Chem. Senses 28: 57-69,
2003
© Oxford University Press 2003
Relationship between Molecular Structure, Concentration and Odor Qualities of Oxygenated Aliphatic Molecules
Centre For Advanced Food Research, College of Science, Technology and the Environment, University of Western Sydney, Locked Bag 1797, Penrith South DC, NSW, Australia, 1797
Correspondence should be sent to: Professor David G Laing, Centre For Advanced Food Research, College of Science, Technology and Environment, University of Western Sydney, Locked Bag 1797, Penrith South DC, NSW, Australia, 1797. e-mail: d.laing{at}uws.edu.au
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Increasing the concentration of an odorant increases the number of receptor cells and glomeruli in the olfactory bulb that are stimulated, and it is commonly acknowledged that these represent increased numbers of receptor types. Currently, it is not known whether a receptor type is associated with a unique quality and a unique molecular feature of an odorant, or its activation is used by the brain in a combinatorial manner with other activated receptor types to produce a characteristic quality. The present study investigated the proposal that a molecular feature common to several aliphatic odorants and known to be the key feature required to stimulate the same mitral cells in the olfactory bulb results in a quality that is common to the odorants. Since the common structural feature may activate a specific receptor type possibly at a similar concentration, the qualities of the odorants were determined at seven concentrations where the lowest and highest concentrations were the detection threshold (DT) and 729DT of each subject. A list of 146 descriptors was used by 15 subjects to describe the qualities of each odorant at each concentration. The results indicate that each of the five odorants was characterized by different qualities and the qualities of four of the odorants changed with changes in concentration. Importantly, no quality common to each of the odorants that had the same molecular feature could be identified and it is proposed that identification of the odorants occurs via a combinatorial mechanism involving several types of receptors.
Key words: humans, olfaction, odour profiles, thresholds, coding
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Relationships between the molecular structure and qualities of odorants have been sought by many workers (Guillot, 1948
; Amoore, 1952,
1967;
Wright, 1954). However, to
date no structure—activity theory or model has been proposed that
accounts for the wide range of odors encountered. In more recent times the
advent of technologies based on genetics, molecular biology and computers, and
new techniques for monitoring the responses of olfactory cells both at the
periphery and centrally, have created new avenues for studying the
relationship between molecular structure and odor qualities. Most prominent of
the new discoveries of relevance to the present study are that humans have
500-1000 types of olfactory receptors (Buck
and Axel, 1991), each receptor cell may carry only one receptor
type (Chess et al.,
1994), cells containing a particular receptor type project their
axons to two glomeruli in the olfactory bulb
(Mombaerts et al.,
1996), and that the responses of the output neurons of the bulb,
the mitral cells, reflect the responses of the receptor cells to which they
are connected via synapses in the glomeruli
(Imamura et al.,
1992). Importantly, many of the recent studies have attempted to
be systematic in their approach to studies of molecular structure and odors,
and have used both homologous and analogous groups of aliphatic odorants.
These studies have provided an insight as to how the olfactory system may code
odors both at the level of the receptor cells and the various structures
within the olfactory bulb. Araneda et al.
(Araneda et al.,
2000), for example, showed that the length and branching of the
carbon chain and the type of functional group were critical for activating a
receptor that responded best to octanal. Indeed molecules without an aldehyde
group failed to activate this receptor. In contrast, Imamura et al.
(Imamura et al.,
1992) found mitral cells in the dorso-medial region of the
olfactory bulb that responded best to aliphatic oxygenated molecules with a
similar carbon chain length. The oxygen-containing moiety in the stimulating
odorants included aldehyde, ketone, ester or acid functional groups, with
alcohols rarely activating these cells. Interestingly, each of the different
types of oxygenated aliphatic molecules of similar chain length have very
different odors. Since the responses of mitral cells are reported to reflect
the responses of the corresponding receptor cells and a receptor cell is
reported to contain only one type of receptor, the different oxygenated
odorants must activate at least one other type of receptor for them to be
discriminated and identified by their own characteristic quality(s). This
argument is supported by the finding of a receptor type by Araneda et
al. (Araneda et al.,
2000), as described above, that did not respond to oxygenated
aliphatic molecules other than aldehydes with a similar carbon chain length.
It is also supported by recent studies by Johnson and Leon
(Johnson and Leon, 2000), who
showed in studies of 2-DG activity in the olfactory bulb that aliphatic
aldehydes were characterized by activation of glomeruli in the lateral and
medial regions.
