Edited by:
Reviewed by:
*Correspondence:
This article was submitted to the journal Frontiers in Integrative Neuroscience.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Understanding the influence of taste perception on food choice has captured the interest of academics, industry, and the general public, the latter as evidenced by the extent of popular media coverage and use of the term supertaster. Supertasters are highly sensitive to the bitter tastant propylthiouracil (PROP) and its chemical relative phenylthiocarbamide. The well-researched differences in taste sensitivity to these bitter chemicals are partially controlled by variation in the
Taste sensitivity is relevant to our everyday lives. It is then no surprise that the role of taste variation in health has captured the interest of industry, health professionals, and the public alike. Substantial interest in recent years has centered on the term supertaster. It is paradoxically a broad superlative and a narrowly defined phenotype, complicating its general use within the public lexicon. This confusion is furthered due to its popular and continued use in the media and popular science communications to describe anyone with a sensitive palate (e.g.,
The most well-researched taste phenotype, the variable bitterness of PROP is largely due to three single nucleotide polymorphisms (SNPs) in the gene
In light of this, the Denver Museum of Nature & Science (Museum) felt both a need and a unique means to address this discordance through public participation in research. One of the historical challenges of human behavioral research on taste has been small sample sizes, coupled with labor-intensive data processing which often leads to conflicting findings such as the FP density-supertasting correlation. The Museum’s permanent health exhibit, Expedition Health, is open 364 days a year and sees over 400,000 visitors during that time; our access to a large cohort of human participants combined with our trained core of volunteers allowed us to conduct research using models of crowdsourcing and citizen science. Executed correctly, these models can advance scientific discovery while concurrently providing a source of awareness and engagement for a general audience via authentic and active research (
Participants in the “Bitter Study” hosted in the Genetics of Taste Lab are Museum guests who elect to enhance their visitor experience by participating in an authentic human genetics research study in the context of taste and health. Our study sample consists of 394 healthy, non-smoking participants who participated in every data collection station and subsequently provide full data sets of the required variables of age, sex, PROP intensity score, FP density, and
Volunteer citizen scientists underwent a 12 week certification program. The program includes detailed trainings on internal quality control for data collection, an online ethics course for working with human populations
DNA was extracted from Epicentre buccal swabs using the Maxwell 16 Buccal Swab LEV DNA Purification Kit and the Promega Maxwell. TAS2R38 was amplified using PCR primers (Forward ACCAATGCCTTCGTTTTCTTGGTGA, Reverse TCACAGCTCTCCTCAACTTGGCA, Invitrogen) and sequenced using the forward primer (High Throughput Genomics Center, Seattle, WA, USA, www.htseq.org). Sequences were analyzed using the program Geneious
We have a three step process in place for our genetic analysis to prevent inaccurate recording of variations to the gene TAS2R38. Step 1, staff scientists uploaded all sequences into the software program Geneious. The sequences were then aligned to the TAS2R38 reference sequence (AY258598) using the program option “Align, Map to Reference.” Following alignment, staff used the “Find Variations/SNPs” option to highlight the variations in the aligned sequences, and “Find Heterozygotes” to highlight heterozygotes at each variation. Step 2, once this preparation was performed by staff scientists, a small number of citizen scientists trained in chromatograph analysis worked in teams of two to record the diplotype for each sample. Citizen scientists only recorded samples where the chromatograph matched the computer program reading at that nucleotide position, and only if it showed one of the three main diplotypes (e.g., for SNP at position 145: G/G, C/C or G/C). Step 3, any samples that showed discrepancy from the chromatograph to the computer assignment, or showed any other nucleotide other than the two known variations were flagged and analyzed by staff scientists.
