UCLA

The capacity for language is one of the most complex and least understood phenomena in neuroscience. Because language is unique to humans which are not amenable to invasive or interventionist experiments, research has focused on studying the neural basis of language subcomponents in other species. One such subcomponent is vocal learning, defined as the ability to acquire new sounds via imitation, and is present in a handful of animal groups including songbirds. The behavioral, anatomical and molecular parallels between human speech learning and zebra finch (Taeniopygia guttata) song learning make songbirds such as the zebra finch a compelling model system for studying learned vocalizations.

My dissertation asks related questions of the molecular and physiological basis of vocal learning through the lenses of the transcription factor FoxP2 and cortico-basal ganglia function, respectively. Mutations in the transcription factor Forkhead box protein 2 (FoxP2) are associated with a specific language disorder in humans. In both humans and songbirds FoxP2 is enriched inthe basal ganglia and, in zebra finch, FoxP2 levels are decreased as a function of vocal practice within the song-dedicated nucleus known as Area X. This downregulation is accompanied by an acute increase in vocal variability as well as the bidirectional regulation of thousands of genes many of which are known targets of FoxP2 in humans. Here, I test whether online regulation of FoxP2 is necessary for vocal learning and whether FoxP2 downregulation is causally related to the enhanced variability following vocal practice. To this end, I pioneered the use of AAV driven overexpression of a biologically relevant gene- a first in the zebra finch- and found that preventing FoxP2 downregulation led to poor song learning and disrupted the acute increase in vocal variability. Based on these findings I suggest that dynamic behavior-linked regulation of FoxP2, rather than absolute levels per se, is critical for vocal learning.

Next, to gain insight into the physiological changes that accompanied the FoxP2 regulation, I asked what physiological manipulations could recapitulate the FoxP2 dependent transition to high variability. To that end, I pioneered the use of designer drugs exclusively activated by designer receptors (DREADDs) -another first in the songbird system- to bidirectionally affect cell excitability. Because of its known role as positive regulator of variability I also focused my attention on the cortical song control nucleus lateral magnocellular nucleus of the nidopallium (LMAN). I found that bidirectional manipulations of either LMAN or basal ganglia Area X lead to bidirectional regulation of vocal variability with higher levels of LMAN activity leading to higher levels of variability whereas higher levels of Area X led to lower levels of variability. These effects on variability where limited to rendition-to-rendition and millisecond timescale moment-to-moment variability but not higher order syntax variability. Based on these observations I suggest that the cortex and striatopallidum reciprocally control vocal variability and that FoxP2-dependent decrease in neuronal excitability may account for practice dependent increases in variability.