This Week in The Journal

Effects of Wnt9a on Hair Cell Fate

Vidhya Munnamalai, Ulrike J. Sienknecht, R. Keith Duncan, M. Katie Scott, Ankita Thawani, et al.

The basilar papilla of birds is homologous to the mammalian cochlea, and like the cochlea, it contains two types of hair cells. Tall hair cells resemble cochlear inner hair cells: they lie along the edge of the papilla nearest the auditory ganglion (the neural edge), and they convey sound information to afferent nerves. Short hair cells, in contrast, resemble cochlear outer hair cells: they are aligned along the papilla's abneural edge, and they receive efferent input from the brain. How hair cells acquire these distinct phenotypes is not fully understood. A study using Wnt activators in mouse cochlear cultures suggested that Wnts promote inner-hair-cell fate and suppress outer-hair-cell fate (Munnamalai and Fekete 2016 Development 143:4003), but which Wnts were involved could not be determined in that pharmacological study.

Because Wnt9a is expressed in the neural half of the developing basilar papilla of chicks, Munnamalai et al. explored the role of this Wnt on chick hair-cell specification in vivo. When Wnt9a was overexpressed early in development, short hair cells failed to develop. Instead, abneural hair cells were morphologically indistinguishable from tall hair cells. Although these abneural hair cells acquired some electrophysiological properties of short hair cells, they also exhibited properties of tall hair cells. Overexpression of Wnt9a also increased the total number of hair cells, making the basilar papilla wider than normal, and it disrupted efferent innervation of the basilar papilla: efferent axons did not extend as far as normal, they developed fewer branches, and in some embryos, they failed to enter the auditory organ. These anatomical changes were accompanied by altered expression of numerous genes, including several genes involved in axon growth.

These results suggest that expression of Wnt9a along the neural edge of the basilar papilla promotes specification of tall-hair-cell fate and—either consequently or independently—influences the growth of efferent axons in chicks. Thus, Wnts contribute to hair-cell fate specification in birds as well as in mammals. Wnt9a is expressed at low levels in mouse cochlea, however, and knocking it out in mice had little effect on cochlea patterning, suggesting a different Wnt directs hair cell fates in mammals.

Responses of Serotonin and Dopamine Neurons to Reward

Weixin Zhong, Yi Li, Qiru Feng, and Minmin Luo

(see pages 8863–8875)

Serotonergic neurons in the dorsal raphe nuclei and dopaminergic neurons in the ventral tegmental area change their firing rate in response to rewards and reward-predicting cues. Thus, both populations are thought to help animals predict and respond appropriately to reward availability. Dopamine neurons fire bursts of spikes when an unexpected reward is received; but if a reward always follows a specific stimulus, the neurons begin to respond only when the stimulus occurs, not when reward is obtained. Therefore, these neurons are thought to encode reward prediction errors that guide learning. The responses of serotonergic neurons to environmental stimuli are more complex: in fact, they respond to punishments as well as rewards. Consequently, the contribution of serotonergic neurons to reward processing remains unclear.

Calcium changes in serotoninergic (left) and dopaminergic (right) neurons in response to tone (black bar) and sucrose delivery (blue bar) on day 1 (top) and day 4 (bottom) of training show that serotonergic neurons begin to respond during the entire period between cue onset and reward receipt, whereas dopaminergic neurons begin to respond to the cue alone. See Zhong, Li, et al. for details.

A key to understanding any neuron's function is to examine its responses in awake, behaving animals. Zhong, Li, et al. did this by expressing a genetically encoded calcium indicator selectively in serotonergic or dopaminergic neurons and recording activity of these neurons as unrestrained mice learned to associate an auditory cue with subsequent reward (sucrose) delivery. Both dopaminergic and serotonergic neurons responded to reward delivery at the start of training, and after numerous trials, both populations began to respond to the auditory cue. The pattern of responses differed in the two populations, however. Consistent with previous work, reward-associated cues elicited transient responses in dopaminergic neurons, and the magnitude of these responses increased across trials, while reward-evoked responses decreased. In contrast, cue-evoked responses in serotonergic neurons were persistent, and they slowly increased throughout the delay period until the reward was received. Moreover, reward-evoked responses remained strong throughout training. Notably, cue- and reward-evoked responses were reduced in both types of neurons when mice were subjected to head restraint during trials, as well as when trials were conducted in a chamber associated with foot shock. Furthermore, neither serotonergic nor dopaminergic neurons responded when an aversive stimulus (quinine) was delivered along with sucrose.

These results—along with the finding that locomotion was reduced when serotonergic activity was elevated—support the hypothesis that activity in serotonergic neurons facilitates waiting for reward in favorable environments. They also indicate that stressful environments might lead to anhedonia (a symptom of depression) by reducing reward-evoked responses in both serotonergic and dopaminergic neurons.