Chem. Senses 24: 429-438,
1999
© Oxford University Press 1999
Olfactory Discrimination Ability and Odor Structure–Activity Relationships in Honeybees
Institut für Medizinische Psychologie, Ludwig-Maximilians–Universität München, Goethestraße 31, D-80336 München 1 Institut für Neurobiologie, Freie Universität Berlin, Königin-Luise-Straße 28–30, D-14195 Berlin, Germany
Correspondence to be sent to: Matthias Laska, Institut für Medizinische Psychologie, Ludwig-Maximilians–Universität, Goethestraße 31, D-80336 München, Germany. e-mail:Laska{at}imp.med.uni-muenchen.de
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Using the training procedure introduced by von Frisch in 1919, we tested the ability of free-flying honeybees to discriminate a conditioning odor from an array of 44 simultaneously presented substances. The stimuli included homologous series of aliphatic alcohols, aldehydes and ketones, isomeric forms of some of these substances, as well as several terpenes and odor mixtures, and thus comprised stimuli of varying degrees of structural similarity to any conditioning odor. We found (i) that the honeybees significantly distinguished between 97.0% of the 1848 odor pairs tested, thus showing an excellent discrimination performance when tested in a free-flying situation with an array of structurally related substances; (ii) a significant negative correlation between discrimination performance and structural similarity of odorants in terms of differences in carbon chain length with all aliphatic substance classes tested; (iii) that both the position and type of a functional group also affected discriminability of odorants in a substance class-specific manner; and (iv) striking similarities in odor structure-activity relationships between honeybees and human and nonhuman primates tested previously on a subset of substances employed here. Our findings demonstrate that the similiarities found in the structural organization of the olfactory systems of insects and vertebrates are paralleled by striking similarities in relative discrimination abilities. This strongly suggests that similar mechanisms of odor coding and discrimination may underlie olfaction in vertebrates and insects.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Animals of most species are capable of discriminating between a variety of odors. This ability is often crucial for the organization of feeding, mating and social communication, as well as for the processes of learning and memory that are associated with these behaviors. Thus, in order to understand these behaviors and processes, it is important to gain insight into the neural mechanisms underlying the discrimination of odors. Several lines of evidence suggest that odor discrimination begins with differential interaction of odor molecules with different types of olfactory receptors (Buck and Axel, 1991
; Hildebrand and
Shepherd, 1997). However, there is still only sparse information as to which
properties of an odor molecule are functional in
determining the degree of interaction with a given receptor, and thus in determining its perceived
odor quality (Ohloff et al.,1991).
In a series of psychophysical studies aimed at assessing the ability to discriminate
between members of homologous series of aliphatic substances, one of us has recently
shown that significant correlations between structural similarities in terms of differences
in carbon chain length and odor quality exist both in human and in nonhuman primates
(Laska and Freyer, 1997; Laska and Teubner, 1998, 1999b) (M. Laska et al., submitted
for publication). It is clear, however, that carbon chain length is only one of presumably
several determinants of the interaction between odor stimuli and olfactory receptors, and
that other molecular properties, such as position or type of functional groups, may also
affect odor quality and thus discriminability.
Considering the striking similarities in the structural organization of the olfactory
systems in vertebrates and insects (Boeckh et al.,1990), it seems
reasonable to assume
that similar principles of odor coding and discrimination apply to both groups of animals. In
order to test this hypothesis we sought to assess olfactory discrimination ability in honeybees,
one of the model organisms for the study of insect olfaction and olfactory learning (Menzel and Müller, 1996), for a set of substances of varying structural
similarity.
Olfactory performance in honeybees was first investigated by von Frisch (von
Frisch, 1919). He simultaneously offered several odorants to free-flying honeybees,
of which one was
associated with a sugar solution as a reward. The bees were strongly attracted by the substance
associated with sugar, and this conditioning remained after the suppression of the reward, even
though the respective positions of the different odors were changed. von Frisch found that the
animals clearly discriminated 28 out of 32 pairs of odors and partially confused the four others,
for which some similarity of odor quality was observed in man. Thus, he concluded that odor
perception in bees might be similar to that of man.
Although numerous studies using different approaches from the behavioral to the
molecular level have investigated various aspects of honeybee olfaction following von
Frisch's pioneering work, and at least some studies have assessed olfactory discrimination
performance in this species (Vareschi, 1971; Pham-Delegue et al.,1993), only a few
have so far directly addressed odor structure–activity relationships in honeybees (Smith and Menzel, 1989; Sachse et al.,1998). However, new findings on olfactory coding at the
level of the antennal lobes (an
equivalent of the vertebrate olfactory bulb) have shown that different odors are represented in
terms of different glomerular activation patterns (Joerges et
al., 1997). These findings make it worthwhile to compare the behavioral
performance with the neurobiological one, to analyse to what extent the neural coding is
translated into a behavioral response.
We decided to take up von Frisch's experimental design to address this
question as it represents the most naturalistic and probably a more challenging situation for the
animals compared with the proboscis extension reflex paradigm which has frequently been used
with honeybees, particularly to assess mechanisms of odor learning (Menzel and
Müller, 1996; Menzel et al.,1996). Training
free-flying bees to odors also allows simultaneous
testing of many odors, thus avoiding the problem of a sequential test procedure as in the
proboscis extension preparation.
