INTERNATIONAL SYMPOSIUM ON OLFACTION AND TASTE

Genetic Variation in Taste Sensitivity

to 6-n-Propylthiouracil and Its Relationship
to Taste Perception and Food Selection
Beverly J. Tepper,a Elizabeth A. White,a Yvonne Koelliker,a
Carmela Lanzara,b Pio d’Adamo,b and Paolo Gasparinib
a
Department of Food Science, School of Environmental and Biological Sciences,
Rutgers University, New Brunswick, New Jersey, USA
b
Department of Reproductive and Developmental Sciences, IRCCS Burlo Garofolo,
University of Trieste, Trieste, Italy

The ability to taste bitter thiourea compounds and related chemicals is a well-known
human trait. The majority of individuals perceive these compounds, typified by the bit-
terness of 6-n-propylthiouracil (PROP) and phenylthiocarbamide (PTC), as moderately-
to-extremely bitter. Approximately 30% of the population is taste blind to these sub-
stances. It has been hypothesized that PROP/PTC tasters are more sensitive to other
bitter tastes, sweet taste, the pungency of chili peppers, the astringency of alcohol, and
the texture of fats. Tasters may also show lower preferences for foods with these taste
qualities than nontasters who show the opposite set of responses (i.e., lower taste sensi-
tivities and higher preferences for these sensory qualities). This pathway is illustrated
in the following model:
PROP Sensitivity → Food Perception → Preference → Selection
Robust associations between PROP status and taste perceptions have been well doc-
umented. However, subsequent links to food preferences and diet selection have been
more difficult to demonstrate. This is not surprising given the complexity of human in-
gestive behavior that is influenced by numerous factors including health attitudes, per-
sonality traits, and cultural norms. Our laboratory has been using PROP screening to
investigate individual differences in the selection of bitter foods, especially bitter tasting
vegetables and fruits that may have long-term health implications. This chapter will dis-
cuss new and recent findings addressing the following issues: 1) whether PROP-related
differences in perception of bitter compounds predict the perception and liking of bitter
foods; 2) the role of bitter taste modifiers; and 3) the influence of personal characteristics
such as food attitudes and cultural background on PROP-related food preferences.

Key words: 6-n-propylthiouracil; bitter taste; taste modification; food preference

Introduction stances that stimulate the other basic tastes.1

Since many bitter substances are either poi-
A vast number of structurally diverse com- sonous or toxic, the ability of humans to de-
pounds elicit bitter taste in humans and many tect even trace amounts of bitterness in foods
bitter substances can be detected at concen- may have conferred important survival ad-
trations roughly 1000-fold lower than sub- vantages throughout human evolution. It is
not surprising, then, that humans exhibit a
strong innate aversion to strong bitter taste.2
Address for Correspondence: Beverly J. Tepper, Ph.D., Rutgers Univer- Nevertheless, there are large individual differ-
sity, Department of Food Science, School of Biological and Environmental
Sciences, 65 Dudley Road, New Brunswick, NJ 08901-8520. Voice: 732- ences in the perception of bitterness among
932-9611x221; fax: 732-932-6776. tepper@aesop.rutgers.edu people that are well known. The best-known
International Symposium on Olfaction and Taste: Ann. N.Y. Acad. Sci. 1170: 126–139 (2009).
doi: 10.1111/j.1749-6632.2009.03916.x  c 2009 New York Academy of Sciences.

126
Tepper et al.: PROP Bitterness & Food Selection 127

example of this variation is the genetic abil- two common haplotypes, PAV (the taster vari-
ity to taste the synthetic compounds phenyl- ant) and AVI (the nontaster variant). PROP-
thiocarbamide (PTC) and 6-n-propylthiouracil sensitive individuals possess one or two domi-
(PROP). nant alleles (PAV/PAV or PAV/AVI), whereas
PTC and PROP taste moderately-to- insensitive individuals are recessive for the trait
intensely bitter to the majority of individuals (AVI/AVI).11,12 The occurrence of other vari-
across the globe.3 Approximately 70% of Cau- ants is either rare (AAV and PVI) or is limited
casians of Western European origin are consid- to specific populations (AAI in Sub-Saharan
ered “tasters” whereas the remaining 30% are Africans).13
taste-blind to these compounds and are consid- Data from human experiments and cell-
ered “nontasters.” Bartoshuk and colleagues4 based assays suggest that this receptor responds
showed that this large group of tasters can to PTC and PROP but not to other structurally
be subdivided into medium tasters and super- related compounds that lack the thiourea (N-
tasters, with the latter group showing the most C = S) moiety.12 This implies that TAS2R38
extreme sensitivity to PTC/PROP. Thus, the may be narrowly tuned and selective for this
population distribution of nontasters, medium specific subset of compounds. This specificity
tasters and supertasters approximates 30%, could have important nutritional implications
50%, and 20%, respectively.4 However, these since high concentrations of goitrogens (typified
percentages vary as a function of age, gender, by the class of compounds known as glucosi-
and ethnicity, which can modify the expres- nolates) in traditional human diets are known
sion of the trait.3–5 The terms “low-taster (or to interfere with the utilization of iodine and
responder)” and “high-taster (or responder)” compromise thyroid function.14 It has been
have sometimes been used,6 but there is no suggested that the PTC polymorphism is con-
general agreement as to the meaning of these served in human beings as a protective mech-
terms. In general, tasters can be distinguished anism against the overconsumption of dietary
from nontasters based on their threshold sen- goitrogens.15 In support of this notion, a recent
sitivities to PTC/PROP. However, threshold study showed that greater sensitivity to PROP
methods cannot separate medium tasters from was associated with increased bitterness per-
supertasters, hence suprathreshold intensity ception from glucosinolate-producing vegeta-
ratings are used for this purpose. Several valid bles but not from nonglucosinolate-producing
and reliable suprathreshold methods are avail- vegetables.16
able and have been used extensively (for re- On the other hand, numerous psychophys-
view see Tepper7 ). The studies described later ical studies suggest that PTC/PROP tasters
in this chapter utilize suprathreshold methods are more sensitive to a variety of bitter sub-
that were developed and tested in our labora- stances found in foods that are unrelated to the
tory and make use of empirically derived cutoff goitrogens including caffeine,17,18 quinine,18–20
scores to identify taster groups.8 Supertasters naringin,21 and isohumolones.22 Negative find-
also have a high density of fungiform papillae ings for caffeine19,23 and quinine23–25 have also
on the anterior surface of the tongue support- been reported.
ing the idea that supertasters are anatomically Food preference studies involving sampled
distinct from medium tasters.4,9,10 foods or reported food likes and dislikes
The gene responsible for variation in PTC/ have generally shown that PROP tasters show
PROP sensitivity is TAS2R38 which resides greater dislike of certain bitter vegetables and
on human chromosome 7.11 Three single nu- strong-tasting fruits. Differences in liking have
cleotide polymorphisms of this gene result been observed for broccoli, brussels sprouts,
in three amino acid substitutions at positions cabbage, kale, asparagus, and spinach,26–30 as
P49A, A262V, and V296I and give rise to well as for grapefruit and grapefruit juice,21,31
128 Annals of the New York Academy of Sciences

