Regulation of ER-Resident Transcription Factor NFE2L1 in HEK293 Cells

Maiko Yasui; Izumi Nagae; Ryoichi Murase; Kozue Uchio-Yamada; Mahmoud Kandeel; Tadayuki Tsujita; Kentaro Oh-hashi

doi:10.1248/bpb.b25-00607

Abstract

Nuclear factor erythroid-derived 2 like 1 (NFE2L1) is reported to be embedded in the endoplasmic reticulum (ER) membrane and subsequently undergo N-glycosylation at several asparagine residues as well as other ER-resident factors including cAMP response element binding protein 3 (CREB3)/ATF6 family members. In this study, we investigated the regulation of NFE2L1 protein expression by treating wild-type HEK293 cells and HEK293 cells deficient in selected ER-associated degradation (ERAD) factors with various reagents. NFE2L1 protein expression in wild-type HEK293 cells was negligible, but MG132/bortezomib treatment induced Endo H-resistant two bands. Suppressor/enhancer of lin-12-like (SEL1L)/hydroxymethylglutaryl-CoA (HMG-CoA) reductase degradation 1 (Hrd1) loss increased NFE2L1 protein expression without any stimuli. In these deficient cells, the band shift of NFE2L1 by MG132 was mostly suppressed. Treatment with the valosin containing protein (VCP) inhibitor CB-5083 increased NFE2L1 expression, but deficiencies in other ERAD-associated factors (ER degradation-enhancing α-mannosidase-like protein 2 (EDEM2), thioredoxin domain-containing protein 11 (TXNDC11), gp78, ring finger protein 5 (RNF5), ring finger protein 185 (RNF185), and USP19) did not affect its expression. Comparing the stability of the two intrinsic NFE2L1, which increases with proteasome inhibition, the higher molecular weight form corresponding to full-length form, was more unstable. Therefore, we constructed NFE2L1 genes with mutations in the site where NFE2L1 is cleaved by DDI2 and in the four asparagine residues where N-glycosylation occurs, and found that the high molecular weight form, especially a hypoglycosylated mutant, tended to be more unstable. Taken together, this study using several ERAD disordered models shows that the regulation of NFE2L1 is different in some ways from the regulation of CREB3/ATF6 family, and these findings implicate the diversity of N-glycosylated protein regulation in the ER.

INTRODUCTION

Approximately one-third of newly synthesized proteins are processed within the endoplasmic reticulum (ER). Not only proteins with modifications such as N- and O-glycosylation but also nonglycosylated proteins are localized in the ER. Among these proteins, N-glycosylated proteins have been the most widely analyzed in terms of protein quality control mechanisms.^1–4) In general, proteins with specific asparagine residues labeled with Glc₃Man₉GlcNAc₂ are folded via the calnexin/calreticulin cycle after the removal of two glucose units. However, after mannose trimming, some of the proteins that fail to fold are escorted out of the ER. These unfolded proteins then undergo ubiquitination and proteasomal degradation.^3,4) This system of removal and degradation of abnormal proteins in the ER is termed ER-associated degradation (ERAD). The most studied ERAD complex on the ER membrane is the Hrd1/homocysteine-induced ER protein (Herp)/Derlin complex scaffolded by SEL1L.^3–5) However, the major degradation pathways of these N-glycosylated proteins have been analyzed by transient transfection of some genes (e.g., an α1-antitrypsin mutant, NHK),^6–10) and our knowledge on the degradation of endogenous ER-resident proteins is limited.^11–14) In the past few years, we have established several HEK293 cell lines deficient in ERAD-related factors using the clustered regularly interspaced short palindromic repeats/CRISPR-associated proteins 9 (CRISPR/Cas9) method.^13,15,16) Among them, we performed comprehensive MS analysis of protein using SEL1L-deficient HEK293 and wild-type cells and reported on the expression regulation of the CREB3/ATF6 family and TXNDC11.^13,15,16) The CREB3/ATF6 family comprises single-transmembrane proteins with similar structures that respond to ER or Golgi stress and undergo cleavage by S1P/S2P after transport to the Golgi apparatus, after which the N-terminal domain migrates to the nucleus to induce the expression of targeted factors.¹⁷⁾ During degradation of CREB3/ATF6 family within ER, EDEM2 is a key factor in their mannose trimming, and TXNDC11 contributes to the stabilization of EDEM2.^16,18) Therefore, the loss of either factor inhibits the degradation of the CREB3/ATF6 family.