The aims of the study were threefold: (i) to provide systematic data on the olfactory discrimination ability of free-flying honeybees for an array of substances of varying structural similarity; (ii) to assess whether any correlations exist between discrimination performance and different structural features of the odorants under investigation; and (iii) by comparing our data with those of earlier studies which employed the same odorants with mammals, to assess whether the similarities found in the structural organization of the olfactory systems of insects and vertebrates are paralleled by similarities in relative discrimination abilities.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Animals
A total of 78 free-flying honeybee workers (Apis mellifera carnica) from one of the colonies maintained at the apiary of the Department of Neurobiology at the Free University of Berlin was used. The hive was 30 m distant from the room in which the experiments were conducted.
Odorants
A total of 44 odorants, including homologous series of aliphatic alcohols, ketones and aldehydes (C4–C10 respectively) was used (Table 1). Further, the set of substances included isomeric forms of aliphatic alcohols and ketones, several terpenes and etheric oils, and one aliphatic ester.
|
The rationale for choosing the substances was to present the bees with a series of compounds of varying degrees of structural similarity to any conditioning odorant. Thus, for example, when conditioned with an aliphatic alcohol, the bee was also presented with aliphatic alcohols of different carbon chain lengths (i.e. other members of the homologous series), aliphatic alcohols with the same or similar carbon chain length but a different position of the functional group (isomeric forms), substances with the same or similar carbon chain length but different oxygen moieties (aliphatic aldehydes and ketones), odorants with the same functional group but a cylic structure (terpene alcohols) as well as etheric oils, i.e. complex mixtures which comprise compounds belonging to the same chemical class as the conditioning stimulus.
Most of the substances are known to be components of floral scents (Knudsenet al.,1993), and several of the odorants (geraniol, citral, isoamyl acetate,
2-heptanone and 2-nonanol) have also been identified as components of honeybee pheromones (Free, 1987).
Experimental procedure
The odor stimuli were presented in opaque 1 l glass bottles. Each bottle contained a circular filter paper with a diameter of 6 cm which was impregnated with 4 µl of an odorant (Table 1). The neck of each bottle was equipped with a tight-fitting Plexiglas tube of 3 cm length and 1.6 cm inner diameter which had a wire mesh at its inner end in order to allow an odor stimulus to emanate from a bottle but to prevent an animal from crawling or flying into the distal part of the bottle.
A total of 48 bottles was arranged horizontally in a rack of 70 x 80 cm in six rows of eight bottles each. To prevent the odors from intermingling in front of the bottles, the rear of the rack was made conical with the center connected to a suction pump which provided a constant and approximately laminar airstream around each bottle neck.
Foraging worker honeybees were trained to approach a feeder baited with a 20% (w/w) sucrose solution situated in proximity to the test apparatus. Regularly returning bees were individually marked, collected from the feeder when approaching it by use of a glass vial, and put into one of the bottles of the apparatus which contained the conditioning odor (S+) and was baited with 60 µl of a 30% (w/w) sucrose solution placed in the neck of the bottle. Usually the bee entered the neck of the bottle voluntarily, ingested the food reward while perceiving the conditioning odor and then returned to its hive. Five such training trials were conducted in order to allow the animal to build a robust association between the food reward and the conditioned odor while taking care to change the position of the S+ bottle in the array between each trial to prevent the occurrence of positional preferences. Following the fifth training trial, the feeder was removed, the position of the bottle containing the conditioned odor was changed again, and on returning from its hive the bee approached the apparatus and started to search for the S+ by hovering in front of the bottle necks. The behavior of the bee was recorded both on videotape and on protocol sheets which showed the position of each stimulus.
Correct choices consisted of animals both landing in the neck of a bottle containing the S+ and failing to land at a bottle containing an S– after hovering in front of the bottle neck for a minimum of 0.5 s. Conversely, errors consisted of landing in the neck of a bottle containing an S- and failing to land at an S+ bottle after hovering in front of the bottle neck for a minimum of 0.5 s.
In order to obtain a sufficient number of decisions, each bee received a total of 10 extinction trials without food reward and 10 food-rewarded test trials which were alternated with the former. In the extinction trials, the array was composed of 43 bottles containing one S- each, one bottle containing the S+ (without a food reward) and four blank bottles without any odor. In the test trials, the array was again composed of the 43 bottles containing one S- each, but with three bottles containing the S+ (with a food reward) and two blank bottles. In order to maximize the number of decisions per bout and to prevent animals from developing a position preference, the food reward in the test trials was restricted to 10 µl of sugar solution (a full crop load of a honeybee is ~60 µl), and an S+ bottle which already had been visited by a bee was only baited again during the same bout when the animal had landed at one of the other two S+ bottles and consumed the food reward provided there. In order to prevent demotivation of the animals, the time allotted to extinction trials 2–10 was restricted to 3 min. The first extinction trial, i.e. the very first trial following the five (food-rewarded) conditioning trials, lasted only 2 min as a longer duration occasionally caused animals to quit searching and reliably returning to the apparatus. Between any of the 20 trials, the arrangement of the bottles was changed according to a pseudo-randomized scheme.
