Visual Arrestin 1 Acts As a Modulator for N-Ethylmaleimide-Sensitive Factor in the Photoreceptor Synapse

Immunohistochemical fluorescent labeling of NSF and Arr1 in the WT mouse retinas. Adult WT mouse retina frozen sections were triple labeled fluorescently with the anti-mouse Arr1 MAb D9F2 (red), anti-rabbit NSF PAb (green), and appropriate secondary antibodies and TO-PRO-3 for the nuclei (blue). The immunoreactive staining pattern of NSF is mainly in the OPL and IPL in DA (a) or LA (b) retinas. The Arr1 MAb-immunoreactive label is predominantly in the inner segment, perinuclear area, and a fraction in the photoreceptor terminal in DA retinas, whereas the Arr1 MAb immunoreactivity is translocated to the outer segment in LA retinas. The Arr1 MAb immunoreactivity is extensively dual localized with NSF immunological staining in the OPL in DA retinas (c; arrows) and only limited dual staining in the OPL in LA retina (d; arrows). OS, Outer segment; IS, inner segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; RGC, retinal ganglion cell layer. Scale bars: a, b, 50 μm; c, d, 10 μm.

Functional analysis of the interaction between Arr1 and NSF. i, Mapping the binding site of Arr1 on NSF. NSF has three functional domains: N domain is for SNARE protein complex binding, D1 domain is for fusion complex disassembly, and D2 domain is for NSF homo-hexamer formation. To define the region in NSF that interacts with Arr1, His₆-tagged truncated segments of NSF with varying lengths (amino acid residues 1-744, 197-744, 1-477, and 1-205^∧478-744) and GST-tagged NSF (amino acids 1-250) and NSF (amino acids 197-250) were constructed, and GST pull-down assay was performed. Bound proteins were detected by immunoblot analysis with anti-mouse His-tag MAb and anti-mouse Arr1 MAb D9F2. Arr1 only bound NSF fragments that included the amino acids 197-250, which was located in the junction of the N and D1 domains. ii, Defining the binding site of NSF on Arr1. GST pull-down assay demonstrated a direct interaction between the N-terminal domain of Arr1 and NSF. GST alone or GST-Arr1 (amino acids 1-191), GST-Arr1 (amino acids 1-370), and GST-Arr1 (amino acids 1-403) coupled to beads were incubated with purified His₆-tagged NSF in the binding buffer containing 2 mm ATP plus 8 mm MgCl₂ or 2 mm ATP-γ-S plus 8 mm MgCl₂. Bound NSF was detected by immunoblot analysis using anti-rabbit NSF antibody. iii, Association of Arr1 with NSF in COS-7 cells. COS-7 cells coexpressing pcDNA4-HisMax-Arr1 and pcDNA3.1-GFP-NSF (amino acids 1-744) or NSF (amino acids 197-250) were processed for immunofluorescence and analyzed by confocal microscopy for the extent of colocalization of pcDNA4-HisMax-Arr1 [b, e, h, red (anti-mouse His-tag MAb); c, f, i, merge] with pcDNA3.1-GFP-NSF (amino acids 1-744) [a, green (GFP); c, merge], NSF (amino acids 197-250) [d, green (GFP); f, merge], and empty vector [g, green (GFP); i, merge].

Arr1–NSF complex formation is enhanced by ATP. GST-hybrid Arr1 proteins (3 mg) were immobilized on glutathione-agarose beads and then incubated with His₆-tagged NSF protein (3 mg) and buffers containing the following: lane 1, 2 mm ATP/8 mm MgCl₂/without EDTA; lane 2, 2 mm ATP/8 mm MgCl₂/with EDTA; lane 3, only 8 mm MgCl₂; lane 4, 2 mm ATP-γ-S/8 mm MgCl₂. Bound proteins were eluted with 20 mm glutathione and detected by immunoblot analysis. The amount of NSF bound was normalized to the amount of Arr1 immunoprecipitated. Results represent the means ± SEM for three independent experiments. **p < 0.01; ***p < 0.001.

