Regulation of the endoplasmic reticulum stress sensor ATF6α through multiple oxidoreductases in the ER Authors: Shota Wada, Kaiku Uegaki, Kazuhiro Nagata, Ryo Ushioda Published: 2026 ABSTRACT SummaryThe unfolded protein response (UPR) maintains cellular homeostasis during ER stress. In this process, activating transcription factor 6α (ATF6α) is activated through several multi-step processes, including transport to the Golgi apparatus, and plays an important role as a UPR sensor. Disulfide bond-mediated dimerization of ATF6α contributes to efficient transport to the Golgi apparatus. However, this regulatory mechanism remains unclear. Here, we show that ER-resident oxidoreductases, including ERdj5 and PDIR, promote C467 dimerization, whereas ERp18 inhibits this process. In particular, ERdj5 facilitates C467 dimerization via Trx2 and reduces C618 dimerization via Trx3, supporting ATF6α activation. Furthermore, ATF6α dimerization stabilizes its structure and prevents its degradation by ER-associated degradation (ERAD). Our findings reveal that multiple oxidoreductases participate in the regulation of ATF6α activation during ER stress. Introduction In eukaryotes, many secretory and membrane proteins are cotranslationally transported into the endoplasmic reticulum (ER) and are folded by ER-resident molecular chaperones and oxidases following stringent protein quality control (PQC).1 Excessive accumulation of unfolded or misfolded proteins causes ER stress, and severe ER stress ultimately leads to cell death and diseases such as Alzheimer’s disease, diabetes, and heart disease.2,3,4,5 One of the most crucial responses to ER stress is the unfolded protein response (UPR), which alleviates stress by modulating protein synthesis, degradation, and folding capacity.6 The UPR is governed by three major ER transmembrane sensors: PKR-like ER kinase (PERK), inositol-requiring protein 1 (IRE1), and activating transcription factor 6 alpha (ATF6α).7,8,9 These proteins are uniquely activated under ER stress conditions.10 Under normal conditions, ATF6α forms a complex with the ER-resident molecular chaperone BiP/GRP78.11,12 During ER stress, BiP dissociates from ATF6α, after which ATF6α is incorporated into COPII vesicles and transported to the Golgi apparatus.13 In the Golgi, ATF6α is sequentially cleaved by site-1 protease (S1P) and site-2 protease (S2P), releasing the cytosolic N-terminal domain (ATF6α-N).14 ATF6α-N then translocates to the nucleus, where it binds to the ER stress response element (ERSE) in the promoter regions of target genes, inducing the expression of molecular chaperones and oxidases that increase the folding capacity of the ER.15 Recent studies have reported redox-dependent regulation of ATF6α activation.10 ATF6α contains two cysteine residues (Cys467 and Cys618) in the ER lumen, each of which forms a disulfide bond-dependent dimer (C467 dimer and C618 dimer).16 The C467 dimer promotes efficient ER-to-Golgi transport, whereas the function and role of the C618 dimer remain unclear.17 Several protein disulfide isomerases (PDIs), including ERp18, PDIR, and ERdj5, have been identified as regulators of ATF6α activation.17,18 ERp18 suppresses ATF6α transport by reducing Cys467, whereas PDIR and ERdj5 promote ATF6α activation. However, the precise molecular mechanisms by which these PDIs regulate ATF6α activation remain to be fully elucidated. In addition to its role in the stress response, ATF6α is associated with various pathological conditions. In humans, ATF6α mutations have been identified in patients with achromatopsia, a congenital disorder characterized by cone photoreceptor dysfunction.19 One such mutation, Y567N, has been shown to impair ATF6α trafficking to the Golgi apparatus, although the underlying mechanism remains elusive. Additionally, these mutations have been associated with hearing loss syndrome due to defective cochlear hair cell development.20 Given its critical role in ER homeostasis and disease pathogenesis, understanding the molecular mechanisms regulating ATF6α activation and redox regulation is essential. Here, we found that ERdj5 regulates ATF6α activation through two functions: oxidation of Cys467 to promote C467 dimerization and reduction of Cys618 to prevent C618 dimerization. Furthermore, we elucidated that the Y567N mutation promotes C618 dimer formation and prevents the formation of the C467 dimer. These findings provide important insights into therapeutic strategies for the treatment of ATF6α-associated diseases. Results ERdj5 promotes ATF6α C467 dimerization via its Trx2 domain ATF6α contains two cysteine residues, Cys467 and Cys618, within its luminal domain, which enables the formation of C467 and C618 homodimers (Figures S1A and S1B).16 The mobility of the C467 and C618 dimers in SDS-PAGE varies depending on their structural features. Previous studies have demonstrated that C467 dimerization is essential for the ER-to-Golgi transport of ATF6α during ER stress.17 However, the molecular mechanisms underlying this dimerization remain unclear. To investigate the regulation of dimerization, we focused on ERdj5, an ER-resident reductase containing the J-domain at its N-terminus, which has been reported to be involved in the activation of ATF6α, but the regulatory mechanism is unclear.18 We examined C467 dimer formation in ERdj5-knockdown (KD) cells transfected with ERdj5 siRNA (Figures 1A and 1B). Since C467 dimerization is known to be induced under ER stress,17 we treated cells with thapsigargin (Tg), an inhibitor of sarco/endoplasmic reticulum Ca2+-ATPase (SERCA), to deplete ER Ca2+ and induce ER stress. Consistent with previous findings, C467 dimerization of ATF6α was observed in control cells under ER stress conditions (Figure 1B). However, in ERdj5-KD cells, C467 dimerization was significantly reduced compared with that in control cells during ER stress (Figure 1C), indicating that ERdj5 promotes C467 dimer formation.Figure 1ATF6α C467 dimerization is regulated by Trx2 domain of ERdj5(A) Knockdown (KD) efficiency of ERdj5. HEK293T cells were transfected with ERdj5 siRNA and analyzed by immunoblotting with an anti-ERdj5 antibody under reducing conditions. Actin was used as a loading control. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2.(B) Formation of C467 dimers via ERdj5. HEK293T cells (wild-type; WT) and ERdj5-KD cells were treated with 1 μM Tg for the indicated periods and analyzed by immunoblotting with anti-ATF6α antibodies under nonreducing conditions. Representative immunoblots from three independent experiments are shown (n = 3).(D) A series of ERdj5 mutants was constructed in this study. One of the CXXC motifs in the Trx domain of ERdj5 was replaced by SXXS.(E) C467 dimerization through ERdj5 mutants. HEK293T cells were transfected with FLAG-tagged WT or mutant ERdj5 and analyzed by immunoblotting with anti-ATF6α and anti-FLAG antibodies under reducing and nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(G) C467 dimerization through the ERdj5 C2/CC mutant. HEK293T cells were transfected with FLAG-tagged WT or mutant ERdj5 and analyzed by immunoblotting with anti-ATF6α and anti-FLAG antibodies under reducing and nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(I) Regulation of C467 dimerization by ERp18 and ERdj5. HEK293T cells were transfected with Myc-tagged ERdj5 and/or FLAG-tagged ERp18 and analyzed by immunoblotting with anti-ATF6α, anti-FLAG, and anti-Myc antibodies under reducing and nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(C, F, H, and J) The intensity of the C467 dimer bands, as shown in (B, E, G, and I). The bands were quantified using ImageJ software and normalized to the total intensity (monomer and C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05, ∗∗p < 0.01). ATF6α C467 dimerization is regulated by Trx2 domain of ERdj5 (A) Knockdown (KD) efficiency of ERdj5. HEK293T cells were transfected with ERdj5 siRNA and analyzed by immunoblotting with an anti-ERdj5 antibody under reducing conditions. Actin was used as a loading control. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2. (B) Formation of C467 dimers via ERdj5. HEK293T cells (wild-type; WT) and ERdj5-KD cells were treated with 1 μM Tg for the indicated periods and analyzed by immunoblotting with anti-ATF6α antibodies under nonreducing conditions. Representative immunoblots from three independent experiments are shown (n = 3). (D) A series of ERdj5 mutants was constructed in this study. One of the CXXC motifs in the Trx domain of ERdj5 was replaced by SXXS. (E) C467 dimerization through ERdj5 mutants. HEK293T cells were transfected with FLAG-tagged WT or mutant ERdj5 and analyzed by immunoblotting with anti-ATF6α and anti-FLAG antibodies under reducing and nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (G) C467 dimerization through the ERdj5 C2/CC mutant. HEK293T cells were transfected with FLAG-tagged WT or mutant ERdj5 and analyzed by immunoblotting with anti-ATF6α and anti-FLAG antibodies under reducing and nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (I) Regulation of C467 dimerization by ERp18 and ERdj5. HEK293T cells were transfected with Myc-tagged ERdj5 and/or FLAG-tagged ERp18 and analyzed by immunoblotting with anti-ATF6α, anti-FLAG, and anti-Myc antibodies under reducing and nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (C, F, H, and J) The intensity of the C467 dimer bands, as shown in (B, E, G, and I). The bands were quantified using ImageJ software and normalized to the total intensity (monomer and C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05, ∗∗p < 0.01). To determine which domain of ERdj5 contributes to C467 dimerization, we mutated the Cys-X-X-Cys (CXXC) motif within each thioredoxin (Trx) domain of ERdj5 (Figure 1D). Among these mutants, the C2/SS mutant, which lacked Trx2 activity, failed to promote C467 dimerization (Figures 1E and 1F). Furthermore, we expressed each ERdj5 mutant at endogenous levels in ERdj5-KO cells generated by CRISPR-Cas9 and confirmed its effect on C467 dimers (Figures S1C and S1D). The amount of C467 dimers was decreased in ERdj5-KO cells compared to WT cells but was increased to the same level as in WT cells in mutants other than C2/SS (Figure S1E). These data suggest that Trx2 of ERdj5 is involved in the formation of C467 dimers. Next, we used a C2/CC mutant in which all CXXC motifs, except for Trx2, were replaced with Ala-X-X-Ala (AXXA) (Figures 1G and 1H). The C2/CC mutant contributed to C467 dimer formation similar to WT, whereas the AA mutant, in which all CXXC motifs were inactivated, failed to promote dimerization. In addition, when the C2/CC mutant was expressed at endogenous levels in ERdj5-KO cells, the amount of C467 dimers increased to the same level as that in WT cells (Figures S1F and S1G). These results indicate that the Trx2 domain is a key determinant of C467 dimerization. Next, we examined the relationship between ERdj5 and other oxidoreductases. First, we focused on the reductase ERp18, which regulates ATF6α activation by reducing Cys467 and inhibiting ATF6α transport17; however, whether ERp18 directly reduces the C467 dimer remains unknown. To investigate this process, we overexpressed ERp18 and ERdj5 individually or in combination. Overexpression of ERdj5 significantly increased C467 dimerization, whereas overexpression of ERp18 alone had no apparent effect. However, in cells coexpressing ERdj5 and ERp18, ERp18 strongly suppressed the C467 dimer formation promoted by ERdj5 (Figures 1I and 1J). These results indicate that ERp18 suppresses the formation of C467 dimers. Next, we focused on the oxidase PDIR, which is known to be involved in ATF6α activation. However, whether PDIR directly oxidizes the C467 dimer remains unknown.18 To clarify how ERdj5 and PDIR control ATF6α activation, we generated ERdj5-KO PDIR-KD cells and examined C467 dimers and ATF6α-N (Figures S1H–S1L). Similar to ERdj5-KD cells, the amount of C467 dimers in ERdj5-KO cells was reduced compared to WT cells (Figures 1C and S1J). Furthermore, the amount of C467 dimers in ERdj5-KO PDIR-KD cells was lower than that in ERdj5-KO cells (Figures S1I and S1J). The amount of ATF6α-N, an indicator of ATF6α activation, was also reduced in ERdj5-KO cells compared to that in WT cells and was further reduced in ERdj5-KO PDIR-KD cells (Figures S1K and S1L). These results suggest that ERdj5 and PDIR regulate the formation of C467 dimers and ATF6α activation. These results suggest that the Trx2 domain of ERdj5 oxidizes the C467 dimer reduced by ERp18 and that its function is complemented by PDIR. It is possible that oxidoreductases, such as ERdj5, ERp18, and PDIR, regulate ATF6α activation in a redox-dependent manner. Electron transfer between ERdj5 and ATF6α To elucidate the mechanism by which ERdj5 oxidizes ATF6α, we first examined the redox states of ATF6α and ERdj5 using PEG-maleimide assay. Because the redox state of the ATF6α monomer varies by cell type,16,21 we observed it in HEK293T cells using a PEG-maleimide assay (Figure 2A). PEG-maleimide reacts specifically with the thiol groups of cysteine residues via its maleimide moiety, thereby increasing the apparent molecular weight of the modified protein. The PEG-modified proteins were then analyzed by SDS-PAGE followed by western blotting to evaluate the redox state of ATF6α. To investigate the redox state of ER luminal cysteines in ATF6α, we used an AAACC mutant in which Cys81, Cys321 (cytosolic domain), and Cys380 (intermembrane domain) were replaced with alanine (Figure S1A). Under untreated conditions, analysis of the AAACC mutant revealed that most of the ATF6α monomers existed in a reduced monomer (Figure 2B). Next, to investigate the redox state of each Trx domain within ERdj5, we constructed C1/CC to C4/CC mutants. In these mutants, the CXXC motif of each thioredoxin-like domain was retained, whereas all other CXXC motifs were replaced with AXXA motifs. The results revealed that the reduced form of C2/CC mutant was present in WT cells but absent in ATF6α-KO cells (Figures 2C and 2D). In contrast, the C3/CC mutant was reduced in ATF6α-KO cells but absent in WT cells (Figure 2D). These findings suggest that the C2/CC mutant accepts electrons from ATF6α, facilitating the oxidation of Cys467 and thereby promoting C467 dimerization, whereas the C3/CC mutant donates electrons to ATF6α (Figure 2E). In other words, ERdj5 may have dual functions as both an oxidase and a reductase for ATF6α.Figure 2Redox states of ERdj5 and ATF6α(A) PEG-maleimide assay. The redox state of proteins is fixed by trichloroacetic acid (TCA) precipitation. The thiol groups of cysteine residues are then modified with PEG-maleimide and detected by SDS-PAGE. This modification causes an upward band shift for proteins in the reduced state, whereas oxidized proteins do not shift. The oxidizing agent diphenyl disulfide (DPS) promotes disulfide bond formation and inhibits PEG-maleimide modification, while the reducing agent dithiothreitol (DTT) reduces disulfide bonds and enhances PEG-maleimide modification.(B) The redox state of ATF6α was determined by nonreducing SDS-PAGE. ATF6α-knockout (KO) cells were transfected with the p3×FLAG-tagged AAACC mutant and analyzed by immunoblotting with anti-FLAG antibodies under nonreducing conditions. The bands from top to bottom indicate the C467 dimer, C618 dimer, reduced monomer and oxidized monomer. Reduced monomer indicates the reduced form of ATF6α; oxidized monomer indicates the oxidized form of ATF6α. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2.(C) The redox state of the ERdj5 mutants was observed using PEG-maleimide. WT or ATF6α-KO cells were transfected with FLAG-tagged ERdj5 mutant and analyzed by immunoblotting with anti-ATF6α and anti-FLAG antibodies under reducing conditions. Representative immunoblots from three independent experiments are shown (n = 3).(D) The intensity of reduced form bonds, as shown in (C). The bands were quantified using the ImageJ software and normalized to the total intensity (reduced form and oxidized form). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05, ∗∗p < 0.01).(E) ERdj5 oxidizes Cys467 of ATF6α via the Trx2 domain, and ATF6α forms the C467 dimer. Redox states of ERdj5 and ATF6α (A) PEG-maleimide assay. The redox state of proteins is fixed by trichloroacetic acid (TCA) precipitation. The thiol groups of cysteine residues are then modified with PEG-maleimide and detected by SDS-PAGE. This modification causes an upward band shift for proteins in the reduced state, whereas oxidized proteins do not shift. The oxidizing agent diphenyl disulfide (DPS) promotes disulfide bond formation and inhibits PEG-maleimide modification, while the reducing agent dithiothreitol (DTT) reduces disulfide bonds and enhances PEG-maleimide modification. (B) The redox state of ATF6α was determined by nonreducing SDS-PAGE. ATF6α-knockout (KO) cells were transfected with the p3×FLAG-tagged AAACC mutant and analyzed by immunoblotting with anti-FLAG antibodies under nonreducing conditions. The bands from top to bottom indicate the C467 dimer, C618 dimer, reduced monomer and oxidized monomer. Reduced monomer indicates the reduced form of ATF6α; oxidized monomer indicates the oxidized form of ATF6α. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2. (C) The redox state of the ERdj5 mutants was observed using PEG-maleimide. WT or ATF6α-KO cells were transfected with FLAG-tagged ERdj5 mutant and analyzed by immunoblotting with anti-ATF6α and anti-FLAG antibodies under reducing conditions. Representative immunoblots from three independent experiments are shown (n = 3). (D) The intensity of reduced form bonds, as shown in (C). The bands were quantified using the ImageJ software and normalized to the total intensity (reduced form and oxidized form). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05, ∗∗p < 0.01). (E) ERdj5 oxidizes Cys467 of ATF6α via the Trx2 domain, and ATF6α forms the C467 dimer. ERdj5 functions as both an oxidase and a reductase to regulate ATF6α for ER-Golgi trafficking Since the C3/CC mutant of ERdj5 retains electrons in ATF6α-KO cells, we examined whether ERdj5 directly interacts with ATF6α using a substrate-trapping ERdj5 mutant designed to form stable mixed disulfide bonds with its substrates.22 The CXXC motif of ERdj5 was replaced with Cys-X-X-Ala (CXXA) in the substrate-trapping mutant to prevent substrate dissociation (Figure 3A). We confirmed the presence of mixed disulfide complexes between the ERdj5 mutants and ATF6α (Figure 3B). Among the ERdj5 CA mutants, the ERdj5 C3/CA mutant, but not the C2/CA mutant, bound most strongly to ATF6α. Furthermore, the C3/CA mutant formed a disulfide bond-mediated complex with ATF6α under nonreducing and reducing conditions (Figure S2A). These results demonstrate that the Trx3 domain of ERdj5 reduces the disulfide bond at Cys618 in ATF6α. Next, we examined whether ERdj5 interacts with Cys467 or Cys618 of ATF6α using immunoprecipitation (Figure 3C). The C3/CA mutant bound to the WT and C467A mutant but not to the C618A mutant, indicating that ERdj5 binds to ATF6α via Cys618. Furthermore, under nonreducing and reducing conditions, immunoblotting revealed that the ERdj5 C3/CA mutant interacted with both the WT and the C467A mutant (Figure S2B). These results demonstrate that the Trx3 domain of ERdj5 reduces the disulfide bond at Cys618 in ATF6α. Next, we investigated the effect of endogenous ERdj5 on C618 dimerization of ATF6α. Compared with the control cells, the C467 dimer in ERdj5-KD cells were decreased, whereas the C618 dimer were increased (Figures 3D and 3E). Based on these results, we concluded that the Trx3 domain of ERdj5 reduces Cys618 in ATF6α.Figure 3ERdj5 regulates ATF6α dimerization through dual function(A) Steps for substrate reduction by reductase and stabilization of the complex by the CA mutant.