With the exception of two substances [linalool and (–)-limonene], all odorants were used as conditioned stimuli, and with the exception of six substances ( n-nonanal, n-decanal, geraniol, citral, lime-blossom oil and clove oil), data from two bees per S+ were recorded (cf. Figure 1). Usually two bees were tested per day while taking care to renew the odor stimuli and to thoroughly clean the Plexiglas tubes in the neck of the bottles for each animal.
|
Data analysis
In assessing performance of the bees, only unequivocal decisions (see above)
were scored. The minimum number of decisions per animal was, 1300, the mean 1708.
Significance levels were determined separately for extinction and
test trials by calculating binomial z-scores corrected for continuity (Siegel
and Castellan, 1988)
from the number of correct and false responses for each individual and stimulus.
Comparisons across tasks were made using Friedman's two-way ANOVA. When differences between tasks were detected, pairwise Wilcoxon's signed-rank tests were applied for related samples to evaluate which tasks were responsible. Correlations between discrimination performance and structural similarity in terms of differences in carbon chain length were evaluated using Spearman's rank correlation coefficient and tested for significance by computing t-values. All tests were two-tailed, and the alpha level was set at 0.05. All data are reported as means ± SD.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
General discrimination performance
Figure 1 summarizes the discrimination performance of the 78 honeybees tested in the extinction trials. Of the 1848 odor pairs, the bees were able to significantly discriminate , 1793, i.e. 97.0%. Only seven odor pairs (i.e. 0.4% of all odor pairs) were not distinguished by both bees trained to a given S+, and 48 odor pairs (2.6% of all odor pairs) were confused by one bee.
When used as the conditioned stimulus (S+), 15 substances were significantly discriminated from all other stimuli, 10 substances were confused with only one of the 43 S-, seven substances with two test odors, nine substances with three, and one substance with four. Thus, all 42 substances employed as S+ were significantly distinguished from >90% of the other stimuli.
Overall discrimination scores in test trials and extinction trials with a given animal were usually very similar and differences between the two only rarely exceeded 5%. Interindividual variability in overall discrimination scores was markedly low, with all 78 bees scoring >90% correct choices in both the extinction trials and the test trials, and the majority of animals even scoring >95% correct. None of the bees failed to correctly identify its conditioning odor at a significant level (diagonal in Figure 1) and none of the bees confused >4 of the 44 odor pairs.
Carbon chain length
Figure 2 summarizes the discrimination performance of the honeybees as a function of differences in carbon chain length. In all four homologous series of substances employed (2-ketones, n-aldehydes, 1-alcohols and 2-alcohols) we found that odor pairs which differed by only one carbon atom were significantly more difficult to discriminate than odor pairs which differed by two or more carbon atoms (Wilcoxon, P < 0.01 for all pairs and all four homologous series). Odor pairs which differed by two or more carbon atoms did not differ significantly from each other in their degree of discriminability (Wilcoxon, P > 0.05 for all pairs and all four homologous series). Accordingly, a highly significant negative correlation between discrimination performance and structural similarity of odorants in terms of differences in carbon chain length was found in all four homologous series (Spearman, P < 0.01). This finding is also illustrated by the fact that 23 out of 27 cases of failure to significantly discriminate between members of a homologous series were due to odor pairs which differed by only one carbon atom (cf. Figure 1). The number of cases in which an animal failed to significantly distinguish its conditioned odor from a member of the same homologous series with a shorter carbon chain was 13—almost identical to the 14 cases in which the carbon chain of the unconditioned odor was longer than the S+.
|
Position of a functional group
The discrimination of an aliphatic alcohol from other stimuli with the same carbon chain length but a different position of the functional alcohol group, i.e. isomeric forms, presented little difficulty to the bees. None of the animals confused any 1-alcohol with a 2-alcohol having the same number of carbon atoms (20 cases), irrespective of whether the conditioned stimulus was an alcohol with a terminal oxygen moiety or an alcohol with a non-terminal alcohol group. The only two cases of failure to significantly discriminate between isomeric forms of aliphatic alcohols occurred with the odor pair 3-pentanol versus 2-pentanol, i.e. two aliphatic alcohols which share a non-terminal alcohol group (cf. Figure 1).
Figure 3 illustrates that the degree of discriminability in odor pairs involving isomeric alcohols, i.e. substances which only differ in the position of the functional group, was at least slightly lower than in odor pairs which not only differed in the position of the functional alcohol group but also differed by one carbon atom. The discriminability of isomeric alcohols was slightly higher, however, than that of alcohols which share the position of the functional group but differ in chain length by one carbon atom. The latter were significantly more difficult to distinguish than isomers that differ by one carbon atom (Wilcoxon, P < 0.01).
|
The discrimination of an aliphatic ketone with seven carbon atoms from other straight-chain heptanones with a different position of the functional keto group led to seven cases of significant discrimination and three cases of failure (2-heptanone versus 4-heptanone, 3-heptanone versus 2heptanone and 3-heptanone versus 4-heptanone), and thus to a markedly higher proportion of confusions compared with the alcohols (cf. Figure 1).
Figure 3 shows that the degree of discriminability in odor pairs involving isomeric ketones, i.e. substances which differ only in the position of the functional group, was as high as in odor pairs which differed not only in the position of the functional keto group but also by one carbon atom. The discriminability of isomeric ketones was significantly higher, however, than that of ketones which share the position of the functional group but differ in chain length by one carbon atom (Wilcoxon, P < 0.05).
