Motor circuits are required to encode a sensory model for imitative learning

Abstract

Premotor circuits help generate imitative behaviors and can be activated during observation of another animal's behavior, leading to speculation that these circuits participate in sensory learning that is important to imitation. Here we tested this idea by focally manipulating the brain activity of juvenile zebra finches, which learn to sing by memorizing and vocally copying the song of an adult tutor. Tutor song-contingent optogenetic or electrical disruption of neural activity in the pupil's song premotor nucleus HVC prevented song copying, indicating that a premotor structure important to the temporal control of birdsong also helps encode the tutor song. In vivo multiphoton imaging and neural manipulations delineated a pathway and a candidate synaptic mechanism through which tutor song information is encoded by premotor circuits. These findings provide evidence that premotor circuits help encode sensory information about the behavioral model before shaping and executing imitative behaviors.

Figure 1. Testing the role of premotor circuits in sensory learning in songbirds

a, Song learning (upper panel) in juvenile male zebra finches comprises a sensory learning phase, during which the pupil memorizes the song of a tutor, and a longer sensorimotor learning phase, during which the pupil uses auditory feedback to match its song to the memorized model. The brain regions important to sensory learning could be restricted to auditory circuits or also require the participation of motor circuits. b, Dorsal view of the zebra finch brain (left panel) and a parasagittal view through the medial forebrain (right panel) showing song premotor circuitry (red), including HVC, and auditory circuitry (blue). A1, primary auditory regions (Field L); CM, caudal mesopallium; A2–3, secondary and tertiary auditory regions; Area X, striatal component of the song system; HVC (used here as a proper name); NIf, nucleus interface of the nidopallium; RA, robust nucleus of the arcopallium; VMNs, vocal motor neurons. c, Schematic of tutor song-contingent disruption of neural activity in the pupil’s brain.

Figure 2. Optogenetic disruption of neural activity in the pupil’s HVC during tutoring impairs copying

a, Dorsal view of the finch brain showing bilateral viral delivery of hChR2-YFP to the song nucleus HVC. b, Parasagittal section through HVC showing neuronal expression of hChR2-YFP immunoreacted with anti-GFP, 11 days after scAAV-hChR2-YFP injection into the same region; scale bar = 100μm; LTV, lateral telencephalic ventricle. c, In vivo extracellular recording of light-evoked action potentials (473nm, 500ms, 10 trials) in the HVC of a juvenile zebra finch injected with HSV-hChR2. d, Sketch of the experimental timeline in which activity in the pupil’s HVC is optogenetically disrupted while the tutor is singing, but not at other times. e, Sonograms of a tutor’s song and the adult songs of two of his pupils, including a control and one that received optogenetic activation of HVC during tutoring (ChR2); scale bar = 200ms; ordinate = 0 – 9 kHz. f, Optogenetic disruption of a juvenile finch’s HVC only when its tutor is singing disrupts subsequent copying of the tutor’s song (green-experimental; black-controls; p = 1.7 × 10⁻⁶; green filled diamond–average for birds raised in isolation from a tutor (n = 3 birds); black filled diamond-average for birds raised with free access to the same tutor used for optogenetic experiments (n = 3 birds); diamond plot whiskers denote 10–90% range of similarity scores for each bird; learning outcomes measured in adulthood).

Figure 3. Tutor song syllable-triggered microstimulation of HVC disrupts copying of the targeted syllable

a, Sketch of the experimental design in which the pupil’s HVC is microstimulated (20μA per HVC, biphasic pulses, 300 μs each phase at 170Hz for 200ms) while the tutor is singing syllable ‘c’. b, Pupils fail to imitate the syllable paired with HVC microstimulation (syllable c; F(4,14) = 7.508, P = 0.001; n = 4 birds; notched box plot whiskers = 1.5 standard deviations). c, Sonograms of the tutor’s song and the adult song of one his pupils that was microstimulated in HVC when the tutor sang syllable ‘c’. Green bar under syllable ‘c’ and scale bar at lower right = 130ms; ordinate = 0 – 9 kHz.

Figure 4. Blocking NMDA receptors in HVC during tutoring prevents spine enlargement and disrupts copying of the tutor song

a, Schematic of in vivo multiphoton imaging of dendritic spines in HVC and the pharmacological blockade of NMDA receptors achieved by injecting D-APV (25mM) into HVC immediately prior to tutoring. b, Example of a stable spine (yellow arrowheads) imaged in HVC before and after tutoring + D-APV; scale bar = 1μm. Spine size did not change when tutoring was preceded by infusion of D-APV (tutoring + D-APV: P = 0.30, 74 dendritic spines from 4 birds; tutoring alone: P = 0.001, 47 dendritic spines from 5 birds, tutoring alone data are provided from Roberts et al. (2010) ). c, Schematic for reversibly blocking NMDA receptors in HVC during tutoring. Upper left: a zebra finch with reverse microdialysis probes bilaterally implanted in HVC. Upper right: treatment groups and tutoring schedule used in these experiments. Lower panel: the timeline of the experiments. d, Infusion of D-APV in HVC during tutoring sessions (green diamond plots), but not during periods of vocal practice (black diamond plots), prevents subsequent copying of the tutor song (P = 0.0001, diamond plot whiskers denote 10–90% range of similarity scores for each bird). d, Sonograms of a tutor’s song and the adult songs of 4 of his pupils in which APV was infused in HVC during (D-APV birds) or immediately after each of five morning tutoring sessions (control bird); scale bar = 200ms; ordinate = 0 – 9 kHz.

Figure 5. Tutor experience is conveyed to HVC from nucleus NIf

a, Timeline for the NIf lesion experiments. b, Lesioning NIf prior to tutoring severely disrupts subsequent imitation of the tutor song (R² = 0.79; n= 9 birds). Birds with >50% of NIf lesioned (n=4 birds) showed severe disruption in tutor song imitation compared with birds with <30% of NIf lesioned (n=5 birds; P = 0.001). See Supplementary Fig. 4 for sonograms of NIf lesioned birds. c, Schematic of NIf inactivation experiments. Upper panel shows the treatment groups and tutoring schedule used in these experiments. Lower panel shows the timeline for the NIf inactivation experiments. d, Sketch of the experimental design in which the pupil’s NIf or Field L1 was microstimulated (20uA per side at 76–170HZ for 200–400ms) while the tutor was singing.. e, Reversible inactivation of NIf (left columns, green bar; 14 nl of 50μM TTX) during but not immediately after tutoring sessions (left columns, gray bar) impairs subsequent copying (P = 0.016; tutor + TTX, n = 7 birds; tutor + saline, n = 5 birds; error bars = s.e.m). Tutor song-triggered microstimulation of NIf (right columns, green bar), but not Field L1 (right columns, gray bar), disrupts subsequent imitation of the tutor song (P = 0.0067, NIf = 3 birds, Field L1 = 4 birds; error bars = s.e.m). f, Sonograms of a tutor’s song and the adult songs of two of his pupils in which NIf was inactivated with TTX either during (TTX) or after (saline) morning tutoring sessions; scale bar = 100ms; ordinate = 0 – 9 kHz. g, Sonograms of a tutor’s song and the adult songs of two of his pupils in which tutor-triggered microstimulation was applied to either Field L1 or NIf during tutoring sessions; scale bar = 100ms; ordinate = 0 – 9 kHz.