and soy products.32 However, findings are not man eating is a complex behavior with multiple
always uniform across the studies. Two reports determinants. These determinants include, but
found no relation between PROP status and ac- are not limited to, age, gender, ethnicity and
ceptance of broccoli,28,33 although one of the culture, prior experiences with foods, personal-
aforementioned studies found a relationship be- ity traits, and attitudes about nutrition, health,
tween PROP status and liking of spinach.28 and body weight. Relatively few studies have ex-
PROP status has also been associated with amined these variables as potential mediators of
greater perceived intensity of other oral sensa- the association between PROP status and nutri-
tions including sweetness,34 irritation from cap- tional endpoints. Studies that have considered
saicin and alcohol,35,36 and the texture of liquid one or more of these variables have generally
fats.9,37 These observations imply that this trait produced positive results.27,42,43
might serve as an index of general taste ability, Our laboratory has been studying the role
an idea that has been suggested previously.38 of the PROP bitter taste phenotype in dietary
This possibility could have broad implications behavior and health outcomes, such as body
for food selection and nutritional status since weight. The following sections will discuss new
numerous studies have reported that PROP and selected, earlier findings from our work ex-
tasters express lower liking for sweet taste34 amining associations between PROP and bitter
(except for PROP-tasting children who show taste perception and selection of bitter foods.
the opposite response39,40 ), salad dressings,41
sweetened milks,42 pungent foods,43 and alco-
holic beverages.44 Does PROP Status Influence the
Several recent investigations of reported food Perception of Chemically Diverse
intakes have also shown that PROP tasters Bitter Substances?
consume fewer vegetables30,45,46 and added
fats27 than nontasters. One study directly mea- Experiments were designed to determine
sured short-term vegetable intake in preschool- whether PROP supertasters would perceive
ers who were given a choice of different veg- more bitterness than nontasters from chemi-
etables as a snack.29 Results showed that taster cally diverse bitter compounds in aqueous so-
children consumed approximately half as many lutions. Five bitter compounds found in foods
vegetables as nontaster children. Furthermore, as well as PROP were tested. Figure 1 shows
32% of the taster children consumed no veg- that PROP supertasters gave higher bitterness
etables during snack time as compared to 8% ratings than nontasters to caffeine, epicatechin,
of the nontaster children. Despite these posi- and PROP across all concentrations as well as
tive findings, there are also studies reporting higher ratings to naringin and L-phenlyalanine
no association between PROP status and liking across most concentrations. PROP status had
and/or intake of these classes of foods.47,48 The only a weak influence on the intensity of qui-
reader is referred to Tepper7 for a comprehen- nine which was rated higher in bitterness by
sive review of the role of PROP status in food supertasters than nontasters at only the high-
preference and dietary behavior. est quinine concentration (0.08 mM). Taste
The lack of concordance among studies sug- intensity for PROP was moderately corre-
gests that the relationship between the ability lated with the ratings for caffeine, epicatechin,
to taste PROP and the perception and liking of and L-phenylalanine (Pearson r = 0.42–0.54;
bitter and other strong-tasting foods is com- P < 0.004 after Bonferroni correction), but not
plex and not completely understood. More- for quinine (r = 0.29; ns) or naringin (r = 0.32;
over, linking PROP status with food selection, ns). These findings suggest that PROP super-
dietary patterns, and ultimately to nutritional tasters perceive greater intensity from a diver-
outcomes may be a difficult task because hu- sity of bitter substances, but the strength of
Tepper et al.: PROP Bitterness & Food Selection 129

Figure 1. Mean (± SE) bitterness intensity of six compounds across concentrations by

PROP nontasters (n = 30) and supertasters (n = 30). The labeled magnitude scale (0 = “barely
detectable” to 100 = “strongest imaginable oral sensation”) was used to collect the ratings.
The compounds were: caffeine, quinine HCl, 6-n-2-propylthiouracil, -(-)epicatechin, naringin,
and L-phenylalanine. All were laboratory grade except for naringin, which was food grade
∗
and was donated by International Flavors and Fragrances, Union Beach, NJ. P < 0.05;
∗∗
P < 0.01.