In this study, we focused on nuclear factor erythroid-derived 2 like 1 (NFE2L1) as another factor remarkably increased in SEL1L-deficient HEK293 cells by our MS analysis. NFE2L1 is a different ER-localized transcription factor regulating oxidative stress and proteasome-related genes.^14,19–21) NFE2L1, also known as NRF1, possesses a single N-terminal transmembrane domain.^20–22) Several research approaches have identified it as an ERAD substrate, and it has been demonstrated that Hrd1 and VCP contribute to its regulation.^14,22,23) Furthermore, after retro-translocation into the cytoplasm, it has been reported that the protein undergoes deglycosylation by NGLY-1 and proteolytic cleavage by DDI2, and subsequently undergoes nuclear translocation.^24–26) However, the involvement of other ERAD-related factors in the regulation of the NFE2L1 protein remains incompletely understood. Then, we established additional cell lines deficient in ERAD-related factors such as gp78, RNF5, and USP19, which have been reported to be associated with Hrd1,^27–29) and analyzed the regulation of NFE2L1 expression by using various inhibitors and these mutant cells. Furthermore, by constructing several mutant NFE2L1 and VCP genes, we evaluated the regulation of both endogenous and overexpressed NFE2L1.

MATERIALS AND METHODS

Materials

CB-5083 (CB), kifunensine (Kif), ML240 (ML), NMS-873 (NMS), nelfinavir (Nelf), and tunicamycin (Tm) were purchased from Cayman Chemical Company (Ann Arbor, MI, U.S.A.). Bortezomib (Bor), MG132 (MG), z-VAD-fmk (z-VAD), and concanamycin A (CMA) were obtained from Kanto Chemical (Tokyo, Japan), Peptide Institute (Osaka, Japan) and Wako (Osaka, Japan), respectively. Cycloheximide (CHX) was obtained from Sigma-Aldrich (St. Louis, MO, U.S.A.). Endoglycosidase H (Endo H) was purchased from New England Biolabs (Ipswich, MA, U.S.A.).

Construction of Plasmids

The gRNAs against human EDEM2 (5′-TTCCGGCTGCTCATCCCGCT-3′ and 5′-GCTGAGGCAGCAGCGCGCAC-3′, EDEM2 KD), human Hrd1 (5′-GGCCAGCCTGGCGCTGACCG-3′ and 5′-CCACAGCCCCGGTCAGCGCC-3′, Hrd1 KD#1), human Hrd1 (5′-CC-GCCATCATCACTGCCGTG-3′ and 5′-CCACAGCCCCGGTCAGCGCC-3′, Hrd1 KD#2), human gp78 (5′-CCCAGCCTCCGCACCTACAC-3′ and 5′-GTGGGCCAGGGGAAGCGCTCG-3′, gp78 KD#1), human gp78 (5′-GTGGGCCAGGGGAAGCGCTCG-3′ and 5′-CGCTGAGGCCCGTGTAGGTG-3′, gp78 KD#2), human RNF5 (5′-GAGGAGGACGGGGGCCCCGA-3′ and 5′-CAAACATATATTACATTCGA-3′, and human RNF185 (5′-AGTGGGAGCAGCAATGGCGC-3′ and 5′-CCCACTGGGCCCCCCTGCAC-3′), RNF5/RNF185 KD), human SEL1L (5′-GAGCTTGGCCTCGGCGTCCT-3′, SEL1L KD#1) and (5′-GCAGCAGCGTCAGCCCTATC-3′, SEL1L KD#2) and human TXNDC11 (5′-CCGGCCGCTGGCGCGCCATG-3′ and 5′-CAGCGCGCAGCCGAGCGCCA-3′, TXNDC11 KD) human USP19 (5′-GCAGAAGGATCGAGCAAACC-3′ and 5′-GTCCTGGGGGCCCTCTCCTT-3′, USP19 KD#1), human USP19 (5′-GTCTCCTTGGGCCTGTGGCAC-3′ and 5′-GTCCTGGGGGCCCTCTCCTT-3′, USP19 KD#2) aligned with tracer RNA were, respectively, inserted into a pcDNA3.1-derived vector containing a U6 promoter.^13,15) To prepare donor genes, a DNA fragment encoding the N-terminal region of human EDEM2 (107 bp from the translation start site), human Hrd1 (97 bp from the translation start site), human gp78 (114 bp from the translation start site), human RNF5 (123 bp from the translation start site), human RNF185 (107 bp from the translation start site), human SEL1L (70 bp from the translation start site), human TXNDC11 (124 bp from 131 bp to 254 bp for TXNDC11 KD), and human USP19 (131 bp from the translation start site) was fused with a puromycin or hygromycin-resistance gene via IRES and inserted into a pGL3-derived vector.^13,15) The hCas9 construct (#41815) used in this study was obtained from Addgene.³⁰⁾ Mouse wild-type NFE2L1 gene was inserted into pcDNA3.1V5/His vector. Mouse wild-type NFE2L1 having a Flag epitope at the N-terminus was constructed in pcDNA3.1V5/His vector. Hypoglycosylated NFE2L1 gene which replaced four asparagine residues with glutamine (N319Q/N331Q/N383Q/N394Q (Q4)) and DDI2 resistant NFE2L1 in which the sequence around the DDI2 recognition site was replaced with alanine (mDDI2) were also inserted into pcDNA3.1V5/His vector. Wild-type and mutant human VCP (R155H, A232E and E305Q/E578Q (DKO)) genes in pcDNA3.1 myc/His vector were prepared from human VCP having EGFP at the C-terminus (Addgene).³¹⁾

