The Drosophila tumor suppressor gene, dlg, is involved in structural plasticity at a glutamatergic synapse

Abstract

Background: Synaptic contacts between neurons and their targets are dynamic entities that can change depending on developmental and functional states of the pre- and postsynaptic cell. However, the molecular factors involved in this plasticity have remained largely unknown. We have demonstrated previously that the Drosophila tumor suppressor gene, discs-large (dlg), is expressed at neuromuscular synapses, and is required for normal synapse structure. A family of dlg homologues is also expressed at mammalian synapses, where they interact with the N-methyl-D-aspartate receptor and ion channels. Here, we provide the first demonstration of the involvement of dlg in structural synaptic plasticity during postsynaptic target growth.

Results: We used a temperature-sensitive dlg allele to demonstrate that there are two stages, late embryogenesis and larval stages, at which dlg is necessary for normal formation of synapses. These stages are coincident with dynamic DLG expression at presynaptic sites in the late embryo, and at postsynaptic regions in the larva. Ultrastructural and confocal analyses reveal that Drosophila neuromuscular junctions undergo a dramatic expansion of the postsynaptic apparatus, which is paralleled by target muscle growth. We show that this process of postsynaptic expansion is partially blocked in dlg mutants.

Conclusions: Our results demonstrate that dlg is required during synapse maturation. We show that dlg is involved in the determination of postsynaptic size during target muscle growth. Because motoneuron targets in the larva are continuously growing, synaptic contacts are structurally plastic, undergoing continuous expansion. We conclude that dlg plays an important role in this form of structural synaptic plasticity.

Figure 1

The body wall muscles of Drosophila larva are innervated by synaptic boutons with different structure, position on the muscle cells, and neurotransmitter content. Depicted are the longitudinal muscle fibers 6, 7, 12 and 13, which are innervated by different complements of motorneurons. Type I motorneurons are different for each of the four muscle cells shown. A single type II motorneuron innervates muscles 12 and 13.

Figure 2

Developmental expression of DLG at type Ib boutons in wild-type. (a,b) type Ib boutons at muscles 12 and 13 in the mid-stage-17 embryo, about 4 h before hatching, double-stained with (a) anti-HRP, and (b) with anti-DLG antibodies. Note that at this stage little DLG label is observed. Arrow in (a) points to filopodia observed in the newly formed boutons at this stage. (c) Type Ib boutons at muscles 12 and 13 in the late-stage-17 embryo, about 2 h before hatching, double-stained with anti-DLG (green) and anti-HRP (red) antibodies. Regions of antigen colocalization appear yellow in the micrographs. Note that DLG immunoreactivity is present inside type I boutons (arrowhead) and is absent from the developing bouton border and filopodia (arrow). (d) Type Ib boutons at muscles 12 and 13 in the early first instar. Note that the DLG label (arrow) is present at the edge of the bouton (yellow), and at the postsynaptic muscle (green; arrow). (e) Wandering third instar type Ib boutons at muscle 12 double-labeled with anti-DLG and anti HRP antibodies. The anti-HRP and anti-DLG label colocalize at the border of the bouton (yellow), and the DLG label extends to the postsynaptic muscle (green). Is: type Is boutons; Ib: type Ib boutons; III: type III boutons; II: type II boutons. Bar = 1 μm in (a–d) and 3 μm in (e).

Figure 3

Wild-type DLG expression at muscles 6 and 7 during larval stages: (a) in early first instar (0 h); (b) during mid-to-late first instar (about 12 h after hatching); (c) in the second instar stage (about 36 h after hatching); and (d) in the third instar wandering stage (about 96 h after hatching). Arrows point to the DLG label surrounding boutons. Bar = 1.5 μm.

Figure 4

Development of the SSR at type Ib boutons in wild-type and in dlg^v59/Df mutant samples. (a) First instar wild-type, (b) second instar wild-type, and (c) early third instar wild-type Ib bouton. (d) First instar dlg^v59/Df, (e) second instar dlg^v59/Df, (f) early third instar dlg^v59/Df type Ib bouton. Abbreviations: bl, basal lamina; b, presynaptic bouton; m, muscle; t, trachea. Short arrow = active zones, long arrow = SSR membrane. Bar = 0.22 μm in (a,d), and 0.44 μm in (b,c,e,f).

Figure 5

Morphometric analysis of wild-type and dlg^v59/Df boutons through development. (a) length of the SSR (L_SSR); (b) volume of the muscle cell (Vm), (c) area of the bouton (Ab). Time axis is measured in hours after hatching. (d) Schematic representation of SSR development during larval stages in wild-type and dlg mutants.

Figure 6

Type Ib boutons in dlg^ts/Df reared at permissive and restrictive temperature. (a–c) electron micrographs of third instar type Ib boutons in (a) dlg^ts/Df at 18 °C, (b) dlg^ts/Df at 29 °C, and (c) wild-type. Note the less extensive SSR in dlg^ts/Df at 29 °C as compared to dlg^ts/Df at 18 °C or wild-type. Short arrows: active zone; longer arrow: SSR; b: presynaptic bouton; m: postsynaptic muscle. (d–f) anti-DLG immunoreactivity at muscle 6 type Ib boutons in (d) dlg^ts/Df at 18 °C, (e) dlg^ts/Df at 29 °C, and (f) in wild-type. Notice that anti-DLG immunoreactivity in dlg^ts/Df at 18 °C is similar to wild-type, and that its extent in dlg^ts/Df at 29 °C boutons is significantly reduced. Bar = 0.44 μm in (a–c), and 5 μm in (d–f).

Figure 7

Percentage of samples with a mutant phenotype at type Ib boutons in dlg^ts/Df mutants pulsed to restrictive temperature at different stages of embryonic and larval development. Number of wild-type and mutant samples for each temperature pulse paradigm are shown next to each bar. The approximate duration of each temperature pulse (tp, in days) is indicated below. The onset and duration of temperature pulses was determined by examining larval molts (see Materials and methods). E, embryonic stage (tp ~ 40 h in dlg^ts/Df and controls); 1st, first instar (tp ~ 3 days in dlg^ts/Df, and 2 days in controls); 2nd, second instar (tp ~ 4 days in dlg^ts/Df, and 3 days in controls); 3rd, third instar (tp ~ 6 days in dlg^ts/Df, and 4 days in controls). Second bar from the top, wandering third instar (tp ~ 4 days in dlg^ts/Df, and 2 days in controls).

Figure 8

Development of innervation during the embryonic period in dlg mutants. (a,b) Stage-16 embryo fillets stained with anti-fasciclin II antibody showing that pathway of motor axons in dlg mutant (b) is indistinguishable from that in wild-type controls (a). (c,d) High-magnification view of stage-16 growth cones stained with anti-Fasciclin II antibody, as they interact with muscles 6, 7, 12 and 13 in wild-type (c) and a dlg^v59/Df mutant embryo (d). The approximate location of muscle fibers is indicated in each figure. (e,f) High-magnification view of growth cones and synaptic boutons at muscles 6, 7, 12 and 13, in stage-17 wild-type (e) and dlg^v59/Df (f) embryos stained with anti-Fasciclin II antibodies. Abbreviations: ISN, intersegmental nerve; SN, segmental nerve; TN, transverse nerve; VG, ventral ganglion; B, brain; Ph, pharyngeal apparatus. Bar = 116 μm in (a,b), and 13 μm in (c–f).