this effect is not identical across compounds. V291I), the last two are in perfect linkage dis-
These results support the notion that PROP equilibrium. Therefore, subjects were grouped
status serves as a general marker for bitter by the first and second variant site. Fifty-three
taste perception rather than a specific marker subjects provided DNA samples and they were
for thiourea compounds as other studies have grouped as having two bitter-insensitive alleles
suggested.18,49 The reasons for these opposing (AV/AV; n = 20) or at least one bitter-sensitive
findings are unknown at this time, hence further allele (AV/PA or PA/PA; n = 28). We note
investigation of this question seems warranted. that a large proportion (10/14) of the super-
Subjects were also characterized for tasters in our sample were also heterozygous
TAS2R38 genotypes and the data were rean- for TAS2R38, which has been reported pre-
alyzed to determine if the results of the study viously.12 This outcome was anticipated since
remained the same. Although there are three variation in TAS2R38 accounts for a large
variant sites in this gene (P49A, V262A, and portion, but not all of the variation in taste
130 Annals of the New York Academy of Sciences

sensitivity to PROP,13,49 and there is signifi- liking of any of the foods by PROP status (see
cant overlap in PROP taste intensity between Fig. 2).
AV/PA and PA/PA groups.12 An additional five There are relatively few studies examining
subjects possessed an uncommon form of the the influence of PROP status on the sensory
gene (AA) and were eliminated from the sta- responses to foods (or model foods) that con-
tistical analyses. Overall, the relationship be- tained the same bitter compounds that were
tween phenotypic group (nontaster or super- tested here. For example, one study reported
taster) and allelic group (AV/AV or AV/PA that supertasters discriminated added caffeine
and PA/PA combined) was strong (Spearman in orange juice and both medium and super-
r = 0.72; P ≤ 0.001). tasters discriminated caffeine in cream cheese
When the bitterness ratings were reanalyzed better than nontasters.50 In another study, caf-
contrasting subjects with the AV/AV form to feine solutions were judged as more bitter by
those with the PA/AV or PA/PA form (com- PROP tasters than nontasters, but the ad-
bined), the results were virtually identical to dition of a non-nutritive sweetener (neohes-
those obtained when subjects were classified as peridin dihydrochalcone) to caffeine solutions
phenotypic nontasters and supertasters. Thus, decreased bitterness intensity and increased
PROP phenotype and TAS2R38 haplotype hedonic scores of both groups to a similar
provided similar information about the bitter- level.51 The latter study also reported no dif-
ness of the samples. ferences among taster groups in the sensory
responses to chocolate (white, milk, or dark
chocolate).51 PROP status did not influence
the intensity or liking of grapefruit juice in the
Does PROP Status Influence present study. These findings conflict with ear-
Responses to These Same Bitter lier observations from our laboratory showing
Substances in Foods? that taster children gave lower hedonic rat-
ings to a grapefruit/orange juice blend at a
A second experiment investigated whether 3:1 ratio,31 but not at a 1:1 ratio.27 In con-
PROP supertasters would perceive great in- trast, another study reported that supertasters
tensity from representative foods that naturally gave lower hedonic rating than nontasters to
contained the same bitter compounds tested naringin solutions lightly sweetened with su-
in the previous study. Liking ratings were also crose (4%) and that greater sensitivity to PROP
collected to determine if PROP-related differ- was associated with lower reported preferences
ences in bitterness perception influenced the for grapefruit juice.21
acceptance of these foods. Mattes52 has argued that genetically medi-
As shown in Figure 2, PROP status influ- ated differences in bitterness perception may
enced the perception of black coffee (represen- not be sufficient to alter food acceptability,
tative of caffeine) and dark chocolate ganache since bitterness represents only one facet of the
(epicatechin), but not tonic water (quinine) or complex sensory profile of a food. Our find-
white grapefruit juice (naringin). Specifically, ings lend partial support to this assertion. su-
supertasters perceived more coffee flavor and pertasters perceived more bitterness than non-
overall flavor, as well as more bitter persis- tasters from most of the compounds we tested
tence from black coffee than nontasters. su- when dissolved in aqueous solutions but not
pertasters also perceived more chocolate fla- when these same substances were present in
vor and more overall flavor from the dark real foods. Moreover, PROP status did not in-
chocolate than nontasters. Despite noticeable fluence acceptability of these foods. As stated
differences in the perception of coffee and previously, it is also important to consider
chocolate, there were no differences in overall the role of cultural norms, eating attitudes
Tepper et al.: PROP Bitterness & Food Selection 131

Figure 2. Mean (± SE) intensity ratings (left panel) and liking ratings (right panel) for
attributes of four bitter foods by PROP nontasters (n = 24) and supertasters (n = 24). Data were
collected using 15-cm linear scales with anchors “none” to “strong” for intensity, and “dislike
extremely” to “like extremely” for acceptance. The foods were: black coffee (Nescafe Classic
Instant Coffee, Nestle USA); 100% juice, no sugar added white grapefruit juice (Ocean Spray,
Lakeville-Middleboro, MA); dark chocolate ganache (standard household recipe); and tonic
∗ ∗∗
water (Schweppes, Dr. Pepper/Seven Up Inc., Plano, TX). P < 0.05; P < 0.01.

and health beliefs in food acceptability, which such as fruits and vegetables, whole grain prod-
were not measured here. These factors will ucts, and soy.53 Thus, the food and pharmaceu-
be discussed in a subsequent section of this tical industries are highly motivated to control
chapter. bitterness in their products. Common meth-
ods for modifying bitterness include selective
plant breeding, the use of encapsulation to iso-
Is There a Role for Bitter Taste late bitter ingredients from the food matrix,
Modification? A Case Study or the addition of salt, sweeteners, and added
flavors.53
Dislike of bitter taste is the primary reason A novel approach to controlling bitterness
that consumers give for rejecting certain foods, is to utilize compounds that directly modify
132 Annals of the New York Academy of Sciences