Establishment of SEL1L-, Hrd1-, gp78-, RNF5/RNF185, EDEM2-, TXNDC11- and USP19-Deficient HEK293 Cells

HEK293 cells deficient for each ERAD factor were established using the CRISPR/Cas9 system as described previously.^13,15,16) In brief, donor genes encoding each N-terminus in a pGL3-derived vector, together with constructs for each gRNA and hCas9 gene, were transfected into HEK293 cells, and cells were selected with the appropriate concentrations of puromycin or hygromycin. Except for establishment of SEL1L-deficient cells, two different constructs for guide RNAs were cotransfected together with the donor gene and hCas9 constructs.

Cell Culture and Treatment

HEK293 cells were maintained in Dulbecco’s modified Eagle’s minimum essential medium containing 5% fetal bovine serum. The cells were seeded in 12-well plates and treated with Bor (0.1, 1, 10 μM), CB (2.5 μM), CHX (20 μg/mL), CMA (50 nM), ML (5 μM), MG (10 μM), Kif (2 μg/mL), Nelf (20 μM), NMS (2.5 μM), Tm (2 μg/mL), z-VAD (20 μM), or vehicle for the indicated times; the expression of the indicated proteins was then measured using Western blot analysis. Transfection of the indicated NEF2L1 genes into HEK293 cells in 12-well plates was performed using a PEI-MAX reagent as described previously.^13,16)

Western Blot Analysis

Cells were lysed in homogenization buffer (20 mM Tris–HCl (pH 8.0) containing 137 mM NaCl, 0.2 mM EDTA (pH 8.0), 10% (v/v) glycerol, 1% (v/v) Triton X-100, 1 mM PMSF, 10 μg/mL leupeptin, and 10 μg/mL pepstatin A) as described previously.^13,15,16) The protein concentrations of the lysates were determined using the Bradford assay, and equal amounts of lysate protein from different samples were separated by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, Burlington, MA, U.S.A.). The membranes were incubated with the indicated antibodies, and the target proteins were detected by enhanced chemiluminescence (ECL) (GE Healthcare, Buckinghamshire, U.K.) or enhanced chemiluminescence plus (ECL-plus) reagents (Life Technologies, Waltham, MA, U.S.A.) using ChemiDoc (BioRad). We used the following antibodies at the indicated dilutions: ATF6α (1 : 2000) (Proteintech, Rosemont, IL, U.S.A.); EDEM2 (1 : 1500) (Novus Biological, Centennial, CO, U.S.A.); CREB3 (1 : 1500) (Proteintech); cAMP response element binding protein 3-like 2 (CREB3L2) (1 : 2400) (Proteintech); glyceraldehyde 3-phosphate dehydrogenase (G3PDH) (1 : 15000) (Proteintech); gp78 (1 : 1200) (Proteintech); Herp (1 : 1000) (Cell Signaling Technology, Beverly, MA, U.S.A.); Hrd1 (1 : 2400) (Proteintech), LC3 (1 : 2400) (MBL, Nagoya, Japan), NFE2L1 (1 : 2400) (Proteintech); p62 (1 : 8000) (MBL); RNF5/RNF185 (1 : 2400) (Abcam, Cambridge, U.K.): SEL1L (1 : 2000) (Abcam); TXNDC11 (1 : 2000) (Proteintech); USP19 (1 : 2400) (Proteintech), V5-epitope (1 : 2400) (MBL). The level of expression of each protein was analyzed using ImageJ software (National Institutes of Health, U.S.A.), and the relative amount of each protein was calculated based on the G3PDH value obtained from an identical sample of cell lysate. Protein expression was normalized as described in the figure legends.

Endoglycosidase Digestions

An aliquot of cell lysate prepared for Western blot analysis was incubated with Endo H (100 U) for 18 h at 37°C, as described previously.³²⁾ As a control, an equal amount of the same cell lysate was incubated without enzyme. Each sample was then analyzed by Western blot analysis using an antibody against NFE2L1.