Filter disks were impregnated with a solution of 0.453 M PROP for the taste test (
Participants’ tongues were temporarily stained blue using ESCO Foods liquid color (deep blue shade), diluted 1:10 with deionized water. Participants steadied their head by placing their chin on fisted hands, with their elbows on the bench. They held a popsicle stick with their unique visitor identification number just to the side of their mouth. A paper disk with a 1 cm circular cut out was placed to the left of the center line of the tongue at the apical tip (
All papillae counts were completed using the free software ImageJ hosted by the National Institutes of Health
Twenty-six counters scored 15 photographs using the standard methodology in the field (
We used the following processes to ensure FP counts analyzed by our team of citizen scientists are usable. We had a total of 1005 photographs that were counted as part of the full study. We used a simple random sample to verify 10% of the photos (
We report a proportion of 0.81 with 95% Confidence Interval of 0.733≤ p≤ 0.887. In addition, after calculating the difference in counts between professional and citizen scientists on the random sample, we find the following numerical summary: the minimum difference between a citizen science counted photograph and a professional scientist counted photograph is 0, the maximum difference observed is 23, the interquartile range is 3.5, the median difference is 3, and the mean difference is 3.77. From these quality control results, we feel confident that the full data set and the statistics derived from citizen science analyzed samples are valid and supported.
To establish the quality control of our community lab we confirmed two key findings from previous studies. First, we used a Student’s
Having established scientific integrity of our crowdsourcing science model through reproducibility, we now were ready to address the unresolved question: does the density of FP on one’s tongue predict the perception of bitterness of the chemical tastant PROP in the phenomenon known as supertasting? First, we employed a simple linear regression model across the entire data set, regressing logged PROP ratings on the predictor FP. We failed to find any evidence that FP associates with PROP response [R-sq = 0.003,
FP density is not predictive of PROP intensity rating.
Significance of factors |
|||
---|---|---|---|
Predictor | |||
Diplotype | PAV/AVI | 0.7221 | <0.0001 |
PAV/PAV | 0.7818 | <0.0001 | |
Sex (male) | -0.1190 | =0.002 | |
Age | -0.0038 | =0.001 | |
FP | 0.0010 | =0.306 |
This result made us reassess how to approach our original question of the role of FP in PROP taste sensitivity and supertasting. Perhaps the positive relationship only exists when one specifically compares FP between subgroups representing different taste sensitivities like the aforementioned population of supertasters. Therefore, under this hypothesis, groups that contain people with low sensitivity to PROP should exhibit less FP than groups that are composed of people with high PROP sensitivity. Previous reports demonstrate that three populations can be separated out when taste ratings of PROP are plotted (
Because this recent report and our present data suggest FP is not related to PROP intensity, the Museum specifically wished to test if FP density differed between categorized taster status subgroups. We selected four methods for determining subgroup assignment that would be used for further analysis. The first method was based solely on the distributions observed in our data set. We categorized participants based on a mixture model of three normal distributions of their logged PROP ratings (
FP density does not differ across taster status groups.
Division method | Subgroup comparison | ||
---|---|---|---|
Mixture model | Group 1 | Group 2 | 0.810 |
Group 1 | Group 3 | 0.137 | |
Group 2 | Group 3 | 0.267 | |
Tertiles | T1 | T2 | 0.899 |
T1 | T3 | 0.651 | |
T2 | T3 | 0.896 | |
NT | MT | 0.926 | |
NT | ST | 0.351 | |
MT | ST | 0.466 | |
Quartiles | Lower | Middle two | 0.737 |
Lower | Upper | 0.414 | |
Middle two | Upper | 0.750 |
Using an advanced model of citizen science in a community-based lab setting, the Genetics of Taste Lab at the Denver Museum of Nature & Science has collected and analyzed population data to assess the putative role of FP density in PROP taste intensity and in the phenomenon known as supertasting. First, our genetic analysis reiterated the well-established relationship between
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The authors wish to thank current and previous volunteer citizen scientists, Teen Science Scholars, interns and staff members in both the Genetics of Taste Lab and on the Expedition Health core team for their preparatory work in supporting the crowdsourcing and our citizen science research model. This study was supported in part by a Science Education Partnership Award from the National Center for Research Resources, National institutes of Health (award number 1R25RR025066).