(B and C) Interaction between ERdj5 and ATF6α. HEK293T cells were transfected with FLAG-tagged WT or mutant ERdj5 and Myc-tagged WT or mutant ATF6α, and the cell lysates were prepared for immunoprecipitation with anti-FLAG. The immunoprecipitates were subjected to SDS-PAGE and analyzed by immunoblotting with anti-FLAG and anti-Myc antibodies under reducing conditions. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2.(D) Reduction of C618 dimers by ERdj5. ATF6α-KO and ATF6α-KO ERdj5-KD cells were transfected with p3×FLAG-tagged WT ATF6α and analyzed by immunoblotting with anti-FLAG and anti-Myc antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(E) The intensity of each band, as shown in (D). The bands were quantified using the ImageJ software and normalized to the total intensity of each band position (monomer, C467 dimer, and C618 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05, ∗∗p < 0.01). ERdj5 regulates ATF6α dimerization through dual function (A) Steps for substrate reduction by reductase and stabilization of the complex by the CA mutant. (B and C) Interaction between ERdj5 and ATF6α. HEK293T cells were transfected with FLAG-tagged WT or mutant ERdj5 and Myc-tagged WT or mutant ATF6α, and the cell lysates were prepared for immunoprecipitation with anti-FLAG. The immunoprecipitates were subjected to SDS-PAGE and analyzed by immunoblotting with anti-FLAG and anti-Myc antibodies under reducing conditions. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2. (D) Reduction of C618 dimers by ERdj5. ATF6α-KO and ATF6α-KO ERdj5-KD cells were transfected with p3×FLAG-tagged WT ATF6α and analyzed by immunoblotting with anti-FLAG and anti-Myc antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (E) The intensity of each band, as shown in (D). The bands were quantified using the ImageJ software and normalized to the total intensity of each band position (monomer, C467 dimer, and C618 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05, ∗∗p < 0.01). We also examined how the reduction in Cys618 in ATF6α affected its transport through C467 dimers. Previously, ATF6α transport was analyzed using PF-429242, which inhibits the protease activity of S1P.17 This method detects Golgi-localized ATF6α (Golgi-ATF6α) and C467 dimers (Golgi-C467 dimer), both of which undergo N-glycan processing and O-glycosylation under ER stress, migrating above their corresponding ATF6α-F and C467 dimer bands (Figure S3A). In addition, we evaluated the glycosylation of the Golgi-C467 dimer with BADAP, an O-glycosylation inhibitor, and Endo H, which cleaves N-glycans from ER-localized proteins. BADAP treatment decreased the amount of Golgi-C467 dimer (Figure 4A, lane 3). Endo H treatment resulted in a downward shift of the bands representing the C467 dimer, C618 dimer, and monomer but did not affect the Golgi-C467 dimer (Figure 4A, lanes 2 and 5). These results suggest that the C467 dimer localizes to the Golgi apparatus during ER stress, which is consistent with previous reports.17Figure 4C618 dimers suppress C467 dimerization and its transport to the Golgi apparatus(A) Evaluation of ATF6α glycan levels. ATF6α-KO cells were transfected with p3×FLAG-tagged WT ATF6α and pretreated with 2 mM BADAP and 50 μM PF429242 (S1P inhibitor) for 1 h prior to the induction of ER stress with 5 μM Tg, and the results were analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. The cell lysates were left untreated or treated with Endo H. g and u indicate glycosylated and unglycosylated proteins, respectively. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2.(B) Transport and C467 dimerization of the ATF6α mutant during ER stress. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or C618A mutant ATF6α, pretreated with 50 μM PF429242 for 1 h prior to the induction of ER stress with 5 μM Tg for 0, 15, and 30 min, and analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(C) The intensity of the C467 dimer bands, as shown in (B). The bands were quantified using the ImageJ software and normalized to the total intensity (monomer, C618 dimer, and C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05).(D) The intensity of the C618 dimer bands, as shown in (B). The bands were quantified using ImageJ, normalized to the total intensity (monomer, C618 dimer, C467 dimer, and Golgi-C467 dimer), and plotted against time. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05).(E) The intensity of the Golgi-C467 dimers, as shown in (B). The bands were quantified using ImageJ, normalized to the total intensity (monomer, C618 dimer, C467 dimer, and Golgi-C467 dimer), and plotted against the time. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05).(F) Interaction between ATF6α and BiP. ATF6α-KO cells were transfected with p3×FLAG-tagged ATF6α and Myc-tagged BiP, and cell lysates were prepared for immunoprecipitation with an anti-Myc antibody. The immunoprecipitates were subjected to SDS-PAGE and analyzed by immunoblotting with anti-FLAG and anti-Myc antibodies under nonreducing and reducing conditions. Representative immunoblots from three independent experiments are shown (n = 3).(G) The intensity of each ATF6α band, as shown in (F). The bands were quantified using ImageJ software and normalized to the BiP (Input/IP) intensity. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05).(H) Transport of ATF6α mutant during ER stress. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α, pretreated with 50 μM PF429242 for 1 h prior to the induction of ER stress with 5 μM Tg for 1 h, and analyzed by immunoblotting with anti-FLAG antibodies under reducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(I) The intensity of the Golgi-ATF6α bands, as shown in (H). The bands were quantified using the ImageJ software and normalized to the total intensity (ATF6α-F and Golgi-ATF6α). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05).(J and K) Immunofluorescence of HeLa cells exogenously expressing GFP-tagged WT or mutant ATF6α. HeLa cells were transfected with N-terminal GFP-fused ATF6α (WT and various mutants). For ER localization analysis, cells were fixed and stained with calnexin (CNX), an ER-resident protein, without Tg treatment (J). For Golgi localization analysis, cells were pretreated with 50 μM PF429242 for 1 h and subsequently treated with 500 nM Tg for 30 min before staining with GM130, a Golgi marker (K). All cells were counterstained with Hoechst 33342 and immunostained with the indicated antibodies. Scale bars, 5 μm.(L) The percentage of cells showing colocalization between GFP-fused ATF6α and GM130, as shown in (K). The means ± standard error of the mean (SEM) of at three independent experiments, each performed with 10 cells. Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05).(M) ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α, pretreated with 5 μM Tg for 0, 30, and 60 min, and analyzed by immunoblotting with anti-FLAG antibodies under reducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(N) The intensity of ATF6α-N bands at 30 min after Tg treatment, as shown in (M). The bands were quantified using ImageJ, normalized to the total intensity (ATF6α-F and ATF6α-N). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05).(O) BiP mRNA levels in ATF6α-KO cells overexpressing WT or mutants under Tg treatment conditions. After treatment with 500 nM Tg for 8 h (+Tg), BiP mRNA levels were quantified by qPCR. The expression levels were normalized to GAPDH and compared to those in untreated cells (−Tg). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05, ∗∗p < 0.01). C618 dimers suppress C467 dimerization and its transport to the Golgi apparatus (A) Evaluation of ATF6α glycan levels. ATF6α-KO cells were transfected with p3×FLAG-tagged WT ATF6α and pretreated with 2 mM BADAP and 50 μM PF429242 (S1P inhibitor) for 1 h prior to the induction of ER stress with 5 μM Tg, and the results were analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. The cell lysates were left untreated or treated with Endo H. g and u indicate glycosylated and unglycosylated proteins, respectively. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2. (B) Transport and C467 dimerization of the ATF6α mutant during ER stress. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or C618A mutant ATF6α, pretreated with 50 μM PF429242 for 1 h prior to the induction of ER stress with 5 μM Tg for 0, 15, and 30 min, and analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (C) The intensity of the C467 dimer bands, as shown in (B). The bands were quantified using the ImageJ software and normalized to the total intensity (monomer, C618 dimer, and C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05). (D) The intensity of the C618 dimer bands, as shown in (B). The bands were quantified using ImageJ, normalized to the total intensity (monomer, C618 dimer, C467 dimer, and Golgi-C467 dimer), and plotted against time. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05). (E) The intensity of the Golgi-C467 dimers, as shown in (B). The bands were quantified using ImageJ, normalized to the total intensity (monomer, C618 dimer, C467 dimer, and Golgi-C467 dimer), and plotted against the time. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05). (F) Interaction between ATF6α and BiP. ATF6α-KO cells were transfected with p3×FLAG-tagged ATF6α and Myc-tagged BiP, and cell lysates were prepared for immunoprecipitation with an anti-Myc antibody. The immunoprecipitates were subjected to SDS-PAGE and analyzed by immunoblotting with anti-FLAG and anti-Myc antibodies under nonreducing and reducing conditions. Representative immunoblots from three independent experiments are shown (n = 3). (G) The intensity of each ATF6α band, as shown in (F). The bands were quantified using ImageJ software and normalized to the BiP (Input/IP) intensity. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05). (H) Transport of ATF6α mutant during ER stress. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α, pretreated with 50 μM PF429242 for 1 h prior to the induction of ER stress with 5 μM Tg for 1 h, and analyzed by immunoblotting with anti-FLAG antibodies under reducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (I) The intensity of the Golgi-ATF6α bands, as shown in (H). The bands were quantified using the ImageJ software and normalized to the total intensity (ATF6α-F and Golgi-ATF6α). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05). (J and K) Immunofluorescence of HeLa cells exogenously expressing GFP-tagged WT or mutant ATF6α. HeLa cells were transfected with N-terminal GFP-fused ATF6α (WT and various mutants). For ER localization analysis, cells were fixed and stained with calnexin (CNX), an ER-resident protein, without Tg treatment (J). For Golgi localization analysis, cells were pretreated with 50 μM PF429242 for 1 h and subsequently treated with 500 nM Tg for 30 min before staining with GM130, a Golgi marker (K). All cells were counterstained with Hoechst 33342 and immunostained with the indicated antibodies. Scale bars, 5 μm. (L) The percentage of cells showing colocalization between GFP-fused ATF6α and GM130, as shown in (K). The means ± standard error of the mean (SEM) of at three independent experiments, each performed with 10 cells. Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05). (M) ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α, pretreated with 5 μM Tg for 0, 30, and 60 min, and analyzed by immunoblotting with anti-FLAG antibodies under reducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (N) The intensity of ATF6α-N bands at 30 min after Tg treatment, as shown in (M). The bands were quantified using ImageJ, normalized to the total intensity (ATF6α-F and ATF6α-N). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05). (O) BiP mRNA levels in ATF6α-KO cells overexpressing WT or mutants under Tg treatment conditions. After treatment with 500 nM Tg for 8 h (+Tg), BiP mRNA levels were quantified by qPCR. The expression levels were normalized to GAPDH and compared to those in untreated cells (−Tg). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05, ∗∗p < 0.01). Next, we evaluated C467 dimerization using the C618A mutant (Figure 4B), which formed the C467 dimer more efficiently without Tg treatment (Figure 4B, lanes 1, 4, and 4C). Moreover, the C618 mutation increased the amount of Golgi-C467 dimer after 15 and 30 min of treatment with Tg and PF-429242, respectively (Figures 4B, 4D, and 4E). These data indicate that the reduction of Cys618 in ATF6α promotes C467 dimerization and facilitates its transport to the Golgi apparatus. Previous studies have shown that ATF6α transport is inhibited by BiP binding.12 Using immunoprecipitation, we found that BiP binds more robustly to the C618 dimer than to the monomer or C467 dimer (Figures 4F and 4G), suggesting that C618 dimerization retains ATF6α in the ER by enhancing its interaction with BiP, thereby preventing its transport to the Golgi apparatus. To further examine the contribution of the C618 dimer to ATF6α transport, we analyzed the Golgi-ATF6α forms of the WT, C618A, and C467A/C618A mutants after Tg and PF-429242 treatments (Figure 4H). ER-Golgi trafficking was lower in the C467A/C618A mutant than in the WT and C618A mutants (Figure 4I). We also overexpressed WT and mutant forms of N-terminal GFP-fused ATF6α in HeLa cells and evaluated the subcellular localization of ATF6α after Tg and PF-429242 treatment using immunofluorescence (Figures 4J–4L). WT and mutants co-localized with the ER marker calnexin under normal conditions, and with the Golgi marker GM130 upon Tg and PF-429242 treatment (Figures 4J and 4K). We calculated the proportion of ATF6α localized to the Golgi apparatus to assess the transport efficiency between WT and mutants (Figure 4L). The WT and C618A mutant showed higher Golgi localization than the C467A and C467A/C618A mutants, and the C618A mutant showed greater localization than the WT. These results suggest that the reduction in the C618 dimer enhances C467 dimerization during ATF6α transport. Next, to investigate the relationship between C467 dimerization and ATF6α activation, we detected ATF6α-N under Tg treatment (Figures 4M and 4N). The WT and C618A mutant showed higher levels of ATF6α-N than the C467A and C467A/C618A mutants, and the C618A mutant showed higher levels than the WT. Furthermore, we overexpressed WT or each mutant in ATF6α-KO cells and quantified the amount of BiP mRNA, a downstream target induced by ATF6α-N, using quantitative PCR (Figure 4O). The WT and C618A showed higher BiP mRNA levels than the C467A and C467A/C618A mutants, with the C618A mutant again showing a greater increase than the WT. These results indicate that C467 dimerization not only promotes ATF6α transport but also enhances the expression of downstream target genes. We showed that the C3 domain of ERdj5 reduces C618 dimer as a reductase, and the C2 domain forms C467 dimer as an oxidase, promoting transport to the Golgi apparatus under ER stress. ERdj5 sequentially performs the dual functions of oxidation and reduction on a single substrate to control ATF6α transport. Achromatopsia-associated ATF6α Y567N mutation locks C618 dimerization and disrupts ER-to-Golgi transport A mutation in ATF6α associated with achromatopsia results from missense mutations in the exon that encodes the ER luminal domain.19 One of these mutations, Y567N, in which tyrosine is mutated to asparagine at position 567, suppresses transport from the ER to the Golgi apparatus. C467 dimerization facilitates ATF6α transport from the ER to the Golgi apparatus, whereas C618 dimerization inhibits this process (Figure 4). Based on these results, we investigated whether the Y567N mutation affected dimerization and transport. Compared with the WT, the Y567N mutant presented increased C618 dimer formation and reduced C467 dimer formation (Figures 5A and 5B). Next, we investigated the relationship between C467 and C618 dimerization using Y567N and Y567N/C618A mutants (Figures 5C and 5D). Compared with the WT, the C618A mutant presented enhanced C467 dimer formation; however, no significant difference in C467 dimer formation was observed between the Y567N and Y567N/C618A mutants, suggesting that Y567N prevents C467 dimerization, regardless of the C618A mutation. Notably, the Y567N and Y567N/C618A mutants formed fewer C467 dimers than the WT.Figure 5Inhibition of C467 dimerization in the ATF6α transport suppressor mutant Y567N(A) Dimerization of Y567N mutant. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or Y567N mutant and analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(B) The intensity of each band, as shown in (A). The bands were quantified using the ImageJ software and normalized to the total intensity (monomer, C618 dimer, and C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗∗p < 0.01).(C) C467 dimerization of Y567N and Y567N/C618A mutants. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α and analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(D) The intensity of the C467 dimer bands, as shown in (C). The bands were quantified using the ImageJ software and normalized to the total intensity (monomer, C618 dimer, and C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05, ∗∗p < 0.01).(E) Transport of the Y567N mutant from the ER to the Golgi apparatus. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α, pretreated with 50 μM PF429242 for 1 h prior to the induction of ER stress with 5 μM Tg for 0, 30, or 60 min, and analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3).(F) Determination of the C467 dimer level in the Golgi apparatus after 60 min of Tg treatment. The intensity of the Golgi-C467 dimer band, as shown in (E). The bands were quantified using ImageJ software and normalized to the total value (monomer, C618 dimer, C467 dimer, and Golgi-C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗∗p < 0.01).(G) Reduced protein levels of Y567N/C618A under ER stress. The intensity of ATF6α-F bands, as shown in (E). The bands were quantified using ImageJ, normalized to the total value (monomer, C618 dimer, C467 dimer, and Golgi-C467 dimer), and plotted against time. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05).(H and I) Interaction between ERdj5 and Y567N mutant. ATF6α-KO cells were transfected with FLAG-tagged WT or mutant ERdj5 and Myc-tagged WT or Y567N mutant ATF6α, and cell lysates were prepared for immunoprecipitation with FLAG or Myc beads. The immunoprecipitates were subjected to SDS-PAGE and analyzed by immunoblotting with anti-FLAG and anti-Myc antibodies under reducing (H) and nonreducing (I) conditions. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2. Inhibition of C467 dimerization in the ATF6α transport suppressor mutant Y567N (A) Dimerization of Y567N mutant. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or Y567N mutant and analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (B) The intensity of each band, as shown in (A). The bands were quantified using the ImageJ software and normalized to the total intensity (monomer, C618 dimer, and C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗∗p < 0.01). (C) C467 dimerization of Y567N and Y567N/C618A mutants. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α and analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (D) The intensity of the C467 dimer bands, as shown in (C). The bands were quantified using the ImageJ software and normalized to the total intensity (monomer, C618 dimer, and C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05, ∗∗p < 0.01). (E) Transport of the Y567N mutant from the ER to the Golgi apparatus. ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α, pretreated with 50 μM PF429242 for 1 h prior to the induction of ER stress with 5 μM Tg for 0, 30, or 60 min, and analyzed by immunoblotting with anti-FLAG antibodies under reducing or nonreducing conditions. Actin was used as a loading control. Representative immunoblots from three independent experiments are shown (n = 3). (F) Determination of the C467 dimer level in the Golgi apparatus after 60 min of Tg treatment. The intensity of the Golgi-C467 dimer band, as shown in (E). The bands were quantified using ImageJ software and normalized to the total value (monomer, C618 dimer, C467 dimer, and Golgi-C467 dimer). The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗∗p < 0.01). (G) Reduced protein levels of Y567N/C618A under ER stress. The intensity of ATF6α-F bands, as shown in (E). The bands were quantified using ImageJ, normalized to the total value (monomer, C618 dimer, C467 dimer, and Golgi-C467 dimer), and plotted against time. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (∗p < 0.05). (H and I) Interaction between ERdj5 and Y567N mutant. ATF6α-KO cells were transfected with FLAG-tagged WT or mutant ERdj5 and Myc-tagged WT or Y567N mutant ATF6α, and cell lysates were prepared for immunoprecipitation with FLAG or Myc beads. The immunoprecipitates were subjected to SDS-PAGE and analyzed by immunoblotting with anti-FLAG and anti-Myc antibodies under reducing (H) and nonreducing (I) conditions. Original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study are contained in Data S2. We also examined whether the Y567N and Y567N/C618A mutants were transported to the Golgi apparatus during ER stress (Figure 5E). Immunoblotting revealed that in both mutants, Golgi-localized C467 dimers were barely detected compared with those in the WT, indicating that their transport was impaired (Figure 5F). We used immunoprecipitation to investigate the interaction between ERdj5 and the Y567N mutant. Immunoprecipitation analysis revealed that the Trx3 domain of ERdj5 bound more strongly to the Y567N mutant than to the WT under both reducing and nonreducing conditions (Figures 5H and 5I). These results suggest that the Y567N mutation interferes with C467 dimer formation, thereby blocking ATF6α transport to the Golgi apparatus. In addition, the Y567N mutation led to an increased accumulation of C618 dimers compared with the WT. ERAD of the ATF6α monomer is prevented by dimerization We observed that the amount of ATF6α-F was lower in the Y567N/C618A mutant than in the Y567N mutant under ER stress (Figure 5G). As ATF6α is known to be degraded by ERAD,23,24,25 we examined whether the reduction in ATF6α-F in the Y567N/C618A mutant was due to monomer degradation. We investigated whether ATF6α dimerization affects its degradation via ERAD using cycloheximide (CHX), a protein synthesis inhibitor. The WT was more resistant to degradation than the C467A/C618A mutant (Figures 6A and 6B). These data indicate that monomeric ATF6α is more susceptible to ERAD. To confirm that ATF6α degradation occurs via ERAD, we treated the cells with MG132, a proteasome inhibitor. MG132 treatment effectively prevented ATF6α degradation, supporting its dependence on ERAD (Figures S4A and S4B). Next, we examined the degradation of Y567N and Y567N/C618A mutants. The Y567N/C618A mutant was degraded more rapidly than the Y567N mutant, indicating that C618 dimerization protects ATF6α from ERAD (Figures 6C and 6D). This degradation was significantly suppressed by MG132 treatment, confirming that it occurred via the ERAD pathway (Figures S4C and S4D).Figure 6Degradation of the ATF6α monomer by ERAD(A and C) ATF6α degradation was evaluated using the protein synthesis inhibitor, cycloheximide (CHX). ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α, treated with 50 μg/mL CHX, and analyzed by immunoblotting with anti-FLAG antibodies under reducing conditions. Representative immunoblots from three independent experiments are shown (n = 3).(B and D) The intensity of the ATF6α-F bands, as shown (A and C). The bands were quantified using the ImageJ software, normalized to the value of actin, and plotted against time. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05). Degradation of the ATF6α monomer by ERAD (A and C) ATF6α degradation was evaluated using the protein synthesis inhibitor, cycloheximide (CHX). ATF6α-KO cells were transfected with p3×FLAG-tagged WT or mutant ATF6α, treated with 50 μg/mL CHX, and analyzed by immunoblotting with anti-FLAG antibodies under reducing conditions. Representative immunoblots from three independent experiments are shown (n = 3). (B and D) The intensity of the ATF6α-F bands, as shown (A and C). The bands were quantified using the ImageJ software, normalized to the value of actin, and plotted against time. The means ± SDs are shown (n = 3). Statistical analysis using a one-tailed test (ns, not significant; ∗p < 0.05). These findings suggest that ATF6α escapes ERAD through C467 dimerization, which stabilizes ATF6α and enables its ER-to-Golgi transport. Conversely, the Y567N mutant, which predominantly forms C618 dimers, is resistant to ERAD, revealing the effects of dimerization on ATF6α stability and its degradation. Discussion In this study, we elucidated the unique mechanism underlying redox-regulated ATF6α trafficking by ERdj5 (Figure 7). Previously, ERdj5 was characterized as a reductase in the ER22,26,27,28; however, we found that the Trx2 domain of ERdj5 promoted the formation of C467 dimers, as shown in Figure 1. The active center Cys-Pro-Pro-Cys (CPPC) motif of the Trx2 domain is characteristic and has not been found in the ER-localized PDI family in mammalian cells.29 In addition, proteins such as tryparedoxin from T. bruci, thioredoxin-h3, 4, and 5 from A. thaliana, and mammalian nucleoredoxin also contain the CPPC motif, but the detailed function of the CPPC motif is unclear.30,31 Previous reports have demonstrated that the Trx3 and 4 domains of ERdj5 have reducing activities, but the Trx2 domain has almost no activity.32 The redox potential of the Trx2 domain could not be determined through titration with glutathione (GSH) because it is extremely resistant to reduction by GSH due to the steric hindrance of the J domain. However, as shown in Figures 1 and 2, the characteristic active center of Trx2 accepts electrons from a specific client and works as an oxidase. In this case, since the redox state of monomeric ATF6α remains predominantly a reduced state, ERdj5 accepts electrons from the ATF6α monomer through the Trx2 domain and promotes C467 dimerization. On the other hand, we showed that the Trx3 domain releases electrons to ATF6α, reducing the disulfide bond formed by Cys618 and suppressing C618 dimerization. Taken together, we revealed the dual functions of ERdj5: reductase activity via the Trx3 domain and oxidative activity via the Trx2 domain.Figure 7ATF6α activation via redox regulationUnder normal conditions, ATF6α is localized in the ER as a C467 dimer, C618 dimer, or monomer, and ATF6α dimerization is regulated by ERdj5, PDIR, and ERp18. Some monomers are subsequently degraded through the ERAD. During ER stress, ATF6α monomer forms a C467 dimer via the Trx2 domain of ERdj5, which is then transported from the ER to the Golgi apparatus. A fraction of monomeric ATF6α is also transported to the Golgi. ATF6α activation via redox regulation Under normal conditions, ATF6α is localized in the ER as a C467 dimer, C618 dimer, or monomer, and ATF6α dimerization is regulated by ERdj5, PDIR, and ERp18. Some monomers are subsequently degraded through the ERAD. During ER stress, ATF6α monomer forms a C467 dimer via the Trx2 domain of ERdj5, which is then transported from the ER to the Golgi apparatus. A fraction of monomeric ATF6α is also transported to the Golgi. Building on this dual regulation, we next examined how these redox events influence the balance between the two disulfide-linked ATF6α dimers. We demonstrated that the C618 dimer suppresses C467 dimer formation and inhibits ATF6α transport (Figure 4). Although ERdj5 reduces the C618 dimer to promote ATF6α transport, the abundance of the C618 dimer is low in HEK293T cells, suggesting that this step is unlikely to be rate limiting for ATF6α activation in this cell line. Because our analyses were primarily conducted in HEK293T cells, it is also important to consider whether this redox-regulated mechanism operates in human tissues under physiological conditions. Previous studies have reported that the small-molecule compound AA147 activates ATF6α in a redox-dependent manner and rescues cone photoreceptor developmental defects in retinal organoids derived from ATF6α-Y567N patients.33,34 These findings indicate that redox-dependent ATF6α activation functions physiologically in human tissues. Taken together, the ERdj5-ATF6α redox regulatory mechanism identified here is not limited to an in vitro specific phenomenon but is likely to contribute to ATF6α activation in human disease-related tissues. We also found that C618 dimers bound more strongly to BiP than monomers and C467 dimers (Figures 4F and 4G). Given