Statistical Analysis

The results are expressed as the mean ± standard error of the mean (S.E.M.). Statistical analyses were carried out using one-way ANOVA followed by Tukey–Kramer test.^13,15)

RESULTS

ER-Resident NFE2L1 Is Degraded in a Proteasome-Dependent Manner

We have recently established SEL1L, EDEM2, and TXNDC11-deficient HEK293 and comparatively analyzed the expression of the CREB3/ATF6 family,^13,15,16) which are ER-localized glycosylated transcription factors and are involved in ER and Golgi stress responses.¹⁷⁾ On the other hand, since SEL1L is one of the key factors in ERAD, we performed a comprehensive MS analysis of protein expression using wild-type and SEL1L-deficient HEK293,^15,16) and found NFE2L1 as one of the candidate factors increased in our SEL1L-deficient HEK293 cells. NFE2L1 is an ER-localized transcription factor that has a transmembrane domain near the N-terminus and undergoes multiple N-glycosylation.^20,22,24,25) NFE2L1 has been reported as an extremely unstable Hrd1-dependent ERAD substrate.^14,22,23) However, the role of other ERAD-related factors in the regulation of NFE2L1 protein remains to be fully elucidated. Therefore, we investigated the protein regulation of NFE2L1 together with CREB3L2, which is rapidly degraded in a proteasome-dependent manner.^13,33) First, we treated wild-type HEK293 cells with the following reagents to analyze the protein expression of NFE2L1 and CREB3L2 (Fig. 1A). MG132 (MG) and concanamycin A (CMA) inhibit protein degradation by the proteasome and lysosome, respectively. Cycloheximide (CHX) inhibits new protein synthesis. Kifunensine (Kif) is an inhibitor of α-mannosidase involved in the mannose trimming during ERAD, while tunicamycin (Tm) inhibits protein N-glycosylation. The NFE2L1 protein level in HEK293 cells under unstimulated conditions was negligible; however, two clear bands were detected after treatment with the proteasome inhibitor, MG132. Considering the previous studies^20,22–26) and our subsequent experiments, one is thought to be the form (II, high molecular weight form) that has been deglycosylated by NGLY-1 in the cytoplasm after being retro-transported from the ER, and the other is the form (III) that was subsequently cleaved by DDI2. Regarding CREB3L2 expression, an intermediate form with a slightly lower molecular weight than the full-length form, and a cleaved form around 50 kDa was increased by MG132 as we previously reported.^13,15) Treatment with kifunensine, which inhibits mannose trimming of N-glycosylated proteins, had no effect on NFE2L1 expression, although it noticeably increased CREB3L2 expression.^7,13)

Fig. 1. ER-Resident NFE2L1 Is Degraded in a Proteasome-Dependent Manner

(A) Wild-type (wt) HEK293 cells were treated with MG132 (MG, 10 μM), CMA (50 nM), CHX (20 μg/mL), kifunensine (Kif, 2 μg/mL), tunicamycin (Tm, 2 μg/mL) or vehicle (con) for 8 h. Representative results of three independent experiments are shown. (B) Wt HEK293 cells were treated with MG132, bortezomib (Bor) at the indicated concentration or vehicle for 6 h. The expression of the indicated proteins was measured as described in Materials and Methods. Each value presented is the mean ± S.E.M. from five independent cultures. The amount of each NFE2L1 protein in cells by treatment with 1 μM MG132 was set at 1. The p-values for statistically significant differences between the indicated groups are described. NFE2L1 (II, III) indicates the non-ER-localized full-length form (high molecular weight form) and cleaved form, respectively. Filled and open arrowheads, respectively, indicate the full-length and cleaved CREB3L2.

Next, in addition to MG132, another proteasome inhibitor, bortezomib, was used to compare the concentration-dependent effects on NFE2L1 and CREB3L2 expression. Both inhibitors tended to increase high molecular weight NFE2L1 expression at higher concentrations (Fig. 1B). On the other hand, each treatment increased intermediate and cleaved CREB3L2, respectively. These results indicate that although both are N-glycosylated transcription factors localized to the ER and ERAD substrates, their regulation is not identical.

NFE2L1 Is Degraded by the Proteasome System via SEL1L/Hrd1 in HEK293 Cells

In our SEL1L-deficient HEK293 cells, there was hardly any change in the expression level of either of Hrd1 and gp78 Ub ligases (Fig. 2A(a)); however, loss of SEL1L increased expression of full-length NFE2L1 as well as CREB3/ATF6 family proteins, especially full-length CREB3 and CREB3L2 proteins (Fig. 2A(b)). In wild-type HEK293 cells, NFE2L1 protein was scarcely detected, although NFE2L1 mRNA expression was observed (Supplementary Fig. 1).

Fig. 2. NFE2L1 Is Degraded by the Proteasome System via SEL1L/Hrd1 in HEK293 Cells

The expression of the indicated proteins in wild-type (wt), SEL1L-deficient (A), Hrd1- and gp78-deficient (B), Hrd1- and RNF5/185 (RNF)-deficient (C) HEK293 was measured as described in Materials and Methods. (C) Wild-type (wt) and the indicated deficient HEK293 cells were treated with MG132 (MG, 10 μM) or vehicle for 6 h. Representative results of three independent experiments are shown. NFE2L1 (I, II, III) indicates the ER-resident glycosylated form, non-ER-localized full-length form (high molecular weight form) and cleaved form, respectively. Filled and open arrowheads, respectively, indicate the full-length and cleaved CREB3/ATF6 family.

