Portal de Periódicos da CAPES

Associative Learning: The Instructive Function of Biogenic Amines

2006; Elsevier BV; Volume: 16; Issue: 20 Linguagem: Inglês

10.1016/j.cub.2006.09.021

1879-0445

Martín Giurfa,

Olfactory and Sensory Function Studies

Biogenic amines like dopamine or octopamine modify neural function at multiple levels, sensitizing or depressing behaviour. Recent studies in insects have now shown that, besides a role in motivational modulation, biogenic amines substitute the reinforcer function in associative learning, thus instructing the nervous system about the relevance of external events. Biogenic amines like dopamine or octopamine modify neural function at multiple levels, sensitizing or depressing behaviour. Recent studies in insects have now shown that, besides a role in motivational modulation, biogenic amines substitute the reinforcer function in associative learning, thus instructing the nervous system about the relevance of external events. The ability to undergo associative learning is widespread among animals and makes it possible for an individual to extract the logical structure of the world. Two major forms of associative learning are usually recognized: in classical conditioning [1Pavlov I.P. Lectures on Conditioned Reflexes. International publishers, New York1927Google Scholar], an animal learns to associate an originally neutral, ‘conditioned’ stimulus (CS) with a biologically relevant, ‘unconditioned’ stimulus (US); in operant conditioning [2Skinner B.F. The Behavior of Organisms. Appleton, New York1938Google Scholar], the animal learns to associate its own behaviour as anticipatory of some reinforcement. Both forms of learning therefore allow an animal reliably to predict reinforcement. How the reinforcement is represented in the central nervous system is a critical issue in the neurobiology of learning. Recent studies in two insect species [3Schwaerzel M. Monastirioti M. Scholz H. Friggi-Grelin F. Birman S. Heisenberg M. Dopamine and octopamine differentiate between aversive and appetitive olfactory memories in Drosophila.J. Neurosci. 2003; 23: 10495-10502PubMed Google Scholar, 4Unoki S. Matsumoto Y. Mizunami M. Participation of octopaminergic reward system and dopaminergic punishment system in insect olfactory learning revealed by pharmacological study.Eur. J. Neurosci. 2005; 22: 1409-1416Crossref PubMed Scopus (167) Google Scholar, 5Riemensperger T. Völler T. Stock P. Buchner E. Fiala A. Punishment prediction by dopaminergic neurons in Drosophila.Curr. Biol. 2005; 15: 1953-1960Abstract Full Text Full Text PDF PubMed Scopus (234) Google Scholar] have shown that two monoamines, octopamine and dopamine, can substitute respectively for the appetitive (reward) and aversive (punishment) reinforcements used in olfactory conditioning. Blocking octopaminergic and dopaminergic neuronal activity abolishes appetitive and aversive learning, showing that these amines are necessary for these learning forms. Using a remote-control technique which allows selected sets of octopaminergic or dopaminergic neurons to be activated by light in Drosophila larvae, Schroll et al.[6Schroll C. Riemensperger T. Bucher D. Ehmer J. Völler T. Erbguth K. Gerber B. Hendel T. Nagel G. Buchner E. et al.Light-induced activation of distinct modulatory neurons substitutes for appetitive or aversive reinforcement during associative learning in larval Drosophila.Curr. Biol. 2006; 16: 1741-1747Abstract Full Text Full Text PDF PubMed Scopus (434) Google Scholar] have now shown that short-term phasic activation of biogenic amine systems is not only necessary but sufficient for substituting reinforcement in associative learning. As the authors reported recently in Current Biology, dopamine mediates aversive reinforcement, and octopamine mediates appetitive reinforcement, in olfactory conditioning of Drosophila larvae. These results underline the power of invertebrate models in the analysis and understanding of the principles governing associative learning and memory at the behavioural, cellular and molecular levels. This power is based on the existence of learning and memory capabilities that, in some cases, are easily amenable to laboratory protocols. Furthermore, invertebrates possess a relatively simple nervous system that makes it possible to retrace associative phenomena to neuronal networks or even single neurons. In the case of insects, the fruit fly Drosophila melanogaster and the honeybee Apis mellifera offer robust conditioning protocols for studying associative learning. Moreover, the neurobiology of their olfactory circuitry is well known [7Heisenberg M. Mushroom body memoir: from maps to models.Nature Rev. Neurosci. 2003; 4: 266-275Crossref PubMed Scopus (874) Google Scholar, 8Menzel R. Memory dynamics in the honeybee.J. Comp. Physiol. A. 1999; 185: 323-340Crossref Scopus (452) Google Scholar]. After short training, these insects learn to associate odorants with aversive or appetitive reinforcers, and after extended training they can solve sophisticated discrimination tasks [9Giurfa M. Cognitive neuroethology: dissecting non-elemental learning in a honeybee brain.Curr. Opin. Neurobiol. 2003; 13: 726-735Crossref PubMed Scopus (173) Google Scholar]. The possibility of studying changes in neuronal activity at different stages of the olfactory circuit as resulting from associative learning is important, because it makes it possible to determine how, where and when learning modifies the representation of the CS and the US in the insect brain. CS processing starts at sensory neurons on the antennae (and also the maxillary palp in the fruit fly), which project to the antennal lobe, where they terminate in morphologically discrete, synapse-dense areas called ‘glomeruli’ (Figure 1). Glomeruli are the functional units of the antennal lobe where sensory neurons expressing the same olfactory receptor synapse onto local interneurons and projecting neurons that convey the olfactory message to two higher-order brain centers, the mushroom bodies and the lateral horns. In addition to classical electrophysiological techniques, optophysiological methods have made it possibe to quantify in vivo the neuronal activity evoked by olfactory stimulation. Use of these methods showed that an odorant is represented by the activation of overlapping sets of olfactory sensory neurons, followed by the activation of overlapping sets of projection neurons, ultimately leading to odorant-specific activation patterns in the mushroom bodies and lateral horn [7Heisenberg M. Mushroom body memoir: from maps to models.Nature Rev. Neurosci. 2003; 4: 266-275Crossref PubMed Scopus (874) Google Scholar, 8Menzel R. Memory dynamics in the honeybee.J. Comp. Physiol. A. 1999; 185: 323-340Crossref Scopus (452) Google Scholar]. Following classical conditioning, qualitative and quantitative changes can be observed in the odorant representation in the insect brain (see [7Heisenberg M. Mushroom body memoir: from maps to models.Nature Rev. Neurosci. 2003; 4: 266-275Crossref PubMed Scopus (874) Google Scholar, 8Menzel R. Memory dynamics in the honeybee.J. Comp. Physiol. A. 1999; 185: 323-340Crossref Scopus (452) Google Scholar, 10Galizia C.G. Menzel R. The role of glomeruli in the neural representation of odours: results from optical recording studies.J. Insect Physiol. 2001; 47: 115-129Crossref PubMed Scopus (146) Google Scholar] for reviews). Less is known about how the US is represented in the insect brain. In the honeybee, where only appetitive learning is amenable to the cellular or molecular level, the activity of a single, identified neuron, VUMmx1 (the ventral unpaired median neuron of the maxillary neuromere 1), can substitute for the reinforcing function of the US [11Hammer M. An identified neuron mediates the unconditioned stimulus in associative learning in honeybees.Nature. 1993; 366: 59-63Crossref PubMed Scopus (491) Google Scholar] in the olfactory conditioning of the proboscis extension response [12Bitterman M.E. Menzel R. Fietz A. Schäfer S. Classical conditioning of proboscis extension in honeybees (Apis mellifera).J. Comp. Psychol. 1983; 97: 107-119Crossref PubMed Scopus (801) Google Scholar]. In this conditioning procedure, a bee learns to associate an odorant (CS) delivered to its antennae, and a reward of sucrose solution (US) delivered to the antennae and mouth pieces [12Bitterman M.E. Menzel R. Fietz A. Schäfer S. Classical conditioning of proboscis extension in honeybees (Apis mellifera).J. Comp. Psychol. 1983; 97: 107-119Crossref PubMed Scopus (801) Google Scholar]. Sucrose solution on the antennae elicits the extension of the proboscis. Thus, a bee having learned the CS–US association extends the proboscis to the presentation of the CS alone. Importantly, although VUMmx1 is activated by sucrose stimulation of antennae and mouth parts, its activation does not elicit proboscis extension by itself [11Hammer M. An identified neuron mediates the unconditioned stimulus in associative learning in honeybees.Nature. 