Drosophila enabled promotes synapse morphogenesis and regulates active zone form and function

Abstract

Background: Recent studies of synapse form and function highlight the importance of the actin cytoskeleton in regulating multiple aspects of morphogenesis, neurotransmission, and neural plasticity. The conserved actin-associated protein Enabled (Ena) is known to regulate development of the Drosophila larval neuromuscular junction through a postsynaptic mechanism. However, the functions and regulation of Ena within the presynaptic terminal has not been determined.

Methods: Here, we use a conditional genetic approach to address a presynaptic role for Ena on presynaptic morphology and ultrastructure, and also examine the pathway in which Ena functions through epistasis experiments.

Results: We find that Ena is required to promote the morphogenesis of presynaptic boutons and branches, in contrast to its inhibitory role in muscle. Moreover, while postsynaptic Ena is regulated by microRNA-mediated mechanisms, presynaptic Ena relays the output of the highly conserved receptor protein tyrosine phosphatase Dlar and associated proteins including the heparan sulfate proteoglycan Syndecan, and the non-receptor Abelson tyrosine kinase to regulate addition of presynaptic varicosities. Interestingly, Ena also influences active zones, where it restricts active zone size, regulates the recruitment of synaptic vesicles, and controls the amplitude and frequency of spontaneous glutamate release.

Conclusion: We thus show that Ena, under control of the Dlar pathway, is required for presynaptic terminal morphogenesis and bouton addition and that Ena has active zone and neurotransmission phenotypes. Notably, in contrast to Dlar, Ena appears to integrate multiple pathways that regulate synapse form and function.

Keywords: Actin; Drosophila; Ena/VASP, Dlar; Receptor protein tyrosine phosphatase; Synapse.

Fig. 1

Presynaptic Ena expression is required to promote neuromuscular junction development. Fluorescence images (a-b) and quantification (c-d) of NMJs from muscle 6/7 in segment A2 of third-instar wandering larvae. Flies expressing UAS-AP₄mito (control; A-A") and UAS-FP₄mito (ena^LOF; B-B") under the control of the neuronal 1407-GAL4 driver are shown stained with horseradish peroxidase (HRP; green, top panels), Futsch (red, middle panels), and with the HRP/Futsch channels merged (yellow, bottom panels). C, Quantification of synaptic 1b and 1 s bouton number in neuronal ena^LOF lines demonstrate a statistically significant decrease relative to control. Expression of UAS-Ena(+) under the control of 1407-GAL4 rescues the loss of bouton number in ena^LOF animals (c). D, Branch number is also significantly decreased ena^LOF. * P < 0.05, as determined by Welch’s t-test; error bars indicate ± s.e.m. of genotype; gray shading indicates ± s.e.m. of control; n ≥ 20 NMJs for all genotypes, scale = 20 μm

Fig. 2

Ena is epistatic to Lar, Sdc, and Abl in NMJ growth. Gain of function (GOF) of the RPTP, Dlar (a), and associated HSPG ligand, Sdc (b), fail to rescue the bouton loss phenotype of UAS-FP₄mito (ena^LOF) flies when combined (DBL). Loss of the putative Ena suppressor Abl in Abl²/Df stE34 increases bouton number (c), supporting the antagonistic interaction of Ena and Abl. Full suppression of the Abl²/Df stE34 phenotype is observed in UAS-FP₄mito (ena^LOF) flies (c). Partial suppression of the Abl⁴/Df stE34 phenotype is observed with haplosufficient enaGC5/+ flies (d) indicating Ena is both downstream of and antagonized by Abl. Bouton number was determined by quantifying 1b and 1 s boutons. All results shown are statistically significant relative to control, with P < 0.05, as determined by Welch’s t-test. Error bars indicate ± s.e.m. of genotype; gray shading indicates ± s.e.m. of control; n ≥ 20 NMJs for all genotypes

