![]() In vertebrates, it has been shown that prediction error coding dopaminergic neurons (DANs) are involved in learning ( Schultz et al., 1997 Schultz, 2016). It assumes that the efficacy of learning is determined by the momentary discrepancy (or error) between the expected and the received reinforcement (i.e., reward or punishment). The prediction error theory ( Rescorla and Wagner, 1972) describes a basic theoretical concept of classical conditioning. A single learning trial can be sufficient to form a stable memory in the fruit fly ( Beck et al., 2000 Krashes and Waddell, 2008 Zhao et al., 2019) and other insect species (see Discussion). Once an association has formed between the CS+ and the respective US, the learned anticipation of the US can be observed in a memory test that enforces a binary choice behavior between the CS+ and the CS– ( Tempel et al., 1983 Tully and Quinn, 1985). In the training phase flies are typically exposed to two odors (differential conditioning) where one odor is perceived alone whereas a second odor is presented together with either reward or punishment. In the temporal domain, our model achieves rapid learning with a step-like increase in the encoded odor value after a single pairing of the conditioned stimulus (CS) with a reward or punishment, facilitating single-trial learning.įruit flies can learn to associate an odor stimulus with a positive or negative consequence, e.g., food reward or electric shock punishment. Simulating the experimental blocking of synaptic output of individual neurons or neuron groups in the model circuit confirmed experimental results and allowed formulation of testable predictions. Subjecting our model to learning and extinction protocols reproduced experimental results from recent behavioral and imaging studies. A distinct set of four MBONs encodes odor valence and predicts behavioral model output. Recurrent modulation of plasticity through projections from MBONs to reinforcement-mediating dopaminergic neurons (DAN) implements a simple reward prediction mechanism. It employs plastic synaptic connections between Kenyon cells (KCs) and MB output neurons (MBONs) in separate and mutually inhibiting appetitive and aversive learning pathways. The model is tailored to the existing anatomic data and involves two circuit motives of central functional importance. Here, we propose a minimalistic circuit model of the Drosophila MB that supports classical appetitive and aversive conditioning and memory extinction. Recent experimental results in Drosophila melanogaster suggest that, after the behavioral extinction of a memory, two parallel but opposing memory traces coexist, residing at different sites within the mushroom body (MB). Insect models have been instrumental in uncovering fundamental processes of memory formation and memory update. Bar plots are presented as mean ± SD n = 10 independent models Download Figure 3-2, TIF file.Įxtinction learning, the ability to update previously learned information by integrating novel contradictory information, is of high clinical relevance for therapeutic approaches to the modulation of maladaptive memories. Boxplots show the median and the lower and upper quartiles, whiskers indicate 1.5 times interquartile range, outliers are marked with + symbol. ![]() The tuned model also achieves a stronger extinction effect in the appetitive learning paradigm ( A, C) however, it does not abolish appetitive memory completely. With the tuned model we can achieve a stronger aversive memory ( B, D) that fits quantitatively the experimental results (Table 1). Further, the factor ρ representing the inhibition of the PAM DAN was set to 0.5 for the aversive learning conditioning paradigm. ![]() In consequence, lateral inhibition between MBONs was appropriately adjusted: MVP2::M6 inhibition was increased to M 6 – = − 0.6 ( 1 + 200 × e ( − M V P 2 × 16 ) ), MV2::V2 inhibition was decreased to V 2 − = − 0.6 ( 1 + 200 × e ( − M V 2 × 13 ) ). In order to fit the model to the asymmetry in the experimental results, the PPL1 output was tuned towards a lower activation threshold with the sigmoid transfer function P P L 1 = 1 ( 1 + 10000 × e − P P L 1 I n p u t × 21 ). ![]() Extended Data Figure 3-2: Independent tuning of the model parameters for the appetitive and aversive pathways allows to match the experimental results more accurately. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |