Alsin colocalizes with GRIP1.A, Subcellular fractionation of alsin, GRIP1, GluR2, and PSD95. Whole-brain lysate was fractionated into several subcellular fractions, including the soluble (S1), nuclear (P1), cytosol (S2), crude synaptosomal fraction (P2), light membrane (LM), layer between 0.8 and 1.0 M sucrose (0.85), synaptic plasma membrane fraction (Syn), PSD, and mitochondria (Mit). PSD95 was used as a marker to indicate PSD fraction. B, GFP-tagged ALS2 and myc-tagged GRIP1 were cotransfected into cultured hippocampal neurons and immunostained with antibodies against GFP and myc, respectively. ALS2 (green) and GRIP1 (red) colocalized in the dendritic spines (arrows, bottom) as revealed by confocal microscopy. The bottom panels are the enlargements of boxed areas in upper panels. Scale bars: B, top, 20 μm; bottom, 5 μm.

Increased ratio of synaptic GluR2-lacking AMPARs versus GluR2-containing AMPARs in ALS2^-/- cortical neurons. A, The evoked AMPAR-mediated responses from a wild-type and ALS2^-/- neuron bathed in normal physiological solution containing additional 10 μM PhTx at different holding potentials (from −60 mV to +40 mV in 10 mV steps). B, The I—V plots of evoked responses from the wild-type and ALS2^-/- neurons shown in A. C, Bar graph shows that rectification of evoked AMPA responses (defined as the ratio of AMPA responses measured at −60 mV and + 40 mV) in WT and KO neurons. Note that rectification inALS2^-/- neurons bathed in normal physiological solution was significantly larger than those in WT neurons bathed in normal physiological solution and in ALS2^-/- neurons bathed in normal physiological solution containing additional 10 μM PhTx (KO/PhTx; WT, 1.47 ± 0.08, n = 28; KO, 2.36 ± 0.25, n=20; KO/PhTx, 1.46 ± 0.09n=27;**p<0.01). D, Evoked AMPAR-(−60 mV) and NMDAR- (+40 mV) mediated responses from wild-type and ALS2^-/- neurons. E, Bar graph shows that the ratio of NMDA and AMPA responses in ALS2^-/- neurons was the same as that in wild-type neurons (WT, 1.52 ± 0.07, n = 31 vs KO, 1.82 ± 0.18, n = 22; p = 0.47). F, Evoked AMPAR-mediated responses from wild-type and ALS2^-/- neurons at different stimulation intensities. G, Plot of the amplitudes of evoked AMPA responses in wild-type and ALS2^-/- neurons against different stimulus intensities. *p < 0.05 (Mann—Whitney rank sum nonparametric test).

ALS2^-/- spinal motor neurons are more vulnerable to AMPA or KA-mediated excitotoxicity. A, Primary cortical neurons were treated with AMPA, and the viability of these neurons was measured by the LDH assay. Each data point represented three independent experiments. Error bars represent SEM. ^#p < 0.001; *p < 0.01. B, Organotypic spinal cord slice cultures of WT (Ba, Bc) and KO lumbar spinal cords (Bb, Bd) were allowed to recover for 1 week before treatment with vehicle (Ba, Bb) or AMPA (25 μM; Bc, Bd) for 2 weeks. The spinal motor neurons were revealed by immunostaining with the SMI32 antibody. C, Bar graph shows the average numbers of SMI 32-positive spinal motor neurons in each cultured spinal cord slice. Error bars represent SEM. ^#p<0.001.D, KA was injected into the ventral horn parenchyma of lumbar spinal cord of wild-type (Da) and ALS2^-/- (Db) mice. Brackets delineate motor neuron columns in the spinal cord. Arrows mark the injection sites. E, Bar graph shows the quantitative analysis of the KA-induced motor neuron loss in the two different mouse genotypes. Error bars represent SD. Scale bars: B, 20μm; D, 260 μm. *p < 0.001.

