Neto-mediated intracellular interactions shape postsynaptic composition at the Drosophila neuromuscular junction

Abstract

The molecular mechanisms controlling the subunit composition of glutamate receptors are crucial for the formation of neural circuits and for the long-term plasticity underlying learning and memory. Here we use the Drosophila neuromuscular junction (NMJ) to examine how specific receptor subtypes are recruited and stabilized at synaptic locations. In flies, clustering of ionotropic glutamate receptors (iGluRs) requires Neto (Neuropillin and Tolloid-like), a highly conserved auxiliary subunit that is essential for NMJ assembly and development. Drosophila neto encodes two isoforms, Neto-α and Neto-β, with common extracellular parts and distinct cytoplasmic domains. Mutations that specifically eliminate Neto-β or its intracellular domain were generated. When Neto-β is missing or is truncated, the larval NMJs show profound changes in the subtype composition of iGluRs due to reduced synaptic accumulation of the GluRIIA subunit. Furthermore, neto-β mutant NMJs fail to accumulate p21-activated kinase (PAK), a critical postsynaptic component implicated in the synaptic stabilization of GluRIIA. Muscle expression of either Neto-α or Neto-β rescued the synaptic transmission at neto null NMJs, indicating that Neto conserved domains mediate iGluRs clustering. However, only Neto-β restored PAK synaptic accumulation at neto null NMJs. Thus, Neto engages in intracellular interactions that regulate the iGluR subtype composition by preferentially recruiting and/or stabilizing selective receptor subtypes.

Fig 1. Neto-β is a novel Neto isoform at the Drosophila NMJ.

(A) Diagram of the Drosophila neto locus. Two isoforms are generated by alternative splicing, Neto-α and Neto-β. They have different cytoplasmic domains, but share highly conserved domains, CUB (complement C1r/C1s, Uegf, BMP1), LDLa (LDL receptor class a) and transmembrane (TM), with Neto proteins from vertebrates and C. elegans. (B) Confocal images of third instar larvae NMJ4 from neto ^null animals rescued with neto-α (neto ³⁶;G14>neto ^α), neto-β (neto ³⁶;G14>neto ^β), and a neto transgene lacking any intracellular part, netoΔ^intra (neto ³⁶;G14>netoΔ^intra). These Neto variants can rescue the viability and NMJ development in neto ^null animals. The Neto-ex signals mark all NMJs, but the Neto-β antibodies specifically label control and neto-β rescued NMJs. (C-D) Confocal images of late embryos ventral muscle fields (indicated in white) labeled for (C) Neto-ex (red), Neto-β (green), HRP (blue) and (D) GluRIIA (red), Neto-β (green), HRP (blue) indicating the presence of Neto-β in the early stages of the larvae development. Genotypes: control (w ¹¹¹⁸); neto ³⁶;G14>neto ^α (neto ³⁶ /Y; G14-Gal4/UAS-neto-A9); neto ³⁶;G14>neto ^β (neto ³⁶, UAS-neto-B6/Y; G14-Gal4/+); neto ³⁶;G14>netoΔ^intra(neto ³⁶ /Y; G14-Gal4/ UAS-netoΔ^intra -H4). Scale bars: (B) 20 μm (C-D) 10 μm; 5 μm in details.

Fig 2. Generation and characterization of neto-β isoform specific alleles.

(A) Schematics of the Minos transposomal elements and the small lesions corresponding to various neto alleles. MB07125 was mobilized to generate precise excision control and isoform specific neto-β alleles, neto ^βshort (short cytoplasmic tail) and neto ^βnull. The breakpoint coordinates are indicated. (B) Diagram of the predicted Neto-β proteins. The bars mark the antigens for Neto-ex, Neto-β1 and Neto-β2 antibodies. (C) Western blot analysis of lysates from S2 cells (left) and larval muscle (right). S2 cells were transfected with empty vector (lane 1), Neto-β (2), Neto-β^short (3), and Neto-α (4) expression constructs and the lysates were compared with muscle extracts from control (5), neto ^βshort (6), and neto ^βnull (7) third instar larvae. Full length (black arrow) and truncated (white arrow) Neto-β are indicated. No specific signal was detected in neto ^βnull animals. (D-E) Confocal images of boutons at NMJ4 of third instar larvae stained with Neto-ex (red) and with either Neto-β2 (D) or Neto-β1 (E) (green). As expected, Neto-ex signals were detected in control (precise excision) and neto alleles. neto ^βshort NMJs show Neto-β2 signals (D) but both neto-β alleles lack Neto-β1 synaptic signals (E). In contrast, Neto-β1 puncta are present at neto ^hypo NMJs. Scale bars: 2 μm.