The discrimination of a non-aliphatic substance, the terpene alcohol geraniol, from its isomer linalool was readily accomplished.
Number of functional groups
1,5-Pentandiol, an aliphatic alcohol with five carbon atoms and functional alcohol groups at both ends of the carbon chain, presented some difficulties for the bees. This substance was confused with several alcohols of the same or similar carbon chain length, whether assigned as S+ or as S-: there were two cases of failure to discriminate , 1,5-pentandiol from 1-hexanol, and one case each with 1-pentanol, 1-nonanol and 3-pentanol (cf. Figure 1).
Type of functional group
The discrimination of aliphatic substances with the same carbon chain length but different oxygen moieties presented little difficulty to the bees. With 1-alcohols, 2-alcohols and ketones used as S+, there were no cases of confusion with odorants sharing the same number of carbon atoms but differing in their respective functional groups (cf. Figure 1). With the exception of one odor pair (1,5-pentandiol versus 2-butanone) there were even no cases of confusion between aliphatic alcohols (used as S+) and any aliphatic ketone or aldehyde at all. Similarly, 2-ketones (used as S+) were never confused with any aliphatic alcohol or aldehyde (cf. Figure 1).
With n-aldehydes used as S+, there were two cases of failure to significantly discriminate between odors on the basis of different oxygen moieties only (n-octanal versus 2octanone and n-decanal versus 2-decanone). Additionally, there were three cases of confusing an aliphatic aldehyde with 1,5-pentandiol (versus n-butanal, n-pentanal and n-nonanal used as S+ respectively) and five cases with alcohols or ketones having a different number of carbon atoms compared with the S+ (n-butanal versus 2-octanol, n-octanal versus 1-butanol, n-octanal versus 3-pentanol, n-nonanal versus 1-decanol and n-heptanal versus 2-butanone). Thus, with the aliphatic aldehydes (used as S+) there were more cases of failure to discriminate between substances with different functional groups ( n = 10) than cases of failure to distinguish between aldehydes, i.e. substances with the same functional group (n = 6).
The discrimination of non-aliphatic substances that are identical in structure except for their oxygen moieties was readily accomplished by all bees. Geraniol (a terpene alcohol), for example, was never confused with citral (syn. geranial, the corresponding terpene aldehyde), and similarly carvone (a terpene ketone) was easily discriminated from limonene (syn. carvene, the corresponding terpene hydrocarbon), irrespective of which member of these odor pairs was used as S+ or as S-.
Cyclic substances
The discrimination of aliphatic substances from cyclic substances led to only a few cases of confusion. Three cases concerned odor pairs which share an alcohol group as their oxygen moieties [2-nonanol versus (+)-menthol, geraniol versus 2-pentanol and 1,5-pentandiol versus (-)-menthol] and three cases involved odor pairs that differ in their functional groups [2-heptanol versus (-)-carvone, n-butanal versus (-)-menthol and (-)-menthol versus 2-nonanone].
The only two cases of failure to discriminate between two cyclic substances concerned a pair of enantiomers, (-)-menthol versus (+)-menthol, i.e. two substances that are identical in molecular structure except for chirality. Interestingly, the bees did not confuse the other two enantiomeric odor pairs (the optical antipodes of limonene and of carvone) employed (cf. Figure 1).
Odor mixtures
The discrimination of complex odor mixtures from other odorants again
presented little difficulty to the bees. The etheric oils employed were never confused with other
multicomponent odorants, irrespective of whether they were assigned as S+ or as S- (cf.
Figure 1). The only three cases of failure to significantly discriminate
between an odor mixture
and a monomolecular substance involved orange oil, which was confused by single bees with
citral, (-)-carvone and (-)-menthol respectively. The first two of these three
substances are known to be components of orange oil (Nursten, 1970).
Pheromone compounds
Five of the odorants used (geraniol, citral, isoamyl acetate, 2-heptanone and
2-nonanol) have been identified as components of honeybee pheromones (Free, 1987). Four of
these substances were only confused with one of the other 43 stimuli [geraniol with 2-pentanol,
citral with orange oil, isoamyl acetate with 1,5-pentandiol and
2-nonanol with (+)-menthol]. With the fifth pheromone compound used, 2-heptanone, there were
five cases of failure to significantly discriminate this substance from other odorants which all
concerned other, structurally similar, ketones (3-heptanone, 4-heptanone, 2-octanone, and two
cases with 2-hexanone; cf. Figure 1).
Readiness to accept a conditioned odor
Although not systematically assessed, it was evident that the vast majority of odor stimuli was readily accepted by the bees as their respective S+ in the course of the conditioning procedure. The only two substances which presented some difficulties in this respect were aliphatic aldehydes, n-nonanal and n-decanal. Several bees obstinately refused to leave the glass vials used to take them from the feeder to the bottle containing one of these substances when assigned as S+ (cf. Materials and methods) and to enter the neck of the bottle, and thus had to be replaced by conspecifics.
