Derlin-1 is a rhomboid pseudoprotease required for the dislocation of mutant α-1 antitrypsin from the endoplasmic reticulum

Abstract

The degradation of misfolded secretory proteins is ultimately mediated by the ubiquitin-proteasome system in the cytoplasm, therefore endoplasmic reticulum-associated degradation (ERAD) substrates must be dislocated across the ER membrane through a process driven by the AAA ATPase p97/VCP. Derlins recruit p97/VCP and have been proposed to be part of the dislocation machinery. Here we report that Derlins are inactive members of the rhomboid family of intramembrane proteases and bind p97/VCP through C-terminal SHP boxes. Human Derlin-1 harboring mutations within the rhomboid domain stabilized mutant α-1 antitrypsin (NHK) at the cytosolic face of the ER membrane without disrupting the p97/VCP interaction. We propose that substrate interaction and p97/VCP recruitment are separate functions that are essential for dislocation and can be assigned respectively to the rhomboid domain and the C terminus of Derlin-1. These data suggest that intramembrane proteolysis and protein dislocation share unexpected mechanistic features.

Figure 1

Derlins belong to the rhomboid family. (a) ClustalW alignment (UniProtKB accession number indicated) of the transmembrane regions of Derlin-1 from Homo sapiens (hs, Q9BUN8), Danio rerio (dr, Q7ZVT9), Caenorhabditis elegans (ce, Q93561), Derlin-2 (hs, Q9GZP9) and Derlin-3 (hs, Q96Q80) from H. sapiens and GlpG from Shewanella trabarsenatis (st, Q0HDA1), GlpG, Yersinia frederiksenii (yf, C4SQG5), and Escherichia coli (ec, P09391). Pink, the active site dyad in GlpG, absent in Derlins. WR and GxxxG motifs are indicated by *. (b–d) Homology model of the Derlin-1 rhomboid domain. The positions of the WR motif (b), GxxxG motif (c) and positively charged residues on the cytoplasmic side of each transmembrane span in the homology model (d) are indicated. N- and C termini of Derlin-1 are located in the cytoplasm. (e) Representative immunofluorescence images of cells expressing N- and C-terminally tagged HA-Derlin-1-S in which the plasma membrane was permeabilized without (digitonin) or with (digitonin + Triton X-100) ER permeabilization. Scale bars are 10 μm. (f) Comparison of the four-pass topology model of Derlin-1 as inferred from the reported topology of the Saccharomyces cerevisiae ortholog Der1 (ref. 8) with the six-pass topology model of Derlin-1 as predicted by homology with GlpG. (g) Derlin-1 spans the ER membrane six times. Immunoblot of lysates treated with or without Endo H from cells expressing Derlin-1-S containing a glycosylation acceptor sequence inserted after the indicated residue.

Figure 2

p97/VCP binding to Derlin-1 through an SHP box is required for efficient dislocation of NHK. (a) Domain organization of Derlin-1-S and Derlin-2-S constructs used in this study. (b) Alignment of SHP boxes from the p97/VCP/Cdc48 binding proteins H. sapiens Derlin-1 and Derlin-2, S. cerevisiae Dfm1, H. Sapiens Ufd1 and p47 and S. cerevisiae Shp1. (c) The SHP box is required for Derlin-1 and Derlin-2 interactions with p97/VCP. Immunoblot of digitonin lysates and S-protein agarose–precipitated material from untransfected (UT) cells or cells expressing wild-type (WT) or mutant Derlin-1-S or Derlin-2-S lacking the SHP box (1–240 and 1–231, respectively). (d–g) Overexpression of Derlin-1 lacking the SHP box results in stabilization of deglycosylated forms of NHK associated with Derlin-1. (d) Immunoblots of digitonin lysates or S-protein agarose–precipitated material from cells coexpressing NHK-HA with the indicated Derlin-1-S variants or a GFP-S control. Deglycosylated forms of NHK-HA are indicated by arrowheads. (e) Immunoblot of four biological replicates similar to those in d, except cells were lysed in SDS lysis buffer. (f) Quantification of the levels of deglycosylated NHK-HA in e as a percentage of total NHK-HA levels. Error bars represent s.e.m. (g) Immunoblot of material from d treated with or without Endo H. (h) Left panel, immunoblots of Triton X-100 lysate from cells coexpressing NHK-HA, wild-type or mutant Derlin-1 or a control protein (ΔCD4) and an shRNA targeting GFP or PNG1. Quantification of data was carried out as in f. n.d., not detectable. To enable detection, Derlin-1-S and mPNG1-S were enriched by S-protein agarose precipitation. Right panel, efficacy of shRNA-mediated PNG1 knockdown. CHO, carbohydrate; AP, affinity precipitation.

Figure 3

The GxxxG motif in the Derlin-1 rhomboid domain is required for the dislocation of NHK. (a) Derlin-1_G176V acts a dominant negative mutant by inhibiting the dislocation of NHK. Immunoblots of digitonin lysates and S-protein agarose–precipitated material from cells coexpressing NHK-HA and the indicated Derlin-1-S variant or a GFP control. Deglycosylated forms of NHK are indicated by arrowheads. (b) Quantification of deglycosylated NHK-HA levels in a as a percentage of total NHK-HA levels. (c) Deglycosylated forms of NHK are exposed to the cytoplasm. Immunoblots of membranes obtained from homogenized cells coexpressing NHK-HA and Derlin-1-S_G176V treated with the indicated concentrations of proteinase K and Triton X-100. The cytoplasmic C-terminal S-tag of Derlin-1-S, or luminal (SEL1L, GRP94 and BiP) markers were used as controls. (d) Quantification of the data from c. (e) Disruption of NHK dislocation is specific to mutation of the Derlin-1 GxxxG motif. Immunoblots of SDS lysates from cells coexpressing NHK-HA with wild-type Derlin-1-S or the indicated GxxxG mutants (G176V or G180V) or mutants outside of the GxxxG motif (G29V or G147V), or a GFP control, treated with or without Endo H. (f) Mutation of the GxxxG motif does not disrupt homo-oligomerization of Derlin-1 or interaction with p97/VCP. Immunoblots of digitonin lysates and S-protein agarose–precipitated material from cells expressing wild-type or mutant Derlin-1-S or a GFP control.

Figure 4

The Derlin-1 WR motif is essential for the dislocation of NHK. (a) Expression of Derlin-1 WR motif mutants results in a reduction in NHK levels in Triton X-100 lysates. Immunoblots of Triton X-100 lysates from cells coexpressing NHK-HA and Derlin-1-S_WT, the indicated Derlin-1-S WR motif mutants (Q51A, W53A, R54A or T57A), Derlin-1-S mutated at a nonconserved tryptophan (W106A), or a GFP control. (b) Overexpression of Derlin-1_R54A results in aggregation and accumulation of deglycosylated NHK. Immunoblots of Triton X-100 soluble and insoluble material. (c) Mutation of the Derlin-1 WR motif does not disrupt interaction with p97/VCP. Arrowheads indicate deglycosylated NHK. (d–f) Expression of Derlin-1 containing mutations in the WR or GxxxG rhomboid motifs stabilizes NHK. (d) Immunoblot of SDS lysates from cells coexpressing NHK-HA and wild-type or mutant Derlin-1-S or a GFP control that were treated with emetine for the indicated durations. (e) Quantification of four independent experiments is shown. Asterisks denotes P ≤ 0.05. Error bars indicate s.e.m. (f) Derlin-1 GxxxG and WR motif mutants are stable and expressed at similar levels.