TRIM32 is an E3 ubiquitin ligase for dysbindin

Abstract

Mutations in the gene encoding tripartite motif protein 32 (TRIM32) cause two seemingly diverse diseases: limb-girdle muscular dystrophy type 2H (LGMD2H) or sarcotubular myopathy (STM) and Bardet-Biedl syndrome type 11(BBS11). Although TRIM32 is involved in protein ubiquitination, its substrates and the molecular consequences of disease-causing mutations are poorly understood. In this paper, we show that TRIM32 is a widely expressed ubiquitin ligase that is localized to the Z-line in skeletal muscle. Using the yeast two-hybrid system, we found that TRIM32 binds and ubiquitinates dysbindin, a protein implicated in the genetic aetiology of schizophrenia, augmenting its degradation. Small-interfering RNA-mediated knock-down of TRIM32 in myoblasts resulted in elevated levels of dysbindin. Importantly, the LGMD2H/STM-associated TRIM32 mutations, D487N and R394H impair ubiquitin ligase activity towards dysbindin and were mislocalized in heterologous cells. These mutants were able to self-associate and also co-immunoprecipitated with wild-type TRIM32 in transfected cells. Furthermore, the D487N mutant could bind to both dysbindin and its E2 enzyme but was defective in monoubiquitination. In contrast, the BBS11 mutant P130S did not show any biochemical differences compared with the wild-type protein. Our data identify TRIM32 as a regulator of dysbindin and demonstrate that the LGMD2H/STM mutations may impair substrate ubiquitination.

Figure 1.

Dysbindin interacts with TRIM32 in yeast and in vivo. TRIM32 and P130S interact with dysbindin in the Y2H system. (A) Schematic representation of the domain organization of TRIM32 and location of the disease-associated mutations and the region used for polyclonal rabbit anti-sera generation. (B) TRIM32, D487N and P130S bait strains were co-transformed with empty prey plasmid (pYesTrp2) or dysbindin prey plasmid. The double transformants were tested for histidine auxotrophy. The previously characterized β-dystrobrevin (β-db) bait is used as a positive control for the dysbindin interaction. (C) Bait strain protein lysates were western blotted with the TRIM32 3293 antibody, showing all three DNA-binding LexA-fusions are synthesized at comparable levels in yeast. (D) TRIM32 and dysbindin interact in mammalian cells. HEK293T cells were transfected with dysbindin and myc-TRIM32, myc-D487N, myc-P130S, myc-RING-mut. Dysbindin co-immunoprecipitated only in the presence of myc-TRIM32 (or any of the mutants), indicating they form a complex in transfected cells. (E) Dysbindin is polyubiquitinated by TRIM32. Myc-dysbindin was immunoprecipitated from HEK293T cells expressing HA-ubiquitin and either TRIM32, D487N, P130S, R394H or RING-mut. Immunoprecipitated dysbindin was probed for the presence of conjugated HA-ubiquitin by western blot. A high molecular weight smear is seen on the HA blot indicative of polyubiquitination. As expected, the RING mutant is unable to ubiquitinate dysbindin. Interestingly, D487N and R394H have impaired ubiquitin ligase activity towards dysbindin, exemplified by the weaker smearing on the HA blot. (F) Effect of TRIM32 on the ubiquitination of PIASy. The ubiquitination of PIASy was assessed under identical condition to those for dysbindin as shown in (E). Co-expression of TRIM32 or TRIM32–D487N had no apparent effect upon the ubiquitination of PIASy. Note that PIASy is modified by endogenous ubiquitin ligases when expressed in HEK293T cells.

Figure 2.

Localization of TRIM32 and dysbindin in skeletal muscle. Longitudinal sections of guinea pig tibialis anterior muscle were stained with the antibodies 3293 (TRIM32), EA-53 (α-actinin), B4 (myomesin) and PA3111A (dysbindin). Primary antibodies were detected using species-specific fluorescent antibody conjugates (Alexa-488 or rhodamine Red-X). TRIM32 and dysbindin colocalize with α-actinin, suggesting that each protein is found on and around the periphery of the Z-line. In addition to the Z-line labelling, dysbindin is also found at the sarcolemma and M-band. Pre-incubation of the anti-TRIM32 antibody with the fusion protein used for immunization blocked labelling in muscle sections but did not interfere with immunoreactivity of α-actinin on the same tissue section. Scale bar is 20 µm.

Figure 3.

Subcellular distribution of TRIM32, dysbindin and PIASy in COS-7 cells. (A) Co-localization of TRIM32–EYFP and dysbindin in COS-7 cells. TRIM32–EYFP and dysbindin were transfected into COS-7 cells and stained for dysbindin using PA3111A (red). Dysbindin is diffusely expressed in the cytoplasm and nucleus. In addition, dysbindin can be found in aggregates that co-localize with the TRIM32–EYFP fluorescence (arrows), which appear yellow in the merged image. Myc-tagged PIASy and TRIM32–EYFP were transfected into COS-7 cells and stained for myc using 9E10 (red). PIASy exhibited a nuclear localization, with TRIM32 residing in cytoplasmic speckles. (B) Localization of TRIM32–EYFP and mutants in COS-7 cells. COS-7 cells were transfected with TRIM32–EYFP, D487N–EYFP, R394H–EYFP and P130S–EYFP. TRIM32 and P130S form discrete cytoplasmic speckles, whereas D487N and R394H are predominantly cytoplasmic. TRIM32–EYFP transfected COS-7 cells were stained with FK1, which detects polyubiquitinated proteins (red). Merging the two images shows that the two antigens predominantly co-localize indicating that TRIM32–EYFP cytoplasmic speckles are associated with polyubiquitinated proteins.

