M), sodium chloride (saltiness, 1.5 × 10−1 M), l-glutamic acid monosodium salt hydrate (umami, 1.0 × 10−1 M), quinine monohydrochloride dihydrate (bitterness, 5.0 × 10−5 M), aluminum sulfate (astringency, 4.4 × 10−4 M), vanillin (vanilla aroma, 6.6 × 10−4 M) and black tea flavour (black tea aroma, 1.0 mL/L).
In order to familiarize the participants with the scale usage, intensity ratings were recorded on a 15 cm visual analog scale with 1 cm indented labels (bottom label of ‘absent’, top label of ‘high’). The participants were first presented a single flight of solutions in International Organization for Standardization (ISO) wine tasting glasses capped with plastic lids. Each glass contained 20 mL of a single stimulus, and was labeled with the appropriate sensation. Instructions for tasting and rating were the following: remove the lid from the glass, take the full amount of solution into the mouth, swirl for 5 s, expectorate the sample, wait at least 5 s for sensation to fully develop, rate the highest intensity experienced for the sensation on the line scale, rinse thoroughly four times with water, wait a minimum of 1 min between samples. After a mandatory 10-min break, a second flight of the same solutions was randomly presented in glasses labeled with 3-digit randomized codes and capped with plastic lids. The same instructions for tasting and rating were followed, and in addition, the participants were asked to correctly identify the sensation. Screening and training were considered completed when the participants correctly identified and rated each stimulus presented in the second flight. An error in correctly identifying any sensation resulted in the participant being invited back for an additional training and screening session. If an error in identification occurred during this additional session, the participant was thanked and excused from the study.
Design and Data Analysis
A restricted randomized block design was used. Replication was blocked, with initial and duplicate assessments presented in randomized order. All possible combinations of (+)-catechin together with each sweetener, bitter blocker and odourant were assessed as binary, ternary or quaternary solutions. Thus, 26 different solutions and one control [(+)-catechin] were presented to all the participants, in duplicate. These 54 samples were presented over nine separate sessions, with each session consisting of two flights of three samples. Forced breaks of 3 and 10 min, along with four water rinses, were enforced between samples and flights, respectively. Samples were removed from the refrigerator 1 h prior to the testing. Twenty milliliter of each sample were then poured into ISO wine glasses and labeled with randomized 3-digit numbers.
The participants’ responses were collected using a computerized program (Compusense c5v4, 111 Farquhar St., Guelph, Ontario, Canada, N1H 3N4). Five 15 cm line scales with 1 cm indented labels (bottom labels absent, top labels high) were presented for each sample on a single screen, and each sample was rated for intensity of bitterness, astringency, sweetness, in-mouth aroma and ‘other’. All evaluations took place under red lighting to mask any visual differences between the samples. For each sample, the following tasting protocol/instructions were strictly adhered to: remove lid from the glass, take the full solution into the mouth, swirl in mouth for 5 s, expectorate the sample, wait 10 s for sensations to fully develop (participants were verbally instructed to wait longer if intensity was still building after 10 s), rate the highest intensity experienced for each sensation, rinse thoroughly with water at least four times during the forced 3-min break, continue onto the next sample only when mouth is completely free from all taste, tactile and in-mouth aroma sensations.
Data Treatment and Statistical Analyses
Statistical analysis was performed using XLSTAT version 2011.1.01 for Apple Macintosh (Addinsoft, USA). Initial analyses were performed to determine individual participant performance and to identify possible outliers. Three criteria were used to assess participant performance for bitterness responses: reproducibility between replicates (coefficient of variation), bitterness sensitivity (intensity ratings for (+)-catechin) and discrimination (p(F) value from ANOVA), as detailed in Gaudette and Pickering (2012). If a participant failed in two or more of these criteria, they were removed from the dataset (n = 2).
