ATP13A1 prevents ERAD of folding-competent mislocalized and misoriented proteins

Abstract

The biosynthesis of thousands of proteins requires targeting a signal sequence or transmembrane segment (TM) to the endoplasmic reticulum (ER). These hydrophobic ɑ helices must localize to the appropriate cellular membrane and integrate in the correct topology to maintain a high-fidelity proteome. Here, we show that the P5A-ATPase ATP13A1 prevents the accumulation of mislocalized and misoriented proteins, which are eliminated by different ER-associated degradation (ERAD) pathways in mammalian cells. Without ATP13A1, mitochondrial tail-anchored proteins mislocalize to the ER through the ER membrane protein complex and are cleaved by signal peptide peptidase for ERAD. ATP13A1 also facilitates the topogenesis of a subset of proteins with an N-terminal TM or signal sequence that should insert into the ER membrane with a cytosolic N terminus. Without ATP13A1, such proteins accumulate in the wrong orientation and are targeted for ERAD by distinct ubiquitin ligases. Thus, ATP13A1 prevents ERAD of diverse proteins capable of proper folding.

Keywords: ER-associated degradation; protein localization; protein topology; quality control; signal sequence; transmembrane proteins.

Figure 1.. Mitochondrial TA proteins mislocalize to the ER via the EMC.

(A) EMC depletion rescues the localization of FLAG-tagged OMP25 (F-OMP25), a mitochondrial TA protein, in ATP13A1 knockout (KO) cells. Immunofluorescence of F-OMP25 (green) and the ER marker TRAPα (magenta) in wildtype (WT) and ATP13A1 KO Flp-In HeLa T-REx cells treated with control siRNAs or siRNAs targeting EMC2, EMC6 or GET1. Scale bar, 10 μm. (B) Pearson’s correlation coefficients (PCC; mean ± sd and individual points for indicated sample size) measuring colocalization of F-OMP25 and TRAPα. ****, p<0.0001; ns, not significant. (C) Knocking out the EMC reduces F-OMP25 insertion into ER-derived rough microsomes (RMs). SDS-PAGE and autoradiography of total (tot.), membrane-inserted (pel.), and soluble (sup.) radiolabeled F-OMP25 from insertion reactions with RMs obtained from WT, ATP13A1 KO, or ATP13A1 and EMC6 double knockout (DKO) Flp-In 293 T-REx cells (left). Ratios of pelleted to soluble (pel.:sup.) F-OMP25 (top right) or a matched reporter containing the TM of the ER TA protein SQS (F-SQS; bottom right) were normalized to values obtained with WT RMs (mean + sem) for 3 replicates. **, p<0.01; *, p<0.05. See also Figure S1.

Figure 2.. SPP cleaves mislocalized mitochondrial TA proteins during ERAD.

(A) Proteasome inhibition stabilizes an SPP-dependent fragment of mislocalized F-OMP25. SDS-PAGE and immunoblotting for the indicated factors in wildtype (WT) and ATP13A1 knockout (KO) cells treated without or with proteasome inhibitors (0.5 μM bortezomib and 0.5 μM epoxomicin; bort./epox.) and/or an inhibitor [5 μM (Z-LL)₂-ketone] of the signal peptide peptidase (SPP). FL, full-length F-OMP25; frag., F-OMP25 fragment. (B) SPP-cleaved F-OMP25 is cytosolic. Immunoblotting of lysates (tot.) of ATP13A1 KO cells expressing F-OMP25 treated with bort./epox. and separated into cytosolic (cyto.) and membrane-bound (memb.) fractions. (C) SPP inhibition stabilizes mislocalized F-OMP25. Fluorescent flow cytometry of F-OMP25 levels normalized to mitotracker staining in WT or ATP13A1 KO cells treated without or with 5 μM (Z-LL)₂-ketone. (D) SPP depletion does not rescue mitochondrial TA protein localization. Pearson’s correlation coefficients (PCC; mean ± sd and individual points for indicated sample size) measuring the colocalization of F-OMP25 and the ER marker TRAPα in WT or ATP13A1 KO cells treated with control (cont.) or SPP siRNAs as in Figure S2G. ****, p<0.0001; ***, p<0.0003. Cont. siRNA samples are a subset of Figure 1B. (E) CEND1 is an ATP13A1-dependent mitochondrial TA protein. Immunofluorescence (left) and PCC (right; mean ± sd and individual points for indicated sample size) showing colocalization of a FLAG-tagged reporter containing the CEND1 TM (F-CEND1, green) and the ER marker TRAPα (magenta) in WT or ATP13A1 KO Flp-In HeLa T-REx cells. Scale bar, 10 μm; ****, p<0.0001. (F) Mislocalized CEND1 is cleaved by SPP. SDS-PAGE and immunoblotting of soluble fractions from semi-permeabilized WT or ATP13A1 KO cells expressing F-CEND1 treated without or with proteasome inhibitors (bort./epox.) and/or (Z-LL)₂-ketone. See also Figures S2, S3, and Supplemental Table 1.

