Edited by: Carmen Sandi, Ecole Polytechnique Federale De Lausanne, Switzerland
Reviewed by: Judith Reinhard, University of Queensland, Australia; Fernando J. Guerrieri, Univesité François Rabelais Tours, France
*Correspondence: Maria Gabriela de Brito Sanchez, Centre de Recherches sur la Cognition Animale, Building 4R3, Université de Toulouse, 31062 Toulouse Cedex 9, France e-mail:
This article was submitted to the journal Frontiers in Behavioral Neuroscience.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Taste plays a crucial role in the life of honey bees as their survival depends on the collection and intake of nectar and pollen, and other natural products. Here we studied the tarsal taste of honey bees through a series of behavioral and electrophysiological analyses. We characterized responsiveness to various sweet, salty and bitter tastants delivered to gustatory sensilla of the fore tarsi. Behavioral experiments showed that stimulation of opposite fore tarsi with sucrose and bitter substances or water yielded different outcomes depending on the stimulation sequence. When sucrose was applied first, thereby eliciting proboscis extension, no bitter substance could induce proboscis retraction, thus suggesting that the primacy of sucrose stimulation induced a central excitatory state. When bitter substances or water were applied first, sucrose stimulation could still elicit proboscis extension but to a lower level, thus suggesting central inhibition based on contradictory gustatory input on opposite tarsi. Electrophysiological experiments showed that receptor cells in the gustatory sensilla of the tarsomeres are highly sensitive to saline solutions at low concentrations. No evidence for receptors responding specifically to sucrose or to bitter substances was found in these sensilla. Receptor cells in the gustatory sensilla of the claws are highly sensitive to sucrose. Although bees do not possess dedicated bitter-taste receptors in the tarsi, indirect bitter detection is possible because bitter tastes inhibit sucrose receptor cells of the claws when mixed with sucrose solution. By combining behavioral and electrophysiological approaches, these results provide the first integrative study on tarsal taste detection in the honey bee.
Taste is a fundamental sensory modality for individual survival as it allows discriminating edible from non-edible items, which may cause significant harm or death (Scott,
Insects, in particular the fruit fly
Yet, the sequencing of other insect genomes has shown that the taste organization of the fruit fly is not shared by all insects as different life styles led to modifications of the gustatory repertoire. In the honey bee
Behavioral and electrophysiological approaches have been used to characterize taste perception in bees (review in De Brito Sanchez,
Despite their long-claimed role in gustation (Frings and Frings,
Free-flying honey bee foragers (female workers),
Bees were mounted individually in small metal tubes from which only their head and fore-tarsi protruded (De Brito Sanchez et al.,
Each subject was checked for intact PER before starting the experiments. This was done by lightly touching the fore tarsi with a toothpick soaked with sucrose solution 1M without subsequent feeding. Care was taken to ensure that the toothpick contacted both the tarsus and the claws (Figure
The gustatory stimuli employed in behavioral experiments were distilled water, sucrose 1M, quinine hydrochloride (1/10/100 mM), salicin (1/10/100 mM) and caffeine (1/10/100 mM). All chemicals were obtained from Sigma-Aldrich (Saint-Quentin Fallavier, France).
We determined whether bitter substances (quinine, salicin, or caffeine) applied on the fore-tarsi exert an inhibitory effect on proboscis extension reflex (PER), an appetitive response triggered by prior tarsal stimulation with sucrose solution 1 M (Figure
We measured the proboscis response during the first 5 s of sucrose stimulation of one fore-tarsus and during the consecutive 5 s in which the second stimulation was delivered to the opposite fore-tarsus (bitter substance). The latter period allowed determining whether or not bees retracted the proboscis. Measuring retraction during the last 5 s of bitter stimulation, in the absence of sucrose, would be inappropriate as retraction could occur simply due to the absence of sucrose and not as a consequence of the bitter stimulation itself.
Three concentrations of bitter substance were consecutively assayed along trials: 1, 10, and 100 mM. The latter corresponds to a highly saturated solution. This increasing sequence was chosen to avoid fast response saturation. To control for possible sensitization induced by sucrose solution, we performed three control trials interspersed between the three bitter-substance trials. In these control trials, bees were stimulated during 10 s with water on one fore-tarsus, and with a dry toothpick during 10 s on the contralateral fore-tarsus. The interstimulus interval was also 5 s.
For each bitter substance, two groups of bees were assayed: for one group, sucrose solution 1 M or water was delivered on the right fore-tarsus and the bitter substance (quinine, salicin, or caffeine) or the dry toothpick on the left fore-tarsus; for the other group, stimulation sites were inversed. Experiments started with a control trial. Control and bitter-substance trials alternated so that each bee was subjected to six trials. The intertrial interval was 10 min.