that BiP has been reported to bind near Cys467 in the ER lumen of ATF6α and inhibit its trafficking,12 these results suggest that ATF6α prevents BiP interactions by forming C467 dimers. In addition, ERdj5 is a member of the J domain protein (JDP) family, which harbors a J domain and promotes the ATPase activity of BiP.35 It is possible that ERdj5 simultaneously promotes the reduction of Cys618 and the dissociation of BiP. Furthermore, it has been reported that ERdj4, a member of the J protein family, functions as a selective IRE1 repressor,36 and a previous report revealed that ERdj4 complements the JDP function of ERdj5.37 Therefore, it would be interesting to elucidate the mechanism of the dissociation of BiP and ATF6α by the JDP family. Moreover, we found that ERdj5, together with PDIR and ERp18, regulates ATF6α activation (Figures 1I, 1J, and S1I–S1L). ERdj5 and PDIR promote the C467 dimer and ATF6α-N upon ER stress, whereas ERp18 suppresses the C467 dimer induced by ERdj5. These findings suggest that ERdj5, PDIR, and ERp18 regulate C467 dimerization and ATF6α activation. Furthermore, comprehensive interactome analysis by mass spectrometry revealed that some PDI family members, such as ERp57 and p5, form mixed disulfide bonds with ATF6α.38 The combination of PDI family members may be involved in the regulation of ATF6α activity. Dithiothreitol (DTT), a reducing agent, is a typical inducer of ER stress and promotes ATF6α activity. Interestingly, the C467 dimer is resistant to DTT and activates ATF6α in a reducing environment.17 Tg and tunicamycin (Tm), which are ER stress inducers, have been reported to shift the redox environment in the ER lumen from an oxidative environment to a reductive environment.39,40 In addition, some oxidases, including ERO1, are upregulated downstream of ATF6α and maintain the homeostasis of the redox environment in the ER.41 Thus, ATF6α may act as a sensor that senses an imbalance between the redox environment and ER stress. The Y567 mutation, which causes achromatopsia, inhibits the transport of ATF6α from the ER to the Golgi apparatus. Furthermore, other achromatopsia-associated ATF6α mutations, including amino acid mutations near Y567, reportedly affect the transport of ATF6α from the ER to the Golgi apparatus.19 However, the mechanism underlying the inhibition of ATF6α transport by the Y567 mutation remains unclear. In this study, we found that the Y567N mutation decreased C467 dimer formation while increasing C618 dimer formation (Figures 5A and 5B). As shown in Figures 3 and 4, reduction of the C618 dimer by ERdj5 promotes C467 dimer formation; however, in the Y567N mutant, C467 dimerization was not enhanced even when the amount of C618 dimer was reduced (Figures 5C, 5D, 5H, and 5I). Moreover, neither the Y567N nor the Y567N/C618A mutant showed increased C467 dimer formation upon Tg treatment (Figures 5E and 5F). These findings suggest that although ERdj5 can access the disulfide bond of the C618 dimer, structural abnormalities caused by the Y567N mutation may prevent ERdj5 from catalyzing the oxidation of Cys467. Understanding the efficient transport mechanisms of ATF6α is expected to aid in the development of treatments for various diseases linked to ATF6α dysfunction, including achromatopsia. ATF6α is rapidly degraded by ERAD, which involves EDEM2, SEL1, and Hrd1.23,24,25 When ATF6α accumulates excessively in SEL1-KO cells, ATF6α is transported to the Golgi apparatus even in the absence of ER stress; therefore, controlling the amount of ATF6α via ERAD is important. As shown in Figures 6A and 6B, we examined the degradation of ATF6α by the WT and C467A/C618A mutants and found that the mutant was more susceptible to degradation by ERAD. Furthermore, the Y567N mutant, which promotes C618 dimerization, was rapidly degraded by ERAD via the C618A mutation (Figures 6C and 6D), indicating that monomers are more susceptible to degradation than dimers. In addition, a previous report showed that IRE1 multimerizes during ER stress, thereby acquiring resistance to degradation by ERAD and becoming active.42 These results suggest that ATF6α acquires structural stability through C467 dimerization and is transported from the ER to the Golgi apparatus. However, ATF6α monomers are known to be transported from the ER to the Golgi apparatus,21 and similar results were observed in our study (Figures 4H and 4I). ATF6α monomers form complexes via noncovalent bonds16; thus, the monomers can also be transported if their structures are stabilized. A previous report showed that decreased ATF6α expression suppresses the proliferation of tumor cells.43,44 Inhibiting ATF6α dimerization and promoting monomer degradation via ERAD may be a therapeutic strategy. Several therapeutic strategies targeting ATF6α have been proposed, including AA147, which induces ATF6α activation by inhibiting the oxidative activity of PDI in the ER, and Ceapin-A7, which promotes ATF6α degradation via the ATP-binding cassette subfamily D member 3 (ABCD3) peroxisomal transporter.45,46 However, regulatory mechanisms based on structural transitions between the monomeric and dimeric forms of ATF6α have not been explored. In this study, we demonstrated that ATF6α dimerization functions as a molecular switch that determines its fate between activation and degradation (Figure 7). Future studies focusing on the disulfide bonds at Cys467 and Cys618 enable the development of more selective small-molecule modulators targeting ATF6α conformational dynamics. Such an approach could provide a therapeutic avenue for achromatopsia, cancer, and other ER stress-related diseases associated with ATF6α dysfunction. Limitations of the study We revealed that ATF6α dimerization is regulated by multiple oxidoreductases, such as ERdj5, PDIR, and ERp18, thereby controlling its activation. However, because this study is mainly based on cultured cell systems, it remains unclear how these oxidoreductases cooperatively regulate the ATF6α activation under physiological or disease conditions. Future studies integrating human tissues and disease models with approaches that quantify oxidoreductase expression levels and visualize dynamic changes in the ER redox environment will enable a more precise characterization of ATF6α dimerization and activation under physiological conditions. Resource availability Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Ryo Ushioda (ryo3ussy3@cc.kyoto-su.ac.jp). Materials availability This study did not generate new unique reagents. Data and code availability •All data reported in this paper will be shared by the lead contact upon request.•Raw data supporting the findings of this study have been deposited at Mendeley Data and are publicly available as of the date of publication. The DOI for the Mendeley Data is listed in the key resources table. Data S1 contains original uncropped western blot images used for quantification (n = 3 independent biological replicates), and Data S2 contains original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study.•This paper does not report original code.•Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. •All data reported in this paper will be shared by the lead contact upon request. All data reported in this paper will be shared by the lead contact upon request. •Raw data supporting the findings of this study have been deposited at Mendeley Data and are publicly available as of the date of publication. The DOI for the Mendeley Data is listed in the key resources table. Data S1 contains original uncropped western blot images used for quantification (n = 3 independent biological replicates), and Data S2 contains original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study. Raw data supporting the findings of this study have been deposited at Mendeley Data and are publicly available as of the date of publication. The DOI for the Mendeley Data is listed in the key resources table. Data S1 contains original uncropped western blot images used for quantification (n = 3 independent biological replicates), and Data S2 contains original uncropped western blot images of at least two independent repeat experiments that support the conclusions of the study. •This paper does not report original code. This paper does not report original code. •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. Acknowledgments The authors thank the members of the Nagata and Ushioda laboratories for their fruitful discussions and critiques. This work was supported by a Grant-in-Aid for Scientific Research (B) (JP22H02622) from JSPS (to R.U.), a Grant-in-Aid for Transformative Research Areas (A) (JP21H05268) from JSPS (to R.U.), International Leading Research (23K20040) from JSPS (to R.U.), the 10.13039/100009619Japan Agency for Medical Research and Development (10.13039/100009619AMED) under grant no JP21GM1410008 (to R.U.), no JP25gm6910021 (to R.U.), the 10.13039/100007449Takeda Science Foundation (to R.U.), and JST 10.13039/501100025019SPRING, Japan (grant no JPMJSP2157 to S.W.). Author contributions S.W. designed and conducted the experiments; S.W., K.U., K.N., and R.U. conceived the project, interpreted the results, and wrote the manuscript. Declaration of interests The authors have no conflicts of interest to declare. STAR★Methods Key resources table REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesMonoclonal ATF6 antibodyProteintechCat# 66563-1-Ig; RRID: AB_2881924Rabbit PolyAb Anti-DNAJC10ProteintechCat# 13101-1-AP; RRID: AB_10638440Anti-PDIR, Rabbit-PolyGENETEX, Inc.Cat# GTX115511; RRID: AB_10619349Anti-actin, clone C4MilliporeCat# MAB 1501; RRID: AB_2223041Anti-CalnexinBD BiosciencesCat# 610524; RRID: AB_397884Anti-GM130BD BiosciencesCat# 610822; RRID: AB_398141Monoclonal