Fig 3. Physiology and morphology defects at neto-β mutant NMJs.

(A–F) Electrophysiological recordings of control (precise excision) and neto-β alleles of third instar animals. Representative traces of mEJPs and EJPs at 0.8 mM Ca²⁺ are shown in (A) and (D), respectively, and the results are summarized in S1 Table. The number of NMJs examined is indicated in each bar. The mEJPs amplitude (B) and frequency (C) were reduced for both neto-β alleles. The EJPs amplitude was reduced in neto ^βshort compared to controls, but was normal at neto ^βnull NMJs (E). Both alleles had increased quantal content, indicating a presynaptic compensatory response (F). The muscle resting potential and the input resistance were not affected. (G-H) Representative confocal images at NMJ4 and NMJ6/7 (segment A3) in third instar larvae of indicated genotypes labeled with HRP. The neto-β mutant NMJs have significantly fewer and larger boutons relative to the muscle area (quantified in I-J). In particular, the volume of distal boutons increases 2.4 and respectively 2.6 fold at neto ^βshort and neto ^βnull NMJs. Error bars indicate SEM. ***; p<0.001, **; p<0.005, *; p<0.05, ns; p>0.05. Scale bars: 20 μm, 2 μm in details.

Fig 4. iGluRs synaptic accumulation is perturbed at neto-β mutant NMJs.

(A) Confocal images of NMJ4 boutons in larvae of indicated genotypes labeled for Brp (red), GluRIIC (green), and HRP (blue) (quantified in B-C). neto-β mutant NMJs have increased number of synaptic contacts. The intensity of the presynaptic active zone marker Brp appears to be normal, but the GluRIIC synaptic signals are reduced at neto-β mutant NMJs compared with control (precise excision). (D) Western blot comparison of GluRIIC protein levels in muscle lysates from neto-β third instar larvae. Tubulin was used as a loading control. (E) Confocal images of NMJ4 boutons in larvae of indicated genotypes labeled for GluRIIA (red), GluRIIB (green), and HRP (blue). The synaptic accumulation of GluRIIA is severely reduced at neto-β mutant NMJs (quantified in C). In contrast, the GluRIIB synaptic signals are slightly increased at neto ^βshort NMJs and significantly reduced (by 31%) at neto ^βnull NMJs. (F) Confocal images of NMJ4 boutons in control and neto-α ^RNAi larvae labeled for GluRIIA (red), GluRIIB (green), and HRP (blue). The GluRIIB signals, but not GluRIIA are reduced at Neto-α-depleted NMJs (quantified in G). (H) Western blot analysis of larval muscle extracts from neto ^null mutants rescued by V5-tagged Neto- and carried through RNAi-mediated knockdown as indicated. The neto- ^RNAi appears to be more effective than the CUB1 ^RNAi in knocking down V5-tagged Neto- relative to the Tubulin control. Error bars indicate SEM. ***; p<0.001, **; p<0.005, *; p<0.05, ns; p>0.05. Scale bars: 2 μm.

Fig 5. neto-β mutants have reduced postsynaptic components.

(A–B) Representative confocal images at NMJ4 (segment A3) in third instar larvae of indicated genotypes labeled for HRP (blue) and PAK (red) (A) or Neto-ex (red) and Dlg (green) (B). The neto-β mutant NMJs have drastically reduced levels of synaptic PAK and significantly diminished Dlg accumulation (quantified in C-D), albeit the protein levels are normal in larval muscles as indicated by Western blot analysis (E). Tubulin was used as a loading control. Error bars indicate SEM. ***; p<0.001. Scale bars: 20 μm, 2 μm in details.

Fig 6. Ultrastructure defects at neto-β mutant boutons.