We studied whether potential PER inhibition by a bitter substance (quinine or salicin) applied on one fore-tarsus can be overcome by sucrose solution 1 M applied on the contralateral fore-tarsus (Figure
A 10 s stimulation with quinine or salicin solution was applied on one fore-tarsus, followed by a 10 s stimulation with sucrose on the contralateral fore-tarsus. The interstimulus interval was 5 s (onset-onset) (Figure
Three concentrations of bitter substance were consecutively assayed along trials: 1, 10, and 100 mM. The latter corresponds to a highly saturated solution. This increasing sequence was chosen to avoid fast response saturation. To control for possible sensitization effects of sucrose solution, we performed three control trials that were interspersed between the three bitter-substance trials. In these control trials, bees were stimulated during 10 s with water on one fore-tarsus, and with a dry toothpick during 10 s on the contralateral fore-tarsus. The interstimulus interval was also 5 s. Experiments started always with a control trial. Control and bitter-substance trials alternated so that each bee was subjected to six trials. The intertrial interval was 10 min.
Two groups of bees (both for antennae-amputated and intact bees) were treated in this way: for one group, the bitter substance (quinine or salicin) of a given concentration or water was delivered on the right fore-tarsus, and sucrose solution 1 M or the dry toothpick on the left fore-tarsus; for the other group, stimulation sites were inversed.
Captured bees were placed in glass vials and cooled down on ice until they stopped moving. They were then mounted individually in Eppendorf tubes (Le Pecq, France) presenting a lateral slid through which a foreleg could be passed. The fixed bee was laid down horizontally, with the leg extended on a lateral support. The leg was fixed to the support by means of adhesive band tape in order to avoid movements. Bees fixed in this way were kept resting during 1 h before the start of the electrophysiological recordings.
The forelegs consist of six segments: the coxa, the trochanter, the femur, the tibia, the tarsus and the pretarsus. Figure
Electrophysiological recordings were performed on chaetic sensillae, which could be easily identified by their external morphology and which were located on the third and fourth tarsomeres of the tarsus, and at the level of the claws of the pretarsus.
The tastants employed were KCl, NaCl, sucrose, quinine hydrochloride (henceforth quinine), salicin and amygdalin. All chemicals were obtained from Sigma-Aldrich (Saint-Quentin Fallavier, France). Depending on the experiment, chemicals were diluted in a solution of KCl 0.1 or 0.01 mM, which was used as contact electrolyte. Solutions were kept at −4°C. To evaluate the effect of bitter compounds on sensillae responding to sucrose (De Brito Sanchez et al.,
A glass electrode with an external diameter of 10–20 μm was placed over a single taste sensillum. Electrodes were pulled from borosilicate glass capillaries. A chlorinated silver wire inserted into the contralateral eye was used as grounded reference electrode. The stimulating electrode was filled with the solution to be assayed (see above). The stimulation electrodes were stored in a humid chamber before use.
Stimuli were applied for 2 s with an interstimulus interval of 1 min. In some experiments stimulation lasted 5 s in order to favor recording of cellular responses to bitter substances such as quinine, which in some insects (e.g.,
We determined whether bitter substances applied on a fore-tarsus induce retraction of the proboscis once a proboscis extension reflex (PER) had occurred due to a prior stimulation with sucrose on the other fore-tarsus (Figure
Figures
In order to refine the analyses on the basis of individual responses, we distinguished three main classes of responses occurring within a trial: “
We analyzed whether stimulation with sucrose solution on a fore-tarsus overcomes a potential inhibitory effect of stimulation with a bitter-substance on the other fore-tarsus and thus triggers PER. Quinine (
Figures
An analysis of individual responses in terms of the three classes of response variations, “
These results might suggest that quinine and salicin are aversive and inhibit PER elicited by concomitant stimulation with sucrose because the average % of PER recorded was low (30% upon quinine and sucrose stimulation, and 24% upon salicin and sucrose stimulation). Yet, the low percentages of responsiveness to sucrose could be due to antennal amputation rather than reflecting an aversive nature of quinine and salicin. Antennal amputation has been shown to decrease sucrose responsiveness upon tarsal stimulation (De Brito Sanchez et al.,
Quinine (
The pattern of responses of intact bees upon stimulation with quinine and sucrose (Figure
