TRIM28 functions as a negative regulator of aggresome formation

Abstract

Selective recognition and elimination of misfolded polypeptides are crucial for protein homeostasis. When the ubiquitin-proteasome system is impaired, misfolded polypeptides tend to form small cytosolic aggregates and are transported to the aggresome and eventually eliminated by the autophagy pathway. Despite the importance of this process, the regulation of aggresome formation remains poorly understood. Here, we identify TRIM28/TIF1β/KAP1 (tripartite motif containing 28) as a negative regulator of aggresome formation. Direct interaction between TRIM28 and CTIF (cap binding complex dependent translation initiation factor) leads to inefficient aggresomal targeting of misfolded polypeptides. We also find that either treatment of cells with poly I:C or infection of the cells by influenza A viruses triggers the phosphorylation of TRIM28 at S473 in a way that depends on double-stranded RNA-activated protein kinase. The phosphorylation promotes association of TRIM28 with CTIF, inhibits aggresome formation, and consequently suppresses viral proliferation. Collectively, our data provide compelling evidence that TRIM28 is a negative regulator of aggresome formation.Abbreviations: BAG3: BCL2-associated athanogene 3; CTIF: CBC-dependent translation initiation factor; CED: CTIF-EEF1A1-DCTN1; DCTN1: dynactin subunit 1; EEF1A1: eukaryotic translation elongation factor 1 alpha 1; EIF2AK2: eukaryotic translation initiation factor 2 alpha kinase 2; HDAC6: histone deacetylase 6; IAV: influenza A virus; IP: immunoprecipitation; PLA: proximity ligation assay; polypeptidyl-puro: polypeptidyl-puromycin; qRT-PCR: quantitative reverse-transcription PCR; siRNA: small interfering RNA.

Keywords: Aggrephagy; CTIF; DCTN1; EEF1A1; EIF2AK2; influenza A virus.

Figure 1.

TRIM28 competes with EEF1A1 and DCTN1 for the binding to CTIF. (A) Silver staining of the GST affinity isolation of MYC-GST, MYC-CTIF(1-53)-GST, and MYC-CTIF(12-53)-GST. HEK293T cells were transiently transfected with one of the plasmids. Two days later, the cells were treated with MG132 for 12 h before cell harvest. The bands analyzed via LC-MS/MS are indicated by numbers. The proteins identified in each band are listed in Table S1. The arrowhead marks the position of an immunoglobulin heavy chain. (B) In vitro GST affinity-isolation assay using recombinant proteins. The extracts of E. coli expressing GST or a GST-fused protein were mixed with the extracts of E. coli expressing 6× His-TRIM28. The mixtures were subjected to GST affinity-isolation experiments. Representative images obtained from two biological replicates (n = 2) are presented. (C) Immunoprecipitation (IP) of endogenous TRIM28. HEK293T cells were either treated or not treated with MG132. After cell lysis, the cell extracts were digested with RNase A and subjected to IP with the anti-TRIM28 antibody; n = 3. (D) Proximity ligation assay (PLA) involving either FLAG or FLAG-TRIM28 and either MYC or MYC-CTIF. The PLA was performed on HeLa cells transiently expressing the indicated proteins. Nuclei were stained with DAPI (blue). Representative images obtained from three biological replicates (n = 3) are presented. Scale bar: 10 μm. (E) IP of FLAG-CTIF using the extracts of cells depleted of endogenous TRIM28. HEK293T cells transiently expressing FLAG-CTIF and depleted of endogenous TRIM28 were treated with MG132 before cell harvest. The cell extracts digested with RNase A were subjected to IP with FLAG M2 affinity gel. After western blotting with the indicated antibodies, the band intensities were quantitated using the ImageJ software. The intensities of co-IPed proteins were normalized to those of immunoprecipitated FLAG-CTIF. The normalized levels from the undepleted cells were arbitrarily set to 1.0. The average values of normalized intensities from three biological replicates are presented at the bottom of each image. Two-tailed, equal-sample variance Student’s t test was carried out to calculate the P values. These values of each blot are provided in Table S2; n = 3. (F) IP of FLAG-CTIF in the extracts of cells depleted of either EEF1A1 or DCTN1. As performed in panel (E), except that the cells were depleted of either endogenous EEF1A1 or endogenous DCTN1; n = 3

Figure 2.

