Synaptic clustering of the cell adhesion molecule fasciclin II by discs-large and its role in the regulation of presynaptic structure

Abstract

The cell adhesion molecule Fasciclin II (FASII) is involved in synapse development and plasticity. Here we provide genetic and biochemical evidence that proper localization of FASII at type I glutamatergic synapses of the Drosophila neuromuscular junction is mediated by binding between the intracellular tSXV bearing C-terminal tail of FASII and the PDZ1-2 domains of Discs-Large (DLG). Moreover, mutations in fasII and/or dlg have similar effects on presynaptic ultrastructure, suggesting their functional involvement in a common developmental pathway. DLG can directly mediate a biochemical complex and a macroscopic cluster of FASII and Shaker K+ channels in heterologous cells. These results indicate a central role for DLG in the structural organization and downstream signaling mechanisms of cell adhesion molecules and ion channels at synapses.

Figure 1. FASII Colocalizes with DLG at Type I Neuromuscular Synapses

(A–F) Confocal micrographs of wild-type type I synaptic boutons in muscles 6 and 7 double stained with (A and B) anti-DLG antibodies and (C and D) anti-FASII antibodies. (E and F) Superimposition of the images in (A)–(D). (B), (D), and (F) are high magnification views of a single confocal slice through type I boutons. The arrow in (A) points to a type Ib bouton branch and the arrowhead to a type Is bouton branch. Bar = 30 μm in (A), (C), and (E); 4.5 μm in (B), (D), and (F).

Figure 2. FASII Is Abnormally Localized at Type I Synaptic Boutons in dlg Mutants

(A–D) Confocal micrographs of type I boutons in muscles 6 or 7 in preparations double stained with anti-FASII (green) and anti-HRP (red) in (A and B) a wild type and (C and D) a dlg^X1-2/Df mutant. (B) and (D) represent the same view as (A) and (C), in which the intensity of the FASII channel (green) has been enhanced to show the abnormal distribution of FASII in dlg^X1-2/Df mutants. (E) Anti-DLG staining of type I boutons in fasII^e76/fasII^eb112. (F) Anti-FASII immunoreactivity in dlg^m52 Sh¹⁰² double mutants. The arrow points to type I synaptic boutons. Notice the abnormal localization of FASII in dlg^X1-2/Df, the normal DLG immunoreactivity in fasII mutants, and the lack of immunoreactivity in dlg^m52 Sh¹⁰². Bar = 10 μm.

Figure 3. Rescue of Abnormal FASII Localization by Targeted Expression of DLG and SAP97 at Pre- and Postsynaptic Sites

Expression of FASII at type I boutons in (A) wild type, (B) dlg^X1-2/Df mutants, (C) dlg^X1-2/Df mutants with targeted expression of DLG to the presynaptic cell (dlg^X1-2 P[GAL4]BG380/Df; UAS-dlg), (D) dlg^X1-2/Df mutants with targeted expression of DLG to the postsynaptic cell (dlg^X1-2/Df; P[GAL4] BG487/UAS-dlg), (E) dlg^X1-2/Df mutants with targeted expression of DLG to both the pre- and the postsynaptic cell (dlg^X1-2 P[GAL4]BG380/Df; UAS-dlg/P[GAL4]BG487), and (F) dlg^X1-2/Df mutants with targeted expression of DLG to both the pre- and the postsynaptic cell (dlg^X1-2 P[GAL4] BG380/Df; +/P[GAL4]BG487; +/UAS-SAP97). The altered localization of FASII in dlg^X1-2 type I boutons is only partially rescued by targeting DLG to either the pre- or the postsynaptic cell but is completely restored by targeting either DLG or SAP97 to both the pre- or the postsynaptic cell. Bar = 5 μm.

Figure 4. Number of Active Zones Is Increased in both dlg^X1-2 and fasII^e76 Mutants

(A–D) Electron microscopical view of a type I bouton in (A) wild type, (B) dlg^X1-2/Df mutant, (C) fasII^e76 mutant, (D) dlg^X1-2 with targeted expression of DLG at both pre- and postsynaptic cells. Note the increase in the number of presynaptic densities in both dlg^X1-2 and fasII^e76 and the rescue of these alterations by targeted expression of DLG in dlg^X1-2/Df pre- and postsynaptic cells. Bar = 0.6 μm.

