Physiology & Behavior 74 (2001) 329 – 337
Relationship of papillae number to bitter intensity of quinine and PROP
within and between individuals
Jeannine F. Delwichea,b,*, Zivjena Buleticb, Paul A.S. Breslinb
a
Department of Food Science and Technology, Ohio State University, 2015 Fyffe Court, Columbus, OH 43210, USA
b
Monell Chemical Senses Center, 3500 Market Street, Philadelphia, PA 19104, USA
Received 18 October 2000; received in revised form 20 March 2001; accepted 7 June 2001
Abstract
Subjects were asked to assess the bitterness of one 6-n-propyl-2-thiouracil (PROP) and two quinine HCl (QHCl) concentrations presented
via filter papers of varying sizes. The number of taste papillae stimulated by these filter papers was counted in each individual. Whole mouth
sensitivity to PROP was determined in a separate session. In support of other demonstrations of spatial summation, these data indicated that
perceived bitterness intensity increased as a function of area of stimulation within subjects. Between subjects, there was a significant trend for
the perceived bitterness of PROP to increase with the lingual density of fungiform papillae, although this trend was highly variable and was
only demonstrable among those who showed at least moderate sensitivity to PROP. On the other hand, the number of stimulated fungiform
papillae failed to account for individual differences in perceived bitterness of QHCl. D 2001 Elsevier Science Inc. All rights reserved.
Keywords: Bitter taste; Papillae; Spatial summation; PROP status
1. Introduction
A recent examination of individual differences in sensitivities to 11 different bitter-tasting compounds revealed two
principal findings [1]. First, sensitivities to these compounds
tended to form distinct clusters across individuals, including
separate sensitivity clusters for 6-n-propyl-2-thiouracil
(PROP) and quinine HCl (QHCl). Second, despite the
idiosyncratic differences in sensitivities to the 11 compounds tested, there were few subjects at the extreme ends
of sensitivity who demonstrated either a tendency to be less
sensitive to all 11 bitter stimuli or more sensitive to all 11.
Although such overall differences in the bitter taste system
have been reported elsewhere (e.g., Refs. [2,3]), an explanation for these apparent alterations in the gain of the bitter
system’s sensitivity has not been determined.
One possible explanation points simply to differences
in the number of taste buds in individual’s mouths, with
more receptor cells potentially resulting in stronger taste
signals. Miller and Reedy [4] demonstrated that the
* Corresponding author. Department of Food Science and Technology,
Ohio State University, 2015 Fyffe Court, Columbus, OH 43210, USA. Tel.:
+1-614-292-6281; fax: +1-614-292-0218.
E-mail address: [email protected] (J.F. Delwiche).
perceived intensities of some taste compounds, including
PROP, were correlated with subject’s lingual taste papillae density. Specifically, they found that subjects with
higher taste papillae densities rated certain taste stimuli
as stronger than did those with lower taste papillae
densities. Therefore, one possible explanation for the
difference in overall bitter sensitivities is a difference
in the density of taste receptors. Other research has
shown that those who are more sensitive to PROP (as
compared to those who are less sensitive to PROP) are
also more sensitive to many oral sensations, including
tastes (sensitivity to sweeteners [5]), irritants (the burn of
capsaicin and alcohol [6,7]), and tactile stimuli (the
viscosity of milkfat [8]). It has been suggested [9,10]
that these differences in sensitivity can be explained by
the difference in the number of taste buds across subjects, and the resultant differences in neural innervation
of the tissue, since each taste bud is richly innervated
with both gustatory and somatosensory fibers [11]. Consistent with the intensity – papillae density correlations,
other researchers have demonstrated that as the area of
taste stimulation is increased (and hence the number of
papillae and buds), the perceived taste intensity increases
[12 –15] within individuals, a phenomenon described as
‘‘spatial summation.’’
0031-9384/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved.