Figure 4.

TRIM32 targets dysbindin for degradation. (A) COS-7 cells were transfected with dysbindin (red) and TRIM32 (green), or dysbindin and the RING mutant (RING-mut) (all untagged in pCIneo). Protein synthesis was blocked by cycloheximide treatment (50 µg/ml) and samples taken every 5 h. The levels of dysbindin in lysates was analysed by quantitative western and plotted graphically (B). The presence of TRIM32 destabilizes dysbindin, increasing its turnover relative to co-expression the RING-mut. Additional bands that might be possible breakdown products of dysbindin can be seen below the main dysbindin band in the presence of TRIM32, but not the RING-mut (asterisk in A). The error bars show the standard error of the mean from three independent experiments. (C) Knockdown of TRIM32 in C2C12s by siRNA increases dysbindin levels. C2C12 cells were transfected on consecutive days with either control siRNA (non-targeting pool) or TRIM32 siRNA. Seventy-two hours after the first transfection lysates were made and levels of TRIM32, dysbindin and tubulin were analysed by quantitative western blot (WB). TRIM32 and dysbindin levels were normalized to tubulin and plotted as a chart (D). The error bars show the standard error of the mean from three independent experiments. The TRIM32-specific J-12 duplex achieved ∼60% knockdown relative to the control siRNA (**P < 0.01 in a one-sample two-tailed t-test). (E) Knockdown of TRIM32 results in a significant increase in dysbindin levels relative to control (*P < 0.05 in a one-sample two-tailed t-test).

Figure 5.

Self-association properties of TRIM32 and mutants. (A) Y2H-mediated analysis of TRIM32 self-association. Co-transformants containing TRIM32, D487N, P130S, R394H bait and prey plasmids or empty bait, and prey plasmids were tested for histidine auxotrophy on media lacking histidine containing 5 mm 3-AT. Only wild-type TRIM32 and P130S could self-associate in yeast. In each case, selective and non-selective media are compared. (B) Wild-type TRIM32, D487N and P130S self-associate in mammalian cells. HEK293T cells were transfected with myc-tagged TRIM32, D487N or P130S and their EYFP-tagged counterparts. EYFP-tagged proteins are only immunoprecipitated in the presence of the myc-tagged protein, indicating that the myc-tagged and EYFP-tagged proteins interact. (C) The LGMD2H/STM mutants failed to interact with wild-type TRIM32 in yeast. Bait and preys were co-transformed into yeast as indicated. TRIM32 is able to self-associate, but failed to interact with either of the LGMD2H/STM mutants. In each case, selective and non-selective media are compared. (D) Co-immunoprecipitation of wild-type TRIM32 with the LGMD2H/STM mutants, D487N and R394H. HEK293T cells were transfected with the constructed as indicated. Proteins were immunoprecipitated from the cell extracts using the anti-myc antibody. The anti-GFP antibody was used to detect the co-immunoprecipitated EYFP-tagged TRIM32. In each case, TRIM32 and the each of the mutants co-immunoprecipitated with myc-tagged wild-type TRIM32. The EYFP-tagged proteins did not immunoprecipitate in the absence of myc-tagged wild-type TRIM32 demonstrating that non-specific binding to the beads was not a confounding factor in this experiment. (E) Dysbindin ubiquitination in cells co-expressing wild-type and mutant TRIM32s. Ubiquitination assays were performed as described above (Fig. 1E). Myc-tagged dysbindin was immunoprecipitated from cells expressing the designated combination of plasmids. In each case, wild-type TRIM32 was co-expressed with either 0.5 µg of TRIM32–EYFP, D487N–EYFP, R394H–EYFP or RING-mut–EYFP. Dysbindin ubiquitination was assessed with the anti-HA antibody as described above. In each reaction, dysbindin is clearly polyubiquitinated by each combination of TRIM32 and mutant. However in extracts from cells expressing myc-TRIM32 and R394H–EYFP, we consistently found a reduction in the levels of the monoubiquitinated form of dysbindin (asterisk). This change was also detected with the anti-dysbindin antibody (asterisk). Inspection of the lysates revealed that similar levels of myc-TRIM32, dysbindin and the EYFP-tagged mutants were produced in each transfection.

Figure 6.

TRIM32 mutants bind to the E2 ubiquitin-conjugating enzyme UbcM3. Myc-tagged TRIM32 (and mutants) were immunoprecipitated from mammalian cells in low stringency RIPA buffer and probed for the presence of co-immunoprecipitated HA-UbcM3 by western blot using the HA antibody. All of the TRIM32 variants could immunoprecipitate UbcM3, with the RING mutant having the most robust interaction.

Figure 7.

Impaired monoubiquitination of D487N. HEK293T cells were transfected with myc-tagged constructs encoding, TRIM32, TRIM32–D487N, TRIM32–RING-mut and TRIM32–P130S as indicated. After 24 h, each protein was immunoprecipitated from cell extracts using the 9E10 antibody. Immune complexes were separated by gel electrophoresis on a 4–20% gradient gel. The gel was stained with colloidal Coomassie blue to visualize protein bands. Bands A1 and A3 were excised and subjected to FTICR-MS. The table shows the percentage of their amino acid sequence covered by peptides identified in MS and the number of unique peptides that match this sequence.