Figure 1.
Expression profile of NESH in brain and cellular localization in hippocampal neurons.
(A) During development, the expression of NESH was examined by immunoblot analysis in postnatal whole brains (from P7 to adult). (B) Adult rat brains were dissected into subregions, and the lysates were subjected to immunoblot analysis using anti-NESH antibody. SPIN90 and Homer1c were used as controls for PSD proteins; β-actin served as a loading control. (C) Hippocampal neuron lysates were fractionated to crude synaptosomal (P2) and cytosolic fractions (S2), and the equal amount of NESH in each fraction was examined by immunoblotting. (D) Cellular localization of NESH was examined in primary cultured hippocampal neuron at 19 DIV. Mature hippocampal neurons were double-stained with antibodies against NESH, SPIN90 or PSD95 and with Alexa Fluor 594-conjugated phalloidin. Boxed regions were magnified for better imaging of co-localization. Arrows indicate co-localized regions of the images. The scale bar represents 20 µm.
Figure 2.
Overexpression of NESH alters dendritic spine morphology.
Hippocampal neurons transfected with myc-NESH at 10–12 DIV were fixed at 16–18 DIV, and spine morphology was examined. GFP was co-transfected to visualize dendritic spines. (A) The images represent the control (empty vector) and NESH-transfected neurons. (B) Spine numbers per µm (spine density) were determined in neurons overexpressing NESH and control neurons (n = 20). (C) Dendritic spines were classified to four groups (mushroom, thin, stubby and branched) based on shape, and spine density (per µm) was determined (n = 20). (D) Analysis of spine head width in NESH-overexpressing and control neurons (n = 20). (E) Analysis of spine length (n = 42 for control, n = 27 for NESH). The data were obtained from three independent experiments. Data are presented as means ± SEM. *p<0.05, **p<0.01, ***p<0.001.
Figure 3.
NESH knockdown causes abnormal morphological changes in dendritic spines.
(A) HEK 293T cells were co-transfected with GFP-NESH and siRNAs and then immunoblotted with anti-GFP antibody after incubation for 48–72 h. (B) Cultured hippocampal neurons were transfected with NESH siRNAs at 10–12 DIV, and NESH knockdown was evaluated by immunoblotting at 16–18 DIV. (C) Knockdown of NESH by si591 was confirmed with immunofluorescence assay in hippocampal neurons. GFP was co-transfected with siRNAs to visualize transfected neurons. White arrows indicate untransfected neurons and arrow heads indicate transfected neurons. (D–H) Morphometric analyses were performed to examine the effects of NESH knockdown in hippocampal neurons (n = 18 neurons for control; n = 19 neurons for NESH siRNA). Hippocampal neurons were transfected with control (scrambled siRNA) or NESH siRNA (si591) at 10–12 DIV and fixed at 16–18 DIV. GFP was co-transfected to visualize dendritic spines. The images were acquired using an Olympus IX81 fluorescence microscope. (D) Fluorescence images of neurons transfected with NESH siRNA or scrambled siRNA (control). (E) Spine density in NESH knockdown and control neurons. (F) Densities of the four types of dendritic spines (mushroom, thin, stubby or branched). (G) Spine head width. (H) Spine length. Data are presented as means ± SEM. *p<0.05, **p<0.01, ***p<0.001.
Figure 4.
Overexpression of NESH prevents synapse formation in hippocampal neurons.
(A) Cultured hippocampal neurons were co-transfected with GFP and myc-NESH at 10–12 DIV and fixed at 16–18 DIV. Empty vector was used as a control, and GFP was used to visualize dendritic spines. The fixed neurons were stained with anti-VAMP2 (presynaptic marker) antibody, anti-GluR1 (subunit of AMPA receptor) antibody or Alexa Fluor 594-conjugated phalloidin. (B) Synaptic densities were analyzed by counting the dendritic spines contacting presynapses marked by VAMP2 staining. (C) GluR1 clusters on dendritic spines were measured in neurons overexpressing NESH and compared with control. (D) F-actin fluorescence intensity ratios (spine vs. shaft). Data were obtained from three independent experiments; n = 20 each for control and NESH-overexpressing neurons. Data are presented as means ± SEM. *p<0.05, **p<0.01.