PII: S 0 0 3 1 - 9 3 8 4 ( 0 1 ) 0 0 5 6 8 - 6
330
J.F. Delwiche et al. / Physiology & Behavior 74 (2001) 329–337
Therefore, as a logical continuation of the above observations, the current investigation tested the hypothesis that
stimulating an equal number of papillae (a rough approximation of taste buds) would result in matched bitter taste
intensities across individuals, including those who differ in
PROP sensitivity and/or quinine sensitivity. This papillary
density explanation of global differences in bitterness sensitivity across subjects has been previously implied but has
never been directly tested.
The following study tested this hypothesis by applying
different-sized filter paper disks soaked in taste solutions
to the anterior-dorsal surface of the tongue, thereby
stimulating different numbers of taste papillae/buds with
the different-sized disks. Although subjects differ in the
number and density of fungiform papillae, the use of
different stimulation areas allows for the stimulation of
overlapping numbers of taste papillae across subjects. To
quantify this difference, the number of fungiform papilla
stimulated by each disk size for every subject was
directly counted twice. Together, these data allowed for
a direct comparison of the number of taste papillae
stimulated and the perceived bitter intensity of quinine
and PROP across subjects.
2. Methods and materials
stimuli presented to subjects via filter paper ‘‘sandwiches’’ (see Fig. 1, bottom). These sandwiches consisted
of four layers. First, a ‘‘support’’ disk (parafilm, diameter
1 1/4 in.) provided stiffness to the ‘‘sandwich’’ and
prevented contact of the solutions with any papillae on
the soft palate. Second was a ‘‘backing’’ disk (Whatman
1, diameter 1 1/4 in.) dipped in deionized water. This
provided a fixed-area, wetted, filter paper stimulus that
would create the same somatosensory experience on all
trials, limiting the subjects’ ability to feel that the filter
paper ‘‘stimulus’’ disk varied in size. Third, a ‘‘diffusion
barrier’’ disk was placed on the ventral-lingual surface of
the ‘‘backing’’ disk. This disk was cut from wax paper
(Reynolds) and was slightly larger than the ‘‘stimulus’’
disks: 0.313, 0.75, 1.00, 1.125, and 1.25 in. diameters.
The ‘‘barrier’’ disk served the purpose of preventing the
diffusion of the stimulus solution onto the ‘‘backing’’ disk
(see Fig. 1, bottom). Finally, the variable-sized ‘‘stimulus’’ disk was placed on the ventral-lingual surface of the
‘‘diffusion barrier’’ after being dipped into the stimulus
solution. Stimulus disks (labeled 1 –5) were cut from filter
paper (Whatman 1) to be 0.25, 0.50, 0.75, 1.00, and 1.25
in. diameters, respectively (see Fig. 1, top), and after
soaking contained approximately 0.01, 0.03, 0.06, 0.09,
and 0.14 ml of liquid, respectively. Thus, all stimulus
sandwiches appeared the same from the dorsal surface
and stimulated the same lingual area yet selectively
2.1. Subjects
Twelve female and eight male paid volunteers, ages
20 – 35, from the Monell Chemical Senses Center and
University of Pennsylvania participated in this study and
gave informed consent on a University of Pennsylvania
IRB approved form. They were instructed neither to eat
nor drink for an hour prior to the scheduled testing session
nor to smoke for 2 h prior to the session. Two female
subjects were excluded based upon their extremely inconsistent responses (one gave responses ranging by 25 points
and the other by 44 points on a 97-point Labeled Magnitude Scale (LMS) across only three ratings of a single
stimulus), and two additional female subjects were
excluded since they failed to follow experimenter instructions on scale usage (as determined at exit interview).
2.2. Experimental procedures
2.2.1. Stimuli
The stimuli in the training session were 0.5, 1.0, 1.5, and
2.0 M sodium chloride (Fisher Scientific). The four stimuli
in the experimental sessions were 1.19 10 4 and
5.95 10 5 M QHCl dihydrate (Fluka), 5.5 10 3 M
PROP (Sigma), and deionized Millipore2 filtered water.
2.2.2. Stimulus ratings
Using a novel technique of our own invention, subjects
were asked to rate the bitter intensity of a series of
Fig. 1. Stimuli: (Top) Stimuli disks of the illustrated sizes were dipped into
the appropriate test solution and placed on the anterior tongue. (Bottom)
Layered stimulus ‘‘sandwiches’’ were constructed with the backing disk
being placed upon a support disk (not shown) of parafilm.
