Perineuronal nets in motor circuitry regulate the performance of learned vocalizations in songbirds

Abstract

The accurate and reliable performance of learned vocalizations (e.g., speech and birdsong) modulates the efficacy of communication in humans and songbirds. Consequently, it is critical to understand the factors that regulate the performance of learned vocalizations. Across taxa, neural circuits underlying motor learning and control are replete with perineuronal nets (PNNs), and we analyzed how PNNs in vocal motor circuitry regulate the performance of learned song in zebra finches. We report that developmental increases in PNN expression in vocal circuitry are associated with developmental increases in song stereotypy. We also document that enzymatically degrading PNNs in the motor nucleus HVC acutely altered song structure (changes in syllable sequencing and production). Collectively, our data reveal a causal contribution of PNNs to the performance of learned behaviors and, given the parallels in the regulation of birdsong and speech, suggest that PNNs in motor circuitry could modulate speech performance.

Fig. 1. Vocal performance (stereotypy) increases with age.

A Representative spectrograms [frequency (kHz; y-axis) vs time (x-axis) with brightness representing amplitude] of three song snippets (1.5 s) of the songs of a juvenile (left, 62 dph) and adult (right, 10 months) zebra finch. Scale bar: 0.5 s. B The acoustic stereotypy of song snippets was significantly higher for adult zebra finches than for juveniles (**p < 0.01). C Representative spectrograms of ten renditions of one syllable type in the songs of the juvenile and adult depicted in A. Scale bar: 0.1 s. D The acoustic stereotypy of individual syllables was significantly greater for adult zebra finches than for juveniles (*p < 0.05). Plotted are average %similarity scores across syllable types within each bird. The gray circles in B and D indicate the data for the juvenile and adult birds depicted in A and C.

Fig. 2. Perineuronal net expression in the song system increases with age.

A Schematic of the song system, highlighting the forebrain areas HVC, RA, LMAN, and Area X that are replete with PNNs. B Representative photomicrographs illustrating a PV neuron surrounded by a PNN (white arrow) and PV neuron without PNNs (yellow arrow) in RA. Scale bar: 20 μm. Age-related changes in C PNN density, D PNN intensity, E the percentage of PV neurons with PNNs, and F PV neuron density in HVC, RA, LMAN, and Area X (*p < 0.05, **p < 0.01).

Fig. 3. Degrading PNNs in HVC affected the ability to perform learned songs but did not affect vocal stereotypy.

A Representative photomicrographs illustrating PNN (green) degradation caused by ChABC (bottom row) and intact PNNs in control birds that received the injection of PEN (top row). The white dashed line outlines HVC in each panel (based on PV expression). Scale bar: 200 μm. B The percentage of HVC with PNN degradation was significantly higher in ChABC birds than in PEN birds (**p < 0.01). C There was no significant difference in PV neuron density in HVC between ChABC and PEN birds (p > 0.25). The similarity of post-surgery song snippets (D) and individual syllables (E) to pre-surgery songs or syllables. Following ChABC infusions, similarity to baseline song snippets was significantly decreased on post-surgery days 2 (n = 8; p < 0.05) and 3 (n = 9; p < 0.01), with a strong trend for post-surgery day 4 (n = 11; p < 0.06). This change was only significantly different on post-surgery day 1 (n = 4; p < 0.01) following PEN infusions. E Syllable structure was significantly different from baseline on a number of days following surgery for both ChABC- and PEN-treated birds. Only days in which birds produced at least 20 song renditions were analyzed and plotted here. **p < 0.01, *p < 0.05, ~p < 0.06. The acoustic stereotypy of song snippets (F) and individual syllables (G) was not affected by PNN degradation (p > 0.05).

Fig. 4. ChABC infusions cause significant changes to syllable probabilities.

A Representative spectrograms of three song snippets (1.5 s) with syllables (white letters) from a PEN bird (pre-surgery). Scale bar: 0.5 s. B Histograms of syllable probabilities within pre-surgery and post-surgery days for a ChABC bird (top row; pink) and a PEN bird (bottom row; blue). C The magnitude of changes in syllable probabilities (relative to baseline song) was significantly higher for ChABC birds than for PEN birds (p < 0.05; see text). Error bars represent mean ± S.E. The solid line connects the means of changes in syllable probabilities among ChABC birds and the dashed line connects that among PEN birds.