the activity of bitter taste cells. The 5′ - be fully elucidated. But strong evidence sug-
nucleotides including adenosine 5′ monophos- gests that AMP inhibits the α-gustducin sig-
phate (AMP), guanidine 5′ monophosphate, naling pathway either at cell-surface receptors
inosine 5′ monophosphate, among others, or at downstream elements of the signaling
are known to block bitter taste perception cascade.54 In the present experiments, AMP
in rodents by inhibiting taste cell signaling specifically reduced the bitterness of caffeine
pathways.54 The exact mechanism by which and (to a lesser extent) quinine, molecules that
this inhibition occurs is unknown. One report are amphipathic and capable of rapidly pene-
in humans showed that AMP reduced the bit- trating membrane.58 Indeed, quinine has been
terness of oral pharmaceuticals as well as qui- shown to activate G proteins directly, bypassing
nine and urea, and was more effective than membrane receptors.59 Caffeine is a potent in-
sodium chloride.55 As part of the same exper- hibitor of phosphodiesterase, a major enzyme
iments described above, we also determined if implicated in gustducin-mediated signal trans-
AMP reduced the bitterness of aqueous bitter duction.60 Thus, it is conceivable that AMP
solutions and bitter foods, and whether super- reduces the bitterness of compounds that ac-
tasters were more sensitive to this effect than tivate a common signaling pathway. Since bit-
nontasters. ter taste is thought to be mediated by multi-
These experiments used adenosine 5′ ple receptor/transduction mechanisms,61 it is
monophosphate sodium salt (Redpoint Bio possible that the other bitter substances acti-
Corp., Cranbury, NJ) henceforth referred to as vate AMP-resistant bitter transduction path-
AMP. Pilot tests showed that AMP at 10 µM ways and therefore AMP was not effective in re-
and 20 µM reduced the bitterness of the test ducing the bitterness of these other compounds.
solutions but did not impart “umami” (savory) This possibility deserves further attention. Our
flavor or other side tastes that are commonly results differ from those of previous studies
associated with the 5′ nucleotides.56,57 Thus, suggesting that AMP acts as a broad-based
AMP was used at concentrations of 0, 10, and bitterness inhibitor in mice and humans.54,55
20 µM. However, we used much lower concentrations
As shown in Figure 3, AMP at both concen- of AMP than in previous experiments, which
trations reduced the bitterness of caffeine at might explain our findings.
all caffeine concentrations. Overall, AMP led There were no significant interactions be-
to a 27% reduction in the bitterness of caf- tween PROP status and AMP concentration
feine. In addition, AMP at 20 µM, but not at on the perception of bitterness in caffeine or
10 µM reduced the bitterness of quinine at the quinine (data not shown). That is, although
highest quinine concentration. AMP did not supertasters gave higher bitterness ratings to
reduce the bitterness of the other compounds caffeine (and to a lesser extent quinine) than
tested. These data suggest that AMP selec- did nontasters (see Fig. 1), both groups experi-
tively inhibited the perceived intensity of the enced proportionally similar reductions in bit-
bitter alkaloids caffeine and quinine, and had ter intensity with AMP. Thus, on a comparative
a more pronounced effect on the former than basis, AMP was as effective in reducing the bit-
on the latter compound. Previous experiments terness of caffeine and quinine for supertasters
with AMP have not investigated its effects on as it was for nontasters. These data imply that
caffeine.55 Thus, to our knowledge, the present PROP status influences an individual’s overall
finding of a robust inhibition of the bitterness bitter taste responsiveness and that AMP in-
of caffeine by AMP is novel, and not previously hibits bitterness through mechanisms that are
reported. distinct from those involved in PROP tasting.
The exact molecular mechanisms by which This distinction is further supported by results
AMP reduces bitterness perception has yet to showing that AMP did not reduce the bitterness
Tepper et al.: PROP Bitterness & Food Selection 133

Figure 3. Mean (± SE) bitterness intensity of six compounds with 0, 10, or 20 µM AMP.
The samples and procedures were the same as those shown in Figure 1. Subjects were also
the same as those shown in Figure 1 with nontasters (n = 30) and supertasters (n = 30)
combined into a single group (total n = 60). ∗∗ P < 0.01.

of PROP when PROP was tasted as a separate did not alter the perception of any of the other
stimulus (Fig. 3). foods. These findings agree with our results in
AMP was also added to the same four foods aqueous solutions showing that AMP strongly
tested previously. Following pilot testing, AMP reduced the bitterness of caffeine solutions. The
was used at either 0 or 100 µM/L of sam- observed potency of AMP in coffee is intriguing
ple in black coffee and chocolate ganache and because coffee contains a variety of other bit-
at either 0 or 350 µM/L of sample in grape- ter compounds (e.g., polyphenols, quinic acid,
fruit juice and tonic water. Thus, two samples caffeic acid) that may make a greater contri-
of each food (with and without AMP) were bution to coffee bitterness than caffeine62 (also
evaluated. When added to black coffee, AMP see Hofmann, this volume63 ). AMP did not al-
reduced mean bitterness ratings from 12.6 ± ter liking ratings of black coffee. However, our
0.4 cm to 10.9 ± 0.5 cm (on a 15-cm subjects did not habitually consume their coffee
scale) as well as persistence of bitterness from black, which might have contributed to the low
10.9 ± 0.6 cm to 9.2 ± 0.6 cm (P ≤ 0.05). AMP overall liking ratings for this sample (< 3 points).
134 Annals of the New York Academy of Sciences

Figure 4. Influence of food adventurousness on liking of food groups as a function of

PROP taster status. Data represent the mean (± SE) number of foods liked per food group by
PROP nontasters (n = 67) and tasters (medium and supertasters combined; n = 165). Liking
responses (“like” or “dislike”) were analyzed by factor analysis to organize the foods into
groups; separate models were calculated for nontasters and tasters. Numbers appearing in
parentheses after the food group labels indicate the total number of foods in the group based
on the factor analysis. Statistical comparisons are between food adventurous and nonfood
adventurous subjects within each taster group. ∗∗ P < 0.01; ∗∗∗ P < 0.001. (Adapted from Ref.
43 with permission from Elsevier.)