NSF ATPase activity assay. The ATPase activity of NSF was measured with a colorimetric assay. Recombinant NSF (0.2 μg/μl) was pretreated with 10 mm NEM, a NSF inhibitor, as a negative control or with increasing concentrations of recombinant Arr1 protein for 10 min at 37°C. The release of inorganic phosphate was measured by adding BIOMOL Green (BIOMOL) and reading the absorbance at 620 nm on a Benchmark Plus microplate reader (Bio-Rad). Arr1 significantly increases the NSF ATPase activity in a dose-dependent manner. Corrections were made by subtracting the NEM-treated control values. Results of the ATPase assay represent the means ± SEM for four independent experiments. **p < 0.01; ***p < 0.001.

SNARE complex disassembly assay. a, Assembly status of SNARE complexes. GST-Syntaxin 4 was immobilized on glutathione-agarose beads and incubated with His₆-tagged VAMP2 and SNAP-25 with (lane 2) or without (lane 1) NSF and α-SNAP for 30 min at 4°C. Proteins precipitated with beads were separated by 12% SDS-PAGE and visualized by Coomassie blue staining. b, Arr1 enhanced NSF disassembly activity in a dose-dependent manner. The NSF disassembly assay was performed by pretreating recombinant His₆-tagged NSF with increasing concentration of Arr1 and then mixing with immobilized GST-Syntaxin 4 with His₆-tagged-α-SNAP, VAMP2, or SNAP-25. Proteins precipitated with glutathione-agarose beads were transferred to PVDF and incubated with (from top to bottom) anti-rabbit NSF PAb, anti-mouse Syntaxin 4 MAb, anti-rabbit SNAP-25 PAb, and anti-mouse VAMP2 MAb and appropriate secondary antibodies. SNARE complexes of ∼250 kDa were detected with all the antibodies listed above.

Quantitative RT-PCR and protein expression levels of NSF, vGLUT1, VAMP2, and EAAT5 in the retinas. a, Expression levels of NSF, VAMP2, EAAT5, and vGLUT1 in WT and Arr1^−/− mouse retina were measured by quantitative RT-PCR. Each column represents the average of three amplification reactions (mean ± SEM), performed on a cDNA sample reverse transcribed from total RNA prepared from 10 pooled retinas using Trizol reagent and transcribed into cDNA with oligo-dT₂₀ using the Superscript III system (Invitrogen). Values for light-adapted WT retinas were set to 1. The NSF, vGLUT1, VAMP2, and EAAT5 mRNA levels were significantly higher in the dark-adapted (D) WT retinas compared with light-adapted (L) WT retinas. In the light, significantly lower expression levels of vGLUT1 and EAAT5 were observed in the Arr1^−/− retinas compared with WT retinas. In the dark, NSF, vGLUT1, VAMP2, and EAAT5 mRNA levels were markedly decreased in the Arr1^−/− retinas compared with the WT retinas. *p < 0.05; **p < 0.01; ***p < 0.001. b, Immunoblots analysis of NSF, vGLUT1, EAAT5, VAMP2, and SNAP-25. Mouse mGAPDH was used as internal control. In the bar graph, the expression level of these proteins is expressed as a ratio to GAPDH expression. Values for light-adapted WT retinas were set to 1. Results represent the means ± SEM for three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001.

Synaptic uptake of FM1-43 revealed a decrease of synaptic activity in the Arr1^−/− mouse retina. Enhanced FM1-43 uptake in the OPL (arrows) and IPL was observed in depolarized (25 mm K⁺) WT retina (a), compared with depolarized Arr1^−/− retina (b). FM1-43 intensity was also significantly reduced in the OPL (arrows) and IPL in non-depolarized WT (c) and Arr1^−/− (d) retinas. Scale bars, 50 μm.