Three separate two-way ANOVAs were then conducted to test the main effects of the following on the perception of (+)-catechin and additional stimuli in aqueous solutions: taste + odour interactions [binary and ternary solutions containing REB, SUC, V and T alone, and in all possible taste-odour combinations were included in the ANOVA model], taste + bitter blocker interactions [binary and ternary solutions containing REB, SUC, HD and CD alone, and in all possible taste-bitter blocker combinations were included in the ANOVA model], bitter blocker + odour interactions [binary and ternary solutions containing HD, CD, T and V alone, and in all possible bitter blocker-odour combinations were included in the ANOVA model]. The dependent variables for each of these analyses were the intensity ratings of bitterness, astringency, sweetness or in-mouth aroma. For each participant, the averaged ratings across both replicates were used. For each initial ANOVA, treatment (solutions corresponding to taste + odour, taste + bitter blocker or bitter-blocker + odour groupings), participant and their interaction were included as the independent variables. ANOVA was repeated with the interaction term removed if not significant (p(F) < 0.05). Tukey’s HSD mean separation tests were used for post-hoc analyses (α = 0.05). To assist in the interpretation of figures, displayed intensity ratings, captured on a 15 cm visual analog scale, were converted to a score out of 100 prior to analyses.
The intensity of vanilla aroma was significantly increased by the addition of sucrose and rebaudioside A (p < 0.05). Sweetener addition, as expected, did not alter the intensity of black tea aroma (p < 0.05). In (+)-catechin containing solutions, the sweetness of sucrose increased with the addition of vanillin and black tea odourants (p < 0.05). However, the addition of these odourants did not affect the sweetness of rebaudioside A (p > 0.05). The odour-induced enhancement of sucrose sweetness did not result in a further suppression of bitterness compared to sucrose alone (p > 0.05). In addition, odourants alone did not alter the bitterness of (+)-catechin (p > 0.05). Compared to control [(+)-catechin alone], a trend of lower bitterness was observed in both binary and ternary solutions with added sweeteners. On average, all binary and ternary treatments (with bitter blockers or odourants) with added sweeteners resulted in a 48 % decrease in bitterness.
Taste/Astringency-Bitter Blocker Interactions
Rebaudioside A in a (+)-catechin solution elicited more sweetness when combined with ß-cyclodextrin than when combined with homoeriodictyol sodium salt (p < 0.0001). However, neither combination was perceived significantly sweeter compared to (+)-catechin + rebaudioside A alone (p > 0.05) (Fig. 2). The astringency of (+)-catechin was significantly decreased by the addition of rebaudioside A + ß-cyclodextrin (p < 0.05), although, this did not result in less astringency compared to rebaudioside A alone (p > 0.05). No other treatments were significant at modifying the astringency of (+)-catechin (p > 0.05). All combinations of bitter blockers + sweeteners decreased the bitterness of (+)-catechin (p < 0.0001), but this was not more effective than sweeteners alone (p > 0.05).
Odour-Bitter Blocker Interactions
The addition of bitter blockers and odourants to (+)-catechin aqueous solutions did not significantly modify the bitterness and astringency of (+)-catechin (p < 0.05). The addition of bitter blockers to (+)-catechin + odourants did not alter the aroma intensity of vanilla or black tea (p < 0.05) (data not shown).
We hypothesized that within (taste-taste) sensory interactions would be more effective at decreasing the bitterness of (+)-catechin than cross-modality (taste-odour) interactions. Overall, the results show that the application of sweeteners is more effective than odourants at decreasing (+)-catechin bitterness. We also anticipated that the sweet-associated odours would suppress bitterness and enhance sweetness; however, odourants had no impact on the perception of bitterness. While vanilla increased sweetness, this did not lead to less bitterness, as predicted. The main cognitive mechanism involved in such cross-modality interactions is the link with prior taste-odour associations. Thus, it is possible that vanilla-induced enhancement of sweetness did not lead to decreased bitterness due to the novelty of the model beverage. A similar result was found by Labbe et al. (2006), where a bitter milk beverage was not perceived as less bitter when a vanilla odourant was added. In addition, as cross-modal interactions are less able to affect perception compared to within-modal interactions (Gillan, 1983), sweetness enhancement by vanilla may not confer a strong enough cognitive effect to also alter the perception of bitterness.
Sucrose-induced enhancement of vanilla aroma is a phenomenon previously reported by Green et al. (2012). We demonstrated the capacity of both sucrose and rebaudioside A to increase vanilla aroma, providing further evidence of this cross-modal enhancement of in-mouth aroma. In addition, rebaudioside A—a plant-derived, alternative sweetener—confers a similar level of aroma boost compared to sucrose, which may be of value in some sweet-tasting food formulations. Sweet taste-induced enhancement of black tea aroma was not observed, most likely due to the lack of prior associations between these sensations in such a beverage system.