The latter findings raise an interesting question regarding whether an odor
quality arises from the activation of a single receptor type or from an odor
molecule interacting with two or more receptor types where it may adopt
different conformations to be successful. Thus, in the case of the receptor
that responded to an aliphatic aldehyde, ketone, ester or acid of similar
carbon chain length (Imamura et
al., 1992), these molecules must have adopted a similar
conformation to activate the receptor. If activation of a single type of
receptor produces a specific odor quality then all the molecules that
activated the latter receptor may have a quality that is common to all.
However, since each odorant must activate another receptor(s) type to allow
each to be discriminated and identified, each must have been able to adopt a
different and unique conformation that could not be achieved by the other
odorants, but that clearly involved their only unique feature, their
functional group. Accordingly, an aim of the present study is to determine
whether aliphatic oxygenated odorants of similar chain length but having a
different functional group (Figure
1) share a common quality that could reflect activation of a
common receptor type(s). If this proposal is correct each odorant should also
be characterized by at least a second non-common quality suggesting that two
or more receptor types were activated. Failure to find a common quality should
support the view that odor quality and identification of an odorant is
achieved by a combinatorial mechanism that produces a composite of the pattern
of activation in memory (Malnic et
al., 1999). To date no study has investigated the possibility
that an odor quality may arise from activation of a single receptor type. The
exception to this could be studies with pheromones particularly in insects,
where it is not unusual for an insect to have a specialist receptor cell type
that has a high sensitivity to a single odorant, a pheromone, and presumably a
single quality is registered by the insect.
|
The present study also examines another aspect of odor quality, namely, the
relationship between quality and concentration. Electrophysiological studies
of aliphatic odorants have shown that an olfactory receptor cell can have a
very narrow response spectrum at low concentrations which enlarges
substantially as the odorant concentration is increased
(Sato et al., 1994).
Thus, it was demonstrated that a receptor cell may respond to only one odorant
at near threshold concentrations, e.g. a C7 alcohol, but to gradually increase
its response spectrum to increasingly shorter or longer chain aliphatic
alcohols in a systematic manner as the concentrations of the odorants are
increased. This finding indicates that the single C7 odorant had a structure
that best fitted the receptor and consequently had the higher probability of
activating the receptor than related molecules used. Accordingly, if
activation of a single receptor type is sufficient to produce a characteristic
odor quality, it is possible that the molecular feature of heptanal, for
example, that activated specific mitral cells
(Imamura et al.,
1992) and the corresponding receptor type, is the same as that
shared by the other carbonyl-containing aliphatic odorants, i.e. ketone, ester
and acid with a similar chain length. A second aim of this study, therefore,
is concerned with determining the odor qualities of the latter odorants over a
range of concentrations from the detection threshold (DT) to high levels to
ascertain whether commonality of odor quality occurs at a specific
concentration as a result of a common receptor type(s) being activated.
Commonality at near threshold levels would support the proposal that a single
receptor type had been activated to produce a single quality, whilst if it
occurred at high concentrations activation of a common set of receptor types
in a combinatorial mechanism for the production of an odor quality(s) would be
more likely. To determine the importance of concentration on the odor
quality(s) of the oxygenated aliphatics, the detection threshold of each
subject was established before profiling of the odor qualities of each odorant
over a range of concentrations was undertaken. Establishing the thresholds of
subjects allowed a comparison of the quality(s) each subject perceived at
equivalent supra-threshold concentrations. With one exception where only one
supra-threshold concentration was used
(Stevens and O'Connell, 1991),
to our knowledge this `equivalence' approach to describing odor qualities has
not been reported.
In summary, the study has two main aims. First, to determine whether aliphatic oxygenated odorants having the same hydrocarbon chain length and a carbonyl group in an equivalent position (Figure 1) share a common quality that could reflect activation of a common receptor type. If this proposal is correct, to be discriminated from each other, each odorant should also be characterized by at least a second non-common quality indicating that two or more receptor types were activated. A second and complementary aim is concerned with determining the odor qualities of the five test odorants over a range of concentrations from the detection threshold to high levels to ascertain whether commonality of odor quality occurs at a specific concentration as a result of a common receptor type(s) being activated.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Subjects
Overall, 37 subjects (33 female and 4 male) aged between 18 and 47 years of age (mean ± SD = 29.7 ± 11.3 years) who were staff or students at the university or were residents from local suburbs, participated in the study. During the measurement of the detection threshold of a single odorant and its quality profile 15 of these subjects participated. Some of the subjects had experience in sensory studies and all received payment for participating. The study was conducted within the air-conditioned sensory facilities of the University.
Odorants
The main test substances (Fluka, Switzerland) and their purities were the oxygenated aliphatic odorants 1-heptanal (pract > 95%), 2-octanone (purum > 97%), 1-heptanoic acid (puriss > 99%), methyl heptanoate (puriss > 99%) and 1-heptanol (purum > 99%). The diluent used was 1,2-propandiol (99.5%, Aldrich, Sydney, Australia).
