Reinforcement rate and the balance between excitatory and inhibitory learning: insights from deletion of the GluA1 AMPA receptor subunit

Abstract

Conditioned responding is sensitive to reinforcement rate. This rate-sensitivity is impaired in genetically modified mice that lack the GluA1 subunit of the AMPA receptor. A time-dependent application of the Rescorla–Wagner learning rule can be used to derive an account of rate-sensitivity by reflecting the balance of excitatory and inhibitory associative strength over time. By applying this analysis, the impairment in GluA1 knockout mice may be explained by reduced sensitivity to negative prediction error and thus, impaired inhibitory learning, such that excitatory associative strength is not reduced during the nonreinforced periods of a conditioned stimulus. The article describes a test of the role of GluA1 in inhibitory learning that requires summing of the associative strengths of cues presented in compound. Mice were trained on a feature negative discrimination of the form A+/AX–. GluA1 knockout mice acquired the discrimination to a similar extent as controls. The inhibitory properties of cue X were verified in a summation test that included a control for nonassociative, external inhibition. The performance of GluA1 knockout mice was similar to that of controls. However, in line with previous findings, GluA1 deletion impaired the precision of timing of conditioned responding. These results provide further evidence that impaired sensitivity to reinforcement rate is not a consequence of impaired inhibitory learning. The results may more readily fit with accounts of rate sensitivity that propose that it reflects encoding of temporal and numeric information rather than being a consequence of changes in associative strength over time.