Figure 5.
NESH knockdown reduces synapse formation and affects the postsynaptic apparatus.
(A) Hippocampal neurons were transfected with the control (scrambled siRNA) or NESH siRNA at 10–12 DIV and stained with anti-VAMP2 antibody, anti-GluR1 antibody or Alexa Fluor 594-conjugated phalloidin at 16–18 DIV. GFP was co-transfected with siRNAs to visualize dendritic spines. (B) Synapse formation per µm was analyzed in NESH knockdown neurons and compared with control (n = 15 for control; n = 19 for NESH siRNA). (C) Numbers of GluR1 cluster per µm on spines (n = 17 for control; n = 15 for NESH siRNA). (D) F-actin fluorescence intensity ratios (spine vs. shaft; n = 15 for control and NESH siRNA). Data are presented as means ± SEM. *p<0.05, **p<0.01.
Figure 6.
NESH interacts directly with filamentous actin via its N-terminal region, but not with monomeric actin.
(A) Schematic diagram showing representations of full-length NESH (amino acids 1–367), N-term (N-terminal half, amino acids 1–229) and C-term (C-terminal half, amino acid 221–367). (B) GST pull-down assays were performed to verify the interaction between NESH and monomer G-actin. GST-fused NESH proteins were incubated with purified monomeric G-actin and then pulled down with glutathione Sepharose beads, after which the bound proteins were detected with anti-actin antibody. GST-SPIN90-C-term served as a positive control. (C) F-actin co-sedimentation assays. Purified NESH proteins were incubated with polymerized F-actin. After separating the supernatant (S) and pellet (P) by ultracentrifugation, co-sedimented proteins were detected by Coomassie Brilliant Blue staining. (D) NESH N-term and C-term in F-actin co-sedimentation assays. Note that NESH N-term only interacts with F-actin.
Figure 7.
Overexpression of NESH N-term inhibits spine maturation and synapse formation.
(A–I) Analysis of spine morphology and synaptic structures in neurons overexpressing NESH N-term or C-term. Cultured hippocampal neurons were co-transfected with myc-NESH truncation mutants (N-term or C-term) and GFP at 10–12 DIV and fixed at 16–18 DIV. GFP was used to visualize dendritic spines. (A) Images showing dendrites from neurons overexpressing NESH N-term and C-term. (B–E) Spine morphology (n = 20 neurons for control; n = 18 for N-term; n = 14 for C-term). (B) Spine density per µm. (C) Density per µm of the four established spine shapes (mushroom, thin, stubby and branched) (D–E) Measurement of spine head width (D) and spine length (E). (F) Transfected neurons labeled at 16–18 DIV with anti-VAMP2 antibody, anti-GluR1 antibody or Alexa Fluor 594-conjuagted phalloidin. (G) Synaptic density measured by counting synaptic contacts with presynapses marked by anti-VAMP2 (n = 16 for control, n = 14 for N-term, n = 15 for C-term). (H) GluR1 clusters per µm on spines (n = 20 for control; n = 18 for N-term; n = 14 for C-term). (I) F-actin fluorescence intensity ratios (spines vs. shafts; n = 16 for control; n = 19 for N-term; n = 17 for C-term). Data are presented as means ± SEM. *p<0.05, **p<0.01, ***p<0.001.
Figure 8.
NESH is involved in actin cytoskeleton rearrangement.
(A) Cos-7 cells were transfected with GFP (control) or GFP-NESH, after which the F-actin was stained with Alexa Fluor 594-conjugated phalloidin to observe F-actin-rich lamellipodia. Mock, untreated condition; Latrunculin A, treated with latrunculin A for 10 min to depolymerize F-actin; Recovery, cells maintained for the indicated times after removing latrunculin A. Note that lamellipodia formation was inhibited in the NESH transfectants, as compared to control. (B) Quantification of F-actin fluorescence intensity in NESH-overexpressing cells, as compared with untransfected or GFP-transfected cells (control) (n>25 for untransfected, control and NESH in each condition). Data are presented as means ± SEM. *p<0.05, **p<0.01. !p<0.05, !!p<0.01, !!!p<0.001. *: NESH vs. control, !: NESH vs. untransfected.