Downregulation of TRIM28 enhances the formation of aggresomes. (A–C) The effect of TRIM28 downregulation on the formation of aggresomes containing MYC-CTIF. HeLa cells either undepleted or depleted of TRIM28 were transiently transfected with a plasmid expressing MYC-CTIF. The cells were treated with either DMSO or MG132 before fixation. (A) Immunostaining of MYC-CTIF (red). Nuclei were stained with DAPI (blue). Scale bar: 10 μm; n = 3. (B) The diameter of an aggresome containing MYC-CTIF. Immunostained images in panel A were quantitated. More than 30 cells were analyzed from each of the three biological replicates. Statistical analysis was performed by one-way ANOVA with post hoc Tukey’s honestly significant difference test; ***, P < 0.00001. (C) The number of aggresomes containing MYC-CTIF per cell in panel A. Two-tailed, equal-sample variance Student’s t test was conducted to calculate the P values; ^#, not significant. (D–F) The effect of TRIM28 or CTIF downregulation on the formation of aggresomes containing polypeptidyl-puro. As performed in panels (A–C), except that HeLa cells depleted of TRIM28, CTIF, or both were treated with MG132 for 12 h and puromycin for 1 h before fixation. (D) Western blotting confirming specific downregulation of endogenous proteins. (E) Immunostaining of polypeptidyl-puro (green). Scale bar: 10 μm; n = 3. (F) Relative percentages of cells containing either an aggresome or dispersed aggregates of polypeptidyl-puro presented in panel (E). Two-tailed, equal-sample variance Student’s t test was carried out to calculate the P values; ^#, not significant; **, P < 0.01

Figure 3.

TRIM28 inhibits a step in aggresome formation. (A) The frequency of GFP-CTIF particles that reach the aggresome. The number of GFP-CTIF particles (aggregates) per cell that reached the aggresome during a span of 30 s was determined. In total, 56 undepleted and 56 TRIM28-depleted cells from three independent experiments were examined. Statistical analysis was performed by one-way ANOVA with post hoc Tukey’s honestly significant difference test. (B–E) The influence of autophagy inhibition on TRIM28-mediated inhibition of aggresome formation. HeLa cells stably expressing GFP-LC3B were depleted of TRIM28, ATG5, or both. One day later, the cells were transiently transfected with a plasmid expressing MYC-CTIF. The cells were treated with MG132 before fixation; n = 3. (B) Western blotting showing specific downregulation of the tested proteins. (C) The number of GFP-LC3B puncta per cell. Immunostained images in panel (D) were quantitated. Two-tailed, equal-sample variance Student’s t test was carried out to calculate the P values. *, P < 0.05; **, P < 0.01. (D) Immunostaining of MYC-CTIF (red) and GFP-LC3B (green). Nuclei were stained with DAPI (blue). Scale bar: 10 μm. (E) The diameter of an aggresome containing MYC-CTIF. Immunostained images in panel (D) were quantitated. Statistical analysis was performed by one-way ANOVA with post hoc Tukey’s honestly significant difference test; ***, P < 0.00001

Figure 4.

The S473 residue is responsible for TRIM28-mediated inhibition of aggresome formation. (A) A schematic diagram of the TRIM28 protein. The mutation sites are indicated by solid circles. Deletion variants of TRIM28 with the N-terminal FLAG-tag are also indicated. R, RING domain; BB, two B-box-type zinc finger domains; CC, coiled-coil motif; H, HP1-binding domain; P, PHD domain; and B, bromodomain. (B) Immunostaining of GFP-CFTR-ΔF508 and FLAG-TRIM28, either WT or its variant. HeLa cells stably expressing GFP-CFTR-ΔF508 were transiently transfected with a plasmid expressing FLAG, FLAG-TRIM28-WT, or its variant. Two days later, the cells were treated with MG132 for 12 h before fixation. The cells were stained with the anti-GFP antibody (green) and an anti-FLAG antibody (red). Nuclei were stained with DAPI (blue). Scale bar: 10 μm; n = 3. (C) Relative distribution of GFP-CFTR-ΔF508. Relative percentages were determined by counting the cells containing either the aggresome or the dispersed aggregates of GFP-CFTR-ΔF508 in the immunostaining images. To accurately assess the effect of exogenously expressed FLAG-TRIM28, only the cells expressing both GFP-CFTR-ΔF508 and exogenous FLAG-TRIM28 were counted. Two-tailed, equal-sample variance Student’s t test was carried out to calculate the P values. **, P < 0.01

Figure 5.