Training effects
Figure 4 shows the discrimination performance of the bees as a function of the number of test and extinction trials respectively. Performance in terms of number of errors per trial improved systematically across trials and thus a significant training effect at the group level was found (Spearman, P < 0.05). Interestingly, the mean number of errors in a given trial did not differ significantly between test and extinction trials (Wilcoxon, P > 0.05 in all trials). The fact that the mean number of errors in extinction trial 1 was lower than in extinction trial 2 can be attributed to the shorter duration of the former (cf. Materials and methods).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The results of the present study demonstrate (i) that honeybees have an excellent olfactory discrimination ability when tested in a free-flying situation with an array of structurally related monomolecular substances; (ii) a significant negative correlation between discrimination performance and structural similarity of odorants in terms of differences in carbon chain length with all substance classes tested; (iii) that both the position and type of a functional group also affected stimulus quality and thus discriminability of odorants, although in a substance class-specific manner; and (iv) striking similarities in odor structure–activity relationships between honeybees and human and nonhuman primates tested previously on a subset of substances employed here.
The excellent performance of A. mellifera found here is in agreement with
earlier studies on odor discrimination using free-flying honeybees (von Frisch, 1919) or conditioning of the proboscis extension reflex (Vareschi, 1971). However, whereas von Frisch's landmark study mainly employed etheric
oils, i.e. complex odor mixtures of high biological significance both in the contexts of nectar
source recognition (Pham-Delegue et al., 1993) and nestmate
recognition (Bowden et al., 1998), we could show that
honeybees are also able to clearly distinguish between a large number of monomolecular
substances in a paradigm
designed to simulate odor-guided foraging behavior. This ability has so far only been shown
employing proboscis extension reflex conditioning, which represents a far more artificial test
situation than our naturalistic experimental approach, and which has the disadvantage of
sequential rather than simultaneous presentation of stimuli. Despite the presumably more
challenging conditions of a free-flying test situation, the honeybees in our study scored an
average of >95% correct choices and thus showed almost exactly the same performance as in
Vareschi's study (95.5% correct choices), which employed restrained animals. It is
remarkable that our experimental set-up which required an animal to perform approach flights
and spatial orientation in addition to the actual discrimination task, and which exposed the bees
to the natural and ever-changing climate obviously did not impair performance.
Odor structure–activity relationships have so far only rarely been tested
systematically in honeybees. We found a significant negative correlation between discrimination
performance and structural similarity of odorants in terms of differences in carbon chain length
with all four homologous series of substances tested (1-alcohols, 2-alcohols, n-aldehydes, and 2-ketones; cf.Figure 2). This finding is in line with an
earlier report of proboscis extension
conditioning with carboxylic acids (Vareschi, 1971). Similar to our
findings, direct neighbors in
this homologous series were more difficult to distinguish for the bees than substance class
members that differed by two or more carbon atoms. In contrast to our findings, however,
Vareschi reported that carboxylic acids with a shorter carbon chain compared with the
conditioned odor led to more confusions than those with a longer carbon chain. More recently, it
has been shown that alcohols of similar carbon chain length elicited similar activity patterns in
the honeybee antennal lobe as visualized by in vivo calcium imaging (Sachse
et al., 1998). In an extension of this study (Sachse et al.,
submitted for publication) the authors found that regular
connections between this molecular property and glomerular activation patterns also occur with
aliphatic aldehydes, ketones, carboxylic acids and alkanes.
The same correlation between differences in carbon chain length and
discriminability as shown here has also been reported in several species of mammals.
Discrimination performance in squirrel monkeys and humans was found to vary systematically as
a function of structural similarity of aliphatic esters (Laska and Freyer, 1997), carboxylic acids (Laska and Teubner, 1998), 1-alcohols, n-aldehydes and 2-ketones (Laska and Teubner, 1999b)
(M. Laska et al., submitted for publication). It has also been reported (Linster and Hasselmo, 1998) that rats generalized between aliphatic aldehydes of
similar carbon chain length whereas
their response to less similar aldehydes was comparable to that for a control odor.
The results of these behavioral studies correspond with electrophysiological
findings which showed that tuning specificities of rodent olfactory receptor neurons correlate
with carbon chain length of aliphatic alcohols (Hirono et al., 1998; Sato et al., 1994) and
aldehydes (Zhao et al., 1998). Similarly, single cell recordings
from the rabbit olfactory bulb showed that molecular receptive ranges for both excitation and
inhibition of single mitral cells
are critically determined by the carbon chain lengths of alcohols (Imamura et al.,1992),
aldehydes (Mori, 1995; Yokoi et al.,1995) and ketones (Mori and Yoshihara, 1995).
Thus, both the behavioral and the electrophysiological findings suggest that the carbon chain length of aliphatic odorants—irrespective of their chemical class—is a critical determinant of the interaction between the stimulus molecule and its receptor in both vertebrates and insects.
With regard to the type of functional group, the second molecular feature investigated here, we found that stimulus quality and thus discriminability of odorants was affected in a substance class-specific manner. Whereas the discrimination of substances with the same or a similar carbon chain length but different oxygen moieties presented no difficulties to the bees when one of the discriminants was an alcohol or a ketone, there were several cases of failure when one of the odors in a given stimulus pair was an aldehyde. This suggests that substances with a functional aldehyde group might interact with a broader range of molecular receptors than substances belonging to one of the other two chemical classes tested.