Future studies will have to determine if AMP and their preference ratings (like or dislike for
is useful for modifying bitterness in caffeine- each food) were analyzed by factor analysis to
containing foods. On the other hand, AMP organize the foods into groups (fruits; vegeta-
may be a useful laboratory tool for prob- bles; nonfat condiments including hot peppers,
ing the molecular and psychophysical basis of raw garlic and onions, pickles, etc.; and alco-
taste. holic beverages). Separate models were devel-
oped for tasters and nontasters. Subjects were
also characterized as being high or low in food
Do Food Attitudes and Cultural adventurousness.
Norms Play a Role? As shown in Figure 4, PROP tasters who
were food adventurous liked more foods per
With a few notable exceptions,30,43,64 studies food group than tasters who were not food
have not examined the role of personal char- adventurous. In contrast, food adventurous-
acteristics in the relationship between PROP ness had no influence on the number of foods
status and food selection, and this involvement liked from each of the food groups for non-
could be significant. For example, we previously tasters. Thus, characterizing subjects by their
identified a variable called “food adventurous- self-described food adventurousness unmasked
ness” (defined as the self-reported frequency of two subgroups of PROP tasters—those who
trying new foods) that helped to clarify the food liked a broad range of foods and those with
preferences of PROP taster groups.43 Briefly, more narrow food preferences. Classifying sub-
232 American consumers completed a 75- jects by PROP status alone or food adventur-
item food preference checklist that was heav- ousness alone did not provide meaningful in-
ily weighted toward bitter and strong-tasting sights into the reported food preferences of
foods. Subjects were classified as nontasters or these individuals. It seems likely that other
tasters (medium and supertasters combined) personal characteristics interact with PROP
Tepper et al.: PROP Bitterness & Food Selection 135

status to influence food likes and dislikes.

Identifying and incorporating these variables
into future studies will lead to a more com-
plete understanding of food preferences and
selection patterns than examining PROP sta-
tus in isolation.
It is noteworthy that most of the available
data on PROP tasting and food preferences
have been collected in mainstream consumer
groups or students living in North America,
Western Europe, or Australia.7 It has been as-
sumed that PROP status has a similar influence
on food preferences across different ethnic and
cultural groups, but this might not be the case.
One study showed no influence of PROP sta-
tus on general food preferences in Tunisians,65
but data are lacking in other cultures. As part Figure 5. Mean (± SE) liking ratings for vegeta-
of an ongoing investigation on the genetics of bles as a function of PROP taster status by partici-
multifactorial diseases, we have been studying pants in the Village of Carlantino (n = 535). Data
the village of Carlantino, a genetically and ge- were collected using a 9-point liking scale (1 = dis-
ographically isolated community located in the like extremely and 9 = like extremely). No significant
differences were found by PROP status.
mountains of Southern Italy. We administered
a food preference questionnaire to 535 volun-
teers from this village who were 15–89 years
of age (mean age = 52.0 ± 0.8 years). The
questionnaire was based on the food preference
checklist from our food adventurousness study
(see above) but modified to include foods com-
monly consumed in Italy. The response cate-
gories of the questionnaire were also expanded
to a 9-point liking scale.
As shown in Figure 5, there was a high de-
gree of liking for vegetables in Carlantino with
ratings > 7 on the 9-point scale for the majority
of the items. However, liking of these foods did
Figure 6. Mean (± SE) liking ratings for pun-
not vary by PROP taster group. We speculate gent foods and alcohol as a function of PROP taster
that in a food culture where the overall ac- status by participants in the Village of Carlantino
ceptance of vegetables is high, the influence of (n = 535). Data were collected using a 9-point liking
PROP status on vegetable liking may be weak scale (1 = dislike extremely and 9 = like extremely).
or absent. We did not collect information on Mean values for individual foods with different super-
script letters are significantly different by PROP taster
food preparation. However, in Italy, vegetables
status. P < 0.001.
are more commonly eaten cooked than raw,
which reduces bitterness. Thus, another poten-
tial explanation for our findings is that culi- A very different picture emerged for the de-
nary traditions overshadowed the influence of gree of liking for pungent and spicy foods, and
genetic taste variation on vegetable acceptance, alcoholic beverages. Figure 6 shows that super-
which has been suggested previously.52 tasters expressed lower liking for most of these
136 Annals of the New York Academy of Sciences