J.F. Delwiche et al. / Physiology & Behavior 74 (2001) 329–337
stimulated different-sized areas of the tongue with taste
solutions. The front edge of all ‘‘stimulus’’ disks was
placed as anteriorly as possible on the dorsal tongue. The
experimenter precut the filter and wax papers disks with a
self-centering variable diameter punch (Precision Brand
Tru Punch), and the stimulus sandwiches were assembled
during the interstimulus interval (ISI) out of the sight of
the subject. At the end of the ISI, the stimuli were placed
on the subject’s tongue tip (covering the anterior-dorsal
surface of the tongue).
Subjects were instructed to draw the tongue into the
mouth, close the mouth, and rate the intensity of bitterness
(relative to all other oral sensations) on a computerized LMS
[16,17]. For the training session, subjects rated the total
intensity of 10 stimuli sandwiches (two randomized blocks
of the four NaCl concentrations and water). On each of the
three experimental sessions, subjects rated the bitter intensity of 20 randomized presentations that consisted of each of
the five sizes of filter paper dipped in each of the four
stimuli (water, PROP, and both concentrations of QHCl).
Between stimulus presentations in the experimental sessions, subjects rinsed once whole mouth with isotonic saline
and four times with deionized water during a 120 s ISI, and
in the training session they rinsed with deionized water at
least four times during a 60 s ISI.
2.2.3. PROP sensitivity
Subjects were classified into sensitive and insensitive
PROP tasters because sensitivities to this compound cluster
into two easily distinguished distributions [18], whereas
sensitivities to quinine do not.
Subjects’ whole mouth PROP sensitivity was determined by having subjects rate the bitterness and total
intensity of 30 ml samples of five concentrations of PROP,
presented at concentration levels that differed by 1/2 log
steps (5.5 10 5, 1.7 10 4, 5.5 10 4, 1.7 10 3,
and 5.5 10 3 M), and deionized water. These ratings
also were made upon a computerized LMS, and subjects
were instructed to rate bitterness such that it was a
component of the total intensity. During the 1 min ISI,
subjects rinsed four times with deionized water. In addition, subjects rated the loudness of six tones (generated by
a Maico Hearing Instruments tone generator, presented via
headphones at 4000 Hz for 2 s at levels 0, 20, 35, 50, 65,
and 80 dB) and the heaviness of six visually identical
weights (opaque, sand-filled jars at levels 225, 380, 558,
713, 870, and 999 g). These ratings were also made on
computerized LMSs that were described as a loudness
scale for the former and a heaviness scale for the latter,
and all judgments were made within the context of the full
range of sensations experienced in life. Because of the
minimal adaptation that occurred for the tone and weight
assessments, the ISI was shortened to 30 s. All stimuli
were rated twice, presented in two blocks of ascending
order. Subjects first rated the intensity of PROP solutions,
then tones, and finally weights.
331
Significant correlations were found between average
ratings for the highest levels of PROP bitter intensity
and loudness (r = .75, P = .001), PROP bitter intensity and
heaviness (r = .64, P = .007), and loudness and heaviness
(r = .63, P = .009). Since these variables should be unrelated, this indicated that the LMS ratings were subject to
bias (reflecting differences in scale usage) and required
standardization across subjects. Thus, to determine a cor-
Fig. 2. Papillae distribution: (Top) Distribution of papillae for a PROPinsensitive subject (whole mouth ratings of two highest concentrations of
PROP averaged 5.73); (bottom) distribution of papillae for a PROPsensitive subject (whole mouth ratings of two highest concentrations of
PROP averaged 52.94).
332
J.F. Delwiche et al. / Physiology & Behavior 74 (2001) 329–337
Table 1
Correlation coefficients of intensity ratings vs. number of papillae
PROP
Disk
Disk
Disk
Disk
Disk
1
2
3
4
5
.237
.812
.756
.688
.479
High quinine
.034
.044
.129
.239
.134
Low quinine
.064
.211
.302
.468
.277
Correlation coefficients in bold italics are significant at P < .05 after
Bonferroni correction for multiple tests.
rection factor, each individual’s average intensity rating for
loudness was divided by the grand mean for loudness
across levels and subjects. The same was done with
heaviness ratings. These two correction factors determined
with loudness and heaviness were then averaged, and each
individual’s ratings for the PROP solutions and all stimulus disks were multiplied by his or her resulting average
correction factor.