In an attempt to quantify odor discrimination using electromyogram recordings
of the proboscis extension reflex with a smaller set of substances (three esters, four aldehydes,
and five ketones and alcohols), it was found (Smith and Menzel, 1989)
that alteration of carbon
chain length had more of an effect on perceptual similarity for alcohols and ketones than for
aldehydes and acetic esters. Our data correspond with this observation as the mean difference in
discrimination scores between 
C1 and C2 odor pairs was larger with alcohols
and ketones than with aldehydes (cf. Figure 2).
Based on single-cell recordings from antennal placodes, honeybee olfactory
receptor neurons could be assigned to 10 reaction groups which at least coarsely corresponded to
different chemical classes (Vareschi, 1971). These findings, however,
have been questioned in more recent studies (Akers and Getz, 1992; Getz and Akers, 1993) which failed to confirm
Vareschi's results employing more sophisticated recording techniques.
Investigations of human discrimination ability as a function of type of
functional group have so far concentrated largely on cyclic rather than aliphatic substances and
generally found perceptual similarity to be odor pairspecific, ranging from easily distinguishable
odor pairs to indiscriminable ones, with no recognizable correlation between odor quality and
this structural property of the stimulus (Ohloff et al., 1991;
Yoshii and Hirono, 1996).
Single-cell recordings from olfactory receptor cells in mice (Sato et al.,
1994)
and from mitral cells in the olfactory bulb of rabbits (Imamura et al., 1992) in response to stimulation with aliphatic substances from different chemical classes,
however, suggest
functional groups to be an important determinant of tuning specificity and thus response
selectivity of mammalian olfactory sensory neurons.
Position of a functional group, the third molecular feature investigated here, was also found to clearly affect odor quality, again in a substance class-specific manner. Whereas the bees had little difficulty in distinguishing between 1-alcohols and 2-alcohols, there were several cases of failure with isomeric forms of ketones. In both classes of substances, however, we found discrimination scores for isomeric odor pairs to be higher than for odor pairs sharing the position of their respective functional group but differing in chain length by one carbon atom (cf. Figure 3). This suggests that odor quality of aliphatic carbohydrons as perceived by honeybees is also critically determined by the position of functional alcohol and keto groups. Similar findings in human and nonhuman primates, which both readily discriminated between isomeric forms of alcohols and ketones, and performed better in these tasks compared with discriminations involving non-isomeric odor pairs which differed by one carbon atom (M. Laska et al., submitted for publication), again indicate that similar, if not identical, mechanisms of odor coding may apply to both vertebrates and insects.
Our finding that the discrimination of cyclic substances from aliphatic
substances led to only few cases of confusion concurs with the idea that overall molecular shape
is another important determinant of differential interaction of odor stimuli with olfactory
receptors (Yoshii and Hirono, 1996). The only cases of failure to
discriminate between two cyclic substances concerned the enantiomers of menthol, i.e.
substances that are identical in structure
except for chirality. The reasons why two other pairs of optical antipodes, limonene and carvone,
were readily discriminated by the bees remain to be revealed. It is interesting to note, however,
that exactly the same pattern of discrimination performance with these three pairs of enantiomers
has been found with human subjects (Laska and Teubner, 1999a). This
suggests that even
phylogenetically distant species might express the same or at least similar types of
enantioselective molecular odor receptors.
It is commonly agreed that the olfactory systems of both vertebrates and insects
are particularly adjusted to process complex odor mixtures as the vast majority of naturally
occurring odors in our environment comprise numerous compounds (Laing et al., 1989). Our
finding that the bees readily discriminated between all etheric oils employed is in line both with
this view and with von Frisch's study (von Frisch, 1919).
However, the fact that at least
some confusions between odor mixtures and monomolecular substances which are known to be
constituents of these mixtures occurred suggests that honeybees—similar to mammals
(Laska and Hudson, 1993)—may at least partially rely on key
compounds, i.e. monomolecular substances that characterize a Geruchsgestalt, to
recognize odor mixtures. This
assumption is supported by recent findings from a coupled gas chromatography–proboscis
extension reflex assay that showed honeybees to selectively respond to certain components of an
extract of oilseed rape floral volatiles after conditioning to the complete extract (Le
Metayer et al., 1997).
A final aspect of the present study is our finding that the discrimination
performance of the bees systematically improved across trials (cf. Figure 4). This is remarkable
considering that numerous studies have shown that honeybees are capable of one-trial associative
learning with odor stimuli (Menzel and Müller, 1996; Menzel et al., 1996), and
considering that our experimental procedure even included five conditioning trials in order to
allow an animal to build a robust association between the food reward and the conditioning odor
(cf. Materials and methods). The ability to correctly respond to a conditioned odor after a single
trial, however, does not necessarily imply that this learning process is sufficient to
reliably discriminate between qualitatively similar odors. Rather, it seems that several
encounters with the odors in question are necessary to fully exploit the discriminatory power of
the olfactory system. Thus, this finding is likely to illustrate the significance of experience for
odor discrimination (Rabin, 1988).