Figure 3.. A reporter system to assay ER membrane protein topology and stability.

(A) Scheme of split fluorescent reporter system to assay protein topology. Flp-In 293 T-REx cells were engineered to express the first 10 β-strands mNeonGreen3K [GFP(1–10)] in the cytosol and the first 10 β-strands of mCherry3V [RFP(1–10)] localized to the ER by the bovine preprolactin signal sequence and a C-terminal KDEL ER retention signal. Reporters incorporated into the Flp-In locus express proteins without or with an N-terminal ER targeting signal sequence (SS) or type II TM followed by the distinct 11^th β-strands of mNeonGreen3K (GFPb11) and mCherry3V (RFPb11) and separated by a P2A ribosome skipping sequence from TagBFP (BFP). (B) Protein products of the reporter system in (A). GFP_cyto and RFP_ER fluorescence depends on reporter protein topology. (C) RFP_ER:BFP (left) and GFP_cyto:BFP (right) ratios of topology reporters of the type II protein ASGR1 (dark orange) or the cytosolic protein DHFR (teal). (D) RFP_ER:BFP (left) or GFP_cyto:BFP (right) ratios of topology reporters of wildtype CHST10 (dark gray) or CHST10 in which His2 and His3 are mutated to glutamines (QQ-CHST10; teal) or arginines (RR-CHST10; dark orange). The N-terminal and TM sequence of CHST10 is shown with His2 and His3 in blue. See also Figure S4.

Figure 4.. ATP13A1 facilitates type II membrane protein topogenesis.

(A) ATP13A1 is required for CHST10 stability. RFP_ER:BFP ratios of wildtype CHST10 (left), RR-CHST10 (middle), or ASGR1 (right) type II protein topology reporters in wildtype cells treated with control siRNAs (control KD; gray; same samples as Figure 3D) or siRNAs against ATP13A1 (ATP13A1 KD; light blue). (B) Reconstitution of ATP13A1-dependent CHST10 topogenesis. SDS-PAGE and autoradiography (left) of radiolabeled CHST10, RR-CHST10, or ASGR1 synthesized in vitro without or with ER-derived rough microsomes (RMs) isolated from wildtype (WT) or ATP13A1 knockout (KO) cells before (tot.) or after carbonate extraction (carb. ext.) to isolate membrane-embedded populations. All proteins contain N-glycosylation sites that are modified in the correct type II topology (right) and a C-terminal FLAG tag used for denaturing immunoprecipitations. FL, full-length substrate; +glyc, glycosylated substrate. (C) Glyc.:FL ratios (mean ± sem) of CHST10 (left) or RR-CHST10 (right) in insertion assays as in (B) containing RMs derived from WT, ATP13A1 KO, or ATP13A1 KO cells re-expressing WT or catalytically inactive (D533A; mut.) ATP13A1, normalized to reactions containing WT RMs for 3 replicates. See also Figure S5.