Training and test procedures
Thresholds
The detection threshold of a subject for an odorant was determined at a
single test session using the staircase method
(Cornsweet, 1962). Eleven
concentrations of each odorant with a dilution factor of 2 between each
concentration were prepared and presented in 250 ml polyethylene squeeze
bottles (20 ml aliquot in each) equipped with a flip-top spout. To minimize
the possibility of spillage each bottle contained odor-free cotton wool which
absorbed the odorant. Subjects were instructed as to the manner of sampling,
and at the start of the first session were allowed time to familiarize
themselves with the handling of the bottles and the sampling technique. Care
was taken that the flip-up spout was only a short distance (1-2 cm) from the
nasal septum during sampling of an odorant in order to maximize the chance
that the stimulus entered both nostrils. In the two-alternative forced-choice
method used for presenting the samples, one bottle contained the odorant and
solvent, the other only the solvent. At each test trial a subject was asked to
sniff both bottles in the order prescribed and to identify `which of the two
bottles produced the stronger sensation/smell'. Between each pair of samples
there was an inter-trial interval of at least 20 s. The sequence of presenting
each pair of bottles was systematically varied between sessions and individual
subjects. No feedback regarding the correctness of a response was given. As is
usual in the staircase procedure, the test stimulus in the first trial was
always the middle dilution level. If the response was incorrect, the next
stimulus was two concentration steps higher. If the responses to two trials at
that concentration were correct, the concentration presented in the next trial
decreased by one step. Two correct trials at that concentration initiated
counting of reversals. The criterion used for establishing a threshold was
that six consecutive reversals had to occur within a range of three dilution
steps. The detection threshold of a subject was calculated as the geometric
mean of the six reversals.
Odor profiles
As in the threshold tests, sniff bottles were used to present the odorants.
Subjects attended one short familiarization session before two test sessions
(replicates) with an odorant. During the familiarization session subjects were
given a list of 146 odour qualities
(Dravnieks, 1985) where each
quality had a 9-point category rating scale ranging from zero (no smell) to
nine (extremely strong smell), and a sniff bottle that contained a weak to
moderate strength of iso-amyl acetate (50 000 p.p.m. extra pure,
banana/fruity; Reidel Haen AG, Germany). The familiarization odorant was
changed to 2-phenyl ethanol (25 000 p.p.m., purum > 99%, floral, rose-like;
Fluka) before the test session with methyl heptanone since the latter odorant
and the acetate were characterized by fruity qualities and subjects may have
tended to use descriptors of the acetate as a result of their exposure to it
immediately before the test session. Subjects were told that a sample bottle
may have as many as 146 odour qualities or none, and that they should use only
odor qualities which appeared relevant to the stimulus being assessed. If they
did not find any appropriate odor qualities the scoresheet should be left
blank. If an odor quality was chosen they should indicate the strength of the
odor on the rating scale.
During each of the two test sessions with a particular odorant, a subject assessed the odor qualities and perceived intensity of 14 samples. These consisted of seven concentrations of the test odorant, e.g. heptanal, at concentrations which were 1, 3, 9, 27, 81, 243 and 729 times the detection threshold (DT) of that particular subject, six `distracter' odorants of moderate perceived intensity and a sample of the solvent propandiol. The `distracters' were odorants that had very different odor qualities to those of the five test odorants and were included because it was observed during pilot studies that subjects very quickly became demotivated to profiling the same odorant at several concentrations. Accordingly, to maximize the chance that subjects attended fully to the qualities of the test stimuli the order of presenting the six `distracters', seven samples of the test stimulus and the solvent was randomized. Furthermore, to assist attention and minimize olfactory adaptation, the 14 samples were presented in two blocks of seven, 30 min apart. The seven samples always contained three of the `distracters' randomly arranged with the other four samples. Each subject received different randomized sequences which varied across odorants and sessions. There was a 1 min interval between the assessment of each of the seven samples and subjects could sniff a sample as many times as necessary to produce a profile. The order of assessing the 146 odor qualities was also randomized across subjects and sessions. The `distracters' were anisole (1000 p.p.m., purum > 99%; Fluka), eugenol (25 000 p.p.m., purum > 99%; Fluka), butanol (25 000 p.p.m.; Fluka, for UV spectroscopy), triethylamine (1000 p.p.m., 99%; Aldrich), furaneol (50 000 p.p.m., 15% in propylene glycol; Dragoco, Sydney, Australia), and galaxolide (250 000 p.p.m., 50% in diethylphthalate; International Flavour and Fragrances, Sydney, Australia).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Odor thresholds
The distributions of thresholds of the individual subjects for each odorant
are given in Figure 2. As
expected, subjects differed widely in their sensitivities for each odorant,
the difference between the most and least sensitive subjects with each of the
odorants was heptanal (x92), methyl heptanoate (x140), octanone
(x488), heptanoic acid (x46.2) and heptanol (x48). Such
differences in sensitivity vindicates the approach adopted in this study to
obtain the responses of individual subjects to the qualities of equivalent
concentrations of an odorant derived from the detection thresholds of the
individuals. Although individuals may also differ in their
intensity—concentration response functions, and alter the perceptual
equivalence of concentrations, it has been shown that subjects with lower
thresholds provide higher intensity estimates of concentration than those with
high thresholds (Laing et al.,
1978), suggesting the present equivalence assumption is a
reasonable one. As regards comparing the mean thresholds with those in the
literature, no thresholds for these odorants in propandiol could be found.