1993; 366: 59-63Crossref PubMed Scopus (491) Google Scholar]. Rather, VUMmx1 converges with the olfactory circuit (CS circuit) at the three sites mentioned above — the antennal lobes, the mushroom bodies and the lateral horn (Figure 1) — thus ‘instructing’ the olfactory system about the presence of a reinforcing stimulus. Indeed, activity of the VUMmx1 neuron substitutes for the US of sucrose, because pairing of an odorant with intracellular stimulation of the neuron (without any sucrose delivery) results in conditioned odour-evoked proboscis extension [11Hammer M. An identified neuron mediates the unconditioned stimulus in associative learning in honeybees.Nature. 1993; 366: 59-63Crossref PubMed Scopus (491) Google Scholar]. VUMmx1 belongs to a group of octopamine-immunoreactive neurons [13Kreissl S. Eichmüller S. Bicker G. Rapus J. Eckert M. Octopamine-like immunoreactivity in the brain and suboesophageal ganglion of the honeybee.J. Comp. Neurol. 1994; 348: 583-595Crossref PubMed Scopus (141) Google Scholar]. Pairing an odorant with injections of octopamine as a substitute for sucrose into the mushroom bodies or the antennal lobes (but not the lateral horn) lobe produced a lasting, learning-dependent enhancement of proboscis extension [14Hammer M. Menzel R. Multiple sites of associative odor learning as revealed by local brain microinjections of octopamine in honeybees.Learn. Memory. 1998; 5: 146-156PubMed Google Scholar]. Thus, octopamine signalling via VUMmx1 is sufficient to substitute for sugar reinforcement in honeybees. This conclusion was confirmed by silencing octopaminergic receptor expression in the honeybee antennal lobe using double-stranded RNA [15Farooqui T. Robinson K. Vaessin H. Smith B.H. Modulation of early olfactory processing by an octopaminergic reinforcement pathway in the honeybee.J. Neurosci. 2003; 23: 5370-5380PubMed Google Scholar]. This treatment inhibited olfactory acquisition and recall, but did not disrupt odorant discrimination. Similarly, Drosophila mutants in which the biosynthetic pathway to octopamine is blocked cannot learn to associate an odour with a sugar reward [3Schwaerzel M. Monastirioti M. Scholz H. Friggi-Grelin F. Birman S. Heisenberg M. Dopamine and octopamine differentiate between aversive and appetitive olfactory memories in Drosophila.J. Neurosci. 2003; 23: 10495-10502PubMed Google Scholar]. They can, however, learn an aversive olfactory discrimination, in which they have to avoid an odorant previously paired with an electric shock [3Schwaerzel M. Monastirioti M. Scholz H. Friggi-Grelin F. Birman S. Heisenberg M. Dopamine and octopamine differentiate between aversive and appetitive olfactory memories in Drosophila.J. Neurosci. 2003; 23: 10495-10502PubMed Google Scholar]. Conversely, transgenic flies in which synaptic output from dopaminergic neurons was blocked are deficient in aversive but not in appetitive learning [3Schwaerzel M. Monastirioti M. Scholz H. Friggi-Grelin F. Birman S. Heisenberg M. Dopamine and octopamine differentiate between aversive and appetitive olfactory memories in Drosophila.J. Neurosci. 2003; 23: 10495-10502PubMed Google Scholar]. Thus, octopamine is necessary for appetitive olfactory learning while dopamine is required for aversive olfactory learning. Similar results were found for olfactory learning in crickets [4Unoki S. Matsumoto Y. Mizunami M. Participation of octopaminergic reward system and dopaminergic punishment system in insect olfactory learning revealed by pharmacological study.Eur. J. Neurosci. 2005; 22: 1409-1416Crossref PubMed Scopus (167) Google Scholar]: pharmacological blocking of octopaminergic receptors impaired the acquisition of appetitive, but not aversive, olfactory learning, while the opposite was found for the blocking of dopaminergic receptors. In the Drosophila brain, dopaminergic neurons capable of substituting and predicting aversive reinforcement have been identified [5Riemensperger T. Völler T. Stock P. Buchner E. Fiala A. Punishment prediction by dopaminergic neurons in Drosophila.Curr. Biol. 2005; 15: 1953-1960Abstract Full Text Full Text PDF PubMed Scopus (234) Google Scholar]. These neurons respond originally to electric shock, and not to the odorant used for conditioning. After conditioning, however, the neurons respond to shock-predictive odorants, as predicted by associative learning theories [1Pavlov I.P. Lectures on Conditioned Reflexes. International publishers, New York1927Google Scholar] which posit that a learned CS can access some of the circuits that were previously engaged only by the US. It has been unclear, however, whether octopamine or dopamine signalling is sufficient for reinforcement processing. In a study that is a technical tour de force, Schroll et al.