Figure 5.. ATP13A1 facilitates signal sequence topogenesis.

(A) ATP13A1 is required for the stability of WNT1 homologs. RFP_ER:BFP ratios of topology reporters of the signal sequence-containing proteins WNT1, its C. elegans homolog EGL-20, an EGL-20 variant with a more hydrophobic signal sequence (EGL-20H), or preprolactin (PRL) in wildtype (WT; gray) or ATP13A1 knockout (KO; light blue) cells. (B) Reconstitution of ATP13A1-dependent signal sequence topogenesis. SDS-PAGE and autoradiography (top) of radiolabeled EGL-20 or EGL-20H synthesized in vitro without or with ER-derived rough microsomes (RMs) isolated from WT or ATP13A1 KO cells before (tot.) or after pelleting (tot.) to enrich RM-associated proteins. All proteins contain a C-terminal FLAG tag used for denaturing immunoprecipitations. Precursor and signal-cleaved substrates were quantified from the pelleted samples (bottom) and normalized to the average WT signal for each substrate. Shown are mean +/− sem of precursor (orange with negative orange error bar), signal-cleaved (gray with positive orange error bar), and total (black error bar) values for 3 replicates. (C) Fluorescence microscopy showing RFP_ER, GFP_cyto, and BFP signal of the EGL-20H topology reporter in untreated WT cells or ATP13A1 KO cells treated with 1 μM MLN7243. Scale bar, 10 μm. See also Figure S6.

Figure 6.. ATP13A1 protects misoriented proteins from ERAD.

(A) Inhibiting the ubiquitin-proteasome system stabilizes misoriented proteins. GFP_cyto:BFP (top) and RFP_ER:BFP (bottom) ratios of the CHST10 type II (left) or WNT1 signal sequence (right) topology reporter in wildtype (WT; gray) or ATP13A1 knockout (KO; light blue) cells treated without or with inhibitors (inhib.) of SPP (purple), proteasomal activity (bort./epox.; light green), the E1 ubiquitin activating enzyme (1 μM MLN7243; dark green), or the p97 AAA-ATPase (1 μM CB-5083; dark blue). (B) Distinct ubiquitin ligases mediate misoriented CHST10 ERAD. GFP_cyto:BFP (left) and RFP_ER:BFP (right) ratios of the CHST10 topology reporter in WT (gray) ATP13A1 KO cells treated with siRNAs to knock down (KD) the indicated factors. Shown are median values and interquartile range of at least two biological replicates of n≥5,000 each. (C) Specific but redundant ubiquitin ligases mediate misoriented protein ERAD. GFP_cyto:BFP ratios of the CHST10 (top) or WNT1 (bottom) topology reporter in ATP13A1 KO cells treated without (light blue) or with E1 inhibitor (dark green), or with siRNAs to knock down either AMFR and RNF185 (purple) or SYVN1 and MARCH6 (pink). (D) Timecourse of misoriented CHST10 ERAD. Translation shut-off reactions of ATP13A1 KO cells expressing the CHST10 topology reporter treated without (light blue) or with siRNAs against AMFR and RNF185 (purple) or E1 inhibitor (dark green). Samples taken at the indicated timepoints were analyzed by immunoblotting for unmodified CHST10, normalized to time = 0 min, and the mean +/− sem for 3 replicates plotted. See also Figure S7.

Figure 7.. ATP13A1 prevents wasteful ERAD of mislocalized and misoriented proteins.

(A) Dislocation by ATP13A1 provides mislocalized mitochondrial TA proteins additional opportunities to insert into the outer membrane of mitochondria (mito.) and protects against SPP-mediated ERAD. (B) Correct topogenesis of certain signal sequences and type II TMs requires ATP13A1, which prevents ERAD of misoriented substrates mediated by ER-resident ubiquitin ligases. Ribosomes present during cotranslational protein translocation and TM insertion are omitted for clarity.