Figure 1.
Identification of dysbindin-1-associated proteins.
A) Expression of the GST and GST-dysbindin-1 in E. coli (asterisk). Purified GST and GST-dysbindin-1 were separated by 10% SDS-PAGE and visualized by Coomassie Brilliant Blue (CBB) staining. B) Proteins identified as dysbindin-1-associated proteins. The purification of GST-dysbindin-1, isolation of lysates from the human neuroblastoma cell line SH-SY5Y, and the pull-down assays were performed as described in the “Materials and Methods.” Eluates of the GST- or GST-dysbindin-1 associated proteins were separated by SDS-PAGE and stained with CBB. Lanes 1 and 2 represent GST and GST-dysbindin-1 alone without cell extract, and lanes 3 and 4 show the associated proteins pulled down by incubating whole-cell extracts from SH-SY5Y cells with GST-dysbindin-1 or GST, respectively. The appropriate portions of the polyacrylamide gel containing the specific protein bands in lane 3 were analyzed by MALDI-TOF-MS as described in the “Materials and Methods.” The arrows indicate the proteins identified by MALDI-TOF-MS with a high score.
Table 1.
Proteins identified as dysbindin-1-associated proteins by mass spectrometry.
Figure 2.
Evaluation of the anti-dysbindin-1 antibody and co-immunoprecipitation of endogenous dysbindin-1 and Ku70/80.
A) The anti-dysbindin-1 antibody recognizes all isoforms of human dysbindin-1. The lysates of COS-7 cells overexpressing each isoform were detected by Western blotting using the anti-dysbindin-1 antibody (center panel), the antigen-absorbed antibody (left panel), or the anti-myc antibody (right panel). In the lanes 8 and 12, there were two bands, lower one of which was thought to be the degraded form of isoform C. B) Endogenous Ku70/80 co-immunoprecipitated with endogenous dysbindin-1. Immunoprecipitation was performed using SH-SY5Y cells with the anti-dysbindin-1 antibody or normal rabbit IgG (negative control). As indicated by the arrowheads, Ku70 and Ku80 immunoprecipitated with dysbindin-1 (indicated by arrows), which was pulled down with the anti-dysbindin-1 antibody (lanes 5–7). In the overexposured immunoblot, two more bands corresponding to isoforms C and B were detected, indicating that the protein levels of these isofoms were much lower than that of isoform A.
Figure 3.
Localization of endogenous dysbindin-1.
A) Localization of endogenous dysbindin-1 in SH-SY5Y cells. SH-SY5Y cells were grown on collagen-coated glass coverslips and immunostained using the anti-dysbindin-1 antibody and the secondary antibody conjugated with Alexa 488 (a-1). Nuclei were visualized by incubating with TOPRO3 (b). Nuclei and dysbindin-1 were merged in c. The boxed area is enlarged in the bottom row (a-2), which indicates existence of endogenous dysbindin-1 in nuclei. Green: dysbindin-1; Blue: TOPRO3; White: phase-contrast. Scale bar, 20 µm except for a-2, 5 µm. B) Subcellular distribution of endogenous dysbindin-1 in SH-SY5Y cells. Cytosolic and nuclear fractions were obtained from SH-SY5Y cells and identified with anti-α-tubulin and anti-lamin B antibodies. Lys: whole lysate; Cyt: cytosolic fraction; Nuc: nuclear fraction. Endogenous dysbindin-1 and Ku localized to both the nuclear and cytosolic fractions.
Figure 4.
Subcellular localization and interaction with Ku70/80 of three isoforms of dysbindin-1.
Protein extracts were prepared from COS-7 cells transfected with the plasmids as indicated. A) Subcellular distribution of three isoforms of dysbindin-1 in transfected COS-7 cells. Cytosolic and nuclear fractions were obtained from COS-7 cells and identified with anti-α-tubulin and anti-lamin B antibodies. Lys: whole lysate; Cyt: cytosolic fraction; Nuc: nuclear fraction. Isoform A and B were localized to both the nuclear and cytosolic fractions but isoform C was not. B) and C) Differences in binding specificities among dysbindin-1 isoforms. Proteins were immunoprecipitated with the anti-myc antibody and detected with the anti-V5 (upper panels) and anti-myc antibodies (lower panels). Ku70/80 were co-immunoprecipitated with isoforms A and B of dysbindin-1 but not with isoform C, indicating that Ku70/80 formed complex with dysbindin-1 in an isoform selective manner.
Figure 5.
Phosphorylation of dysbindin-1 by DNA-PK complex.
A) In vitro phosphorylation of dysbindin-1 by DNA-PK complex. The purified DNA-PK complex and GST-dysbindin-1 were incubated with [γ-32P]-labeled ATP as described in the “Materials and Methods.” These samples were subjected to 10% SDS-PAGE and stained with CBB (left panel). The right panel shows the uptake of [γ-32P] ATP by phosphorylation of dysbindin-1. As shown in the left panel, the amount of BSA (lanes 1, 2), GST (lanes 3, 4), GST-dysbindin-1A (lanes 5–8), B (lanes 9, 10), and C (lanes 11, 12) were equal. Lanes 7–12 in the right panel show that the three isoforms of dysbindin-1 were phosphorylated by the DNA-PK complex. B) Phosphorylation of dysbindin-1 isoforms A and B in mammalian cells. Protein extracts were prepared as described in “Materials and Methods”. These samples were separated by Phos-tag SDS-PAGE and detected with the anti-dysbindin antibody. The phosphorylation levels of V5-dysbindin-1A and B were higher than that of C. Lanes 1 and 2: empty vector; lanes 3 and 4: V5-dysbindin-1A; lanes 5 and 6: V5-dysbindin-1B; lanes 7 and 8: V5-dysbindin-1C. The immunoblot after the normal SDS-PAGE showed that the amounts of dysbindin-1 were not altered after the dephosphorylation procedure by AP.
Figure 6.
Schematic representation of the human dysbindin-1 gene and its three isoform structures.
The gene encoding dysbindin-1 located at chromosomal locus 6p22.3, and its several genetic variations are associated with schizophrenia. These variations (SNPs and haplotypes) are located in intron or promoter regions, and almost all are located in the N-terminus of the gene. Red asterisks indicate major SNPs. Three isoforms (A, B, and C) of human dysbindin-1 were reported by NCBI. Isoform A encodes the longest isoform, and isoform B contains an additional segment in the coding region compared to isoform A. Isoform C contains an alternate splice site in the 5′ coding region and uses a downstream start codon, compared to isoform A; this isoform has a shorter N-terminus compared to isoform A.