Figure 5. Morphometric Analysis of Type I Boutons in dlg^X1-2, fasII^e76, and dlg^X1-2 Mutants with Targeted Expression of DLG or SAP97 to Pre- and/or Postsynaptic Cells

(A) Number of active zones, (B) cross-sectional area of type I bouton midlines, and (C) number of active zones normalized by the cross-sectional bouton area. Number of samples used in this analysis are WT = 22, dlg/DF = 19, dlg/Y = 23, dlg/Df pr DLG = 12, dlg/Df pt DLG = 11, dlg/Df pr + pt DLG = 17, dlg/Df pr + pt SAP97 = 17, fas2/Y = 19, and fas2 dlg/Y = 19. pr = presynaptic; pt = postsynaptic.

Figure 6. In Vitro Interactions between DLG and SAP97 and FASII

(A) Wild type and dlg^X1-2 body wall muscles extracts were immunoprecipitated with anti-DLG antibodies and the immunoprecipitate sequentially immunoblotted for FASII and DLG as shown. Note the presence of FASII in the wild-type anti-DLG immunoprecipitate and its absence in dlg^X1-2. Positions of molecular size markers are shown in kilodaltons. (B) Dose–response curves of C-terminal 98-amino-acid residues of FASII binding to the PDZ domains in DLG and SAP97. H6-FASII was bound at 100 nM and incubated with various concentrations of GST-fusion proteins (GST, GST-DLG-PDZ1–2, GST-DLG-PDZ2, GST-DLG-PDZ3, GST-SAP97-PDZ1–2, GST-SAP97-PDZ2, GST-SAP97-PDZ3) that were serially diluted (1:3) starting at 5 μM. Binding of GST fusion proteins was detected with an anti-GST antibody and monitored as absorptionat 405 nm with a microplate reader after an alkaline phosphatase–dependent color reaction. The data were fit with the Hill equation (Abs = Abs_max/1+[EC₅₀/{X}]ⁿ) using a nonlinear least square algorithm. (C) Yeast two-hybrid assay of FASII C-terminal 10 amino acid tail to the PDZ domains of DLG and members of mammalian SAP90/PSD-95 family. The semiquantitative assay is based on the induction of yeast reporter genes HIS3 and β-gal. HIS3 activity was measured by the percentage of colonies growing on histidine-lacking medium (+++, >60%; ++, 30%–60%; +, 10%–30%; -, no significant growth) and β-gal activity by determining the time taken for colonies to turn blue in X-gal filter lift assays at room temperature (+++, <1 hr; ++, 1–3 hr; +, 3–8 hr; -, no significant β-gal activity).

Figure 7. Coimmunoprecipitation of FASII with DLG and Shaker

COS7 cells were triply transfected with DLG and either wild-type (V) or mutant (A) forms of Shaker channel Kv1.4 and FASII, as indicated. In the mutants, the C-terminal valine residue (V) of the protein has been mutated to alanine (A): Shaker V, wild-type C-terminus: -ETDV; Shaker A, -ETDA; FASII V, wild-type C-terminus: -NSAV; FASII A, -NSAA. Equal amounts of each cell lysate were immunoprecipitated with Shaker (Kv1.4) antibodies. The immunoprecipitates were then sequentially immunoblotted for Shaker, DLG, and FASII, as shown. “Input” lanes were loaded directly with 10% of the transfected cell lysate used for the immunoprecipitation. As shown in lane 5, FASII coimmunoprecipitates with Shaker only when wild-type FASII and wild-type Shaker are cotransfected with DLG. Positions of molecular size markers are shown in kilodaltons.

Figure 8. Clustering of FASII and Shaker by DLG in Heterologous Cells

(A and B) COS7 cell cotransfected with FASII and Shaker (Kv1.4) and visualized with anti-FASII (A) and anti-Kv1.4 (B) antibodies by double-label immunofluorescence. (C and D) COS7 cell triply transfected with FASII, Shaker (Kv1.4), plus DLG and visualized with anti-FASII antibodies (C) and anti-Kv1.4 antibodies (D). When expressed together in the absence of DLG, FASII and Shaker are distributed diffusely in the cell with some perinuclear accumulation (A and B). When expressed together in the presence of DLG, both FASII and Shaker are redistributed into plaque-like clusters in which both proteins are exactly colocalized (C and D). Thus, DLG can organize coclusters of FASII and Shaker. Bar = 10 μm.