Neural mechanisms of song memory formation in juvenile zebra finches

Abstract

There are many parallels between the acquisition of spoken language in human infants and song learning in songbirds, at the behavioural, neural, genetic and cognitive levels. Both human infants and juvenile songbirds are able to imitate sounds from adults of the same species (often their parents), called vocal learning. Vocal learning is a relatively rare ability in the animal kingdom that even appears to be absent in our closest relatives, non-human primates. In contrast, almost 5,000 avian species are vocal learners. Research is often conducted in the zebra finch (Taeniopygia guttata), a songbird that has a very stereotyped song. Birdsong learning provides an excellent model for vocal learning and auditory memory formation, because it is a behaviour that comes very natural to the juvenile songbird, and can easily be experimentally manipulated. Juvenile songbirds imitate the song of any adult that sings in close proximity, which, under natural circumstances, is usually their father. In the lab we can raise juveniles with another unrelated male, which will lead to song imitation. The male that the juvenile imitates is generally called a ‘tutor’. Zebra finches memorise tutor song in a sensitive phase from about three weeks to two months after hatching, and start imitating when they are about one month old, gradually shaping it into adult-like song until they are about three months old. I studied early stages of song memorisation in juvenile zebra finches that were raised with their father for the first one-and-a-half month and in juveniles that received only two or ten days of exposure to a tutor song. I used immunocytochemistry to quantify an immediate early gene called Zenk, that is used a marker for neuronal activation. I specifically looked at neuronal activation in response to tutor song playbacks and during sleep in the NCM (nidopallium caudomediale), a higher-order auditory brain region. The NCM is involved in song perception, recognition and memory and might functionally be similar to Wernicke’s area of humans, one of the regions in the ‘language brain network’. During linguistic processing in human brains, regions in the left side of the brain are often higher activated than the regions in the right hemisphere. I demonstrated similar lateralisation for tutor song in juvenile male zebra finches. Furthermore, we demonstrated significant song imitation in juvenile males after only two days of tutor exposure. I also studied the role of sleep in tutor song memorization, as sleep is one aspect important for successful learning. I found that sleep might indeed be involved in song learning, and it is likely to be most important immediately after the first song learning opportunity. Thus, in addition to behavioural, genetic and neural parallels that were found between speech in humans and song in zebra finches, our findings suggest that perception of human speech and birdsong show similar patterns of lateralised brain activation. Songbird research will allow to further study the functions of sleep and lateralisation for memory formation, and a possible link between sleep and lateralisation characteristics and individual differences in learning performances.