ANTI-FLAG M2 antibody produced in mouse (F1804)Sigma-AldrichCat# F1804; RRID: AB_262044Monoclonal ANTI-FLAG M2 antibody produced in mouse (F3165)Sigma-AldrichCat# F3165; RRID: AB_259529Antibody for the detection of FLAGTM conjugated proteins (rabbit)ROCKLANDCat# 600-401-383; RRID: AB_219374Anti-Myc-tag mAbMBLCat# M192-3; RRID: AB_11160947Anti-Myc-tag pAbMBLCat# 562-5; RRID: AB_591116Alkaline Phosphatase AffiniPure Goat Anti-Mouse IgG (H+L)Jackson ImmunoResearchCat# 115-055-146; RRID: AB_2338538Alkaline Phosphatase AffiniPure Goat Anti-Rabbit IgG (H+L)Jackson ImmunoResearchCat# 111-055-144; RRID: AB_2337953Goat anti-Mouse IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 555InvitrogenCat# A-21424; RRID: AB_141780Goat anti-Rabbit IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 546InvitrogenCat# A-11035; RRID: AB_2534093Chemicals, peptides, and recombinant proteinsDMEM, high glucose, pyruvateThermo Fisher ScientificCat# 11995040Penicillin-Streptomycin Mixed Solutionnakarai tesqCat# 26253-84Fetal Bovine Serum USA origin, suitable for cell cultureSigma-AldrichCat# 12103C-500MLPEI MAX[R] - Transfection Grade Linear Polyethylenimine Hydrochloride (MW 40,000)POLYSCIENCES,INC.Cat# 24765-100Lipofectamine RNAiMAX Transfection Reagent - Delivery of siRNAThermo Fisher ScientificCat# 13778150Dithiothreitol (DTT)nakarai tesqCat# 14128-62BCIP-NBT Solution (Ready To Use)nakarai tesqCat# 19880-84Separating Gel Buffer Solution (∗4)WakoCat# 192-11041Stacking Gel Buffer Solution (∗4)WakoCat# 192-1105130w/v% Acrylamide/Bis Mixed Solution (37.5:1)WakoCat# 016-15915N, N, N', N'-Tetramethylethylenediaminenakarai tesqCat# 33401-72Ammonium Peroxodisulfatenakarai tesqCat# 02627-34Blocking Onenakarai tesqCat# 03953-95PF429242 dihydrochlorideMedChem ExpressCat# 2248666-66-0Thapsigargin (Tg)FUJIFILMCat# 205-17283SUNBRIGHT ME-020MANOF AMERICA CORPORATIONCat# 883993-35-9N-ethylmaleimide (NEM)nakarai tesqCat# 15512-24Cycloheximide (CHX)nakarai tesqCat# 06741-91MG132ADQCat# A11043-100PrimeSTAR HS DNA PolymeraseTakara BioCat# 07493-344% Paraformaldehyde Phosphate Buffer SolutionFUJIFILMCat# 163-20145Normal Goat SerumabcamCat# ab7481Hoechst 33342nakarai tesqCat# 19172-51COSMOGEL(R) Ig-Accept Protein Gnakarai tesqCat# 02198-22THUNDERBIRD(R) Next SYBR(R) qPCR MixTOYOBO CO., LTDCat# 20498-74Deposited dataMendeley dataThis paperhttps://doi.org/10.17632/dh4c3prcbk.1OligonucleotidesStealthTM RNAi Negative Control Medium GC Duplex #2This paperN/AStealth siRNA targeting sequence for ERdj5:5’-UAAGCUCCCUAUUCCAUUGU-3’This paperN/AsiRNA targeting sequence for PDIR:Sence: 5’-CCUCGCUCAUUGAGAGAAUCU-3’Antisense: 5’-AUUCUCUCAAUGAGCGAGGAG-3’This paperN/APrimers for C81A:Forward: 5’-ACATCAACAACCAAATCGCTACAGT-3’Reverse: 5’-GTAATATCTTTAACTGTAGCGATTTG-3’This paperN/APrimers for C321A:Forward: 5’-AAATCGAGAATCCGCTGCTCAGTC-3’Reverse: 5’-TCTTCTTGCGAGACTGAGCAGCGGA-3’This paperN/APrimers for C380A:Forward: 5’-CAAAGCGAAGAGTTGTCGCTGTGAT-3’Reverse: 5’-CCAATACTATCATCACAGCGACAA-3’This paperN/APrimers for C467A:Forward: 5’-CCCCACCTCCTGCTCAGC-3’Reverse: 5’-TAATTAGGGGCTGAGCAGGAG-3’This paperN/APrimers for C618A:Forward: 5’-ATGATGCAGATTGACGCTCAG-3’Reverse: 5’-TCCATCACCTGAGCGTCAA-3’This paperN/APrimers for Y567N:Forward:5’-CCGCAGAAGGGGAGACACATTTAATGTTGT-3’Reverse:5’-TGATCCCTTCGAAATGACACAACATTAAATG-3’This paperN/APrimers for Myc-tag:Forward:5’-CTATAGGGAGACCCAAGCTGGCTAGCATGGGGGAGCCGGCTGGGG-3’Reverse:5’-GTTTAAACGGGCCCTCTAGATTACAGGTCCTCCTCTGAGATCAGCTTCTGCTCTTGTAATGACTCAGGGATGGTGCTGA-3’This paperN/APrimers for qRT-PCR GAPDH:Forward: 5’-TCCTTGGAGGCCATGTGGGCCAT-3’Reverse: 5’-TGATGACATCAAGAAGGTGAA-3’This paperN/APrimers for qRT-PCR BiP:Forward: 5’-CATGGTTCTCACTAAAATGAAAG-3’Reverse: 5’-GCTGGTACAGTAACAACTG-3’This paperN/APrimers for sgRNA target of ATF6α:Forward: CACCGAAAGGCTTGGGCTGAATTGAReverse: AAACTCAATTCAGCCCAAGCCTTTCThis paperN/APrimers for sgRNA target of ERdj5:Forward: CACCGTAAGAAAGAGAATGGGAGTCReverse: AAACGACTCCCATTCTCTTTCTTACThis paperN/ARecombinant DNAp3xFLAG-ATF6 in pCMV-3xFLAG-7.1AddgeneCat# 11975p3xFLAG-ATF6/WT in pcDNA3.1This paperN/Ap3xFLAG-ATF6/C467A in pcDNA3.1This paperN/Ap3xFLAG-ATF6/C618A in pcDNA3.1This paperN/Ap3xFLAG-ATF6/C467A/C618A in pcDNA3.1This paperN/Ap3xFLAG-ATF6/C467A/AAACC in pcDNA3.1This paperN/Ap3xFLAG-ATF6/Y567N in pcDNA3.1This paperN/Ap3xFLAG-ATF6/Y567N/C618A in pcDNA3.1This paperN/AMyc-ATF6/WT in pcDNA3.1This paperN/AMyc-ATF6/C467A in pcDNA3.1This paperN/AMyc-ATF6/Y567N in pcDNA3.1This paperN/ApEGFP-ATF6 in pEGFP-C3AddgeneCat# 32955pEGFP-ATF6/C467A in pEGFP-C3This paperN/ApEGFP-ATF6/C618A in pEGFP-C3This paperN/ApEGFP-ATF6/C467A/C618A in pEGFP-C3This paperN/ApCMV BiP-Myc-KDEL-wtAddgeneCat# 27164FLAG-ERdj5/WT in pcDNA3.1Ushioda et al.22N/AMyc-ERdj5/WT in pcDNA3.1Ushioda et al.22N/AFLAG-ERdj5/AA in pcDNA3.1Ushioda et al.22N/AFLAG-ERdj5/C1CC in pcDNA3.1Hagiwara et al.32N/AFLAG-ERdj5/C2CC in pcDNA3.1Hagiwara et al.32N/AFLAG-ERdj5/C3CC in pcDNA3.1Hagiwara et al.32N/AFLAG-ERdj5/C4CC in pcDNA3.1Hagiwara et al.32N/AFLAG-ERdj5/C1CA in pcDNA3.1This paperN/AFLAG-ERdj5/C2CA in pcDNA3.1This paperN/AFLAG-ERdj5/C3CA in pcDNA3.1This paperN/AFLAG-ERdj5/C4CA in pcDNA3.1Ushioda et al.22N/AFLAG-ERdj5/C1SS in pcDNA3.1This paperN/AFLAG-ERdj5/C2SS in pcDNA3.1Thie paperN/AFLAG-ERdj5/C3SS in pcDNA3.1Uegaki et al.47N/AFLAG-ERdj5/C4SS in pcDNA3.1Uegaki et al.47N/AERp18/Flag in pcDNA3.1Tsutsumi et al.48N/ACRISPER/ATF6α in px330This paperN/ACRISPER/ERdj5 in px459Uegaki et al.47N/ASoftware and algorithmsImageJNIHhttps://imagej.nih.gov/ij/CHOPCHOPLabun et al.49https://chopchop.cbu.uib.no/Zenn 2009 softwareZEISShttps://www.zeiss.com/microscopy/de/home.html REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesMonoclonal ATF6 antibodyProteintechCat# 66563-1-Ig; RRID: AB_2881924Rabbit PolyAb Anti-DNAJC10ProteintechCat# 13101-1-AP; RRID: AB_10638440Anti-PDIR, Rabbit-PolyGENETEX, Inc.Cat# GTX115511; RRID: AB_10619349Anti-actin, clone C4MilliporeCat# MAB 1501; RRID: AB_2223041Anti-CalnexinBD BiosciencesCat# 610524; RRID: AB_397884Anti-GM130BD BiosciencesCat# 610822; RRID: AB_398141Monoclonal ANTI-FLAG M2 antibody produced in mouse (F1804)Sigma-AldrichCat# F1804; RRID: AB_262044Monoclonal ANTI-FLAG M2 antibody produced in mouse (F3165)Sigma-AldrichCat# F3165; RRID: AB_259529Antibody for the detection of FLAGTM conjugated proteins (rabbit)ROCKLANDCat# 600-401-383; RRID: AB_219374Anti-Myc-tag mAbMBLCat# M192-3; RRID: AB_11160947Anti-Myc-tag pAbMBLCat# 562-5; RRID: AB_591116Alkaline Phosphatase AffiniPure Goat Anti-Mouse IgG (H+L)Jackson ImmunoResearchCat# 115-055-146; RRID: AB_2338538Alkaline Phosphatase AffiniPure Goat Anti-Rabbit IgG (H+L)Jackson ImmunoResearchCat# 111-055-144; RRID: AB_2337953Goat anti-Mouse IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 555InvitrogenCat# A-21424; RRID: AB_141780Goat anti-Rabbit IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 546InvitrogenCat# A-11035; RRID: AB_2534093Chemicals, peptides, and recombinant proteinsDMEM, high glucose, pyruvateThermo Fisher ScientificCat# 11995040Penicillin-Streptomycin Mixed Solutionnakarai tesqCat# 26253-84Fetal Bovine Serum USA origin, suitable for cell cultureSigma-AldrichCat# 12103C-500MLPEI MAX[R] - Transfection Grade Linear Polyethylenimine Hydrochloride (MW 40,000)POLYSCIENCES,INC.Cat# 24765-100Lipofectamine RNAiMAX Transfection Reagent - Delivery of siRNAThermo Fisher ScientificCat# 13778150Dithiothreitol (DTT)nakarai tesqCat# 14128-62BCIP-NBT Solution (Ready To Use)nakarai tesqCat# 19880-84Separating Gel Buffer Solution (∗4)WakoCat# 192-11041Stacking Gel Buffer Solution (∗4)WakoCat# 192-1105130w/v% Acrylamide/Bis Mixed Solution (37.5:1)WakoCat# 016-15915N, N, N', N'-Tetramethylethylenediaminenakarai tesqCat# 33401-72Ammonium Peroxodisulfatenakarai tesqCat# 02627-34Blocking Onenakarai tesqCat# 03953-95PF429242 dihydrochlorideMedChem ExpressCat# 2248666-66-0Thapsigargin (Tg)FUJIFILMCat# 205-17283SUNBRIGHT ME-020MANOF AMERICA CORPORATIONCat# 883993-35-9N-ethylmaleimide (NEM)nakarai tesqCat# 15512-24Cycloheximide (CHX)nakarai tesqCat# 06741-91MG132ADQCat# A11043-100PrimeSTAR HS DNA PolymeraseTakara BioCat# 07493-344% Paraformaldehyde Phosphate Buffer SolutionFUJIFILMCat# 163-20145Normal Goat SerumabcamCat# ab7481Hoechst 33342nakarai tesqCat# 19172-51COSMOGEL(R) Ig-Accept Protein Gnakarai tesqCat# 02198-22THUNDERBIRD(R) Next SYBR(R) qPCR MixTOYOBO CO., LTDCat# 20498-74Deposited dataMendeley dataThis paperhttps://doi.org/10.17632/dh4c3prcbk.1OligonucleotidesStealthTM RNAi Negative Control Medium GC Duplex #2This paperN/AStealth siRNA targeting sequence for ERdj5:5’-UAAGCUCCCUAUUCCAUUGU-3’This paperN/AsiRNA targeting sequence for PDIR:Sence: 5’-CCUCGCUCAUUGAGAGAAUCU-3’Antisense: 5’-AUUCUCUCAAUGAGCGAGGAG-3’This paperN/APrimers for C81A:Forward: 5’-ACATCAACAACCAAATCGCTACAGT-3’Reverse: 5’-GTAATATCTTTAACTGTAGCGATTTG-3’This paperN/APrimers for C321A:Forward: 5’-AAATCGAGAATCCGCTGCTCAGTC-3’Reverse: 5’-TCTTCTTGCGAGACTGAGCAGCGGA-3’This paperN/APrimers for C380A:Forward: 5’-CAAAGCGAAGAGTTGTCGCTGTGAT-3’Reverse: 5’-CCAATACTATCATCACAGCGACAA-3’This paperN/APrimers for C467A:Forward: 5’-CCCCACCTCCTGCTCAGC-3’Reverse: 5’-TAATTAGGGGCTGAGCAGGAG-3’This paperN/APrimers for C618A:Forward: 5’-ATGATGCAGATTGACGCTCAG-3’Reverse: 5’-TCCATCACCTGAGCGTCAA-3’This paperN/APrimers for Y567N:Forward:5’-CCGCAGAAGGGGAGACACATTTAATGTTGT-3’Reverse:5’-TGATCCCTTCGAAATGACACAACATTAAATG-3’This paperN/APrimers for Myc-tag:Forward:5’-CTATAGGGAGACCCAAGCTGGCTAGCATGGGGGAGCCGGCTGGGG-3’Reverse:5’-GTTTAAACGGGCCCTCTAGATTACAGGTCCTCCTCTGAGATCAGCTTCTGCTCTTGTAATGACTCAGGGATGGTGCTGA-3’This paperN/APrimers for qRT-PCR GAPDH:Forward: 5’-TCCTTGGAGGCCATGTGGGCCAT-3’Reverse: 5’-TGATGACATCAAGAAGGTGAA-3’This paperN/APrimers for qRT-PCR BiP:Forward: 5’-CATGGTTCTCACTAAAATGAAAG-3’Reverse: 5’-GCTGGTACAGTAACAACTG-3’This paperN/APrimers for sgRNA target