Thresholds for the odorants have been reported in water to be substantially
lower than found here (Fazzalari,
1978). However, this is not surprising given the different
solvent/air partition coefficients that produce higher headspace
concentrations when the odorants are in water compared to propandiol.
|
Odor qualities
For each of the 14 stimulus conditions (seven concentrations of the test
odorant plus one solvent and six distracters), the relative contribution of a
descriptor was based on a weighted score that was dependent on the frequency
of use of a descriptor and perceived intensity ratings
(Jinks and Laing, 2001). In
these calculations the data from all subjects at a particular level relative
to the detection threshold (`equivalent concentration') of each subject were
combined. The relative contribution of a descriptor, therefore, was calculated
by (i) examining the data of individual subjects separately and adding the two
replicate perceived intensity ratings of each descriptor (if a descriptor was
not selected it was given a score of zero), e.g. if a descriptor was chosen
twice by a subject and given ratings of 5 and 6, a value of 11 would be
obtained; (ii) multiplying the sum of the ratings for a descriptor by the
number of times a non-zero rating was chosen, e.g. 11 x 2 = 22; (iii)
adding the totals across all subjects for individual descriptors relative to
the detection threshold at the specific level; and (iv) multiplying the summed
value for a descriptor with the total number of times the descriptor was
chosen over all subjects. The two multiplication steps allowed descriptors
that were chosen consistently by the same subject and by many subjects to
achieve high scores, thereby indicating the relative contribution of the
descriptor to the overall quality of the stimulus. The highest possible
weighted score was 16 200. The weighting system, therefore, ensured that both
the quality and intensity of an odorant contributed to the estimation of its
importance, and reduced the possibility of a descriptor being considered
important on the basis of having a high-perceived intensity but low selection
frequency. The final weighted score for a quality was obtained by subtracting
the value recorded for the solvent, from the weighted score calculated for the
odorant. The data included all descriptors used by any subject, even if used
only once by only one subject. Applying similar criteria to that used
previously (Jinks and Laing,
2001) only the descriptors with the five highest weighted scores
for each odorant at each concentration were used as indicators of odor quality
(Figure
3,Figure
3,Figure 3). The
most striking features of the data in
Figure
3,Figure
3,Figure 3 are as
follows.
|
|
|
(i) For each odorant there is a clear trend to different descriptors as the concentration changed from low to high levels. Even at a concentration which was 27DT when an odorant can clearly be perceived, the main quality(s) was not necessarily the one(s) that dominated at higher concentration levels. For example, although `oily' and `pineapple' are the main descriptors of heptanal and methyl heptanoate at 729DT, respectively, only `oily' is included as one of the five main descriptors of heptanal at 27DT, but is not the dominant one. Furthermore, examination of the profiles of each of the five odorants shows that even at 81DT which is substantially above the detection threshold, no quality which is the dominant one at 729DT is the dominant one at this lower level.
(ii) The odorants differ in the set of descriptors each has at the highest
concentrations indicating that they smell differently as is found during the
casual sampling of each. Heptanal was characterized by oily, unpleasant
qualities, methyl heptanoate was like pineapple, sweet and citrus, octanone
was sweet, chemical and non-citrus, heptanoic acid was like paint and
chemical, and heptanol was citrus and sweet. Interestingly, although the
concentration range was substantial, i.e. 729DT, the change in some of the
qualities from 243DT to 729DT suggests that an asymptote in perceived
intensity was not reached with any of the odorants. Other evidence
(Berglund et al.,
1978) supports this suggestion since no asymptotes were reached
with ranges of 891 and 5238 with the odorants butanol and hydrogen sulphide,
respectively.
