This Week in The Journal

Serotonergic Fibers Release Glutamate in Amygdala

Ayesha Sengupta, Marco Bocchio, David M. Bannerman, Trevor Sharp, and Marco Capogna

Serotonergic neurons in the brainstem densely innervate the amygdala, where they regulate affective states. Depleting serotonin selectively in the amygdala increases social interaction and impairs fear conditioning in rats, for example (Johnson et al. 2015 Pharmacol Biochem Behav 138). Elucidating the effects of serotonergic projections on amygdala circuits is difficult, however, because serotonin is released at both synaptic and nonsynaptic sites, and it can exert inhibitory or excitatory effects on different neuronal classes. Furthermore, some serotonergic neurons co-release glutamate, although whether this occurs in the amygdala has been unknown. Therefore, studies examining the effects of exogenous serotonin on amygdala neurons are unlikely to provide a complete picture of how physiological activation of serotonergic afferents regulates amygdala function.

A gradient of Shh (from bottom to top in right panel) causes β-actin levels to increase on the side of the growth cone nearer the Shh source. β-actin levels are lower in growth cones exposed to a control gradient (left). See Sengupta et al. for details.

To address this question, Sengupta et al. expressed channelrhodopsin in serotonergic neurons of the dorsal raphe nucleus and optically stimulated terminals in the basal amygdala in mouse brain slices. The effects differed depending on stimulation frequency and postsynaptic cell type. In some interneurons, single light pulses evoked fast EPSCs that were completely blocked by a glutamate receptor antagonist. These EPSCs were depressed during high-frequency (10–20 Hz) pulse trains, and in some cells, they were replaced by slow IPSCs mediated by 5-HT_1A receptors. Some interneurons exhibited only slow serotonergic IPSCs during high-frequency stimulation, while others exhibited slow EPSCs mediated by 5-HT_2A receptors. Unlike interneurons, principal cells exhibited neither fast nor slow EPSCs. Instead, high-frequency stimulation evoked slow IPSCs in these neurons. In addition, high-frequency stimulation increased the frequency of GABAergic IPSCs in principal neurons, and when current was injected to elicit principal-cell spiking, stimulation of serotonergic fibers reduced the firing rate.

These results suggest that serotonergic fibers in the amygdala release glutamate that excites subsets of interneurons under baseline conditions. When the spike rate of serotonergic neurons increases—during reward anticipation and receipt, for example (Luo et al. 2016 Neurobiol Learn Mem 135:40)—they release serotonin, which exerts multiple effects that decrease spiking of principal neurons. As we learned last week, however, some conditions, including nerve injury, alter serotonin receptor expression in the amygdala, and this may cause serotonin to increase principal neuron activity (Ji et al. 2017 J Neurosci 37:1378).

Sonic Hedgehog Induces Local Protein Translation

Léa Lepelletier, Sébastien D. Langlois, Christopher B. Kent, Kristy Welshhans, Steves Morin, et al.

(see pages 1685–1695)

Sonic hedgehog (Shh), a secreted molecule, influences all stages of nervous system development, from the initial differentiation of progenitor pools in the neural tube to axon guidance and synaptogenesis. Shh helps specify neuron fates through the so-called canonical pathway, in which activation of Gli transcription factors promotes expression of cell-type-specific genes. In contrast, Shh guides axons through Gli-independent non-canonical pathways involving activation of Src-family kinases (SFKs). In the developing spinal cord, for example, Shh secreted by the ventral floorplate attracts axons of commissural neurons by inducing a gradient of SFK activity in the growth cone (Ferent and Traiffort 2015 Neuroscientist 21:356).

How polarized activation of SFKs causes growth cones to turn toward a Shh source has been unclear. Lepelletier et al. reasoned that the mechanism might overlap with that mediating axon guidance by netrin-1, another secreted molecule. Specifically, they hypothesized that Shh promotes SFK-dependent phosphorylation of zipcode binding protein 1 (ZBP1), an RNA-binding protein. ZBP1 is involved in transport and translational regulation of various mRNAs, including that encoding ß-actin. When phosphorylated, ZBP1 releases its cargo, enabling local translation. The resulting increase in ß-actin levels would promote filopodial protrusion, the first step in growth cone turning.

Consistent with this hypothesis, Lepelletier et al. found that Shh-induced turning of rat commissural axons in vitro required local protein translation. Exposure to Shh increased levels of ß-actin mRNA and protein preferentially on the side of the growth cone nearer the Shh source. This accumulation was blocked by inhibiting SFKs or protein synthesis. Shh also increased phosphorylation of ZBP1 in the growth cone, and overexpressing a phosphorylation-resistant form of ZBP1 prevented Shh-induced turning. Moreover, both knocking out ZBP1 in mice and overexpressing phosphorylation-resistant ZBP1 in chick embryos impaired guidance of spinal commissural axons in vivo. Notably, this phenotype resembled that of mice lacking Shh receptors.

This work suggests that Shh regulates gene expression not only at the transcriptional level, via the canonical pathway, but also at the translational level, through a noncanonical pathway involving phosphorylation of ZBP1. The results also emphasize the importance of RNA-binding proteins and local protein syntheses in axon guidance. Similar mechanisms may contribute to Shh-dependent synaptic plasticity at later developmental stages.