(A–C) Electron micrographs of type Ib boutons in third instar larvae of indicated genotypes. The upper panels show entire boutons; the active zones (AZ, arrows), mitochondria (m), and subsynaptic reticulum (SSR, brackets) are indicated. The neto-β mutant boutons have numerous synaptic contacts, but their active zones often have abnormal T-bar structures (B’-B’”, C’-C’”), including closely spaced, distorted, fused, and floating T-bars. The PSDs length, quantified in serial section (D), is significantly reduced at neto-β mutant boutons. The boutons appear enlarged and the SSR area and thickness diminished in both neto-β mutants (quantified in E-H). Error bars indicate SEM. ***; p<0.001, **; p<0.005. Scale bars: 1 μm, 200 nm in details.

Fig 7. PAK synaptic recruitment is impaired at neto-β mutant NMJs.

(A) Representative confocal images of ventral muscle fields in late embryos (21 hours after egg laying) and early first instar larvae (2 hours after hatching) of indicated genotypes labeled for Brp (red), PAK (green), and HRP (blue). The synaptic PAK signals are weak at neto-β mutant NMJs, but normal at the muscle attachment sites. (B) Confocal images of NMJ4 in third instar larvae of indicated genotypes labeled for PAK (red), GluRIIA (green) and HRP (blue). Muscle expression of a constitutively membrane bound form of PAK (G14> pak ^myr) does not rescue the GluRIIA synaptic abundance at neto ^βshort NMJs compared to control (G14 /+). PAK signals remain diffuse and localize perisynaptically in the absence of an intact Neto-β (see bouton details). (C) Confocal images of NMJ4 in control (w ¹¹¹⁸) and various pak heteroallelic third instar larvae labeled for Neto-ex (red), PAK (green) and HRP (blue). Lack of PAK does not impact the synaptic distribution of Neto-β. Scale bars: (A) 10 μm; (B-C) 20 μm, 2 μm in details.

Fig 8. Neto-β, but not Neto-α, restores PAK recruitment and mEJP amplitude at neto ^null NMJs.

(A) Confocal images of larval NMJ4 labeled for Neto-ex (red), PAK (green) and HRP (blue). Neto-β, but not Neto-α, restores PAK synaptic accumulation over a large range of concentrations tested. (B) Western blot comparison of Neto levels in muscle extracts from control (first lane), and neto ^null larvae rescued with neto-α transgenes (low, medium, and high expression) (magenta gradient), or neto-β transgenes (low, medium, high, and very high expression)(blue gradient). Arrows indicate unprocessed and processed Neto-α (magenta) and Neto-β (blue). (B’)—low exposure. Tubulin was used as a loading control. (C-H) Electrophysiological recordings of neto ^null NMJs rescued by various levels of Neto-α or Neto-β. Data are reported relative to controls matched by rearing at 18°C (empty bars) or 25°C (hatched bars). Representative traces of mEJPs and EJPs are shown in (C) and (F), respectively. The mEJPs amplitude is reduced at NMJs rescued by low and medium levels of Neto-α (D). The mEJPs frequency appears less dependent on Neto levels/isoforms, but is significantly increased in larvae reared at 18°C (E). The EJPs amplitudes are largely normal (G), likely due to subtle, but significant increases in quantal content at Neto-α-rescued NMJs (H). (I) Confocal images of NMJ4 boutons (segment A3) in control (w ¹¹¹⁸) and GluRIIA ^SP16/Df (GluRIIA ^SP16/ Df(2L)cl ^h4) third instar larvae labeled for Neto-ex (red), PAK (green), and GluRIIA (blue). PAK is normally present at GluRIIA mutant synapses, indicating that the synaptic recruitment PAK does not depend on GluRIIA. Genotypes: control (G14-Gal4/+); neto ^{α low} (neto ³⁶ /Y; G14-Gal4/UAS-neto-A9), reared at 25°C); neto ^{α med} (neto ³⁶ /Y; G14-Gal4/UAS-neto-A3, 18°C); neto ^{α high} (neto ³⁶ /Y; G14-Gal4/UAS-neto-A3, 25°C); neto ^{β low} (neto ³⁶, UAS-neto-B6/Y; G14-Gal4/+, 18°C); neto ^{β med} (neto ³⁶, UAS-neto-B6/Y; G14-Gal4/+, 25°C); neto ^{β high} (neto ³⁶ /Y; G14-Gal4/UAS-neto-B3, 18°C); neto ^{β very high} (neto ³⁶ /Y; G14-Gal4/UAS-neto-B3, 25°C). Error bars indicate SEM. ***; p<0.001, **; p<0.005, *; p<0.05, ns; p>0.05. Scale bars: 20 μm, 2 μm in details and (I).