gp78 functions downstream of Hrd1 to promote degradation of misfolded proteins of the endoplasmic reticulum

Abstract

Eukaryotic cells eliminate misfolded proteins from the endoplasmic reticulum (ER) via a conserved process termed ER-associated degradation (ERAD). Central regulators of the ERAD system are membrane-bound ubiquitin ligases, which are thought to channel misfolded proteins through the ER membrane during retrotranslocation. Hrd1 and gp78 are mammalian ubiquitin ligases homologous to Hrd1p, an ubiquitin ligase essential for ERAD in Saccharomyces cerevisiae. However, the functional relevance of these proteins to Hrd1p is unclear. In this paper, we characterize the gp78-containing ubiquitin ligase complex and define its functional interplay with Hrd1 using biochemical and recently developed CRISPR-based genetic tools. Our data show that transient inactivation of the gp78 complex by short hairpin RNA-mediated gene silencing causes significant stabilization of both luminal and membrane ERAD substrates, but unlike Hrd1, which plays an essential role in retrotranslocation and ubiquitination of these ERAD substrates, knockdown of gp78 does not affect either of these processes. Instead, gp78 appears to act downstream of Hrd1 to promote ERAD via cooperation with the BAG6 chaperone complex. We conclude that the Hrd1 complex forms an essential retrotranslocation module that is evolutionarily conserved, but the mammalian ERAD system uses additional ubiquitin ligases to assist Hrd1 during retrotranslocation.

FIGURE 1:

gp78 forms a transient interaction with Hrd1. (A) Schematic diagram of the domain structure of gp78, UbxD8, and UBAC2. (B) Hrd1 and gp78 bind to different interactors with different affinities. Control transfected HEK293T cells or cells overexpressing FLAG-Hrd1 or FLAG-gp78 were lysed in the CHAPS lysis buffer. Proteins immunoprecipitated with FLAG beads were analyzed by immunoblotting. (C and D) Endogenous interaction of Hrd1 with gp78. (C) 293T cells were lysed in a CHAPS-containing lysis buffer. The cell extracts were subject to immunoprecipitation by either control antibody or two affinity-purified Hrd1 antibodies. Where indicated, a fraction of the cell lysate was analyzed directly by immunoblotting. (D) 293T-derived cell extracts were subject to immunoprecipitation by the indicated antibodies.

FIGURE 2:

Characterization of the gp78-UbxD8-UBAC2 complex. (A) Stoichiometry of the gp78-UbxD8-UBAC2 complex. The indicated proteins were overexpressed in HEK293T cells. Cells were lysed in NP40 lysis buffer, and proteins immunoprecipitated with anti-Myc antibodies were analyzed by quantitative immunoblotting. (B) Schematic diagram of the UBXD8 variants used in the pull-down study. All constructs contain a FLAG tag at the amino terminus. (C) Subcellular localization of the UbxD8 variants. COS7 cells expressing the indicated UbxD8 variants were stained by FLAG antibodies. Scale bar: 5 μm. (D) Cells overexpressing FLAG-tagged UBXD8 constructs were lysed in a NP40-containing lysis buffer. Proteins coprecipitated by FLAG beads were analyzed by immunoblotting. (E) The transmembrane domain of UbxD8 is sufficient to bind UBAC2.

FIGURE 3:

Both gp78 and Hrd1 are required for ERAD of luminal and membrane substrates. (A) Diagram illustrating the model ERAD substrates used in this study. (B) Both gp78 and UbxD8 are required for degradation of MHC 1-147. Cells cotransfected with indicated shRNA constructs and a plasmid expressing MHC 1-147 were treated with cycloheximide for the indicated time points. Cells were directly lysed in the Laemmli buffer, and the whole-cell extracts were analyzed by immunoblotting. Graph on the right represents the quantification of the experiment. (C) Guiding sequence used to create hrd1 knockout CRISPR cell. The PAM sequence and the target sequence are colored in red and blue, respectively. Red arrow indicates the predicted Cas9 D10A cutting site. (D) Hrd1 is required for the degradation of MHC 1-147. Cycloheximide chase was performed in control CRISPR and hrd1 knockout CRISPR cells. Where indicated, plasmids expressing WT Hrd1 or a catalytically inactive Hrd1 (C1C3) mutant were cotransfected with MHC 1-147. Whole-cell extracts were analyzed by immunoblotting. (E) Verification of the gp78 CRISPR cells by immunoblotting. (F and G) gp78-deficient CRISPR cells do not have ERAD defects. The steady-state level of MHC 1-147 in either the parental HEK293T cells or the indicated CRISPR clones was analyzed by immunoblotting. Where indicated, cells were treated with the proteasome inhibitor MG132 (10 μm, 15 h). (G) The indicated CRISPR cells transfected with a plasmid expressing MHC 1-147 were treated with cycloheximide for the indicated time points. Cells were directly lysed in the Laemmli buffer, and the whole-cell extracts were analyzed by immunoblotting.