(iii) Unlike some odorants which have a single dominant quality and smaller
secondary qualities, e.g. carvone is largely spearmint, naphthalene is
mothballs, cis-3-hexenal is cut-grass, methyl salicylate is
wintergreen, none of the five odorants had a weighted score for a quality as
high as found in an earlier study with four odorants that were chosen because
one quality dominated at moderate to high concentrations
(Jinks and Laing, 2001). This
suggests that subjects in the present study found it difficult to find a
single descriptor that adequately described these fairly nondescript odors so
that across the subject group a variety of descriptors were chosen with few
being selected by more than about half the subjects. Consequently, they made
use of several complementary descriptors to describe the qualities they
perceived. Thus, with heptanoic acid the descriptors `paint' and `varnish' and
possibly `chemical' may have been the closest descriptors to a particular
quality that did not have a descriptor in the list that provided an exact
description. With heptanal, almost all the main descriptors at the two highest
concentrations were of unpleasant sensations and again may have been the
closest ones to the quality(s) perceived. Importantly, the main descriptors
found here were similar to those reported by Dravnieks
(Dravnieks, 1985). For
example, `oily/fatty' and `sickening' were common to both studies for
heptanal, and `citrus/lemon/orange' and `fragrant/sweet' were common for
heptanol.
(iv) At the lower concentrations the main descriptors were generally of pleasant qualities. With heptanal, methyl heptanoate and octanone, each had several main descriptors of the group fragrant/floral/sweet/vanilla/cool/light suggesting that they may share some quality(s) at low concentrations. In contrast, the descriptors of low concentrations of the non-carbonyl odorant heptanol suggested green/floral/fruity qualities may exist, as is suggested for heptanoic acid, the latter being linked to the green qualities cucumber/herbal/crushed grass and floral qualities of rose/violets/floral.
Since the results above suggested there was a change in the quality(s) of each odorant as the concentration increased above the detection threshold, a principal component analysis (PCA) was conducted to explore the factor structure of each odorant and concentration and to compare the factor structures amongst the stimulus conditions. In brief, the analysis provided information about the relative similarity of each of the descriptor profiles and a more quantitative assessment of the existence of a quality change. Thus, a low similarity between the profiles of an odorant at low and high concentrations would indicate a change in quality had occurred, whilst high similarity would indicate there was little change in the qualities. In this analysis all the data from the list of 146 descriptors were included for each odorant and concentration, but not for the distracters. Since moderate correlations were obtained between some factors (<0.45) a non-orthogonal rotation of the components was conducted to provide the best solution. This analysis found nine rotated components had an eigen value of >1. To ascertain the extent of the similarities between odorants and between the seven concentrations of an odorant, the components with the greatest loadings (>0.2) for an odorant at each concentration were tabulated (Table 1). This showed that for heptanal the components at 3DT were different to those at 9-81DT which were different to the single common component at 243 and 729DT; for methyl heptanoate the same component at DT and 3DT differed in part from those at 9-27DT, following which there was a transition in the components which stabilised at 81 and 243DT, changing again as the concentration was increased to 729DT; for octanone the same main component was present at concentrations from 3 to 729DT; for heptanoic acid the main component was the same at DT and 3DT but changed to a new one at 9DT, which was the main component for all concentrations above this; for heptanol the two lowest concentrations had the same single component, there was a transition of components over the range 9-81DT, and a single main common component at 243 and 729DT. Since the data used in the PCA were composed of odor quality and intensity information, it seems likely that the changes in components that occurred as the concentration changed primarily represented changes in the quality(s) perceived. Indeed, there was no indication that any of the components specifically represented intensity or hedonics. Accordingly, quality changes appear to have occurred with each odorant except octanone, with the data suggesting that three changes in quality may have occurred with heptanal, methyl heptanoate and heptanol, and a single change with heptanoic acid.
|
As regards the components of the odorants at and near to the level of the detection threshold, no component was common across the odorants indicating that a common quality for several of the odorants at these levels did not exist. The finding that each odorant at the lowest concentration had a unique component or set of components (2-octanone) indicates that over the two replicates individuals varied in their sensitivity sufficiently to sense odorants on at least one of these occasions. If their sensitivity had not varied it would be expected that for all five odorants no component would have been recorded since by definition no quality can be perceived at the detection threshold.
The data in Table 1 also show that there is no evidence for commonality of quality across the odorants at any of the concentration levels. Different components were important to different odors. For example, at the highest concentration no component was common across the odorants. These results are in agreement with the quality profiles in Figure 3,Figure 3,Figure 3 that suggest each odor has its own unique set of profiles across the concentrations used, that is different from that of the other odorants.
