Auditory learning in an operant task with social reinforcement is dependent on neuroestrogen synthesis in the male songbird auditory cortex

Abstract

Animals continually assess their environment for cues associated with threats, competitors, allies, mates or prey, and experience is crucial for those associations. The auditory cortex is important for these computations to enable valence assignment and associative learning. The caudomedial nidopallium (NCM) is part of the songbird auditory association cortex and it is implicated in juvenile song learning, song memorization, and song perception. Like human auditory cortex, NCM is a site of action of estradiol (E2) and is enriched with the enzyme aromatase (E2-synthase). However, it is unclear how E2 modulates auditory learning and perception in the vertebrate auditory cortex. In this study we employ a novel, auditory-dependent operant task governed by social reinforcement to test the hypothesis that neuro-E2 synthesis supports auditory learning in adult male zebra finches. We show that local suppression of aromatase activity in NCM disrupts auditory association learning. By contrast, post-learning performance is unaffected by either NCM aromatase blockade or NCM pharmacological inactivation, suggesting that NCM E2 production and even NCM itself are not required for post-learning auditory discrimination or memory retrieval. Therefore, neuroestrogen synthesis in auditory cortex supports the association between sounds and behaviorally relevant consequences.

Keywords: Audition; Estradiol; Neuroestrogen; Nongenomic; Vocal learning; Zebra finch.

Figure 1.

(a) Auditory discrimination GO/NO-GO procedure. Birds initiate all trials. Solid arrows indicate trajectories that yield consequences (reward or punishment); dotted arrows indicate trajectories that do not yield consequences. Possibilities after a GO-trial are indicated by blue arrows, and the ones after a NO-GO-trial by red arrows. The bird has 2 seconds to respond after the stimulus is played. Reward consists of the activation of the polarized glass and visual access to the other bird for 6 seconds, while punishment consists of a loud burst of white noise and the switch being inoperant for 16 seconds. Lack of response to stimuli does not yield consequences. (b) Behavioral testing timeline. Go/No-go testing happened daily for 4 hours. Once birds reached learning criterion (>70% correct on two consecutive days) during training, treatments and the Learning stage started. Sound stimuli were switched (absolute and relative frequencies; see Methods) when birds reached criterion, except for during the Discrimination stage when learned tones were maintained. Shaded boxes represent the treatment stages in Figures 2, 5 (Learning), 3 and 6 (Discrimination).

Figure 2.

NCM fadrozole (FAD) injections impair performance in an auditory learning task. Performance was analyzed using combined (left), GO (middle) or NO-GO trials (right). (a-c) First vs last day whole-day performances. FAD decreased combined trial performance (a). There was a trend for a decrease in hit rates on the first day (b) and a trend for an overall reduction in rejection rates (c). (d-f) Logistic regressions. FAD decreased the area under the curve of the combined-trial performance (d). There was a trend for a reduction in the intercept of the GO trials (e). FAD reduced the area under the curve of the NO-GO trial performance (f). Insets in d-f are all from one representative subject. Dots around 0 and 1 represent incorrect and correct responses respectively, which are the data used for the logistic regressions. (g-i) Reversal learning (first 50 trials of day one versus last 50 of previous day). FAD impaired performance on the first day of combined trials (g) and of GO trials (h), but not NO-GO trials (i). *p < 0.05; **p < 0.01; ^#p < 0.1.

Figure 3.

Discrimination of previously learned tones is not impaired by NCM fadrozole (FAD) and baclofen/muscimol (BM) injections. BM was administered in 3 increasing volumes: 100, 200 and 500 nL. Neither drug affected discrimination of previously learned tones. Daily performance (different colors are different animals) is shown in (a); logistic regressions in (b, c). Insets in (a-c) show GO and NO-GO trial performances.

Figure 4.

Oral fadrozole (FAD) administrations before auditory learning task. Performance was analyzed using combined (left), GO (middle) or NO-GO trials (right). (a-c) First vs last day whole-day performances. FAD decreased combined-trial whole-day performance (a). Hit rates were unchanged, but (b) there was a trend for an overall reduction in rejection rates (c). (d-f) Logistic regressions. FAD did not affect any learning curves parameters, except for producing a trend for a higher intercept in the GO trials due to FAD. Insets in d-f are all from one representative subject. Dots around 0 and 1 represent incorrect and correct responses respectively, which are the data used for the logistic regressions. (g-i) Reversal learning (first 50 trials of day one vs last 50 of previous day). FAD did not affect combined trials (g), but there were trends for an increase in hit rates on the first day (h) and for a decrease in rejection rate on the first day (i). *p < 0.05; ^#p < 0.1.

Figure 5.

Discrimination of previously learned tones is not impaired by oral fadrozole. Whole-day combined-trial performance is shown in (a); combined-trial logistic regressions in (b, c). Insets show GO- and NO-GO-trial performances.