FIGURE 4:

Hrd1 but not gp78 is required for ubiquitination of ERAD substrates. (A and B) The model ERAD substrates MHC 1-147-FLAG (A) and TCRα-YFP-FLAG (B) were cotransfected with a construct expressing HA-tagged ubiquitin in either control or hrd1 CRIPSR cell. The cells were treated with either DMSO as a control or with the proteasome inhibitor MG132 (10 μM, 15 h). Substrates immunoprecipitated from the cell extracts under denaturing conditions were analyzed by immunoblotting. The anti-HA blot reveals ubiquitinated substrate and the anti-FLAG blot shows the nonubiquitinated glycosylated and deglycosylated substrates. (C) As in A, except that cells treated with the indicated shRNA constructs were used. (D) As in B, except that cells treated with the indicated shRNA constructs were used. (E) Endoglycosidase H treatment of the indicated cell extract reveals the glycosylation pattern of MHC 1-147 in the indicated CRISPR cells that have been exposed to a proteasome inhibitor.

FIGURE 5:

Hrd1 but not gp78 is involved in substrate retrotranslocation. (A) Schematic illustration of the split-GFP–based retrotranslocation assay. (B) Hrd1 is required for the retrotranslocation of MHC 1-147. Plasmids expressing s11-tagged MHC 1-147 and GFP S1-10 were cotransfected into control or hrd1 CRISPR cells. Cells were treated with DMSO (control) or 10 μM proteasome inhibitor MG132 for 15 h. Top, NP40-soluble lysates were analyzed by immunoblotting. p97 was used as a loading control. Bottom, fluorescence intensity in the same cells was measured. The graph is from three independent experiments. (C and D) MHC 1-147 still undergoes retrotranslocation in BAG6 and gp78 knockdown cells. Similar to B, shRNA construct or CRISPR cell line was used to knockdown or knockout the indicated gene without MG132 treatment. Both NP40-soluble and NP40-insoluble fractions were analyzed. Bar graph represents the average of three independent experiments. (E and F) gp78 knockdown with MG132 treatment. Bar graph represents the average of three independent experiments.

FIGURE 6:

gp78 functions downstream of Hrd1 to promote substrate solubility in ERAD. (A) Depletion of Hrd1 under gp78 knockdown condition reduces substrate ubiquitination. The indicated CRISPR cells were cotransfected with MHC 1-147-FLAG-S11–expressing plasmid together with the indicated shRNA constructs and then treated with MG132 (10 μM, 15 h). MHC 1-147-FLAG-S11 was immunoprecipitated under denaturing conditions and analyzed by immunoblotting. Both ubiquitinated (HA blot) and deglycosylated forms of MHC class I heavy-chain 1-147 (HC) were reduced upon further knockout of Hrd1 in gp78 knockdown cells (lanes 2 and 4). (B and C) Fluorescence intensity in BAG6 or gp78 knockdown cells was reduced upon further depletion of Hrd1 (p < 0.01, n = 3). F/P, fluorescence intensity normalized against protein levels. Cell extracts prepared from a fraction of the cells were analyzed by immunoblotting to verify the protein level (bottom panels). (D) gp78 acts at a postubiquitination step in degradation of MHC 1-147. Cells expressing MHC 1-147 together with HA-tagged ubiquitin and the indicated shRNA constructs were lysed in a NP40-containing lysis buffer. After both the NP40-soluble and NP40-insoluble fractions were obtained, MHC 1-147 was immunoprecipitated from these fractions under denaturing conditions and analyzed by immunoblotting. (E) Both gp78 and BAG6 knockdown causes MHC 1-147 to aggregate in cells. All images were acquired with the same laser setting using a Zeiss LSM 780 confocal microscope. Scale bar: 5 μm.

FIGURE 7:

The functional relationship between gp78 and Hrd1. Hrd1 is the essential retrotranslocation regulator conserved in yeast and mammalian cells, whereas gp78 serves an assisting role downstream of Hrd1 and possibly another ubiquitin ligase in mammalian cells. gp78 may promote ERAD by maintaining the functionality of the BAG6 complex.