The five main qualities of each of the distracter odorants when these odorants were included in a session with each of the test odorants are given in Table 2. The two most important findings are that (i) the main qualities of the distracters appear to be largely unchanged regardless of the test odorant used. If not listed in the top five, the main qualities were almost always within the top ten (not shown here); and (ii) when the qualities of the test odorants and distracters are compared, there appears to be no influence of one on the other. Thus, no descriptor common to one odorant or distracter appeared in a profile of the other because one or the other had just been assessed. In other words, the strategy of incorporating distracter odorants to assist subjects in attending to the qualities of a particular test odorant that was presented several times at different concentrations was successful. Other evidence for this is the changing qualities of the test odorants as their concentrations were altered.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The present study investigated two proposals. First, whether four odorants that were characterized by a seven-membered hydrocarbon chain with a carbonyl group in an equivalent position shared a common odor quality (s) that is absent in the non-carbonyl analogue heptanol. Secondly, if a common quality(s) was shared by the carbonyl odorants, whether this was perceived at a specific concentration(s) relative to the detection threshold. The results indicated that no quality could be identified as common to the carbonyl-containing odorants, the five odorants differed in their major qualities (Figure 3,Figure 3,Figure 3, Table 1), and changes in qualities occurred with four of the odorants as their concentration was varied (Table 1).
The absence of a quality common to each of the carbonyl-containing odorants
does not support the proposal that a structural feature of a molecule rather
than the whole molecule can produce a quality. The driving force for this
proposal were the findings by Imamura et al.
(Imamura et al.,
1992) that (i) a single mitral cell could be activated by the four
different types of carbonyl-containing odorants used here, whose common
structural feature was the same hydrocarbon chain length and a carbonyl group
in the equivalent position in each; and (ii) the responses of a mitral cell
reflected those of the corresponding receptor. Since each of the odorants is
characterized by different odor qualities, and provided as assumed here that
the human olfactory system operates similarly to other mammals, it now seems
more likely that activation of a common mitral cell(s) provides only part of
the spatial code that is used to identify each odorant. As mentioned in the
introduction, each odorant must activate at least one other type of receptor
to be identified suggesting that a combination of inputs from several receptor
types is required.
One clear outcome of the present study was the change in the type of odor
qualities that occurred with four of the odorants as their concentration was
increased. Qualities that were reported at low concentrations were usually not
sensed at high concentrations and vice versa. Such changes have been reported
by perfumers and flavorists for a small proportion of odorants
(Arctander, 1969). Indeed, some
odors which have been described as unpleasant at high concentrations are
different and pleasant at low levels. For example, indole is putrid at high
concentrations and has a floral odor at low concentrations. Gross-Isseroff and
Lancet (Gross-Isseroff and Lancet,
1988) also reported that changes in quality occur as the
concentration is changed. Recently an insight to the underlying mechanisms of
quality change was provided by Johnson and Leon
(Johnson and Leon, 2000), who
reported that the patterns of glomeruli in the rat which were activated by
pentanal and 2-hexanone, homologues of two of the odorants used here, shifted
as the concentrations of the odorants increased. The shifts in location were
described as being as large as those found between the locations of glomeruli
for different odorants. Both of these odorants were chosen by Johnson and Leon
because they have been reported by humans to have different qualities at
different concentrations (Arctander,
1969). In contrast, the other odorants used in their study,
namely, pentanoic acid, methyl pentanoate and pentanol, all homologues of the
odorants used here, showed no shift in location of the glomeruli, simply
increases in the number of glomeruli in the regions activated by low
concentrations of the odorants. The results of the present study where changes
in quality occurred as the concentration was altered suggests that shifts in
the position of activated glomeruli would occur if they were the stimulating
odorants.
The most obvious differences in quality found here between the odorants
occurred at the highest concentrations. This was not unexpected since informal
sampling at these concentrations had indicated that all five are easily
distinguished, and reported descriptions also indicate that they are
different. The ease of discrimination is in accord with the greater
probability that more receptor types would be activated at the higher
concentrations as is indicated by the corresponding increase in the number of
glomeruli activated (Johnson and Leon,
2000). Since each odorant is characterized by a different
functional group, they could be expected to activate different types of
receptors which may represent different qualities, or the receptor outputs
could be combined to produce a single or several qualities characteristic of
each odorant. As regards the present data, even allowing for semantic
redundancy, e.g. `paint' and `varnish' ((heptanoic acid), `pineapple', `sweet'
and `fruity' (methyl heptanoate), it seems likely that each odorant is
represented by more than a single quality arising from more than a single
combinatorial mechanism (Malnic et
al., 1999).