Fig. 3

Synaptic Abl overgrowth phenotype is epistatic to Dlar and Sdc and requires the catalytic activity of Abl in the pre-synaptic compartment. Third-instar LOF mutants of Dlar (a) and Sdc (b), exhibit decreased bouton number in muscle 6/7 NMJs compared to Canton-S wild-type controls. The phenotypes of Lar and Sdc LOF mutants were suppressed by Abl^LOF (a-b), indicating that Abl is downstream of the Dlar pathway. Expression of UAS-Abl(+) under the control of the neuronal 1407-GAL4 driver rescued the bouton gain phenotype observed in Abl^LOF animals to levels observed in Canton-S flies (c). This indicates that pre-synaptic Abl is necessary and sufficient in synapse morphogenesis. Expression of kinase-dead Abl (UAS-Abl(K-N)) pre-synaptically failed to rescue Abl^LOF phenotypes (c), further supporting the requirement for Abl catalytic activity in synaptogenesis. D, Schematic of the Dlar signaling pathway. Bouton number was determined by quantifying 1b and 1 s boutons. * P < 0.05, n.s. indicates not significant, as determined by Welch’s t-test; error bars indicate ± s.e.m. of genotype; gray shading indicates ± s.e.m. of control; n ≥ 20 NMJs for all genotypes

Fig. 4

Presynaptic Ena regulates active zone structure. Electron micrographs of type 1b synaptic boutons at the 6/7 NMJ from flies expressing UAS-AP₄mito (control, a) and UAS-FP₄mito (ena^LOF, b) under the control of the neuronal 1407-GAL4 driver were obtained to analyze gross, qualitative ultrastructure (a,b) and to quantify active zone area (c,d). Qualitative comparison revealed no catastrophic differences in SSR (pink shading) or bouton (yellow shading) morphology and/or size in ena^LOF (b) compared to controls (a). To determine quantitative phenotypes, mean active zone area was calculated by adding length of the electron dense region multiplied by the thickness of the serial sections (100 nm) for all sections spanning the active zone (c). D, Mean active zone area is significantly increased in ena^LOF. M indicates mitochondria; ** P < 0.01, as determined by Welch’s t-test; error bars indicate ± s.e.m. of genotype; gray shading indicates ± s.e.m. of control; n = 3 animals for all genotypes; scale bar = 500 nm

Fig. 5

Presynaptic Ena function regulates spontaneous but not evoked glutamate release. a, Current clamp recordings from muscle 6 (abdominal segments 3 and 4) revealed similar EJP amplitude and kinetics between the AP₄mito control and FP₄mito under the control of the neuronal 1407-GAL4 driver (top), yet very distinct spontaneous mEJPs (bottom). The mean EJP amplitude was not altered by ena^LOF (b), whereas mEJP frequency (c) and amplitude (d) were both significantly increased. D’, An distribution of mEJP amplitude (an alternate depiction of data in d) shows a shift to the right in ena^LOF animals with abnormally high mEJPs (indicated by five-pointed star), and the mean mEJP amplitude was significantly increased (indicated by filled arrows). Results corresponding to control are depicted with a black arrow and blue distribution; results corresponding to ena^LOF are depicted with a gray arrow and orange distribution). * P < 0.05, as determined by Welch’s t-test, n = 3 animals and 6 NMJs for control, n = 4 animals and 7 NMJs for ena^LOF

Fig. 6

Ena is required to regulate clustering of synaptic vesicles, but not average vesicle size at the T-bar. a-g, Analysis of electron micrographs of type 1b synaptic boutons at the 6/7 NMJ. Representative image of T-bar AP₄mito control (a) and FP₄mito (ena^LOF, b) under control of the neuronal 1407-GAL4 driver. The dashed line (a,b) indicates 200 nm radius from the center of the T-bar. A significant increase in average SV number is observed in FP₄mito animals within this region (ena^LOF, c). d-f, Abnormally-shaped and enlarged SVs within 200 nm from the electron dense adhesive contact of the active zone (indicated by white brackets) were observed in FP₄mito (ena^LOF, white arrow heads, e-f) in contrast to controls (d). Although average area of synaptic vesicles is unchanged in FP4mito (ena^LOF) animals, the distribution of SV area (g) indicates rare large vesicles in these animals (orange distribution), which are not observed in control (blue distribution). * P < 0.05, as determined by Welch’s t-test; error bars indicate ± s.e.m. of genotype; gray shading indicates ± s.e.m. of control; n = 3 animals for all genotypes; scale = 100 nm