This Week in The Journal

doi:10.1523/JNEUROSCI.twij.40.33.2020

Serotonergic Innervation of Drosophila Olfactory Areas

Kaylynn E. Coates, Steven A. Calle-Schuler, Levi M. Helmick, Victoria L. Knotts, Brennah N. Martik, et al.

Neuromodulators have widespread effects in the brain, but they are produced by a relatively small number of neurons. These neurons can project to multiple brain regions, where they can have different effects depending on the number of synapses they form, the types of neurons they target, and the receptors expressed by those target neurons. Consequently, the effects of individual neuromodulatory neurons are difficult to discern. This problem is easier to approach in the small, yet complex brain of Drosophila.

Serotonin influences olfactory processing in flies by acting in the antennal lobe, which receives input from olfactory sensory neurons, as well as in the mushroom bodies and lateral horn, which influence learned and innate olfactory behaviors, respectively. Coates et al. used a whole-brain electron microscopy dataset to reconstruct the primary neurite and map synapses formed by one of a pair of serotonin neurons—the contralaterally projecting, serotonin-immunoreactive deutocerebral neurons (CSDns)—that provide the sole serotonergic input to the antennal lobes and lateral horn, as well as innervate the mushroom bodies.

Like in other invertebrate neurons, the long primary neurite of CSDn serves as both axon and dendrite. This neurite innervates the ipsilateral antennal lobe, extends to the ipsilateral protocerebrum where it arborizes in the lateral horn and mushroom body calyx, then extends contralaterally to innervate the same structures in that hemisphere. Presynaptic and postsynaptic sites were intermingled along CSDn branches in all innervated areas except the ipsilateral antennal lobe and protocerebral antler, where only postsynaptic sites formed.

The density of synapses per neurite length was similar in all areas, and where presynaptic sites formed, they constituted ∼45% of all synapses. Input to CSDn differed across antennal lobe glomeruli: whereas no input was received in some glomeruli, in others it arose primarily from olfactory receptor neurons, local interneurons, or projection neurons. Regardless, CSDn output primarily targeted two specific types of local interneurons across glomeruli. CSDn also received differential input from specific cell types in the lateral horn, and it received previously unrecognized input from cells in the protocerebrum. This detailed mapping of CSDn connectivity will guide future work to uncover how serotonin acts in different structures to shape olfactory behaviors.

Reconstruction of a CSDn neuron, showing arborization in the antennal lobes, antler, superior lateral protocerebrum, mushroom body calyx, and lateral horn. See Coates et al. for details.

Information Flow in Lateral Cortex of Inferior Colliculus

Alexandria M. H. Lesicko, Stacy K. Sons, and Daniel A. Llano

(see pages 6328–6344)

The inferior colliculus (IC) is an important hub in the auditory system. Nearly all ascending auditory information passes through the IC, where it may be shaped by descending input from the auditory cortex. But the IC also receives ascending and descending input from visual and somatosensory areas, as well as from the amygdala, basal ganglia, and superior colliculus. Thus, the IC is thought to integrate auditory and nonauditory input to guide responses to relevant auditory stimuli. (Gruters and Groh, 2012, Front Neural Circuits 6:96). Yet previous work by Lesicko et al. showed that auditory and somatosensory inputs to at least one region of the IC—the lateral cortex (LCIC)—were largely segregated: whereas somatosensory inputs targeted modules that express high levels of GAD67 and other markers, auditory inputs primarily targeted the matrix between these modules. How then is information integrated?

Lesicko et al. addressed this question by recording postsynaptic responses in GAD67-expressing (presumably inhibitory) and GAD67-lacking (presumably excitatory) neurons in module and matrix regions while using focal glutamate uncaging to activate surrounding neurons in both subregions. Local input to matrix cells came predominantly from other matrix cells; for GAD67^– cells, this was almost exclusively inhibitory, whereas excitatory and inhibitory input was balanced for GAD67⁺ cells. Most local input to module cells also came from inhibitory matrix cells, but a subset of GAD67^– module cells received substantial input from other module cells.

Additional experiments showed that GAD67^– matrix cells were responsible for nearly all projections from the LCIC to other IC subnuclei and to the superior colliculus. In contrast, GAD67^– module cells provided the vast majority of LCIC input to the medial subdivision of the thalamic medial geniculate body (mMGB).

These data suggest the LCIC can treat auditory information differently depending on how it will be used. Information transmitted from matrix to modules may be integrated with somatosensory information before being transmitted to the mMGB, which is involved in fear conditioning and shaping responses based on context. In contrast, matrix cells might transmit relatively pure auditory information to the superior colliculus for use in orienting toward salient stimuli. Future work should determine whether LCIC projections to other targets arise from matrix or modules.