The greatest similarity between the main qualities of the odorants occurred at the lower concentrations. Whether this similarity was a consequence of the perceived intensities of the odorants being low resulting in subjects using descriptors that tended to represent qualities which were generally pleasant per se is not known. However, the data from the PCA do not support the existence of a common quality across the odorants that have a common structural feature, at these or any other concentration.
Another interesting finding of the study was that the dominant quality at
the highest concentration was sometimes identified at low concentrations,
though not as the dominant one. Since there is a greater chance that a quality
sensed at low concentrations is activating a single receptor type
(Sato et al., 1994),
the identification of `oily' (heptanal) and `sweet' (octanone) at
concentrations of 9DT and DT, respectively, suggests these major
distinguishing qualities may each activate a highly specific receptor type.
This is particularly so for octanone since no quality change was detected by
the PCA. In contrast, the `pineapple' quality of methyl heptanoate which was
not sensed until 81DT is more likely to have resulted from the combined inputs
of several receptor types with no specific receptor signaling this
quality.
Finally, the absence of descriptors that indicated trigeminal stimulation
particularly at the higher concentrations, was unexpected. The trigeminal
nerve has been shown to modulate, namely inhibit, receptor cell activity
(Bouvet et al., 1987),
which could be expected to reduce or block perception of one or more of the
odor qualities of an odorant. Furthermore, although human trigeminal
thresholds can occur at substantially higher concentrations than olfactory
thresholds e.g., the trigeminal thresholds of methyl ethyl ketone and furfural
were reported to be about 2 log units higher than the olfactory thresholds
(Doty, 1975), the difference
can be much less as occurs with acetic acid, propionic acid and amyl acetate
where the differences were 30, 50 and 200 times
(Walker and Jennings, 1991).
Both of these studies, therefore, suggest that at least with the three highest
concentrations used in the present study, trigeminal stimulation and possibly
confounding of qualities could have occurred. However, there is little
evidence that this happened. Although many trigeminal descriptors such as
`sharp', `acid', `vinegar', `pungent', `cool', `ammonia', `alcohol-like',
`etherish', `eucalyptus', `spicy', `kerosene', `petrol' were included in the
list of 146 used to profile the odorants, none appeared as major quality
contributors. Importantly, none of the main descriptors for heptanal, methyl
heptanoate, or heptanol appear to describe trigeminal activation. The only
possibilities occurred with octanone where the descriptor `chemical' was used
and with heptanoic acid where three of the main descriptors were `paint',
`chemical' and `varnish', all of which are candidates. However, with both
octanone and heptanoic acid `chemical' was used with all stimuli from as low
as nine times the detection threshold (9DT). Since it seems unlikely that
these odorants would have stimulated the trigeminal nerve at a concentration
as low as this, there is little evidence that stimulation of the trigeminal
nerve produced qualities that may have confounded the descriptions of
olfactory-relevant qualities for each of the odorants.
In summary, the data obtained in the present study indicate that the four carbonyl-containing odorants that have a common structural feature have qualities which make them readily distinguishable from each other and heptanol. No common quality for these odorants was identified that could be attributed to their common structural feature. Four of the odorants exhibited changes in quality(s) as their concentration was increased above the detection threshold, suggesting that the recruitment of new receptor types occurred. Finally, the data suggest that a combinatorial mechanism is used by the olfactory system for coding the identity of each of the odorants.
|
Acknowledgments |
|---|
The authors wish to thank the participants for their excellent cooperation and the Australian Research Council and Givaudan Pty Ltd for funding the study. The research was approved by the University of Western Sydney Human Ethics Research Committee.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Amoore, J.E. (1952) The stereochemical specificities of human olfactory receptors. Perf. Essent. Oil Rec., 43,321 -323.
Amoore, J.E. (1967) Specific anosmia: a clue to the olfactory code. Nature,214 , 1095-1098.[CrossRef][Medline]
Araneda, R.C., Kini, A.D. and Firestein, S. (2000) The molecular receptive range of an odorant receptor. Nat. Neurosci., 3,1248 -1255.[CrossRef][Web of Science][Medline]
Arctander, S. (1969) Perfume and Flavor Chemicals (Aroma Chemicals). Stefan's Arctander's Publications, Las Vegas.
