Article Figures & Data
Figures
- Figure 1.
Expression of CTTNBP2 in the brain. A, Schematic of genomic structure and splicing forms of CTTNBP2 transcripts. Arrowheads point to the first and second splice donor sites at the end of exon 4. Arrows denote the positions of RT-PCR primers a–c. Primer a hybridizes before the first splicing site, primer b binds in exon 5, and primer c corresponds to the sequence in intron 4. B, Alignment of the predicted C-terminal amino acid sequences of the long, short, and intron forms of CTTNBP2. The corresponding alternative splice sites are indicated by arrowheads. Amino acid residues in the gray box are encoded by exon 5. Residues in the black box are encoded by intron 4. For the long and short forms, the amino acid sequence encoded by exon 5 results from a frameshift caused by usage of a different spice site. For the long form, the amino acid sequence beyond residue 715 is omitted. C, RT-PCR using mRNA purified from E14.5 mouse brain. D, Domain structures of the three CTTNBP2 forms. The lengths of the three forms are indicated, as is the region (498∼625 aa) used as the immunogen for the production of specific antibodies. CC, Coiled-coil; P, proline-rich; Ank, ankyrin repeat. E, Adult rat brain extract and whole-cell lysates prepared from COS cells transfected with CTTNBP2 isoforms were immunoblotted for CTTNBP2, suggesting that the short form predominates in the brain.
- Figure 2.
CTTNBP2 interacts with cortactin. A, Colocalization of fluorescence protein-tagged CTTNBP2 and cortactin in living cells. GFP-cortactin and mCherry-CTTNBP2 (mCherry-BP2) transiently cotransfected COS cells were replated on glass coverslips 4 h before recording. Images were recorded every 5 s for ∼10 min. A representative image at a single time point is shown. The insets are the local enlargement of the area indicated by the arrow and illustrate the colocalization of CTTNBP2 and cortactin surrounding an intracellular vesicle. B, Enlarged time-lapse images corresponding to the region indicated by the arrow in A. C, Coimmunoprecipitation of cortactin and CTTNBP2 from rat brain with anti-cortactin and nonimmune rabbit IgG. The precipitates were immunoblotted (IB) with CTTNBP2 or cortactin antibodies as indicated. The arrowhead indicates the position of CTTNBP2 (left) or cortactin (right). D, Overlapping distribution of CTTNBP2 and F-actin at the cell cortex (arrowheads). COS cells expressing Myc-tagged CTTNBP2 (Myc-BP2) were fixed and stained with anti-Myc, DAPI (to label nuclei), and phalloidin (to label F-actin). Scale bars: A, 10 μm; B, 2 μm; D, 20 μm.
- Figure 3.
Synaptic distribution of CTTNBP2 in neurons. A, Rat hippocampal neurons were transfected with Myc-tagged CTTNBP2 (Myc-BP2) and GFP at DIV 12 and immunostained with Myc antibody at DIV 18. Bottom panels show higher magnification of dendrites. B, Specificity of CTTNBP2 antibody in immunostaining. Myc-tagged CTTNBP2-transfected COS cells were analyzed by double immunostaining with Myc antibody and CTTNBP2 antibody. C, Synaptic distribution of CTTNBP2 (BP2) in rat hippocampal neurons. Fluorescence staining using PSD-95 and CTTNBP2 antibodies and phalloidin was performed at DIV 23. D, E, Immunohistochemistry of adult mouse hippocampus with CTTNBP2 antibody. D, Low-magnification image covering a part of CA1 and the dentate gyrus of the hippocampus. E, High-magnification image of the dentate gyrus. CTTNBP2 and VGAT antibodies were used for fluorescence immunostaining. F-actin and nuclei were labeled with phalloidin and DAPI, respectively. The enlarged individual images are also shown in the bottom panel. VGAT immunoreactivity is not obviously overlapping or adjacent to the CTTNBP2 puncta. O, Stratum oriens; R, stratum radiatum; L, stratum lacunosum; M, stratum moleculare; DG, dentate gyrus; H, hilus. In C, yellow arrowheads in the individual panels and white arrowheads in the merged panels highlight examples of colocalization. In E, arrowheads indicate the colocalization of CTTNBP2 and F-actin, while arrows denote the positions of VGAT alone. Scale bars: A–C, 20 μm; D, 200 μm; E, 30 μm.
- Figure 4.