items in comparison to nontasters. These re- the examination of multiple taste phenotypes
sults are intriguing because these foods stimu- will provide a more complete understanding
late trigeminal sensations. Capsaicin from chili of human eating behavior than a single taste
pepper and isothiocyantes from raw garlic and phenotype.
onion elicit oral irritation, and alcohol produces Finally, a recent report70 questioned whether
astringency.66 These results support the general PROP sensitivity is the best general index of
premise of this chapter, that PROP status plays heightened sensitivity to oral sensations. In a
a role in the acceptance of strong-tasting foods commentary to that paper, Reed71 coined the
that elicit oral sensations outside the realm of phrase “general supertaster” to describe indi-
bitter taste. Interestingly, food adventurousness viduals who respond with heightened sensitivity
had no influence on food acceptance in Car- to all or most stimuli, without regard to their
lantino volunteers. genetic sensitivity to PROP. This novel concept
uncouples the term “supertaster” from its pre-
sumed genetic link with PROP. Indeed, emerg-
Conclusions and Future Directions ing evidence suggests that greater responsive-
ness to PROP as well as greater responsiveness
More than four decades ago, early re- to thermal taste (a taste phantom associated
searchers suggested that PROP status played with warming and cooling of the tongue)72 both
a role in guiding general food preferences such serve as markers for enhanced global sensory
that tasters had more overall food dislikes than acuity.73 The concept of general supertasting
nontasters.67 Since then, steady progress has is intriguing and awaits additional empirical
been made in establishing the links between support.
the ability to taste PROP and a variety of nutri-
tional outcomes including food preference and Acknowledgments
selection, and dietary intake.7 However, the
large number of contradictory reports in this The authors thank Dr. Danielle Reed and
literature has raised doubts about the signifi- Ms. Kirsten J. Mascioli of the Monell Chemi-
cance of this phenotype in human nutrition.52 cal Senses Center for performing the genotype
The viewpoint advanced here is that many of analyses. These studies were supported, in part,
the factors influencing the pathway between by Redpoint Bio Corp. (formerly Linguagen
the ability to taste PROP and nutritional end- Corp.), Cranbury, NJ (to B.J.T.) and the Italian
points are not presently known, and efforts to Ministry of Research and the Italian Ministry
identify these factors should be an ongoing goal of Health (to P.G.).
for progress in the field to continue. Disclosures: The authors have no other fi-
Until recently, TAS2R38 was considered the nancial relationship with Redpoint Bio Corp.
only bitter-taste gene that exhibits prominent Redpoint Bio Corp. holds a patent on the
phenotypic variation in humans. Consequently, use of the common nucleotide, adenosine 5′
all of the research on genetic variation in bit- monophosphate (CAS # 61-19-8) as a bitter
ter taste was focused on this trait. However, blocker in foods. Redpoint Bio Corp. has no
data are accumulating on the contribution of patent claim on the synthesis or experimental
other taste genes to individual differences in use of this agent.
taste perception and food selection. Recent ex-
amples include TAS2R43 and TAS2R44 which Conflicts of Interest
have been associated with the bitter taste of sac-
charin,68 and TAS2R16 which has been impli- These studies were supported, in part,
cated, along with TAS2R38, in behavioral re- by Redpoint Bio Corp. (formerly Linguagen
sponses to alcohol.69 Thus, it seems likely that Corp.), Cranbury, NJ. The authors have no
Tepper et al.: PROP Bitterness & Food Selection 137

other financial relationship with Redpoint Bio 14. Gaitan, E. 1989. Dietary Goitrogenesis. CRC Press. Boca
Corp. that holds a patent on the use of the com- Raton, FL.
mon nucleotide, adenosine 5’ monophosphate 15. Greene, L.S. 1974. Physical growth and develop-
ment, neurological maturation, and behavioral func-
(CAS # 61-19-8) as a bitter blocker in foods. tioning in two Ecuadorian Andean communities in
Redpoint Bio Corp. has no patent claim on the which goiter is endemic. Am. J. Physical. Anthropol. 41:
synthesis or experimental use of this agent. 139–152.
16. Sandell, M.A. & P.A. Breslin. 2006. Variability in
a taste-receptor gene determines whether we taste
References toxins in food. Curr. Biol. 16: R792–R794.
17. Hall, M.J., L.M. Bartoshuk, W.S. Cain, et al. 1975.
1. Meyerhof, W. 2005. Elucidation of mammalian bitter PTC taste blindness and the taste of caffeine. Nature
taste. Rev. Physiol. Biochem. Pharmacol. 154: 37–72. 253: 442–443.
2. Glendinning, J.I. 1994. Is the bitter rejection response 18. Delwiche, J.F., Z. Buletic & P.A. Breslin. 2001.
always adaptive? Physiol. Behav. 56: 1217–1227. Covariation in individuals’ sensitivities to bitter
3. Guo, S.W. & D.R. Reed. 2001. The genetics of compounds: evidence supporting multiple recep-
phenylthiocarbamide perception. Ann. Hum. Biol. 28: tor/transduction mechanisms. Percept. Psychophys. 63:
111–142. 761–776.
4. Bartoshuk, L.M., V.B. Duffy & I.J. Miller. 1994. 19. Leach, E.J. & A.C. Noble. 1986. Comparison of bit-
PTC/PROP tasting: anatomy, psychophysics, and sex terness of caffeine and quinine by a time-intensity
effects. Physiol. Behav. 56: 1165–1171. procedure. Chem. Senses 11: 339–345.
5. McAnally, H., R. Poulton, R. Hancox, et al. 2007. 20. Chang, W.I., J.W. Chung, Y.K. Kim, et al. 2006. The
Psychosocial correlates of 6-n-propylthiouracil (prop) relationship between phenylthiocarbamide (PTC)
ratings in a birth cohort. Appetite 49: 700–703. and 6-n-propylthiouracil (PROP) taster status and
6. Drewnowski, A., A. Kristal & J. Cohen. 2001. Genetic taste thresholds for sucrose and quinine. Physiol. Behav.
taste responses to 6-n-propylthiouracil among adults: 51: 427–432.
a screening tool for epidemiological studies. Chem. 21. Drewnowski, A., S.A. Henderson, C.S. Hann, et al.
Senses 26: 483–489. 1997. Taste responses to naringin, a flavonoid, and
7. Tepper, B.J. 2008. Nutritional implications of ge- the acceptance of grapefruit juice are related to ge-
netic taste variation: the role of PROP sensitivity netic sensitivity to 6-n-propylthiouracil. Am. J. Clin.
and other taste phenotypes. Annu. Rev. Nutr. 28: 367– Nutr. 66: 391–397.
388. 22. Mela, D.J. 1990. Gustatory perception of isohumu-
8. Zhao, L., S.V. Kirkmeyer & B.J. Tepper. 2003. A lones: influence of sex and thiourea taster status.
paper screening test to assess genetic taste sensitiv- Chem. Senses 15: 485–490.
ity to 6-n-propylthiouracil. Physiol. Behav. 78: 625– 23. Mela, D.J. 1989. Bitter taste intensity: the effect of
633. tastant and thiourea taster status. Chem. Senses 14:
9. Tepper, B.J. & R.J. Nurse. 1997. Fat perception is 131–135.
related to prop taster status. Physiol. Behav. 61: 949– 24. Schifferstein, H.N. & J.E.R. Frijters. 1991. The per-
954. ception of the taste of KCl, NaCl, and quinine HCl
10. Essick, G.K., A. Chopra, S. Guest, et al. 2003. Lingual is not related to PROP sensitivity. Chem. Senses 16:
tactile acuity, taste perception, and the density and 303–317.
diameter of fungiform papillae in female subjects. 25. Yokomukai, Y., B.J. Cowart & G.K. Beauchamp.
Physiol. Behav. 80: 289–302. 1993. Individual differences in sensitivity to
11. Kim, U.K., E. Jorgenson, H. Coon, et al. 2003. Po- bitter-tasting substances. Chem. Senses 18: 669–
sitional cloning of the human quantitative trait locus 681.
underlying taste sensitivity to phenylthiocarbamide. 26. Drewnowski, A., S.A. Henderson, A. Levine, et al.
Science 299: 1221–1225. 1999. Taste and food preferences as predictors of
12. Bufe, B., P.A. Breslin, C. Kuhn, et al. 2005. The dietary practices in young women. Pub. Health Nutr.
molecular basis of individual differences in phenylth- 2: 513–519.
iocarbamide and propylthiouracil bitterness percep- 27. Keller, K.L., L. Steinmann, R.J. Nurse, et al. 2002.
tion. Curr. Biol. 15: 322–327. Genetic taste sensitivity to 6-n-propylthiouracil in-
13. Wooding, S., U.K. Kim, M.J. Bamshad, et al. 2004. fluences food preference and reported intake in
Natural selection and molecular evolution in PTC, preschool children. Appetite 38: 3–12.
a bitter-taste receptor gene. Am. J. Hum. Genet. 74: 28. Turnbull, B. & E. Matisoo-Smith. 2002. Taste sensi-
637–646. tivity to 6-n-propylthiouracil predicts acceptance of
138 Annals of the New York Academy of Sciences