Similarly to antenna-amputated bees (Figure
The analysis of individual responses of intact bees in terms of the
Thus, the low percentages of responsiveness to sucrose were not due to antennal amputation as the same percentages were obtained in experiments in which bees conserved their antennae. To determine whether the two bitter substances did indeed exert a partial inhibitory effect on PER upon sucrose stimulation due to their aversive nature, we performed a final control experiment with antenna-amputated bees in which we replaced the first stimulation with a bitter substance by stimulation with water on one fore tarsus, followed by stimulation with sucrose solution on the opposite fore tarsus (Figure
No significant differences were found between the subgroup that received water on the left fore tarsus (and sucrose on the right fore tarsus) and that receiving water on the right fore tarsus (and sucrose on the left fore tarsus) [
Bees did not show PER to water alone upon contact with one fore-tarsus, while they showed it when sucrose contacted the other fore tarsus. The % of PER upon simultaneous stimulation with water and sucrose on opposite fore tarsi was, however, only 40% so that responses of the water group did not differ significantly from those of the quinine and the salicin groups even if the lack of significance was marginal [
The lower levels of PER registered in all experiments (Figures
We recorded responses of gustatory receptor neurons located in chaetic sensilla of the third and fourth tarsomeres of the foreleg, upon 2-s stimulation with sucrose 1 M, quinine solution 1 and 10 mM, salicin 1 mM, KCl 0.1 mM and NaCl 100 mM. The contact electrolyte used for all solutions was KCl 0.1 mM which proved to be effective and did not elicit significant spiking activity
Figure
To determine whether responses to tastants could be distinguished from those to KCl in terms of their temporal course, we analyzed normalized responses along four consecutive bins of 500 ms each, starting at the onset and finishing at the offset of stimulation (Figure
The previous results suggest that responses to sucrose and to bitter substances of gustatory receptor neurons located in chaetic sensilla of the third and fourth tarsomeres of the foreleg were due to a receptor cell responding to KCl at very low concentrations.
To determine whether a KCl-receptor cell existed within the sensilla studied, we performed an experiment aimed at establishing a dose-response curve to KCl. We quantified responses to 2-s stimulations with five increasing concentrations of KCl: 0.01, 0.1, 1, 10, and 100 mM. Responses were standardized to responses to KCl 100 mM which induced in all cases maximal responsiveness. Figure
Responses increased significantly with KCl concentration [
To determine whether tarsomere sensilla host specific gustatory receptor neurons tuned to bitter substances and sucrose, we aimed at establishing dose-response curves for quinine, on the one hand, and for sucrose on the other hand.
To establish a dose-response curve for quinine, we varied quinine concentration but kept constant the concentration of the contact electrolyte (KCl 0.1 mM) in order to determine whether responses increased progressively, consistently with the presence of a bitter receptor within the sensilla studied, or remained constant if they were due to the KCl cell. We recorded responses to KCl 0.1 mM and to solutions of quinine 0.1, 1, 10, and 30 mM, all containing KCl 0.1 mM. Stimulations lasted 5 s due to the potential long latency of a putative “quinine cell” (Jørgensen et al.,
Responses normalized to those induced by the contact electrolyte increased significantly with quinine concentration [Figure
To establish a dose-response curve for sucrose, we diminished the concentration of the contact electrolyte to 0.01 mM to avoid interferences from the salt in the sucrose responses and to better visualize these responses. We varied sucrose concentration but kept constant the concentration of KCl in order to determine whether responses increased progressively because of the presence of a sucrose receptor cell, or remained constant because they were due to the KCl cell. We recorded responses to KCl 0.01 mM and to solutions of sucrose 1, 10, 100, and 1 M, all containing KCl 0.01 mM. Stimulations lasted 1 s. Figure
Responses normalized to those induced by the contact electrolyte did not vary significantly with sucrose concentration over four orders of magnitude (Figure
We finally aimed at determining the sensitivity of chaetic sensilla located on the claws to perform a comparative analysis with those recorded on the tarsomeres. We used KCl 0.01 mM as contact electrolyte and tested the effect of KCl, sucrose, quinine, amygdalin, and mixtures of sucrose and quinine, and of sucrose and amygdalin. Responses were normalized to response levels obtained for the contact electrolyte.