Next, we tried Hrd1- and gp78-deficient HEK293 cells to evaluate the protein expression of NFE2L1 together with the CREB3/ATF6 family members (Fig. 2B). Hrd1 deficiency had a similar effect on the expression of full-length NFE2L1, CREB3, and CREB3L2 as SEL1L deficiency. In contrast, gp78 deficiency had little effect on them. though it is reported that gp78 controls the protein quality control within ER together with Hrd1.²⁷⁾ Given that RNF5 has also been reported to interact with Hrd1,²⁸⁾ we also established HEK293 cells deficient in both RNF5 and RNF185, structurally similar ER-localized ubiquitin ligases. However, the RNF5/RNF185 deficiency showed no effect on Hrd1, NFE2L1, or CREB3L2 expression (Fig. 2C).

Deficiency of EDEM2 and TXNDC11 Does Not Cause NFE2L1 Accumulation in HEK293 Cells

Although both NFE2L1 and CREB3/ATF6 family members are reported to undergo N-glycosylation,^{14,17,18,24,25)} our results showed that NFE2L1 is not affected by the mannosidase inhibitor kifunensine (Fig. 1A). We then investigated the effect of the N-glycosylation inhibitor tunicamycin on the increase in NFE2L1 expression in wild-type and Hrd1-deficient cells. In addition, both cells were treated with MG132 for 8 h to examine NFE2L1 and CREB3L2 expression (Fig. 3A). As shown in Figs. 1 and 3A, two forms of NFE2L1 were observed in the Western blot of wild-type HEK293 cells after MG132 treatment. In Hrd1-deficient cells, both NFE2L1 and CREB3L2 were abundant even in the unstimulated state. In Hrd1-deficient cells, the molecular weight of CREB3L2 protein decreased upon treatment with MG132 or Tm, while the molecular weight of NFE2L1 protein remained largely unchanged.

Fig. 3. Deficiency of EDEM2 and TXNDC11 Does Not Cause NFE2L1 Accumulation in HEK293 Cells

(A) Wild-type (wt) and Hrd1-deficient (#2) HEK293 cells were treated with MG132 (MG, 10 μM), tunicamycin (Tm, 2 μg/mL) or vehicle (con) for 8 h. (B) The indicated lysates were incubated with or without Endo H as described in Materials and Methods. (C) Wild-type (wt) and EDEM2 (E2)- or TXNDC11 (TX11)-deficient HEK293 cells were treated with MG132 (MG, 10 μM), kifunensine (Kif, 2 μg/mL) or vehicle (con) for 8 h. The expression of the indicated proteins was measured as described in Materials and Methods. Results of three (A, C) and two (B) independent experiments are shown. NFE2L1 (I, II, III) indicates the ER-resident glycosylated form, non-ER-localized full-length form (high molecular weight form) and cleaved form, respectively. The I* are derived by Endo H digestion of the ER-resident glycosylated form. Filled and open arrowheads, respectively, indicate the full-length and cleaved CREB3L2.

Therefore, the lysates of the above-mentioned wild-type and Hrd1-deficient cells were treated with the endoglycosidase Endo H, and NFE2L1 expression was then evaluated (Fig. 3B). Two bands of NFE2L1 protein in MG132-treated wild-type HEK293 cells did not shift due to Endo H treatment. On the other hand, in Hrd1-deficient HEK293 cells, a band shift due to Endo H was observed both with and without MG132 treatment, and the molecular weight of the resulting deglycosylated NFE2L1 did not match that in either MG132-treated wild-type HEK293 cells. We therefore investigated the contribution of mannose trimming to NFE2L1 proteolysis using EDEM2- and TXNDC11-deficient HEK293 cells. Similar to the results obtained with kifunensine, the deficiency of EDEM2 and TXNDC11 caused the accumulation of CREB3L2 but did not affect NFE2L1 expression. Furthermore, the two NFE2L1 proteins induced by MG132 in each deficient cell were equivalent to those in wild-type cells (Fig. 3C). The fact that the effects of EDEM2/TXNDC11 deficiencies were observed only in CREB3L2 suggests that NFE2L1 and CREB3L2 proteins may follow different pathways when reaching the SEL1L/Hrd1 complex.