bitter-tasting spinach in 3–6-y-old children. Am. J. 42. Hayes, J.E. & V.B. Duffy. 2008. Oral sensory phe-
Clin. Nutr. 76: 1101–1105. notype identifies level of sugar and fat required for
29. Bell, K.I. & B.J. Tepper. 2006. Short-term veg- maximal liking. Physiol. Behav. 95: 77–87.
etable intake by young children classified by 6-n- 43. Ullrich, N.V., R. Touger-Decker, J. O’Sullivan-
propylthoiuracil bitter-taste phenotype. Am. J. Clin. Maillet, et al. 2004. PROP taster status and self-
Nutr. 84: 245–251. perceived food adventurousness influence food pref-
30. Dinehart, M.E., J.E. Hayes, L.M. Bartoshuk, et al. erences. J. Am. Diet. Assoc. 104: 543–549.
2006. Bitter taste markers explain variability in veg- 44. Lanier, S.A., J.E. Hayes & V.B. Duffy. 2005. Sweet
etable sweetness, bitterness, and intake. Physiol. Behav. and bitter tastes of alcoholic beverages mediate al-
87: 304–313. cohol intake in of-age undergraduates. Physiol. Behav.
31. Tepper, B.J., K.L. Keller & N.V. Ullrich. 2003. Ge- 83: 821–831.
netic variation in taste and preferences for bitter and 45. Basson, M.D., L.M. Bartoshuk, S.Z. Dichello,
pungent foods: implications for chronic disease risk. et al. 2005. Association between 6-n-propylthiouracil
In Challenges in Taste Chemistry and Biology. T.L. Hof- (PROP) bitterness and colonic neoplasms. Dig. Dis.
mann et al., Eds.: 60–74. American Chemical Society Sci. 50: 483–489.
Symposium Series 867. Washington, DC. 46. Sacerdote, C., S. Guarrera, G.D. Smith, et al. 2007.
32. Akella, G.D., S.A. Henderson & A. Drewnowski. Lactase persistence and bitter taste response: instru-
1997. Sensory acceptance of Japanese green tea mental variables and mendelian randomization in
and soy products is linked to genetic sensitiv- epidemiologic studies of dietary factors and cancer
ity to 6-n-propylthiouracil. Nutr. Cancer 29: 146– risk. Am. J. Epidemiol. 166: 576–581.
151. 47. Drewnowski, A., S.A. Henderson & J.E. Cockroft.
33. Anliker, J.A., L.M. Bartoshuk, A.M. Ferris, et al. 1991. 2007. Genetic sensitivity to 6-n-propylthiouracil has
Children’s food preferences and genetic sensitivity to no influence on dietary patterns, body mass indexes,
the bitter taste of 6-n-propylthiouracil (PROP). Am. or plasma lipid profiles of women. J. Am. Diet. Assoc.
J. Clin. Nutr. 54: 316–320. 107: 1340–1348.
34. Yeomans, M., B.J. Tepper, J. Rietzschel, et al. 2007. 48. Timpson, N.J., M. Christensen, D.A. Lawlor, et al.
Human hedonic responses to sweetness: role of 2005. TAS2R38 (phenylthiocarbamide) haplotypes,
taste genetics and anatomy. Physiol. Behav. 91: 264– coronary heart disease traits, and eating behavior in
273. the British women’s heart and health study. Am. J.
35. Prescott, J. & N. Swain-Campbell. 2000. Responses to Clin. Nutr. 81: 1005–1011.
repeated oral irritation by capsaicin, cinnamaldehyde 49. Hansen, J.L., D.R. Reed, M.J. Wright, et al. 2006.
and ethanol in PROP tasters and nontasters. Chem. Heritability and genetic covariation of sensitivity to
Senses 25: 239–246. PROP, SOA, quinine HCl, and caffeine. Chem. Senses
36. Duffy, V.B., A.C. Davidson, J.R. Kidd, et al. 2004. Bit- 31: 403–413.
ter receptor gene (TAS2R38), 6-n-propylthiouracil 50. Prescott, J., J. Soo, H. Campbell, et al. 2004. Re-
(PROP) bitterness and alcohol intake. Alcohol Clin. sponses of PROP taster groups to variations in sen-
Exp. Res. 28: 1629–1637. sory qualities within foods and beverages. Physiol. Be-
37. Hayes, J.E. & V.B. Duffy. 2007. Revisiting sugar-fat hav. 82: 459–469.
mixtures: sweetness and creaminess vary with phe- 51. Ly, A. & A. Drewnowski. 2001. PROP (6-n-
notypic markers of oral sensation. Chem. Senses 32: propylthiouracil) tasting and sensory responses to caf-
225–236. feine, sucrose, neohesperidin dihydrochalcone and
38. Olson, J.M., M. Boehnke, K. Neiswanger, et al. 1989. chocolate. Chem. Senses 26: 41–47.
Alternative genetic models for the inheritance of 52. Mattes, R.D. 2004. 6-n-propylthiouracil taster status:
the phenylthiocarbamide taste deficiency. Genet. Epi- dietary modifier, marker or misleader? In Genetic Vari-
demiol. 6: 423–434. ation in Taste Sensitivity. J. Prescott & B.J. Tepper, Eds.:
39. Keller, K.L. & B.J. Tepper. 2004. Inherited taste 229–250. Marcel Dekker, New York.
sensitivity to 6-n-propylthiouracil in diet and body 53. Drewnowski, A. & C. Gomez-Carneros. 2000. Bitter
weight in children. Obes. Res. 12: 904–912. taste, phytonutrients, and the consumer: a review.
40. Mennella, J.A., M.Y. Pepino & D.R. Reed. 2005. Ge- Am. J. Clin. Nutr. 72: 1424–1435.
netic and environmental determinants of bitter per- 54. Ming, D., Y. Ninomiya & R.F. Margolskee. 1999.
ception and sweet preferences. Pediatrics 115: e216– Blocking taste receptor activation of gustducin in-
e222. hibits gustatory responses to bitter compounds. Proc.
41. Tepper, B.J. & R.J. Nurse. 1998. PROP taster status Natl. Acad. Sci. USA 96: 9903–9908.
is related to the perception and preference for fat. 55. Keast, R.S. & P.A. Breslin. 2002. Modifying the bit-
Ann. N. Y. Acad. Sci. 855: 802–804. terness of selected oral pharmaceuticals with cation
Tepper et al.: PROP Bitterness & Food Selection 139