[6Schroll C. Riemensperger T. Bucher D. Ehmer J. Völler T. Erbguth K. Gerber B. Hendel T. Nagel G. Buchner E. et al.Light-induced activation of distinct modulatory neurons substitutes for appetitive or aversive reinforcement during associative learning in larval Drosophila.Curr. Biol. 2006; 16: 1741-1747Abstract Full Text Full Text PDF PubMed Scopus (434) Google Scholar] have addressed this question in Drosophila larvae, an insect model that has become popular for the study of associative olfactory learning. This popularity is explained by the simple, yet adult-like nervous system of larvae, and the possibility to study learning while exploiting the neurogenetic tools available in the fruit fly. Schroll et al.[6Schroll C. Riemensperger T. Bucher D. Ehmer J. Völler T. Erbguth K. Gerber B. Hendel T. Nagel G. Buchner E. et al.Light-induced activation of distinct modulatory neurons substitutes for appetitive or aversive reinforcement during associative learning in larval Drosophila.Curr. Biol. 2006; 16: 1741-1747Abstract Full Text Full Text PDF PubMed Scopus (434) Google Scholar] used transgenically expressed channelrhodopsin-2, a light-activated cation channel, to stimulate populations of octopaminergic or dopaminergic neurons by means of light and substitute for appetitive or aversive US function, respectively. Such an approach is feasible because the larval cuticle is essentially transparent. In an olfactory discrimination assay performed in a Petri dish [16Gerber B. Hendel T. Outcome expectations drive learned behaviour in larval Drosophila.Proc. Roy, Soc. Lond. B. 2006; (in press)PubMed Google Scholar], larvae were trained to discriminate two odorants: one paired either with appetitive (fructose) or aversive (salt) US, and a second odorant without any US. Then, the larvae were tested in a dual-choice situation with both odorants: the previously reinforced odorant and the non-reinforced one. Larvae trained under an appetitive regime (with fructose as the US) moved towards the previously reinforced odorant in the test, searching for the appetitive reinforcement that is absent from the Petri dish. Larvae trained under an aversive regime (with salt as the US) avoided the previously reinforced odorant in the test to escape the aversive US in the Petri dish. But the most striking result was that replacing the US by blue-light-induced activation of dopaminergic or octopaminergic neurons yielded similar results: activation of octopaminergic neurons substituted for the reinforcing function of the appetitive US, while activation of dopaminergic neurons substituted for the reinforcing function of the aversive US [6Schroll C. Riemensperger T. Bucher D. Ehmer J. Völler T. Erbguth K. Gerber B. Hendel T. Nagel G. Buchner E. et al.Light-induced activation of distinct modulatory neurons substitutes for appetitive or aversive reinforcement during associative learning in larval Drosophila.Curr. Biol. 2006; 16: 1741-1747Abstract Full Text Full Text PDF PubMed Scopus (434) Google Scholar]. Besides the technical highlight of ‘remote controlling’ defined subsets of neurons non-invasively, the paper by Schroll et al.[6Schroll C. Riemensperger T. Bucher D. Ehmer J. Völler T. Erbguth K. Gerber B. Hendel T. Nagel G. Buchner E. et al.Light-induced activation of distinct modulatory neurons substitutes for appetitive or aversive reinforcement during associative learning in larval Drosophila.Curr. Biol. 2006; 16: 1741-1747Abstract Full Text Full Text PDF PubMed Scopus (434) Google Scholar] shows that octopamine and dopamine are sufficient to substitute for different reinforcement functions in Drosophila olfactory associative learning. The authors conclude that these two modulatory systems are “causative for opposite types of learning in insects”. Though attractive, this conclusion has nevertheless to be taken cautiously as it has been shown to hold so far only for two insect species: the fruit fly [3Schwaerzel M. Monastirioti M. Scholz H. Friggi-Grelin F. Birman S. Heisenberg M. Dopamine and octopamine differentiate between aversive and appetitive olfactory memories in Drosophila.J. Neurosci. 