Local BMP signaling: A sensor for synaptic activity that balances synapse growth and function

Abstract

Synapse development is coordinated by intercellular communication between the pre- and postsynaptic compartments, and by neuronal activity itself. In flies as in vertebrates, neuronal activity induces input-specific changes in the synaptic strength so that the entire circuit maintains stable function in the face of many challenges, including changes in synapse number and strength. But how do neurons sense synapse activity? In several studies carried out using the Drosophila neuromuscular junction (NMJ), we demonstrated that local BMP signaling provides an exquisite sensor for synapse activity. Here we review the main features of this exquisite sensor and discuss its functioning beyond monitoring the synapse activity but rather as a key controller that operates in coordination with other BMP signaling pathways to balance synapse growth, maturation and function.

Keywords: BMP signaling; Local BMP signaling; Positive feedback loop; Sensor for synapse activity; Synapse assembly.

Fig. 1

Fly motor neurons receive and integrate multiple BMP signaling pathways. (A) Diagram of various BMP pathways signaling to fly motor neurons. At synaptic terminals, Gbb/ BMP7 binding to BMP receptors on the motor neuron membrane triggers canonical and non-canonical BMP signaling. Gbb is secreted from both muscle (retrograde signaling—marked here “R”) and motor neurons (autocrine signaling—labeled “A”). During canonical BMP signaling, the high-order BMP/BMPR (Gbb/Wit/Tkv/Sax) complexes are endocytosed and transported to the motor neuron soma where they phosphorylate Mad to regulate various transcriptional programs required for NMJ growth and function. During non-canonical signaling, Wit, the BMPRII, signals through LIMK1 to regulate synapse stability. This pathway does not involve Mad or Medea. In addition, the BMPRs Wit, Tkv and Sax, but not Gbb, enable the synaptic/local BMP signaling. (B) Dissection of a third instar larva along the dorsal side exposes the highly stereotyped body wall muscles labeled with phalloidin (blue) and imaged by confocal microscopy. The anti-horseradish peroxidase antibodies (HRP, magenta) label neuronal membranes. Bruchpilot (Brp, red) is a synaptic scaffold which marks the active zones. In response to canonical BMP signaling, pMad (green) accumulates in the motor neuron nuclei within the ventral nerve cord (VNC), the fly equivalent of mammalian spinal cord. In addition, pMad also accumulates in motor neuron synaptic terminals (right upper panels—muscle 6/7 NMJ). Synaptic/junctional pMad forms discrete puncta that co-localize with the active zone scaffold, Brp. Scale bars: 100 μm (larval fillet), 10 μm (others).

Fig. 2

Synaptic recruitment marks the end of a long journey for iGluRs. The hetero-tetrameric iGluR NMJ complexes must assemble in ER before they could be delivered to muscle surface. Once on the membrane, iGluRs form complexes with Neto and together bind to scaffolds and motors which ensure trafficking to postsynaptic densities, juxtaposing the active zones. Synaptic iGluRs form large aggregates that are further stabilized through interactions with postsynaptic density components. Additional trans-synaptic interactions (such as Neurexin-Neuroligin) keep active zones and postsynaptic densities in register, further minimizing the distance that the neurotransmitter must travel to reach the postsynaptic receptors.

Fig. 3

pMad mirrors the postsynaptic type-A receptors. Confocal images of NMJ boutons from third instar larvae of the indicated genotypes labeled for the obligatory auxiliary subunit Neto (blue), which marks both type-A and type-B glutamate receptors, GluRIIA (green), the glutamate receptor subunit specific for type-A receptors, and pMad (red). Synaptic pMad follows the distribution and intensity of GluRIIA-positive signals: It increases with the muscle overexpression of GluRIIA and becomes undetectable at GluRIIA mutant NMJs. Depletion of GluRIIB in the postsynaptic muscle triggers an increase of synaptic GluRIIA levels and therefore increased synaptic pMad.

Fig. 4

Synaptic pMad localizes at the active zone. (A, B) 3D structured illumination microscopy (3D-SIM) images of NMJ boutons from third instar larvae labeled for Brp- an active zone scaffold (green), pMad (red) and the obligatory auxiliary subunit Neto (blue). (C) High magnification view of a single synapse profile (from panel B). (D) Side view of a surface rendered volume of the synapse shown in panel (B). (E) Electron micrograph of a single synapse illustrating the characteristic T-bar structure juxtaposing the postsynaptic density. The anti-Brp monoclonal antibody recognizes the tip of the T-bar, hence the ring appearance. The anti-Neto antibodies recognize the extracellular CUB1 domain within the synaptic cleft and mark the postsynaptic densities. The pMad signals concentrate in between Brp and Neto, closer to Neto and form thin discs suggestive of a layer of pMad parallel to the presynaptic membrane.

Fig. 5

Model for how Neto-mediated trans-synaptic interactions relay the status of postsynaptic type-A receptors to presynaptic BMP/BMPR complexes. Neto and iGluRs traffic together at synaptic locations. Neto also has two extracellular BMP-interacting CUB domains that may localize BMP activities and anchor the presynaptic BMP/BMPR complexes at active zones via trans-synaptic interactions. These complexes phosphorylate Mad locally and induce pMad accumulation at the synaptic junction. Inactivation of type-A receptors induces conformational changes and dissociation of these trans-synaptic complexes.

Fig. 6

Disruption of presynaptic pMad reduces the levels of postsynaptic type-A receptors. (A, B) Confocal images of NMJ4 boutons from control and third instar larvae with a phosphomimetic Mad variant overexpressed in motor neurons (N > Mad^S25D). Neuronal expression of Mad^S25D reduces the accumulation of synaptic pMad and GluRIIA (type-A receptors) and increases the GluRIIB synaptic accumulation. The anti-horseradish peroxidase (HRP-blue) labels neuronal membranes. (C) The positive feedback loop model

Fig. 7

Model for type-A receptor stabilization via local BMP pathway. Neto in complexes with active type-A receptors localizes BMP activities, promoting the formation of presynaptic BMP/BMP receptor complexes (step 1). These complexes function as “accumulation centers” for stabilizing type-A receptors at nascent synapses (step 2). Dissociation of these local complexes terminates further incorporation of type-A receptors and allows for recruitment of type-B subtypes, which mark mature synapses (steps 3–4).