Taken together, our findings demonstrate that the similarities found in the structural organization of the olfactory systems of insects and vertebrates are paralleled by striking similarities in relative discrimination abilities. This strongly suggests that similar mechanisms of odor coding and discrimination may underlie olfaction in these two groups of animals. In order to further corroborate this hypothesis and to draw further conclusions as to the nature and generality of odor structure–activity relationships across the animal kingdom, it seems worthwhile to follow a comparative approach including both phylogenetically distant and closely related species, and to test their discrimination abilities using the same sets of stimuli.
|
Acknowledgments |
|---|
We thank Nadine Pochanke and Silvia Lee for their help in collecting data. This work was supported by a Heisenberg fellowship awarded to M.L. by the Deutsche Forschungsgemeinschaft, and by the SFB 515 grants to R.M. and M.G.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Akers, R.P. and Getz, W.G. (1992) A test of identified response classes among olfactory receptor neurones in the honey-bee worker. Chem. Senses , 17 , 191–209.
Boeckh, J., Distler, P., Ernst, K.D., Hösl, M. and Malun, D. (1990) Olfactory bulb and antennal lobe. In Schild, D. (ed.), Chemosensory Information Processing. Springer, Berlin, pp. 201–227.
Bowden, R.M., Williamson, S. and Breed, M.D. (1998) Floral oils: their effect on nestmate recognition in the honey bee, Apis mellifera. Insect. Soc. , 45 , 209–214.
Buck, L. and Axel, R. (1991) A novel multigene family may encode odorant receptors. Cell, 65 , 175–187.[Web of Science][Medline]
Free, J.B. (1987) Pheromones of Social Bees. Comstock Press, Ithaca, NY.
Frisch, K. von (1919) Über den Geruchssinn der Bienen und seine blütenbiologische Bedeutung. Zool. Jb., Abt. Allg. Zool. Physiol. , 37 , 1–238.
Getz, W.M. and Akers, R.P. (1993) Olfactory response characteristics and tuning structure of placodes in the honey bee, Apis mellifera. Apidologie , 24 , 195–217.
Hildebrand, J.G. and Shepherd, G.M. (1997) Mechanisms of olfactory discrimination: converging evidence for common principles across phyla. Annu. Rev. Neurosci. , 20 , 595–631.[Web of Science][Medline]
Hirono, J., Sato, T., Kataoka, Y., Ueda, H. and Kaetsu, I. (1998) Tuning specificities to various odorants in mouse olfactory receptor cells. Chem. Senses , 23 , 233.
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.
Joerges, J., Küttner, A., Galizia, C.G. and Menzel, R. (1997) Representation of odours and odour mixtures visualized in the honeybee brain. Nature , 387 , 285–288.
Knudsen, J.T., Tollsten, L. and Bergström, L.G. (1993) Floral scents—a checklist of volatile compounds isolated by head-space techniques. Phytochemistry , 33 , 253–280.[Web of Science]
Laing, D.G., Cain, W.S., McBride, R.L. and Ache, B.W. (eds) (1989) Perception of Complex Smells and Tastes. Academic Press, Sydney.
Laska, M. and Freyer, D.
(1997)
Olfactory discrimination ability for aliphatic esters in squirrel monkeys and humans. Chem. Senses
, 22
, 457–465.
Laska, M. and Hudson, R. (1993) Discriminating parts form the whole: determinants of odor mixture perception in squirrel monkeys, Saimiri sciureus. J. Comp. Physiol. A , 173 , 249–256.[Medline]
Laska, M. and Teubner, P.
(1998)
Odor structure–activity relationships of carboxylic acids correspond between squirrel
monkeys and humans. Am. J. Physiol.
, 274
, R1639–R1645.
Laska, M. and Teubner, P.
(1999a)
Olfactory discrimination ability of human subjects for ten pairs of enantiomers. Chem. Senses
, 24
, 161–170.
Laska, M. and Teubner, P.
(1999b)
Olfactory discrimination ability for homologous series of aliphatic alcohols and aldehydes. Chem. Senses, 24
, 263–270.
Le Metayer, M., Mario n-Poll, F., Sandoz, J.C., Pham-Delegue, M.H.,
Blight, M.M., Wadhams, L.J., Masson, C. and Woodcock, C.M.
(1997)
Effect of conditioning on discrimination of oilseed rape volatiles by the honeybee: use of a
combined gas chromatography–proboscis extension behavioural assay. Chem. Senses
, 22
, 391–398.
Linster, C. and Hasselmo, M.E. (1998) Generalization between n -aliphatic aldehydes in the rat. Chem. Senses , 23 , 641.
Menzel, R. and Müller, U. (1996) Learning and memory in honeybees: from behavior to neural substrates. Annu. Rev. Neurosci. , 19 , 379–404.[Web of Science][Medline]
Menzel, R., Hammer, M., Müller, U. and Rosenboom, H. (1996) Behavioral, neural and cellular components underlying olfactory learning in the honeybee. J. Physiol. Paris , 90 , 395–398.[Web of Science][Medline]
Mori, K. (1995) Relation of chemical structure to specificity of response in olfactory glomeruli. Curr. Opin. Neurobiol. , 5 , 467–474.[Web of Science][Medline]
Mori, K. and Yoshihara, Y. (1995) Molecular recognition and olfactory processing in the mammalian olfactory system. Prog. Neurobiol. , 45 , 585–619.[Web of Science][Medline]
Nursten, H.E. (1970) Volatile compounds: the aroma of fruits. In Hulme, A.C. (ed.), The Biochemistry of Fruits and their Products. Academic Press, London, pp. 239–268.