of ATF6α:Forward: CACCGAAAGGCTTGGGCTGAATTGAReverse: AAACTCAATTCAGCCCAAGCCTTTCThis paperN/APrimers for sgRNA target of ERdj5:Forward: CACCGTAAGAAAGAGAATGGGAGTCReverse: AAACGACTCCCATTCTCTTTCTTACThis paperN/ARecombinant DNAp3xFLAG-ATF6 in pCMV-3xFLAG-7.1AddgeneCat# 11975p3xFLAG-ATF6/WT in pcDNA3.1This paperN/Ap3xFLAG-ATF6/C467A in pcDNA3.1This paperN/Ap3xFLAG-ATF6/C618A in pcDNA3.1This paperN/Ap3xFLAG-ATF6/C467A/C618A in pcDNA3.1This paperN/Ap3xFLAG-ATF6/C467A/AAACC in pcDNA3.1This paperN/Ap3xFLAG-ATF6/Y567N in pcDNA3.1This paperN/Ap3xFLAG-ATF6/Y567N/C618A in pcDNA3.1This paperN/AMyc-ATF6/WT in pcDNA3.1This paperN/AMyc-ATF6/C467A in pcDNA3.1This paperN/AMyc-ATF6/Y567N in pcDNA3.1This paperN/ApEGFP-ATF6 in pEGFP-C3AddgeneCat# 32955pEGFP-ATF6/C467A in pEGFP-C3This paperN/ApEGFP-ATF6/C618A in pEGFP-C3This paperN/ApEGFP-ATF6/C467A/C618A in pEGFP-C3This paperN/ApCMV BiP-Myc-KDEL-wtAddgeneCat# 27164FLAG-ERdj5/WT in pcDNA3.1Ushioda et al.22N/AMyc-ERdj5/WT in pcDNA3.1Ushioda et al.22N/AFLAG-ERdj5/AA in pcDNA3.1Ushioda et al.22N/AFLAG-ERdj5/C1CC in pcDNA3.1Hagiwara et al.32N/AFLAG-ERdj5/C2CC in pcDNA3.1Hagiwara et al.32N/AFLAG-ERdj5/C3CC in pcDNA3.1Hagiwara et al.32N/AFLAG-ERdj5/C4CC in pcDNA3.1Hagiwara et al.32N/AFLAG-ERdj5/C1CA in pcDNA3.1This paperN/AFLAG-ERdj5/C2CA in pcDNA3.1This paperN/AFLAG-ERdj5/C3CA in pcDNA3.1This paperN/AFLAG-ERdj5/C4CA in pcDNA3.1Ushioda et al.22N/AFLAG-ERdj5/C1SS in pcDNA3.1This paperN/AFLAG-ERdj5/C2SS in pcDNA3.1Thie paperN/AFLAG-ERdj5/C3SS in pcDNA3.1Uegaki et al.47N/AFLAG-ERdj5/C4SS in pcDNA3.1Uegaki et al.47N/AERp18/Flag in pcDNA3.1Tsutsumi et al.48N/ACRISPER/ATF6α in px330This paperN/ACRISPER/ERdj5 in px459Uegaki et al.47N/ASoftware and algorithmsImageJNIHhttps://imagej.nih.gov/ij/CHOPCHOPLabun et al.49https://chopchop.cbu.uib.no/Zenn 2009 softwareZEISShttps://www.zeiss.com/microscopy/de/home.html Experimental model and study participant details Cell lines HEK293T cells (ATCC CRL-3216; RRID: CVCL_0063; female origin) were obtained from the American Type Culture Collection (ATCC), and HeLa Kyoto cells (RRID: CVCL_1922; female origin) were obtained from Dr. Tamotsu Yoshimori (Osaka University, Japan). Authentication of HEK293T cells was performed by ATCC. HEK293T cells tested negative for mycoplasma in ATCC. Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco) supplemented with 10 fetal bovine serum (FBS) (Sigma-Aldrich, Cat#12103C) and 1% penicillin/streptomycin at 37°C in a humidified incubator with 5% CO2. HEK293T ERdj5 knockout and ATF6α knockout cell lines were generated from parental HEK293T cells using CRISPR/Cas9-mediated genome editing. Knockout efficiency was confirmed by immunoblotting. sgRNA sequences are listed in Table. Cells were used within 15 passages after thawing. Method details Construction of plasmids The ERdj5 WT, AA, C1/SS, C2/SS, C3/SS, C4/SS, C1/CC, C2/CC, C3/CC, C4/CC, C1/CA, C2/CA, C3/CA, C4/CA and ERp18 constructs were constructed as previously described.26,47,48 p3×FLAG-ATF6α, pEGFP-ATF6α, BiP-Myc and px330 were purchased from Addgene. ATF6α C467A, C618A, C467A/C618A, Y567N, Y567N/C618A, and AAACC were introduced using p3×FLAG-ATF6α as a template. The primers used are listed in Table, and the PCR products were amplified using the PrimeSTAR Mutagenesis Basal Kit (TaKaRa). To generate the ATF6α-Myc construct, primers listed in Table were designed with p3×FLAG-ATF6α as a template. The PCR product was amplified with the PrimeSTAR Mutagenesis Basal Kit and then digested with Nhe1 and Xba1, and the amplified fragment was inserted between the Nhe1 and Xba1 restriction sites in pcDNA3.1 (+). The ATF6α-KO vector was constructed using the CRISPR/Cas9 system according to Zhang’s protocol.50 Target sequences were investigated at the ChopChop site, and primers listed in Table were designed and synthesized using custom DNA oligonucleotides from Invitrogen.49 The primers were annealed and phosphorylated using T4PNK in a thermal cycler. Annealed and phosphorylated target sequences were digested with BbsI. The digested target sequences were inserted into the px330 vector. Cell culture and transfection HEK293T and HeLa cells (HeLa Kyoto strain) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (v/v) fetal bovine serum and antibiotics (100 U/mL penicillin and 100 μg/mL streptomycin) in humidified air containing 7% CO2 at 37°C. Cells were seeded at 0.8×106 cells/mL in dishes and cultured. Twenty-four hours later, the plasmid DNA was transfected with polyethylenimine (PEI) MAX (Polysciences, Inc.). PEI MAX was added to Opti-MEM (Gibco) and incubated for 5 min at room temperature. Plasmid DNA was added, vortexed, and allowed to stand for 20 min at room temperature, after which the entire mixture was placed dropwise in a dish. siRNA transfection was performed using Lipofectamine® RNAiMAX (Invitrogen) according to the manufacturer’s protocol. The target sequences of the siRNAs are listed in Table. Cell lysis, immunoprecipitation, and immunoblotting The cells were washed with ice-cold phosphate-buffered saline (PBS) and lysed with 1% Triton X-100 buffer (20 mM HEPES [pH 7.5], 20 mM N-ethylmaleimide (NEM), 150 mM NaCl, 10% glycerol, 1 mM EDTA, and a protease inhibitor cocktail). After rotation for 1 h at 4°C, the extracts were cleared by centrifugation at 13,000 × g for 10 min at 4°C. The samples were adjusted to contain 1× Laemmli buffer supplemented with 20 mM DTT and subjected to SDS‒PAGE, followed by immunoblotting. For immunoprecipitation, the cells were washed and harvested in PBS containing 20 mM NEM. The cell pellets were incubated with 10% ice-cold trichloroacetic acid (TCA) for 30 min on ice, followed by centrifugation at 10,000 rpm for 30 min at 4°C. The precipitated proteins were washed three times with ice-cold acetone. Air-dried precipitates were dissolved in 0.1 M phosphate buffer (20 mM NEM and 1% SDS). The cleared lysates were diluted with 0.1% SDS in phosphate buffer without SDS, incubated with 50 μL of ANTI-FLAG M2 Affinity Gel (Sigma‒Aldrich) for 1 h at 4°C with gentle shaking, and then centrifuged at 1,000 ×g for 1 min at 4°C to remove the supernatant. Sepharose beads were collected by centrifugation and washed three times with lysis buffer. Immunoprecipitated proteins were subjected to SDS‒PAGE followed by immunoblotting. For immunoblotting, proteins were transferred to a nitrocellulose membrane (Cytiva), which was blocked with Blocking One (Nacalai Tesque) for 60 min at room temperature. The primary antibodies were diluted in PBST or TBST and incubated overnight at 4°C. Alkaline phosphatase-conjugated secondary antibodies were used for detection, typically at a 1:1,000 dilution. Primary and secondary antibodies were used (Table). The blots were scanned using a GT-X820 system (Epson). Band intensities were quantified with ImageJ software (Fiji). Immunofluorescence After HeLa cells were transfected with GFP-fused ATF6α, they were incubated for 24 h and then treated with or without 500 μM Tg for 30 min. The cells were subsequently washed with PBS, fixed with 2% paraformaldehyde (PFA) for 10 min at room temperature, permeabilized with 0.2% Triton X-100 in PBS for 10 min at room temperature, and blocked with 1% normal goat serum for 1 h. The cells were incubated with primary and secondary antibodies, as listed in Table. The nucleus was stained using Hoechst 33342. Confocal images were obtained using an LSM700 confocal laser scanning microscope (Carl Zeiss) equipped with an α Plan-Apochromat 63×/1.4 NA oil immersion objective lens and analyzed using ZEN 2009 software (Carl Zeiss). All image analyses were performed using ImageJ. The ER and Golgi apparatus areas were defined by Calnexin and GM130 signals, respectively. The proportion of ATF6α localized to the Golgi apparatus was quantified by measuring the overlap between the ATF6α- and the GM130-positive areas. Quantitative real-time PCR (qRT-PCR) The expression levels of BiP under Tg treatment condition were determined using real-time PCR. Total RNA was isolated from ATF6α-KO cells overexpressing WT or mutants either untreated or treated with 500 μM Tg for 8 h, using the RNeasy Mini Kit (Qiagen). cDNA synthesis was performed using ReverTra Ace (TOYOBO). qRT-PCR was performed using the THUNDERBIRD® SYBR™ qPCR Mix (TOYOBO). The primer sequences used are listed in Table. GAPDH was used as the internal control. Redox states HEK293T cells were collected in PBS, left untreated or treated with 10 mM DTT or 1 mM DPS, and precipitated with 10% (v/v) TCA on ice for 20 min. After centrifugation at 12,000 rpm, the cell pellets were washed three times with acetone and completely dried. To measure the protein redox states in vivo, proteins in pellets were lysed and modified by incubation in 0.1 M phosphate buffer containing 20 mM NEM or 10 mM mPEG2000-mal (Nichiyu) and 1% SDS for 60 min at room temperature. The samples were subjected to western blotting after SDS‒PAGE. Deglycoside assay ATF6α-KO cells overexpressing ATF6α were harvested in PBS containing 20 mM NEM and solubilized in the ATF6α extraction buffer. Endo H was used according to the manufacturer's protocol (New England Biolabs). The samples were analyzed by SDS‒PAGE under reducing or nonreducing conditions. Generation of ATF6α-KO cells pX330, containing the sgRNA target sequence of ATF6α or ERdj5, was transfected into HEK293T cells, followed by incubation. The cells were cloned by limiting dilution 24 h later, and the KO of ATF6α or ERdj5 was confirmed by immunoblotting. Quantification and statistical analysis Quantitative data is presented as the mean ± standard deviation (SD) or standard error of the mean (SEM). ImageJ was used for data analysis and statistics. One-tailed Student’s t tests were used for comparison of means of data between two groups.