2003; 23: 10495-10502PubMed Google Scholar, 5Riemensperger T. Völler T. Stock P. Buchner E. Fiala A. Punishment prediction by dopaminergic neurons in Drosophila.Curr. Biol. 2005; 15: 1953-1960Abstract Full Text Full Text PDF PubMed Scopus (234) Google Scholar, 6Schroll C. Riemensperger T. Bucher D. Ehmer J. Völler T. Erbguth K. Gerber B. Hendel T. Nagel G. Buchner E. et al.Light-induced activation of distinct modulatory neurons substitutes for appetitive or aversive reinforcement during associative learning in larval Drosophila.Curr. Biol. 2006; 16: 1741-1747Abstract Full Text Full Text PDF PubMed Scopus (434) Google Scholar] and crickets [4Unoki S. Matsumoto Y. Mizunami M. Participation of octopaminergic reward system and dopaminergic punishment system in insect olfactory learning revealed by pharmacological study.Eur. J. Neurosci. 2005; 22: 1409-1416Crossref PubMed Scopus (167) Google Scholar]. In the honeybee, only half of the story is known — the substitution of appetitive reinforcement by octopamine — because there is no protocol for studying aversive learning in the laboratory in a way that allows analyses at cellular and molecular levels to be made in parallel to behavioural studies. So the role of biogenic amines in aversive learning in bees has not been tested yet. Thus, besides regulating behaviour and motivation, the activity of modulatory, aminergic neurons serves, in the insect species tested so far, as a value system in associative learning phenomena — as a system that allows the ordering, prioritizing and assigning of ‘good’ or ‘bad’ labels to odorants. This instructive function is, in principle, different from that proposed in models which posit that biogenic amines control behavioural motivation by modifying neural function at multiple levels, thereby enhancing or depressing ongoing behaviour in appropriate ways and contexts [17Huber R. Amines and motivated behaviors: a simpler systems approach to complex behavioral phenomena.J. Comp. Physiol. A. 2005; 191: 231-239Crossref Scopus (57) Google Scholar, 18Libersat F. Pfluger H.-J. Monoamines and the orchestration of behavior.Bioscience. 2004; 54: 17-25Crossref Scopus (115) Google Scholar]. Schroll et al.[6Schroll C. Riemensperger T. Bucher D. Ehmer J. Völler T. Erbguth K. Gerber B. Hendel T. Nagel G. Buchner E. et al.Light-induced activation of distinct modulatory neurons substitutes for appetitive or aversive reinforcement during associative learning in larval Drosophila.Curr. Biol. 2006; 16: 1741-1747Abstract Full Text Full Text PDF PubMed Scopus (434) Google Scholar] provide an important result for this distinction: while the activity of octopaminergic or dopaminergic neurons substitutes for appetitive or aversive reinforcement, respectively, it does not change odour perception or locomotor activity. Thus, these monoamines do not up-regulate or down-regulate these behaviours in a non-specific way, rather they act specifically as an instructive element in associative learning phenomena. Octopamine and dopamine do not always substitute for appetitive and aversive reinforcement, respectively. In the mollusc Aplysia, for instance, the US pathway uses dopamine as transmitter both in classical and operant appetitive conditioning and direct application of dopamine can mimic appetitive reinforcement [19Brembs B. Lorenzetti F.D. Reyes F.D. Baxter D.A. Byrne J.H. Operant reward learning in Aplysia: neuronal correlates and mechanisms.Science. 2002; 296: 1706-1709Crossref PubMed Scopus (205) Google Scholar]. In that sense, the reinforcing function of dopamine in Aplysia is more similar to the role of dopamine in the mammalian brain, where it appears to mediate appetitive reinforcement, at least in the context of motor learning [20Mirenowicz J. Schultz W. Preferential activation of midbrain dopamine neurons by appetitive rather than aversive stimuli.Nature. 1996; 379: 449-451Crossref PubMed Scopus (621) Google Scholar]. Further studies involving other species are necessary to explain the apparent discrepancy between the insects tested so far on one hand, and Aplysia and mammals on the other hand, with respect to the reinforcing role of dopamine.