and anion series of salts. Pharm. Res. 19: 1019– 64. Goldstein, G.L., H. Daun & B.J. Tepper. 2007. In-
1026. fluence of PROP taster status and maternal variables
56. Yamaguchi, S. & K. Ninomiya. 2000. Umami and on energy intake and body weight of pre-adolescents.
food palatability. J. Nutr. 130: 921S–926S. Physiol Behav. 90: 809–817.
57. Yamaguchi, S. & A. Kimizuka. 1979. Psychometric 65. Pasquet, P., B. Oberti, J. El Ati, et al. 2002. Relation-
studies on the taste of monosodium glutamate. In ships between threshold-based PROP sensitivity and
Glutamic Acid: Advances in Biochemistry and Physiology. food preferences of Tunisians. Appetite 39: 167–173.
L.J. Filer, Jr., S. Garattini, M.R. Kare, et al., Eds.: 66. Lawless, H.T. & H. Heymann. 1998. Sensory Evaluation
35–54. Academic Press. New York. of Food: Principles and Practices. Chapman & Hall. New
58. Peri, I., H. Mamrud-Brains, S. Rodin, et al. 2000. York.
Rapid entry of bitter and sweet tastants into lipo- 67. Glanville, E.V. & A.R. Kaplan. 1965. Food prefer-
somes and taste cells: implications for signal trans- ences and sensitivity of taste for bitter compounds.
duction. Am. J. Physiol. Cell Physiol. 278: C17–C25. Nature 205: 851–852.
59. Naim, M., R. Seifert, B. Nurnberg, et al. 1994. Some 68. Pronin, A.N., H. Xu, H. Tang, et al. 2007. Spe-
taste substances are direct activators of G-proteins. cific alleles of bitter receptor genes influence human
Biochem. J. 297(Pt 3): 451–454. sensitivity to the bitterness of aloin and saccharin.
60. Rosenzweig, S., W. Yan, M. Dasso, et al. 1999. Curr. Biol. 17: 1403–1408.
Possible novel mechanism for bitter taste mediated 69. Wang, J.C., A.L. Hinrichs, S. Bertelsen, et al. 2007.
through cGMP. J. Neurophysiol. 81: 1661–1665. Functional variants in TAS2R38 and TAS2R16 in-
61. Behrens, M. & W. Meyerhof. 2006. Bitter taste re- fluence alcohol consumption in high-risk families of
ceptors and human bitter taste perception. Cell Mol African-American origin. Alcohol Clin. Exp. Res. 31:
Life Sci. 63: 1501–1509. 209–215.
62. McCamey, D.A., T.M. Thorpe & J.P. McCarthy. 70. Lim, J., L. Urban & B.G. Green. 2008. Measures of
1990. Coffee bitterness. In Developments in Food Sci- individual differences in taste and creaminess percep-
ence, Vol. 25. R.L. Rouseff, Ed.: 169–182. Elsevier. tion. Chem. Senses 33: 493–501.
Amsterdam. 71. Reed, D.R. 2008. Birth of a new breed of supertaster.
63. Hofmann, T. 2009. Identification of the key bitter Chem. Senses 33: 489–491.
compounds in our daily diet is a prerequisite for the 72. Cruz, A. & B.G. Green. 2000. Thermal stimulation
understanding of the hTAS2R gene polymorphisms of taste. Nature 403: 889–892.
affecting food choice. Ann. N. Y. Acad. Sci. Interna- 73. Bajec, M.R. & G.J. Pickering. 2008. Thermal taste,
tional Symposium on Olfaction and Taste. 1170: PROP responsiveness, and perception of oral sensa-
116–125. tions. Physiol. Behav. 95: 581–590.