Alsin interacts with GRIP1.A, Schematic representation of ALS2 and GRIP1 expression constructs. Conserved domains within ALS2 include an amino-terminal RLD, a middle DH and PH-like domain, and a carboxyl-terminal VPS9-like domain in conjunction with an upstream membrane occupation and recognition nexus (MORN) motifs. GRIP1 contains seven PDZ domains. B, Myc-tagged GRIP1 was coexpressed with HA-tagged RLD, DH/PH, and VPS9 domains of alsin in HEK293 cells. The presence of GRIP1 in the immunoprecipates was detected with an anti-myc antibody. C, The expression of myc-GRIP1 in the total cell lysates was detected by an anti-myc antibody and the expression of HA-RLD (lane 1), HA-DH/PH (lane 2), HA-VPS9 (lane 3), and HA-GFP (lane 4) in the total cell lysates were detected by an anti-HA antibody. D, The GST-tagged RLD domain of alsin was coexpressed with HA-tagged PDZ1–3, PDZ4–6, and full-length GRIP1 in HEK293 cells. The full-length GRIP1 and its PDZ1–3 but not PDZ4–6 were pulled down together with the RLD domain of alsin and detected by an anti-HA antibody. E, The expression of GST-RLD of alsin in total cell lysates was detected by an anti-GST antibody and the expression of HA-PDZ1–3 (lane 1), HA-PDZ4–6 (lane 2), and HA-GRIP1 (lane 3) in total cell lysates was detected by an anti-HA antibody. F, The P2 fraction of cerebellar homogenates prepared from ALS2 wild-type (+/+) (lane 1) or knock-out (-/-) (lane 2) mice was incubated with purified HA-tagged GRIP1 and pulled down by anti-HA affinity matrix. In a separate control, the P2 fraction of cerebellar homogenate prepared from ALS2 wild-type mice (+/+) (lane 3) was pulled down by anti-HA affinity matrix (preincubated with nontransfected HEK293 cell lysate) alone. The presence of alsin in the immunoprecipate was detected by Western blot using the polyclonal alsin antibody. G, GRIP1-containing protein complex was immunoprecipitated from wild-type (lanes 5, 6) and ALS2^-/- (lanes 7, 8) whole-brain lysates by a GRIP1 specific rabbit polyclonal antibody. As a control, normal rabbit IgG was used in parallel with GRIP1 antibody (lane 9). The presence of alsin in brain lysates and immunoprecipitates was detected by Western blot using the alsin polyclonal antibody.

Increased accumulation of GRIP1 in perikaryon and decreased accumulation of GRIP1 in synapses in ALS2^-/- spinal cord motor neurons. A, Representative images of GRIP1 immunostaining revealed a discrete and punctate staining pattern of GRIP1 in the soma of ALS2^-/- (KO, Ad) and wild-type spinal motor neurons (WT, Aa). Spinal motor neurons were revealed by immunostaining with SMI32 antibody (Ab, Ae). B, Representative images of PICK1 immunostaining showed a similar staining pattern as GRIP1 in the soma of ALS2^-/- (Bd) and wild-type spinal motor neurons (Ba). Spinal motor neurons were revealed by immunostaining with SMI32 antibody (Bb, Be). C, The bar graph shows that the accumulation of GRIP1, but not PICK1, is significantly increasedin the soma of ALS2^-/- spinal motor neurons (n=10) as compared with the wild-type controls (n=9;***p<0.001). The fluorescence signal of GRIP1 or PICK1 is normalized with SMI32 and represented as an arbitrary unit. D, Representative images of wild-type (left) and ALS2^-/- (right) spinal motor neurons (top) and their dendrites (bottom) immunostained with GRIP1 (green) and synaptophysin (synap; red). E, The bar graph shows mean values (arbitrary unit) of GRIP1 and synaptophysin fluorescence intensities after background subtraction at synapses (WT, n = 19 vs KO, n = 20). Error bars represent SEM. *p < 0.05. Scale bars: A, B, 20 μm; D, low magnification images, 20 μm; high magnification images, 5 μm.

Decreased presentation of GluR2 at the plasma membrane of ALS2^-/- cortical neurons after AMPA treatment. A, Presentation of glutamate receptors at the cell surface of KO neurons after AMPA treatment. Cortical neurons at 21 d in vitro were treated with either 100 μM AMPA (lanes 3, 5) or saline (lanes 2, 4) for 5 min before biotinylation. Equal dilution of unpurified (total) and NeutrAvidin-affinity purified proteins (surface) were analyzed by quantitative immunoblotting with anti-GluR1, GluR2, GluR3, GluR2/3, NR1, ALS2, and βIII-tubulin antibodies, respectively. B, Bar graph shows that the surface fractions of GluR1, GluR2, GluR3, GluR2/3, and NR1 in WT and KO neurons after the treatment of saline. C, Bar graph shows that the cell surface presentations of GluR1, GluR2, GluR3, GluR2/3, and NR 1 in WT and KO neurons after the treatment of AMPA. **p<0.01. Errorbars represent SEM.