Deficiency of USP19 Does Not Affect NFE2L1 Expression in HEK293 Cells

Protein homeostasis is thought to be controlled by the balance between ubiquitination and deubiquitination. Among the several deubiquitinating enzymes present in the ER membrane, USP19 has been reported to be involved in the regulation of NFE2L1 and Hrd1.^29,34) Therefore, we established two lines of USP19-deficient HEK293 cells using the CRISPR/Cas9 method to analyze the involvement of USP19 in NFE2L1 and CREB3L2 expression.^13,15,16) Bortezomib caused an increase in the expression and processing of NFE2L1 and CREB3L2 proteins, along with an increase in low molecular weight USP19 protein (Fig. 4). In the USP19-deficient HEK293 cells, although both USP19 proteins were deficient, there was no effect on NFE2L1 or CREB3L2 expression. This CRISPR/Cas9 analysis revealed the presence of two USP19 proteins with different molecular weights in HEK293 cells. However, unlike previous reports,³⁴⁾ it was unexpected that USP19 did not participate in NFE2L1 stabilization.

Fig. 4. Deficiency of USP19 Does Not Affect NFE2L1 and CREB3L2 Degradation in HEK293 Cells

Wild-type (wt) and USP19-deficient (#1 and 2) HEK293 cells were treated with bortezomib (Bor, 0.1 or 10 μM) or vehicle (con) for 6 h. The expression of the indicated proteins was measured as described in Materials and Methods. Each value presented is the mean ± S.E.M. from three independent cultures. The amount of each USP19 and NFE2L1 protein in wild-type (wt) cells by treatment with 10 μM bortezomib was set at 1. The p-values for statistically significant differences between the indicated groups are described. NFE2L1 (II, III) indicates the non-ER-localized full-length form (high molecular weight form) and cleaved form, respectively. Filled and open arrowheads, respectively, indicate the full-length and cleaved CREB3L2.

VCP Suppression Increases NFE2L1 Expression and Attenuates Its Processing in HEK293 Cells

It has been reported that VCP, a member of the AAA ATPase family of proteins, plays an important role in extracting abnormal proteins from the ER into the cytoplasm.^22,31,35,36) On the other hand, VCP is known to have various functions such as membrane fusion, DNA replication, and lysosomal degradation,^37,38) and several VCP inhibitors/activators have been developed to date.^39–42) Therefore, we examined the effects of three structurally different VCP inhibitors on NFE2L1 and CREB3L2 expression. As shown in Fig. 5A, ML240 had little effect on the expression of either protein even at a concentration of 5 μM despite the increase in LC3-II protein expression. VCP has been reported to be involved in lysosomal control in addition to ERAD.^31,37,38) Its inhibition by ML240 may have contributed to the increase in LC3-II. However, it remains unclear why it did not show any ERAD inhibition, such as increased NFE2L1 or CREB3L2 expression. NMS-873 and CB-5083 at 2.5 μM significantly induced GADD153 to the same extent, but the increase in full-length NFE2L1 and CREB3L2 proteins was more pronounced with CB-5083. CB-5083 also induced cleaved CREB3L2. The differing effects of the three inhibitors on each protein expression may be due to differences in their inhibition mechanisms,⁴²⁾ but off-target effects cannot be ruled out.

Fig. 5. VCP Suppression Increases NFE2L1 Expression and Inhibits Its Processing in HEK293 Cells

(A) Wild-type (wt) HEK293 cells were treated with ML240 (ML, 5 μM), NMS-873 (NMS, 2.5 μM), CB-5083 (CB, 2.5 μM) or vehicle (con) for 6 h. Each value presented is the mean ± S.E.M. from four independent cultures. The amount of the indicated proteins in cells by treatment with CB-5083 was set at 1. The p-values for statistically significant differences between the indicated groups are described. B) Forty-two hours after transfection of NFE2L1-V5/His together with the indicated VCP genes or empty vector (mock), cells were treated with bortezomib (Bor, 1 μM) or vehicle (con) for an additional 6 h. The expression of the indicated proteins was measured as described in Materials and Methods. Each value presented is the mean ± S.E.M. from three independent cultures. The amount of each NFE2L1 protein in the bortezomib-treated cells transfected with empty vector was set at 1. The p-values for statistically significant differences between the indicated groups are described. NFE2L1 (I, II, III) indicates the ER-resident glycosylated form, non-ER-localized full-length form (high molecular weight form) and cleaved form, respectively. Filled and open arrowheads, respectively, indicate the full-length and cleaved CREB3L2.

To clarify the contribution of VCP, wild-type and three mutant VCPs were co-transfected with V5-tagged NFE2L1, and their effects on NFE2L1 expression and processing were observed (Fig. 5B). Under normal conditions, a dominant negative mutant VCP, E305Q/E578Q (DKO), showed faint V5-tagged NFE2L1, which is considered to be an ER-localized form. In contrast, the DKO mutant VCP significantly reduced the two NFE2L1 bands induced by bortezomib treatment. Co-transfection of wild-type and other mutant VCPs slightly increased bortezomib-induced NFE2L1 protein compared with cells transfected with an empty vector, but their effects were not significant. ALS-associated VCP mutations (R155H, A232E) are reported to affect the lysosomal system³¹⁾; therefore, it is reasonable that only the DKO affects NFE2L1 expression. However, the reason for the minimal increase in NFE2L1 expression by the DKO mutant under the unstimulated condition remains unclear.