Berglund, B., Berglund, U. and Lindvall, T. (1978) Separate and joint scaling of perceived odor intensity of n-butanol and hydrogen sulphide. Percept. Psychophys., 23,313 -320.[Web of Science][Medline]
Bouvet, J.F., Delaleu, J.C. and Holley, A. (1987) Olfactory receptor cell function is affected by trigeminal nerve activity. Neurosci. Lett.,77 , 181-186.[CrossRef][Web of Science][Medline]
Buck, L.B. and Axel, R. (1991) A novel multigene family may encode odorant receptors: a molecular basis for odorant recognition. Cell, 65,175 -187.[CrossRef][Web of Science][Medline]
Chess, A., Simon, I., Cedar, H. and Axel, R. (1994) Allelic inactivation regulates olfactory receptor gene expression. Cell, 78,823 -834.[CrossRef][Web of Science][Medline]
Cornsweet, T.N. (1962) The staircase method in psychophysics. Am. J. Psychol.,75 , 485-491.[CrossRef][Web of Science][Medline]
Doty, R.L. (1975) Intranasal trigeminal detection of chemical vapors by humans. Physiol. Behav.,14 , 855-859.[CrossRef][Medline]
Dravnieks, A. (1985) Atlas of Odor Character Profiles. ASTM, Philadelphia, PA.
Fazzalari, F.A. (ed.) (1978) Compilation of Odor and Taste Threshold Data. ASTM, Philadelphia, PA.
Gross-Isseroff, R. and Lancet, D. (1988)
Concentration-dependent changes of perceived odor quality.Chem. Senses
, 13,191
-204.
Guillot, M. (1948) Anosmies partielles et odeurs fondamentales. C. R. Acad. Sci.,226 , 1307-1309.
Imamura, K., Mataga, N. and Mori, K.
(1992) Coding of odor molecules by mitral/tufted cells in
rabbit olfactory bulb: I. Aliphatic compounds. J.
Neurophysiol., 68,1986
-2002.
Jinks, A. and Laing, D.G. (2001) The analysis of odor mixtures by humans: evidence for a configurational process. Physiol. Behav., 72,51 -63.[CrossRef][Medline]
Johnson, B.A. and Leon, M. (2000) Modular representations of odorants in the glomerular layer of the rat olfactory bulb and the effects of stimulus concentration. J. Comp. Neurol., 422,496 -509.[CrossRef][Web of Science][Medline]
Laing, D.G., Panhuber, H. and Baxter, R.I.
(1978) Olfactory properties of amines and n-butanol.Chem. Senses Flav.
, 3,149
-166.
Malnic, B., Hirono, J., Sato, T. and Buck, L.B. (1999) Combinatorial receptor codes for odors.Cell , 713-723.
Mombaerts, P., Wang, F., Dulac, C., Chao, S., Nemes, A., Mendelsohn, M., Edmondson, J. and Axel, R. (1996) Visualizing an olfactory sensory map. Cell,87 , 675-686.[CrossRef][Web of Science][Medline]
Sato, T., Hirono, J., Tonoike, M. and Takebayashi, M.
(1994) Tuning specificities to aliphatic odorants in mouse
olfactory receptor neurons and their local distribution. J.
Neurophysiol., 72,2980
-2989.
Stevens, D.A. and O'Connell, R.J. (1991)
Individual differences in thresholds and quality reports of human subjects
to various odors. Chem. Senses, 16,57
-67.
Walker, J.C. and Jennings, R.A. (1991) Comparison of odor perception in humans and animals. In Laing, D.G., Doty, R.L. and Breipohl, W. (eds), The Human Sense of Smell. Springer-Verlag, Berlin, pp. 261-280.
Wright, R.H. (1954) Odour and molecular vibration. 1. Quantum and thermodynamic considerations. J. Appl. Chem. (Lond.), 4,611 -615.
Accepted November 22, 2002
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Sanz, T. Thomas-Danguin, E. H. Hamdani, C. Le Poupon, L. Briand, J.-C. Pernollet, E. Guichard, and A. Tromelin Relationships Between Molecular Structure and Perceived Odor Quality of Ligands for a Human Olfactory Receptor Chem Senses, September 1, 2008; 33(7): 639 - 653. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Armstrong, I. Hutchinson, D. G. Laing, and A. L. Jinks Facial Electromyography: Responses of Children to Odor and Taste Stimuli Chem Senses, July 1, 2007; 32(6): 611 - 621. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. A Wright, M. G.A Thomson, and B. H Smith Odour concentration affects odour identity in honeybees Proc R Soc B, November 22, 2005; 272(1579): 2417 - 2422. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. C. Araneda, Z. Peterlin, X. Zhang, A. Chesler, and S. Firestein A pharmacological profile of the aldehyde receptor repertoire in rat olfactory epithelium J. Physiol., March 15, 2004; 555(3): 743 - 756. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Wilson Rapid, Experience-Induced Enhancement in Odorant Discrimination by Anterior Piriform Cortex Neurons J Neurophysiol, July 1, 2003; 90(1): 65 - 72. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||