Astringency was significantly decreased by ß-cyclodextrin + rebaudioside A. It is possible that encapsulation of (+)-catechin by ß-cyclodextrin decreased the interaction of (+)-catechin and proline-rich salivary proteins in the oral cavity. Further reduction of astringency may have resulted through cognitive suppression of rebaudioside A. Thus, physical chemistry, peripheral physiology and central cognitive effects may all have contributed to the decrease in the astringency of (+)-catechin. Similarly, the application of bitter blockers and sweet-eliciting compounds results in a significant reduction in the bitterness of (+)-catechin, with both oral physiological and central cognitive effects implicated. While it is known that sucrose decreases bitterness via a central cognitive effect (Kroeze and Bartoshuk 1985), it is unclear whether rebaudioside A is effective through the same mode of action, and thus, further investigation is warranted. These bitter blockers decrease bitterness via oral physiological mechanisms, with ß-cyclodextrin likely encapsulating the (+)-catechin (Szejtli 1988), and homoeriodictyol sodium salt hypothesized to interfere with the extracellular portion of TAS2Rs (Ley et al. 2005).
Yamazaki et al. (2014) showed that catechin bitterness is transduced predominately by T2R39. Therefore, it is possible that the bitterness results reported here may not hold true for bitter compounds that activate other bitter receptors. Finally, the application of these bitter blockers did not alter the perceived aroma intensity of odourants, and the perception of sweeteners was not changed, suggesting that ß-cyclodextrin and homoeriodictyol sodium salt may be used in functional beverages with minimal alteration to these other sensations important to perceived quality and acceptance. Several of the stimuli used in this study (e.g. (+)-catechin and rebaudioside A) elicit more than one orosensation, and have therefore not been used in classic psychophysical studies. We encourage further basic and applied research on these more complex ingredients, particularly given their interest to the functional food industry.
Sensory interactions involving taste and smell are an integral part of the perception and modification of flavour. In functional food formulations, the addition of bitter blockers are an additional flavour modifying strategy that may contribute to and/or impact the various sensory interactions elicited in these matrices. Overall, the addition of ß-cyclodextrin and homoeriodictyol sodium salt to solutions containing sweeteners and odourants does not effect the perception of these stimuli. In contrast, sweeteners in particular, and vanilla aroma impact the flavour profile of (+)-catechin solutions by suppressing bitterness and enhancing sweetness, respectively. Future work could apply these results to an array of model and real functional foods in order to determine the best formulation for flavour optimization.
Dr. Jakob Ley of Symrise AG (Holzminden, Germany) is thanked for the donation of homoeriodictyol sodium salt. We are grateful to all the participants who participated in the sensory panels. We acknowledge and thank OMAFRA (Sustainable Production Systems Research Program) and The American Wine Society Educational Foundation for their financial assistance.
Compliance with Ethical Standards
The study was funded by an OMAFRA (Sustainable Production Systems Research Program) grant to the last author, and an American Wine Society Educational grant to the first author.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individual participants included in the study.
Conflict of Interest
The authors declare that they have no competing interests.
- Breslin PAS, Beauchamp GK (1995) Suppression of bitterness by sodium: Variation among bitter taste stimuli. Chem Sens 20:609–623Google Scholar
- Breslin PAS, Beauchamp GK (1997) Salt enhances flavour by suppressing bitterness. Nature 387:563Google Scholar
- Delwiche J, Heffelfinger AL (2005) Cross-modal summation in taste and smell. J Sens Stud 20:512–525Google Scholar
- Gomez-Carneros C, Drewnowski A (2000) Bitter taste, phytonutrients, and the consumer: a review. Am J Clin Nutr 72:1424–1435Google Scholar
- Lesschaeve I, Noble AC (2005) Polyphenols: factors influencing their sensory properties and their effects on food and beverage preferences. Am J Clin Nutr 81:330S–335SGoogle Scholar
- Pickering GJ, Haverstock G, DiBattista D (2006) Evidence that sensitivity to 6-n-propylthiouracil (PROP) affects perception of retro-nasal aroma intensity. J Food Agric Environ 4:15–22Google Scholar