Adrenergic signaling gates astrocyte responsiveness to neurotransmitters and control of neuronal activity

Abstract

How astrocytes regulate neuronal circuits is a fundamental, unsolved question in neurobiology. Nevertheless, few studies have explored the rules that govern when astrocytes respond to different neurotransmitters in vivo and how they affect downstream circuit modulation. Here, we report an unexpected mechanism in Drosophila by which G-protein coupled adrenergic signaling in astrocytes can control, or "gate," their ability to respond to other neurotransmitters. Further, we show that manipulating this pathway potently regulates neuronal circuit activity and animal behavior. Finally, we demonstrate that this gating mechanism is conserved in mammalian astrocytes, arguing it is an ancient feature of astrocyte circuit function. Our work establishes a new mechanism by which astrocytes dynamically respond to and modulate neuronal activity in different brain regions and in different behavioral states.

Fig 1:. Tyramine gates the response to other neurotransmitters via the Oct-TyrR.

(A) Schematic demonstrating the ex vivo preparation used to image larval ventral nerve cord (VNC) astrocytes in an intact nervous system. (B) VNC astrocytes before and after exposure to tyramine with GCaMP6s fluorescence pseudo-colored to demonstrate the increase in calcium level across nearly all astrocytes in response to tyramine. (Scale bar = 100 μm) (C) Example trace of a field of VNC astrocytes responding to tyramine with a calcium influx but showing no response to dopamine as well as demonstration of the quantification method for the tyramine response. (D) Quantification of astrocyte response to various neurotransmitters (NT). Note that astrocytes respond robustly to octopamine and tyramine but not to other neurotransmitters. (E) Example traces of astrocytes responding to tyramine followed by a second neurotransmitter (or vehicle control, Veh). Note that astrocytes respond to dopamine after tyramine stimulation but not to dopamine alone. (F) Quantification of astrocyte calcium responses shows that astrocytes respond to all tested neurotransmitters following tyramine exposure but not following dopamine exposure. We term this response to neurotransmitters after tyramine exposure “gating.” (WO = washout; Ach = acetylcholine, Glu = glutamate, Tyr = tyramine, Gab = GABA) (G) Schematic demonstrating a hypothesized pathway of gating via Oct-TyrR. (H) Quantification of astrocyte calcium responses showing that the gating of all tested neurotransmitters is dependent on Oct-TyrR but not dependent on calcium influx via the TRP channel TrpA1. This suggests that gating does not occur due to calcium influx similar to that induced by Oct-TyrR activation. (I) Example traces of TrpA1-expressing astrocytes responding to AITC exposure but not to subsequent glutamate or dopamine exposure. (J) Adding tyramine at the same time as dopamine (Tyr + Dop) does not lead to a larger calcium response than tyramine alone. In contrast, adding glutamate or acetylcholine with tyramine leads to a significantly larger calcium response than tyramine alone. * Indicates p-value < 0.05; details of statistical comparisons and exact p-values in Table S1. All error bars represent SEM. Red dots in bar graphs correspond to traces chosen as example.

Fig 2:. Dopamine gating is dependent on G_αi signaling.

(A) Schematic showing hypothesized mechanism of gating via the G-protein signaling downstream of Oct-TyrR. (B) Mean trace (thick line) and 95% confidence interval (thin lines) of cAMP levels as measured by the FRET ratio of the cAMP indicator cAMPFIRE-H in astrocytes (n=5). Tyramine exposure causes a decrease in cAMP levels which aligns with its hypothesized function as a G_αi GPCR and this decrease is prevented by knocking down Galphai. (C) Quantification of astrocyte calcium response demonstrates that the gating of all NTs is dependent on G_αi signaling but cAMP modulation is only sufficient to gate dopamine. (D) Example traces of astrocytes treated with the adenylyl cyclase inhibitor SQ22536 responding to dopamine but not glutamate. (E) Quantification of astrocyte calcium responses shows that dopamine gating is dependent on Dop2R but not Dop1R1. Dopamine gating can also be achieved by first treating astrocytes with CMPD101 – which inhibits kinase-mediated internalization of receptors – and this gating is similarly dependent on Dop2R. (F) Example traces of astrocytes responding to dopamine but not glutamate following exposure to CMPD101. (G) Schematic showing the hypothesized mechanism that links tyramine exposure to dopamine gating. (H) Traces (without TTX) of Ddc⁺ dopamine neuron activity. Bath application of dopamine normally inhibits Ddc⁺ neuron activity but becomes excitatory after pre-exposure to CMPD101. Knocking down Dop2R in astrocytes reverts the effects of CMPD101 and bath application of dopamine becomes inhibitory once again. (I) Quantification of neuronal activity from experiments outlined in 2H. (J) Schematic showing hypothesized role of astrocyte gating of Dop2R on the response of dopaminergic neurons to bath application of dopamine (K) Diagram of larval righting assay. (L) Larval crawling analysis shows that Dop2R and Krz manipulations in astrocytes do not affect baseline larval locomotion. (M) Quantification of latency to right of larvae turned to their posterior side (righting). Knocking down Dop2R in astrocytes with or without Gal80 expression in neurons leads to slower righting. Further, krz knockdown and overexpression have bidirectional effects on larval righting that align with the hypothesized role of Krz in internalizing Dop2R. * Indicates p-value < 0.05; details of statistical comparisons and exact p-values in Table S1. All error bars represent SEM. Red dots in bar graphs correspond to traces chosen as example.

Fig 3:. Gating of dopamine is conserved in mammals.

(A) Schematic showing method of calcium imaging in primary rat astrocytes. (B) Example image of primary rat astrocytes loaded with the calcium indicator Fluo-4. (Scale bar = 20 μm) (C) Schematic of NE and drug selectivity for α1 vs α2 adrenergic receptors. (D) Example traces of rat astrocyte calcium responses. Both NE and the α2 adrenergic agonist UK14304 can gate the response to dopamine. (E) Quantification of astrocyte calcium responses to various drugs and dopamine suggest that the cellular mechanism of gating is shared between Drosophila and mammalian astrocytes. Namely, dopamine gating can be mediated by G_αi adrenergic stimulation, cAMP modulation, and CMPD101-mediated inhibition of internalization. NE exposure can also gate glutamate and there is a trend towards NE gating acetylcholine (F) Example astrocyte calcium responses to SQ22536 + dopamine but not dopamine alone shown via pseudo colored Fluo-4 fluorescence. (Scale bar = 20 μm) (G) Example trace of rat astrocytes responding to dopamine following SQ22536 mediated inhibition of adenylyl cyclase but not to dopamine alone nor to dopamine following adenylyl cyclase activation via forskolin (H) Example images of astrocytes (phase) with or without treatment with CMPD101 for 30 minutes and then stained with an antibody against the extracellular domain of DRD2 (DRD2 ED; green) (Scale bar = 10 μm; brightness of the DRD2 ED staining enhanced in the example control image to demonstrate that astrocyte staining is at or near background levels). (I) Quantification of primary rat astrocytes stained for the extracellular domain of DRD2 shows that manipulations that can gate the dopamine response in Drosophila and rat also lead to increased externalization of DRD2. * Indicates p-value < 0.05; details of statistical comparisons and exact p-values in Table S1. All error bars represent SEM. Red dots correspond to traces chosen as example.