Evaluation of Bortezomib Responsiveness Using Mutant NFE2L1

Since the band intensity of NFE2L1 detected differed depending on the concentration of proteasome inhibitors (Fig. 1B), we finally examined the stability of the two NFE2L1 proteins that appear during bortezomib treatment. First, we analyzed endogenous NFE2L1 protein together with CREB3L2 protein in cells cultured for 1.5, 3, and 6 h after adding a protein synthesis inhibitor (CHX) to the bortezomib-treated cells (Fig. 6A). After 6 h of treatment with CHX, the high molecular weight form almost completely disappeared, whereas the low molecular weight form showed only a slight decrease. On the other hand, the bands of intermediate and cleaved CREB3L2, which increased by bortezomib treatment, were rapidly downregulated by CHX treatment. This result indicates that proteasome activity recovered after the removal of bortezomib. To analyze differences in the stability of two bortezomib-induced NFE2L1 proteins furthermore, we constructed four NFE2L1 genes with mutations at the N-terminal region cleaved by DDI2 and four N-glycosylated asparagine predicted by Uniprot (Fig. 6B). As expected, NFE2L1 with a mutation at the DDI2 cleavage site (mDDI2) showed suppression of low molecular weight form expression under bortezomib treatment, while high molecular weight form increased. NFE2L1 with four N-glycosylated residues mutated to glutamine (Q4) had a lower molecular weight than the wild-type. Furthermore, regardless of the presence or absence of mutations in the DDI2-cleavage site, the hypoglycosylated high molecular weight form (Q4) was relatively unstable after treatment with CHX (Fig. 6B). Therefore, this suggests the existence of a certain mechanism that recognizes uncleaved NFE2L1 in cytoplasm and selectively degrades it. The instability of the hypoglycosylated high molecular weight form (Q4) may indicate the importance of the deglycosylation process mediated by NGLY-1 although structural effects caused by amino acid substitution cannot be ruled out.

Fig. 6. Evaluation of Bortezomib-Induced Processing of Intrinsic and Transfected NFE2L1

(A) Six hours after treatment with bortezomib (Bor, 1 μM) or vehicle, cells were treated with CHX (20 μg/mL) for 0, 1.5, 3 and 6 h. The expression of the indicated proteins was measured as described in Materials and Methods. Each value presented is the mean ± S.E.M. from three independent cultures. The amount of each NFE2L1 protein in cells just after pretreatment with bortezomib was set at 1. The p-values for statistically significant differences between the indicated groups are described. (B) Schematic structure of wild-type and mutant NFE2L1. Thirty-six hours after transfection of the indicated NFE2L1-V5/His genes into HEK293 cells, cells were treated with bortezomib (Bor, 1 μM) or vehicle (con) for 6 h. A bortezomib-treated group was cultured for an additional 6 h with CHX (20 μg/mL). The expression of the indicated proteins was measured as described in Materials and Methods. Each value presented is the mean ± S.E.M. from four independent cultures. The amount of each high molecular weight NFE2L1 protein in wild-type NFE2L1 expressing cells just after pretreatment with bortezomib was set at 1 (upper graph). The amount of each high molecular weight NFE2L1 protein in each NFE2L1 expressing cells just after pretreatment with bortezomib was respectively set at 1 (lower graph). Only the non-ER-localized high molecular weight form (II) and cleaved form (III) derived from wild-type NFE2L1 in proteasome deficiency are indicated. Filled and open arrowheads, respectively, indicate the full-length and cleaved CREB3L2.

DISCUSSION

NFE2L1 is a member of the CNC-bZIP family of transcription factors that contains highly conserved CNC and basic leucine zipper domains.^19,20) NFE2L1 is reported to be involved in the induction of several proteasome subunit genes to protect against proteasome defects.^20,21) NFE2L1 is a transcription factor localized on the ER membrane, and its protein level has been reported to be regulated by the ERAD machinery.^14,20,22,23) In particular, Hrd1 and VCP dependency has been reported in the retro-translocation of NFE2L1 into the cytoplasm.^14,22,23) Furthermore, several studies have reported the mechanisms for NFE2L1 transcriptional activation, including the recognition of ubiquitination in the cytoplasm, deglycosylation by NGLY-1, and protein cleavage by DDI2.^24–26,43) However, the roles of other ERAD-related factors in the regulation of the NFE2L1 protein remain incompletely understood. Recently, we established SEL1L-deficient HEK293 cells and elucidated the degradation of other ER-localized CREB3/ATF6 family of transcription factors.¹⁵⁾ We also reported a role for the EDEM2-interacting protein TXNDC11, which is involved in the mannose trimming and degradation of CREB3/ATF6 family proteins.^13,15,18) ATF6, CREB3/Luman, and CREB3L2 have similar structures and are stabilized by SEL1L and TXNDC11 deficiency, but their degradation rates are markedly different.^13,17) Our comprehensive MS analysis also suggested that SEL1L deficiency increases NFE2L1 in addition to the CREB3/ATF6 family.^13,15) Therefore, based on previous findings,^13,15,16) we analyzed NFE2L1 as another factor using various inhibitors and cells lacking several ERAD-related factors.