Based upon the average of each subjects’ bitterness
ratings of PROP at 5.5 10 3 and 1.7 10 3 M, the
subjects were divided into two groups. The five subjects
who gave a mean rating below 10.2 (between ‘‘weak’’ and
‘‘moderate’’) were classified as PROP insensitive, while
the remaining 11 subjects, whose individual mean ratings
were above 13.3 (‘‘moderate’’ up to ‘‘very strong’’), were
classified as PROP sensitive. We classified the subjects
based on an examination of the distribution of the PROP
bitterness intensity results, which indicated a ‘‘natural’’
break in the subjects’ bitterness ratings, rather than using a
predetermined bitterness level criterion. Repeated-measures
analysis of variance confirmed that these two groups
differed significantly ( P = .04) in their ratings of the
stimulus disks dipped in PROP during the separate experimental test sessions. Thus, our classification scheme was
predictive of performance on a different taste-rating task.
Furthermore, the ratings on these two tasks (for
5.5 10 3 M PROP presented both via whole mouth
and via the largest disk) were significantly correlated after
standardization (r = .81, P = .0001).
2.2.4. Taste papillae counts
The subject’s taste papillae were counted twice by a
single rater who was unaware of the subjects’ PROP
classification. To aid in this count, the experimenter first
dyed the dorsal surface of the subject’s tongue with blue
food coloring (Durkee, Tone Brothers, Ankeny, IA),
applying the dye with a cotton-tipped applicator (Puritan
Hardwood Products LP, Guilford, ME). After swallowing
once, the subject then pressed the tongue against a stiff
Plexiglas sheet (1.5 mm thick, 85 40 mm), upon which
circles had been drawn with diameters corresponding to
those of the filter paper disks. In this procedure, the taste
papillae appear as pink circles against a blue background.
The location of each of the fungiform papillae (within
the disk boundaries) was marked with a dot on the slide
with a fine-tipped permanent marker (0.2 mm). Fig. 2
contains scanned images of such slides for two individuals (one from either PROP sensitivity group). For each
subject, the number of marked taste papillae that fell
within a disk’s boundaries was counted at each disk size
for both assessments. Marks that fell exactly upon the
disk border were included in the count of the next
largest disk, and those falling on the border of the
largest disk were not included. Repeated-measures analysis of variance (ANOVA) indicated that there was no
Fig. 3. Number of papillae: Average number of papillae ± standard deviation for each disk size is shown for each group (PROP-sensitive subjects represented by
filled circles and PROP-insensitive subjects by open circles).
J.F. Delwiche et al. / Physiology & Behavior 74 (2001) 329–337
significant difference between the first and second counts
( P = .95). The average of the two counts was used as the
measure of each subject’s number of taste papillae at
each disk size.
2.2.5. Data analysis
After standardization with the correction factor determined for each subject (see PROP sensitivity), histograms
of individual ratings were plotted for all stimulus disks.
Contrary to expectations [16,17], the values were not
lognormally distributed for either quinine concentration.
Dividing the subjects into two distributions, sensitive and
333
insensitive, eliminated the lognormal appearance of the
whole PROP distribution. Accordingly, no further adjustment to the data was necessary. Thus, repeated-measures
ANOVAs and Scheffe’s pairwise post hoc comparisons were
performed on the standardized data to test for significant
differences and interactions between PROP sensitivity
groups, stimuli, and area of stimulation. Additionally, the
Pearson product moment correlation coefficients for intensity rating vs. number of papillae were determined for each
compound and filter paper size. In order to reduce type I
errors, a Bonferroni correction was made to all P-values by
multiplying each by 15.