Knockdown of CTTNBP2 impairs dendritic spine density and reduces spontaneous neuronal activity. A, Knockdown of CTTNBP2 assayed in COS cells cotransfected with BP2-miR or Ctrl-miR and Myc-tagged wild-type CTTNBP2 (Myc-BP2) or a CTTNBP2 silent mutant resistant to BP2-miR (Myc-BP2-resc). Whole-cell extracts were immunoblotted with Myc antibody and tubulin antibody. B–F, The effect of expression of the BP2-miR in cultured neurons. Rat hippocampal neurons were transfected at 12 DIV with BP2-miR or Ctrl-miR. GFP expressed by the miRNA vectors highlights the transfected cells. Neurons were harvested at DIV 18 for analysis. B, BP2-miR reduces the endogenous CTTNBP2 protein level in rat hippocampal neurons. Transfected neurons were fixed and immunostained with CTTNBP2 antibody. C, CTTNBP2 knockdown decreases spine density and reduces spine width. Representative images of GFP signal are shown. D, Quantification of the number of protrusions per 10 μm of dendrites (left) and width and length of protrusions (right two panels). A total of 20 neurons were collected from two independent experiments for each group; >50 dendrites and 400 spines for each group were assayed. E, mEPSCs were recorded on transfected hippocampal neurons. F, G, Quantification of mEPSC frequency (F) and amplitude (G) in transfected neurons (n = 23 for Ctrl-miR and n = 27 for BP2-miR). H, CTTNBP2 participates in the in maintenance of dendritic spines. Cultured rat hippocampal neurons were transfected with Ctrl-miR and BP2-miR at 20 DIV and were fixed for staining at 26 DIV. Error bars indicate mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.005. Scale bars: B, 20 μm; C, 2 μm.
- Figure 5.
Cortactin functions downstream of CTTNBP2 in the regulation of dendritic spine density. A, Coimmunoprecipitation of cortactin and CTTNBP2 (BP2) mutants. Whole-cell extracts of COS cells transfected with cortactin and wild-type or mutant CTTNBP2 (PA1, P540A/P543A; PA2, P599A/P602A) were precipitated with CTTNBP2 antibody. Immunoblotting (IB) was then performed to assess the presence of cortactin and CTTNBP2 in the precipitates. The arrowhead indicates the position of cortactin (top) or CTTNBP2 (bottom). The asterisk indicates a nonspecific signal. B–D, Rat hippocampal neurons were transfected with control miRNA (Ctrl-miR) or CTTNBP2 miRNA (BP2-miR) along with the Myc-tagged CTTNBP2 silent mutant (BP2-resc), the PA1 mutant (BP2-PA1-resc), or cortactin at 12 DIV. Neurons were harvested for staining with GFP and Myc-tag antibodies at DIV 18. B, Cortactin and the CTTNBP2 silent mutant restore the spine density in CTTNBP2 knock-down neurons. GFP signals were used to outline dendrite and spine morphology. Scale bar, 2 μm. C, Quantification of the density of protrusions. Eighteen neurons and >50 dendrites were assayed for each group. D, The width and length of protrusions assessed from >400 spines in each group. Error bars indicate mean ± SEM. *p < 0.05; ***p < 0.005.
- Figure 6.
CTTNBP2 stably resides at dendritic spines after glutamate stimulation. At 21∼24 DIV, cultured rat hippocampal neurons were stimulated with glutamate (50 μm), NMDA (100 μm), AMPA (100 μm), and KCl (75 μm) for 15 min (A), 15 min (C), or 2 min plus 8 min of recovery (B) in normal growth medium. A, Triple staining with PSD-95 antibodies, CTTNBP2 (BP2) antibodies, and phalloidin. Yellow arrowheads in the individual panels and white arrowheads in the merged panels point to the puncta containing overlapping PSD-95 and CTTNBP2. B, C, Distributions of PSD-95, CTTNBP2, F-actin, and cortactin along dendrites with or without stimulation. Due to technical limitations, we were unable to visualize PSD-95, CTTNBP2, cortactin, and F-actin in the same neurons. Images shown in B and C were not captured from the same cells. Scale bars, 5 μm.
- Figure 7.
