Skip to main content
. 2021 Jun 11;15:661225. doi: 10.3389/fnbeh.2021.661225

FIGURE 3.

FIGURE 3

Models of appetitive Pavlovian conditioning in crickets that account for the change from a US value-sensitive CR to a habitual one with extension of training. (A) A model of Pavlovian conditioning (Terao et al., 2015) that consists of four types of neurons: “CS” neurons that code the CS, two classes of OA or DA neurons (“OA1/DA1” neurons and “OA2/DA2” neurons) that code appetitive/aversive US, and “CR” neurons that produce the CR. The “OA1/DA1” neurons or “OA2/DA2” neurons receive inhibitory or excitatory synapses from “CS” neurons, the efficacy of which is strengthened by pairing of the CS and the US and by resulting simultaneous activation of their pre- and postsynaptic neurons (assuming Hebbian plasticity). The efficacy of excitatory synapses from “CS” neurons to “CR” neurons is strengthened when “CS” neurons and “OA1/DA1” neurons are activated at the same time (assuming Kandelian plasticity, Kandel, 2001), and “CR” neurons are activated when “CS” neurons and “OA2/DA2” neurons are activated at the same time (shown as AND gate). (B) A part of the model is shown for highlighting the roles of “OA2” neurons for execution of appetitive CR. (C) In this model, we revised our previous model in (B) to account for the finding that the CR is sensitive to US devaluation after standard training but not after extended training (Mizunami et al., 2019) and that the CR is initially specific to the condition of illumination under which the cricket received training, but the specificity is lost after extended training (Sato et al., 2021). We assume that activation of “OA2” neurons does not occur when the animal is satiated with the US or when the test is performed outside the context of training. Hence, a CR does not occur after US devaluation or outside the context of training after standard training. We also assume that the efficacy of “CS-CR” synapses is further enhanced by extended training, so that activation of “CR” neurons occurs without activation of “OA2” neurons, and hence the CR occurs after US devaluation or outside the context of training. (A,B) Modified from Mizunami et al. (2018). (C) Modified from Sato et al. (2021).