Fig. 4. Bitterness intensity vs. log (area): average bitterness intensity ratings for all compounds (high concentration quinine = filled triangle; low
concentration quinine = open triangle; PROP = filled circle; water = open square). (Top) For all subjects; (middle) for PROP-sensitive subjects; (bottom) for
PROP-insensitive subjects.
334
J.F. Delwiche et al. / Physiology & Behavior 74 (2001) 329–337
3. Results
Repeated-measures ANOVA on the number of counted
papillae stimulated by the different sized disks and across
the two PROP sensitivity groups indicated that while the
number of papillae stimulated were significantly different
between almost all disks (ANOVA: P < .0001; Scheffe’s
< 0.05 for all disk pairs, excluding comparisons between
Disks 3 and 4 and Disks 4 and 5), there was no
significant difference in the number of papillae between
PROP classification groups ( P = .22) nor a significant
interaction between disk size and PROP sensitivity classification ( P = .39). However, there was a significant
correlation of PROP bitterness ratings and number of
papillae for Disks 2 – 4 (see Table 1), in support of Miller
and Reedy [4]. Fig. 3 shows that on average the insensitive subjects had fewer papillae than did the sensitive
ones, but due to the large variance within each group (see
standard deviations, Fig. 3) and the relatively low power
of the ANOVA due to the small size of the insensitive
group, this was not a significant difference despite the
observation of a significant correlation between number of
Fig. 5. Bitterness intensity vs. area and number of papillae: average bitterness intensity ratings for each disk vs. area (left) and vs. group average number of
papillae (right) for PROP-sensitive subjects (filled circles) and PROP-insensitive subjects (open circles). (Top) For PROP (left column, PROP sensitive:
m = 0.16, r = .96, P = .01; PROP insensitive: m = 0.01, r = .58, P = .31); (middle) for high concentration quinine (left column, PROP sensitive: m = 0.16, r = .95,
P = .13; PROP insensitive: m = 0.16, r = .92, P = .025); (bottom) for low concentration quinine (left column, PROP sensitive: m = 0.18, r = .91, P = .03; PROP
insensitive: m = 0.16, r = .99, P = .001).
J.F. Delwiche et al. / Physiology & Behavior 74 (2001) 329–337
papillae and PROP bitterness for the majority of the disk
sizes (Table 1). For the purposes of this study, however,
all that is relevant to test the hypothesis (that the number
of papillae and/or taste buds stimulated accounts for
differences in bitterness sensitivity across subjects) is a
large variability in bitterness ratings among subjects and
large variability in the lingual fungiform density, both of
which we obtained.
Fig. 4 shows the average intensity ratings for each
compound as a function of the log of disk area for all
subjects (top panel) and subdivided into PROP sensitive
(middle panel) and insensitive subjects (bottom panel).
Repeated-measures ANOVA of the dependent factors of
intensity ratings across compounds (PROP, high quinine,
and low quinine) and disk size (diameter 0.25 –1.25 in.)
failed to find an overall main effect for the independent
factor of PROP sensitivity group ( P = .47), although a
significant difference in intensity ratings was found across
compounds ( P < .001) and disk sizes ( P < .001). Additionally, the ANOVA revealed a significant interaction of
sensitivity group with test compound ( P = .014), and compound and disk size ( P = .006), but not between sensitivity
group and area ( P = .70) or between sensitivity group,
compounds, and disk size ( P = .16).
To more clearly understand the driving forces of these
significant differences and interactions, separate repeatedmeasures ANOVAs (with disk size as a dependent factor
and sensitivity group as an independent factor) were
performed for each compound. These revealed that there
was a significant difference between PROP sensitivity
groups for intensity ratings of PROP ( P = .043), of course,
but not for either concentration of quinine or water
( P > .05). Additionally, no significant interaction was
found between disk size and sensitivity group for either
concentration of quinine or water ( P > .05). PROP-sensitive subjects rated all PROP disks as more bitter than did
insensitive subjects, but Scheffe’s test revealed that this
difference between groups was only significant for Disks
4 and 5.
Ratings differed significantly across disk sizes for both
levels of quinine and PROP ( P < .015) but not for water
( P > .12; see Fig. 4, top). Scheffe’s tests revealed that for
PROP, the largest disk was rated significantly higher than
the smallest ( P < .015). For the high concentration of
quinine, Disk 1 was rated significantly lower than all but
Disk 2 ( P < .01), while Disk 2 was rated significantly lower
than Disk 5 ( P < .01). For the lower concentration of
quinine, Disk 5 was rated significantly higher than all but
Disk 4 ( P < .04), while Disk 4 was rated as being significantly higher than Disks 1 and 2 ( P < .005).
In Fig. 5, the bitter intensity ratings averaged for each
sensitivity group were plotted against area (left column) and
the average number of papillae stimulated (right column) in
separate rows for each compound. For PROP (top row),
intensity ratings increased with increasing disk area for
those sensitive to PROP (slope = 0.16, r = .96, P = .01) but
335
not for those insensitive to PROP (slope = 0.01, r = .58,
P = .31). When plotted against the number of papillae
stimulated, it is clear that when the number of papillae is
matched, sensitivity groups still differ greatly in intensity
ratings of PROP. In contrast, for both concentrations of
quinine, not only was there no significant difference in
intensity ratings (as reported above) but also the slopes of
these functions of intensity ratings plotted against area (left
column) were all very similar to that found for the PROPsensitive group.
4. Discussion
4.1. Effects of number of papillae within subjects
Overall, the perceived intensity of a particular concentration increases with increasing area of stimulation and the
accompanying increase in number of papillae stimulated
within individuals. However, for PROP, this only held true
for subjects who were inherently sensitive to PROP. Not
surprisingly, those insensitive to PROP did not demonstrate
spatial summation for this compound.
Fig. 4 confirmed the findings of other investigators
[12 –15] by demonstrating that spatial summation of taste
is a robust phenomenon. Not only was a significant
increase in intensity found with increasing disk size for
both concentrations of QHCl for all subjects, but also for
PROP, although only for those sensitive to PROP. Furthermore, for quinine, a comparison of functions plotting
intensity vs. area yielded slopes (0.16 – 0.18) very similar
to that reported by Smith (0.20; [15]). That the high and
low QHCl stimuli have the same slopes for perceived
intensity vs. number of papillae further indicates (1) that
spatial summation of bitter taste occurs with quinine, (2)
that it occurs via a similar mechanism for both sensitivity
groups, (3) that the number of papillae stimulated within
each unit area is a component of this summation, and (4)
that the spatial summation function is invariant across
concentrations of a given compound.
It is also interesting to note that the slopes of the
spatial summation functions for PROP (among those
sensitive to PROP) and quinine (among both PROP
sensitivity groups) are virtually the same. Smith [15]
reported that different taste stimuli, selected to represent
different taste qualities, had very different spatial summation slopes ranging from 0.07 for NaCl to 0.20 for
quinine. The fact that these two very different bitter
compounds, QHCl and PROP, apparently acting upon
different receptors and/or transduction sequences, share
the same slope may indicate that spatial summation is a
function of the type of cell that is activated (bittersensitive cells in this case). Recent evidence by Adler et
al. [19] suggests that more than one type of bitter taste
receptor is expressed on the same bitter sensitive receptor
cells (cf. Ref. [20]).
336
J.F. Delwiche et al. / Physiology & Behavior 74 (2001) 329–337
4.2. Effects of number of papillae between subjects
As shown in Fig. 5, when members of different PROP
sensitivity groups are matched for number of stimulated
papillae (by varying stimulation area), those insensitive to
PROP remain insensitive to PROP. They appear to be
lacking some critical taste transduction component for the
compound PROP. Thus, stimulating an equal number of
taste papillae or taste buds will not make PROP taste
equivalently bitter to them. PROP-insensitive subjects do
not experience even moderate bitterness from PROP when
tasting it whole mouth, while those sensitive to PROP can
experience moderate bitterness from PROP with the stimulation of far fewer papillae and taste buds (those contained
within only 5 mm2 of anterior tongue). This point is
further illustrated by Fig. 2, which shows spatial distributions of taste papillae that are very similar for both a PROPsensitive and a PROP-insensitive subject who nonetheless
differ drastically in their whole mouth assessments of the
two highest concentrations of PROP, giving average ratings
of 52.94 and 5.73, respectively.
