Allison Doupe

Members & Research Interests

Allison J. Doupe
Professor, Psychiatry
UCSF Box 0444, HSE-818 818
campus_phone: (415)476-6446
campus_phone2: (415)476-6415
fax: (415)476-4929
E.: ajd@phy.ucsf.edu

Our laboratory is interested in how the nervous system mediates behavior, especially complex behaviors that must be learned. Birdsong provides a very useful model system for the study of these issues, with particular parallels to human speech learning. Song is an intricate motor act that is learned by young birds in two distinct phases, both of which depends on the animalÕs auditory experience: first the bird memorizes a tutor song (sensory learning); later, the bird begins to sing, and, in a manner analogous to the acquisition of speech by human infants, uses auditory feedback to refine and correct its vocalizations until it produces an approximate match of the memorized tutor song (sensorimotor learning). There are critical periods for sensory learning of song, just as there are for some types of human learning. Moreover, a discrete set of brain areas, called the song system, controls song learning and production. Both the song system and the adult song behavior are sexually dimorphic, and are regulated by sex steroids. Finally, song learning and singing are highly social behaviors, and are modulated by social cues. All of these features give birdsong the potential to shed light on the neural basis of learning, and on factors which control and limit learning.

At present the laboratory is focussed on several issues. For one, we are interested in a particular song circuit, the anterior forebrain (AF) pathway. Numerous behavioral studies have suggested that this specialized basal ganglia circuit plays an essential role in song learning. Using a variety of physiological, behavioral, anatomical, and theoretical techniques, we are studying how the different features of song are represented in this network, how the animal's auditory experience and vocal learning shape its neuronal properties, and what the crucial function of this pathway might be.

In electrophysiological studies in anesthetized juvenile birds, we have shown that the neurons in this pathway respond to a variety of sounds. As the bird learns its song, however, these same neurons become highly selective: they respond robustly to the sound of the bird's own song, and weakly or not at all to very similar songs of conspecific individuals or even to the bird's own song played in reverse. The temporally and acoustically complex auditory response properties of these neurons suggest that they encode a neural representation of song, formed and perhaps used during learning. Systematic manipulations of the stimulus, the birdÕs own song, have allowed us to analyze the response properties of these very selective neurons in a quantitative manner. Behavioral manipulations have also been useful: altering the the birdÕs vocal organ before onset of singing, so that the bird can never produce a good copy of the tutor song that it memorized, has demonstrated that both the tutor and the birdÕs own song are important for shaping the properties of these neurons. These cells may in fact be involved in comparing these two stimuli during the process of learning.

Since the bird normally hears its own song only during singing, we have also begun recording from this pathway in awake, behaving birds. This has demonstrated that AF neurons are also highly active when the bird sings, and carry signals related to the motor act of singing as well as auditory responses. Behavioral experiments, using lesions of this pathway, suggest that these neurons may be required any time the bird changes its song, even in adulthood, and may encode an error signal when auditory feedback does not match the intended vocal output. Moreover, the activity of these neurons is dramatically affected by different social settings (singing alone vs. singing to a companion), raising the possibility that this circuit may also be involved in social modulation of singing and song learning.

Ongoing experiments include i) recording simultaneously from this pathway and its target song motor nucleus, in awake and anesthetized birds at various stages of learning, to understand how this circuit interacts with and affects the motor pathway; ii) altering auditory feedback and recording neural activity to find where and the songbird brain encodes error signals; iii) in vitro brain slice studies of the AF to assess the synaptic changes underlying song learning; and iv) tracking down the source of the social modulation of activity in this pathway, as well as how the altered activity affects behavior .

A second and new, although related, area of interest is the auditory pathway afferent to the song system (the Field L complex), which must ultimately contain neurons contributing both to song selectivity and tutor song template memorization, and might also be the source of innate predispositions of birds to learn only the songs of their own species. We are analyzing auditory neurons at different stages of Field L using complex song stimuli and quantitative analysis of receptive field properties. Ongoing projects include i) mapping of sequential auditory stations in Field L using the birdÕs own song and conspecific stimuli, and ii) investigating Field L responses to songs of the birdÕs own species and well as selected heterospecific songs, in both adult and juvenile birds.

Wheeler Center Members