Nursten, H.E.(1977) The important volatile flavour components of foods. In Birch, G.G., Brennan, J.G. and Parker, K.J. (eds), Sensory Properties of Foods. Applied Science Publishers, London, pp. 151–166.
Ohloff, G., Winter, B. and Fehr, C. (1991) Chemical classification and structure–odour relationships. In Müller, P.M. and Lamparsky, D. (eds), Perfumes—Art, Science and Technology. Elsevier, London, pp. 287–330.
Passy, M.J. (1892) L'odeur dans la série des alcohols. C. R. Soc. Biol. , 4 , 447–449.
Pham-Delegue, M.H., Bailez, O., Blight, M.M., Masson, C.,
Picard-Nizou, A.L. and Wadhams, L.J.
(1993)
Behavioural discrimination of oilseed rape volatiles by the honey bee Apis mellifera. Chem. Senses
, 18
, 483–494.
Rabin, M.D. (1988) Experience facilitates olfactory quality discrimination. Percept. Psychophys. , 44 , 532–540.[Web of Science][Medline]
Sachse, S., Rappert, A., Galizia, C.G. and Menzel, R. (1998) Alcohols of similar carbon chain length elicit similar activity patterns in the honeybee brain. Eur. J. Neurosci. , 10 (Suppl.10) , 358.
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.
Siegel, S. and Castellan, N.J. (1988) Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, New York.
Smith, B.H. and Menzel, R. (1989) The use of electromyogram recordings to quantify odorant discrimination in the honey bee, Apis mellifera. J. Insect Physiol., 35 , 369–375.
Vareschi, E. (1971) Duftunterscheidung bei der Honigbiene—EinzelzellAbleitungen und Verhaltensreaktionen. Z. Vergl. Physiol. , 75 , 143–173.
Yokoi, M., Mori, K. and Nakanishi, S.
(1995)
Refinement of odor molecule tuning by dendrodendritic synaptic inhibition in the olfactory
bulb. Proc. Natl Acad. Sci. USA
, 92
, 3371–3375.
Yoshii, F. and Hirono, S.
(1996)
Construction of a quantitative three-dimensional model for odor quality using comparative
molecular field analysis (CoMFA). Chem. Senses
, 21
, 201–210.
Zhao, H., Ivic, L., Otaki, J.M., Hashimoto, M., Mikoshiba, K. and Firestein, S.
(1998)
Functional expression of a mammalian odorant receptor. Science
, 279
, 237–242.
Accepted April 24, 1999
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  E. K. Mwilaria, C. Ghatak, and K. C. Daly Disruption of GABAA in the Insect Antennal Lobe Generally Increases Odor Detection and Discrimination Thresholds Chem Senses, March 1, 2008; 33(3): 267 - 281. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Laska Olfactory Discrimination Ability for Aliphatic C6 Alcohols as a Function of Presence, Position, and Configuration of a Double Bond Chem Senses, November 1, 2005; 30(9): 755 - 760. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Linster, S. Sachse, and C. G. Galizia Computational Modeling Suggests That Response Properties Rather Than Spatial Position Determine Connectivity Between Olfactory Glomeruli J Neurophysiol, June 1, 2005; 93(6): 3410 - 3417. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Chaline, J.-C. Sandoz, S. J. Martin, F. L.W. Ratnieks, and G. R. Jones Learning and Discrimination of Individual Cuticular Hydrocarbons by Honeybees (Apis mellifera) Chem Senses, May 1, 2005; 30(4): 327 - 335. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Heiling, K. Cheng, and M. E. Herberstein Exploitation of floral signals by crab spiders (Thomisus spectabilis, Thomisidae) Behav. Ecol., March 1, 2004; 15(2): 321 - 326. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ditzen, J.-F. Evers, and C. G. Galizia Odor Similarity Does Not Influence the Time Needed for Odor Processing Chem Senses, November 1, 2003; 28(9): 781 - 789. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Linster, B. A. Johnson, A. Morse, E. Yue, and M. Leon Spontaneous versus Reinforced Olfactory Discriminations J. Neurosci., August 15, 2002; 22(16): 6842 - 6845. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Deisig, H. Lachnit, and M. Giurfa The Effect of Similarity between Elemental Stimuli and Compounds in Olfactory Patterning Discriminations Learn. Mem., May 1, 2002; 9(3): 112 - 121. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Sachse and C. G. Galizia Role of Inhibition for Temporal and Spatial Odor Representation in Olfactory Output Neurons: A Calcium Imaging Study J Neurophysiol, February 1, 2002; 87(2): 1106 - 1117. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Linster, B. A. Johnson, E. Yue, A. Morse, Z. Xu, E. E. Hingco, Y. Choi, M. Choi, A. Messiha, and M. Leon Perceptual Correlates of Neural Representations Evoked by Odorant Enantiomers J. Neurosci., December 15, 2001; 21(24): 9837 - 9843. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Laska, S. Ayabe-Kanamura, F. Hubener, and S. Saito Olfactory Discrimination Ability for Aliphatic Odorants as a Function of Oxygen Moiety Chem Senses, April 1, 2000; 25(2): 189 - 197. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T Valentincic, J Kralj, M Stenovec, A Koce, and J Caprio The behavioral detection of binary mixtures of amino acids and their individual components by catfish J. Exp. Biol., January 11, 2000; 203(21): 3307 - 3317. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||