However, while spatial summation clearly occurs
within individuals, which indicates that increasing the
number of stimulated papillae increases the perceived
intensity, differences in the number of papillae between
individuals do not fully account for all differences in the
perceived intensity of quinine. It is has been noted that
there is a slight positive correlation between the number
of taste buds per papillae and the number of papillae
across subjects. On average, subjects at the lowest end of
sensitivity to PROP have 2.2 taste buds per papillae and
the most highly sensitive subjects have on average 6.8
taste buds per papillae [9]. This threefold increase in the
number of taste buds also fails to account for the differ-
Fig. 6. Bitterness intensity vs. number of papillae: average ± standard error
of individual bitterness ratings given by four representative individuals
(S5 = filled circles, S7 = open circles, S9 = filled triangles, and S14 = open
triangles) to high concentration quinine vs. number of papillae for each disk
size. Some symbols have been slightly displaced for visual clarity. PROPinsensitive subjects are represented by open symbols and PROP-sensitive
subjects by filled symbols.
ences in sensitivity across subjects. As illustrated in Fig.
6, the intensity of quinine between subjects is somewhat
independent of the number of papillae. While the two
least sensitive subjects in Fig. 6 (S5 and S7) seem to lie
on the same Area Intensity function, the other two
clearly do not. A vertical cross-section at 100 papillae
demonstrates that the perceived bitterness of quinine
varies widely. Similarly, while subject 9 has a very high
density of papillae, this subject is not the most sensitive to
quinine. PROP, on the other hand, tended to show a
stronger and significant correlation of intensity as a
function of papillae density among PROP-sensitive subjects (see Table 1). For Disks 2 –4, the number of papillae
was significantly correlated with PROP bitterness intensity
( P < .05, .01, and .001, respectively).
Earlier research found that PROP nontasters had significantly fewer papillae than did PROP supertasters [9].
Although a similar trend was clearly seen in this study with
PROP-sensitive and -insensitive subjects, when the sample
population was divided into two groups, the difference
between them was not significant. Both Figs. 2 and 3
illustrate the overlap in number of papillae. However, the
goals of this experiment did not require a large sample
size, and the resultant low statistical power of the current
investigation, due to both the relatively small number of
subjects assessed in the current study and the tremendous
amount of intersubject variance in number of papillae
found within subject groups in both investigations [9],
may account for the divergent findings. In addition, in
their study, Bartoshuk et al. [9] reduced some variance by
dividing subjects into three categories of PROP sensitivity
vs. the two divisions employed in the present study,
allowing them to compare the two extremes of sensitivity.
They also reported that those in the middle sensitivity
category had widely varying numbers of fungiform papillae. Nevertheless, the present study did demonstrate a
significant difference in perceived bitter intensities between
sensitivity groups for the PROP stimulus and a significant
overall correlation between sensitivity to bitterness from
PROP and the number of papillae present with the majority
of disk sizes.
In summary, the hypothesis tested in this study, that the
number of papillae (or taste buds) stimulated accounts for
differences in bitterness sensitivity across subjects, fails to
account for individual differences in sensitivity to PROP
or to quinine. If, however, one examines the group of
subjects who are sensitive to PROP, then the number of
papillae seems to predict bitterness of PROP rather well.
If the extreme ends of the sensitivity distributions are
examined, the differences in taste papillae density may
play a more significant role in taste sensitivity for bitter
compounds in general. As previously reported, subjects
representative of the distal ends differ in general in their
sensitivities to 11 different bitter compounds [1]. Despite
the limits of the number of stimulated papillae at predicting bitterness between subjects, the number of stimulated
J.F. Delwiche et al. / Physiology & Behavior 74 (2001) 329–337
papillae had tremendous predictive power when restricted
within a subject.
Acknowledgments
We thank Dr. Gary K. Beauchamp and the anonymous
reviewers for their thoughtful comments on earlier version
of this manuscript. This work was funded by NIH grants
R29 DC02995 (P.A.S.B) and 5 F32 DC00384-02 (J.F.D).