Figure
Figure
In the case of claw sensilla, a different pattern of responses was obtained. Figure
The differences between tarsomere and claw sensilla can also be visualized by comparing the temporal course of their responses to the tastants assayed. To this end, we analyzed normalized responses along the 5 s of stimulation (from stimulus onset to offset) (Figure
In the case of claw sensilla (Figure
The tarsal taste of the honey bee has remained until now mostly unexplored (De Brito Sanchez,
Unilateral stimulation of the tarsi with sucrose (Figure
The reciprocal experiment (Figures
Experiments 1 and 2 suggest that responsiveness to sucrose in conditions of dual tarsal stimulation with sucrose and a different tastant depends on the primacy and nature of the tastant assayed. Sucrose solution delivered first elicits immediate PER which cannot then be inhibited by any other substance delivered afterwards on a different tarsus. In these conditions, sucrose acts as a “winner takes-all” stimulus. When a different tastant is first perceived by one tarsus, sucrose has no such effect when delivered to the other tarsus, thus showing that temporal primacy is also important. In this case, sucrose still elicits appetitive responses but these are significantly diminished, thus revealing an unspecific central inhibition. This indicates that a process of central integration takes place, probably at the level of the thoracic ganglion. Because different legs were used in these experiments, differences in responses can only be due to gustatory cross-comparison and inhibition between opposite tarsi (Dethier and Bowdan,
Our electrophysiological recordings revealed that chaetic sensilla of the claws and of the tarsomeres differed in their response to sucrose solution 1 M. Tarsomere sensilla responded to sucrose solution in a similar way as to the contact electrolyte KCl (Figures
These conclusions were confirmed by the different responses obtained upon stimulation with mixtures of sucrose and bitter substances: while tarsomeres sensilla responded similarly to sucrose and to mixtures of sucrose and bitter substances (see Figure
Four substances that are perceived as bitter by humans—quinine, salicin, caffeine, and amygdalin—and that induce avoidance in the fruit fly (Meunier et al.,
Irrespectively of the concentration used, bitter substances delivered to one tarsus were unable to repress PER if sucrose was delivered to other fore tarsus, even if a lower level of PER was observed in these cases (Figures
We were unable to detect specific electrophysiological responses to quinine, salicin, and amygdalin in electrophysiological recordings of chaetic sensilla located both on the tarsomeres and claws of the fore legs. The responses to bitter substances recorded at the level of tarsomere sensilla were, in fact, responses of a gustatory receptor cell triggered by the contact electrolyte KCl (0.1 mM). The fact that in these sensilla, no variations in responses to quinine were recorded over 3 orders of magnitude (Figure
The lack of an evident aversive effect of bitter substances is consistent with the report of von Frisch who mentioned that bitter solutions have no noticeable effect on the bees' ingestion of sugar solutions (Von Frisch,
Chaetic sensilla on the tarsomeres exhibited high responsiveness to NaCl 100 mM (Figure
High sensitivity to saline solutions was also supported by the fact that tarsomere sensilla responded to extremely low KCl concentrations. In experiments 3 and 5, the contact electrolyte chosen was KCl 0.1 mM because higher concentrations of KCl (e.g., 10 mM) have been used in previous works as contact electrolyte to study gustatory responses in other appendages such as the antennae of bees and moths without inducing significant neural activity (De Brito Sanchez et al.,
Contrarily to sucrose sensitivity, saline sensitivity was higher on the tarsi than on the antennae as shown by the difference of three orders of magnitude between the concentrations of KCl used as efficient contact electrolytes in either case [10 mM for the antennae (De Brito Sanchez et al.,
Our results provide an integrative view of gustatory tarsal detection in the honey bee, where the gustatory modality has received less attention than other modalities such as vision or olfaction (De Brito Sanchez,
Two main lines of future research emerge from our findings. On the one hand, it is imperative to provide a better characterization of molecular gustatory receptors present on single gustatory receptor cells hosted in gustatory sensilla of the tarsi. The genome of the honey bee (The Honeybee Genome Sequencing Consortium,
On the other hand, our results indicate that gustatory information from different legs is subjected to cross-comparison and can trigger, depending on which kind of taste is first perceived, a central excitatory state (sucrose 1st, non-sucrose 2nd) or a central inhibitory state (non-sucrose 1st, sucrose 2nd). This result raises the question of the mechanisms of central taste processing, for instance at the level of the thoracic ganglion where information from opposite tarsi converges. Further studies such as those performed in the thoracic ganglia of the locust (Newland et al.,
Finally, the fact that the taste modality has been relatively neglected compared to other modalities such as vision and olfaction, which have been intensively studied in the honey bee (Galizia et al.,
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We thank two anonymous reviewers for comments and corrections on a previous version of our manuscript. This work was supported by the University Paul Sabatier (Travel Grants to Maria Gabriela de Brito Sanchez and Martin Giurfa), the CNRS, The Institut Universitaire de France (Martin Giurfa), the Xishuangbanna Tropical Botanic Garden, Chinese Academy of Sciences (Fanglin Liu) and Zhejiang University (Songkun Su).