GFP-cortactin is more mobile in the absence of CTTNBP2. Rat hippocampal neurons were transfected at 12 DIV with GFP or GFP-cortactin along with control miRNA (Ctrl-miR) or CTTNBP2 miRNA (BP2-miR). The plasmid expressing miRNA coexpresses mCherry, which was therefore used to label miRNA-expressing neurons. FRAP measurements were performed at 18 DIV. A, Expression of GFP-cortactin and mCherry in transfected neurons. B, FRAP of GFP-cortactin in the spines of Ctrl-miR or BP2-miR transfected neurons. Images depict the same spine before (Pre) and 0, 1, 3, 5, 10, 15, and 25 s after photobleaching. The arrowheads indicate the time of photobleaching. Scale bar, 1 μm. C, FRAP analysis of GFP and GFP-cortactin over the 35 s period after photobleaching. The average of the fluorescence intensities of 10 scans acquired before photobleaching was set to 100%, and the curves were fit with one-phase exponential equations. A total of 10 neurons for each group and three spines for each neuron were analyzed. Error bars indicate mean ± SEM.
- Figure 8.
Knockdown of CTTNBP2 reduces the dendritic spine distribution of cortactin. Cultured rat hippocampal neurons were transfected with Ctrl-miR(mCherry) or BP2-miR(mCherry) at 12 DIV and fixed for staining at 18 DIV using cortactin antibody. Cortactin was visualized by Alexa Fluor 488. mCherry signals were used to outlined neuronal morphology. A, Representative images of the distribution of cortactin along dendrites. The 3-pixel-width lines in the enlarged images indicate the paths for line scanning, which start from the tip of dendritic spines and then cross the dendritic shaft. Scale bar, 5 μm. B, Quantitative analysis of cortactin distribution by line scan. More than 35 spines collected from seven neurons were analyzed for each group. The means ± SEM of fluorescence intensity along the path from dendritic spine to dendritic shaft are provided.
- Figure 9.
Model of the function of CTTNBP2 at dendritic spines. A, Protein–protein interactions of CTTNBP2 and cortactin and the potential regulatory signals of actin cytoskeleton dynamics. Cortactin interacts with the Arp2/3 complex through its N-terminal region and binds F-actin via the middle repeat domains. The C-terminal SH3 domain of cortactin is the binding site for CTTNBP2. Src and ERK phosphorylation of cortactin influences the activity of cortactin. In addition to directly binding to the F-actin cytoskeleton, cortactin may also regulate F-actin branching through the Rac1-IRSp53-WAVE pathway. CTTNBP2NL, a molecule sharing similarity with CTTNBP2, associates with the PP2A protein complex, which contains PP2A, striatins, Mob3, and FAM40A/B. Striatins are highly enriched at dendritic spines (Gaillard et al., 2006), and the Mob protein family has been shown to regulate synapse formation in Drosophila (Schulte et al., 2010). FAM40A and FAM40B have recently been shown to modulate the actin cytoskeleton (Bai et al., 2011). It is unclear whether CTTNBP2 also associates with the PP2A complex in neurons; if so, the PP2A complex likely dephosphorylates the cortactin–CTTNBP2 complex and regulates F-actin dynamics. B, CTTNBP2 regulates cortactin distribution at dendritic spines and thus maintains spine structure. The presence of CTTNBP2 targets cortactin to the dendritic spines. When CTTNBP2 is knocked down, cortactin is not efficiently targeted to dendritic spines, leading to shrinkage of the spines. When cortactin is overexpressed in CTTNBP2 knock-down neurons, the ectopic distribution of cortactin into the dendritic spines maintains the spine structure and rescues the defects caused by CTTNBP2 knockdown. Thus far, it is unclear what mechanism or molecule anchors CTTNBP2 at the dendritic spines.
Tables
GenBank accession no. Clone ID Source tissue Containing full exon 4 Containing intron Splicing form BC141407 IMAGE 9056020 Brain Yes No Long AK173254 Pancreatic islet, adult Yes No Long BC068156 IMAGE 30362957 Brain No No Short BQ769661 IMAGE 5697958 Brain, E12.5 No No Short CB526439 IMAGE 6848778 Brain, embryo No No Short AK032356 RIKEN 6430526E05 Olfactory bulb, adult No No Short CB244938 IMAGE 6833056 Brain, embryonic No No Short AK028980 RIKEN 4732477G22 Skin, 10 d neonate Yes Yes Intron BQ961104 IMAGE 6439978 Mammary gland tumor, 5-month-old female Yes Yes Intron