References
[1] Delwiche JF, Buletic Z, Breslin PAS. Covariation in individuals’ sensitivities to bitter compounds: evidence supporting multiple receptor/
transduction mechanisms. Percept Psychophys 2001;63:761 – 76.
[2] Lawless HT. The taste of creatine and creatinine. Chem Senses Flavour 1979;4(3):249 – 58.
[3] Schiffman SS, Gatlin LA, Frey AE, Heiman SA, Stagner WC,
Cooper DC. Taste perception of bitter compounds in young and
elderly persons: relation to lipophilicity of bitter compounds. Neurobiol Aging 1994;15(6):743 – 50.
[4] Miller IJ, Reedy FE. Variations in human taste bud density and taste
intensity perception. Physiol Behav 1990;47(6):1213 – 9.
[5] Bartoshuk LM. Bitter taste of saccharin related to the genetic ability to
taste the bitter substance 6-n-propylthiouracil. Science 1979;205
(4409):934 – 5.
[6] Bartoshuk LM, Cunningham KE, Dabrila GM, Duffy VB, Etter L,
Fast KR, Lucchina LA, Prutkin JM, Snyder DJ. From sweets to hot
peppers: genetic variation in taste, oral pain, and oral touch. In:
Bell GA, Watson AJ, editors. Tastes and aromas. Sydney: University
of New South Wales Press, 1999. pp. 12 – 22.
337
[7] Pelchat ML, Danowski S. A possible genetic association between
PROP-tasting and alcoholism. Physiol Behav 1992;51(6):1261 – 6.
[8] Tepper BJ, Nurse RJ. Fat perception is related to PROP taster status.
Physiol Behav 1997;61(6):949 – 54.
[9] Bartoshuk LM, Duffy VB, Miller IJ. PTC/PROP tasting: anatomy,
psychophysics, and sex effects. Physiol Behav 1994;56(6):1165 – 71.
[10] Zuniga JR, Davis SH, Englehardt RA, Miller IJ, Schiffman SS,
Phillips C. Taste performance on the anterior human tongue varies
with fungiform taste bud density. Chem Senses 1993;18(5):449 – 60.
[11] Whitehead MC, Ganchrow JR, Ganchrow D, Yao B. Organization of
geniculate and trigeminal ganglion cells innervating single fungiform
taste papillae: a study with tetramethylrhodamine dextran amine labeling. Neuroscience 1999;93(3):931 – 41.
[12] Bujas Z, Ostojcic A. La sensibilité gustative en fonction de la surface
excitée. Acta Insttuti Psychologici, Universitatis Zagrebensis 1941;13:
1 – 20.
[13] Hara S. Interrelationship among stimulus intensity, stimulated area
and reaction time in the human gustatory sensation. Bull Tokyo
Med Dent Univ 1955;2:147 – 58.
[14] McBurney DH. A note on the relation between area and intensity in
taste. Percept Psychophys 1969;6:250.
[15] Smith DV. Taste intensity as a function of areas and concentration:
differentiation between compounds. J Exp Psychol 1971;87:163 – 71.
[16] Green BG, Shaffer GS, Gilmore MM. Derivation and evaluation of a
semantic scale of oral sensation magnitude with apparent ratio properties. Chem Senses 1993;18(6):683 – 702.
[17] Green BG, Dalton P, Cowart B, Shaffer G, Rankin K, Higgins J.
Evaluating the ‘Labeled Magnitude Scale’ for measuring sensations
of taste and smell. Chem Senses 1996;21:323 – 34.
[18] Reed DR, Bartoshuk LM, Duffy V, Marino S, Price RA. Propylthiouracil tasting: determination of underlying threshold distributions
using maximum likelihood. Chem Senses 1995;20(5):529 – 33.
[19] Adler E, Hoon MA, Mueller KL, Chandrashekar J, Ryba NJP, Zuker
CS. A novel family of mammalian taste receptors. Cell 2000;100:
693 – 702.
[20] Caicedo A, Roper SD. Taste receptor cells that discriminate between
bitter stimuli. Science 2001;291:1557 – 60.