EIF2AK2-mediated S473 phosphorylation of TRIM28 inhibits the formation of aggresomes containing MYC-CTIF. (A) IP of FLAG-TRIM28-WT or FLAG-TRIM28^S473A. HEK293T cells transiently expressing MYC-CTIF and FLAG, FLAG-TRIM28-WT, or FLAG-TRIM28^S473A were either transfected or not transfected with poly I:C. The cell extracts were digested with RNase A and subjected to IP with FLAG M2 affinity gel; n = 3. (B and C) The impact of EIF2AK2 on aggresome formation. Either WT or eif2ak2 KO MEFs transiently expressing MYC-CTIF and FLAG, FLAG-TRIM28-WT, or its variant were transfected with poly I:C. The cells were treated with MG132 for 12 h before cell fixation; n = 3. (B) The diameter of aggresomes containing MYC-CTIF. Statistical analysis was performed by one-way ANOVA with post hoc Tukey’s honestly significant difference test; ***, P < 0.00001. (C) Immunostaining of MYC-CTIF (red) and FLAG-TRIM28 or its variant (green). Nuclei were stained with DAPI (blue). Scale bar: 10 μm

Figure 6.

Poly I:C–induced TRIM28 phosphorylation hinders the formation of aggresomes containing GFP-CFTR-ΔF508. HeLa cells stably expressing GFP-CFTR-ΔF508 were treated with a control siRNA or TRIM28 siRNA. One day later, the cells were transfected with a plasmid expressing siRNA-resistant FLAG-TRIM28[R] (WT, S473A, or S473E) or a plasmid expressing only FLAG, which served as a negative control. One day later, the cells were either mock-transfected or transfected with poly I:C. The cells were treated with MG132 for 12 h before cell fixation; n = 3. (A) Western blotting validating specific downregulation of endogenous TRIM28 by means of siRNA and the expression of FLAG-TRIM28[R]-WT, FLAG-TRIM28[R]^S473A, or FLAG-TRIM28[R]^S473E at a level comparable to that of endogenous TRIM28. (B) Immunostaining of GFP-CFTR-ΔF508 (green) and exogenously expressed FLAG-TRIM28[R]-WT, FLAG-TRIM28[R]^S473A, or FLAG-TRIM28[R]^S473E (red). Nuclei were stained with DAPI (blue). Scale bar: 10 μm. (C) Relative percentages of cells containing either aggresomal or dispersed GFP-CFTR-ΔF508. To accurately assess the effect of exogenously expressed FLAG-TRIM28, only the cells expressing both GFP-CFTR-ΔF508 and exogenous FLAG-TRIM28 were counted. Two-tailed, equal-sample variance Student’s t test was carried out to calculate the P values. *, P < 0.05; **, P < 0.01

Figure 7.

Proliferation of the influenza A virus (IAV) lacking NS1 depends on TRIM28. (A and B) The effect of infection with IAV strain PR8-WT on aggresomes containing MYC-CTIF. MEFs either undepleted or depleted of TRIM28 were either mock-infected or infected with PR8-WT for 12 h at a multiplicity of infection of 2; n = 3. (A) Immunostaining of MYC-CTIF (red). Scale bar: 10 μm. (B) The diameter of an aggresome containing MYC-CTIF. Statistical analysis was performed by one-way ANOVA with post hoc Tukey’s honestly significant difference test; ***, P < 0.00001. (C and D) The effect of PR8-ΔNS1 infection on aggresomes containing MYC-CTIF. As performed in panels A and B, except that the cells were infected with strain PR8-ΔNS1. (C) Immunostaining of MYC-CTIF (red). Scale bar: 10 μm. (D) The diameter of an aggresome containing MYC-CTIF. ***, P < 0.00001. (E) The time course of relative expression levels of IAV HA mRNA in the cells infected with strain PR8-WT. MEFs either undepleted or depleted of TRIM28 were infected with PR8-WT at a multiplicity of infection of 2 and harvested at the indicated time points after infection. The HA mRNA level was normalized to that of Gapdh mRNA at each time point. Two-tailed, equal-sample variance Student’s t test was carried out to calculate the P values; n = 3; ^#, not significant. (F) The time course of relative expression levels of IAV HA mRNA in the cells infected with PR8-ΔNS1; n = 3; *, P < 0.05; **, P < 0.01

Figure 8.

The proposed model for the regulation of aggresome formation mediated by TRIM28. Depending on relative amounts of TRIM28 and the extent of EIF2AK2-induced S473 phosphorylation of TRIM28, CTIF may form one of two alternative complexes: CTIF-TRIM28 (a complex inactive in terms of aggresome formation) and the CED complex (a complex active in terms of aggresome formation). See the Discussion section for details