In this study, NFE2L1 mRNA was detectable in unstimulated wild-type HEK293 cells, but NFE2L1 protein was barely detected (Fig. 1, Supplementary Fig. 1). However, detection of NFE2L1 following long exposure showed that the size of the faint band was different from that detected following MG132 treatment (Supplementary Fig. 2). It has been reported that asparagine residues in cytosolic NFE2L1 are converted to aspartic acid during deglycosylation by NGLY-1.^24,25) In parallel, its N-terminal region is cleaved by the aspartic protease DDI2.^20,26,43) Considering the molecular weights of each molecule, the bands observed with MG132 treatment are thought to originate from these two processes. This is consistent with the results that cell lysates from wild-type cells and Hrd1-deficient cells under unstimulated condition were sensitive to Endo H, while lysates from MG132-treated wild-type cells were resistant (Fig. 3B, Supplementary Fig. 2). On the other hand, ER-resident NFE2L1 is not affected by EDEM2 or TXNDC11 deficiency (Fig. 3C), suggesting that the N-glycosylation on NFE2L1 was not associated with the ER protein quality control mechanism mediated by EDEM2/TXNDC11. Therefore, N-glycosylation on NFE2L1 may have some other role within the ER lumen. It is also interesting that CREB3L2 in Hrd1-deficient cells shifted downward 8 h after Tm treatment, while NFE2L1 expression remained unchanged (Fig. 3A). Since NGLY-1 and DDI2 are cytosolic factors, the retro-translocation of mature NFE2L1 from the ER to the cytoplasm appears to be strongly dependent on Hrd1. This contrasts with the rapid processing of full-length CREB3L2. Supporting this idea, NFE2L1 expression in Hrd1-deficient cells remained high even after 24 h treatment with CHX, though CREB3L2 completely disappeared (Supplementary Fig. 3). Treatment of Hrd1-deficient cells with Tm for 24 h resulted in a decrease in high molecular weight NFE2L1 in parallel to increase in low molecular form. In contrast, Tm caused the disappearance of glycosylated full-length CREB3L2 even in Hrd1-deficient cells, while its unglycosylated form increased remarkably. These findings suggest that Hrd1 dependency for the degradation of ER-localized glycosylated NFE2L1 is stronger than that for glycosylated CREB3L2. Low molecular weight NFE2L1 and CREB3L2 are both thought to be unglycosylated newly synthesized forms, but the mechanism by which they accumulate in Hrd1-deficient cells, and whether they were transported to the ER membrane or remained in the cytoplasm, are unclear.

Multiple Ub ligases exist on the ER membrane, and the expression levels of ER-resident proteins are controlled by the balance between ubiquitination and deubiquitination.^{3,29,43–45)} In the Ub ligases and deubiquitinase (DUB) examined in this study, it is reported that gp78 protein is not only regulated by Hrd1,⁴⁴⁾ but also co-operate with VCP.^35,45) It has also been reported that RNF5 directly binds to Hrd1 and promotes its degradation by the proteasome.²⁸⁾ Furthermore, it has been reported that USP19, a deubiquitinase, is involved not only in NFE2L1 expression but also in Hrd1 expression.^29,34) Therefore, it was expected that constitutive deficiency of gp78, RNF5 or USP19 would affect the expression of NFE2L1 or CREB3L2; however, we did not observe any relationships (Figs. 2B, 2C, 4). It is unclear why gp78 and USP19 deficiencies did not have the expected effects. These results may be due to differences in cell types and culture conditions. Furthermore, the difference between transient knockdown by shRNA and constitutive deficiency by CRISPR/Cas9 may be contributing to these discrepancies. In other words, while these factors involved in ERAD typically act cooperatively, the ERAD machinery may have adapted to the constitutive deficiency of individual ERAD component by forming different ERAD complexes. In the future, clarifying the relationships between each ER-resident Ub ligase/DUB and their substrate specificity by several approaches (e.g., conditional knockdown and overexpression) will lead to a deeper understanding of protein homeostasis in the ER, including NFE2L1 and CREB3/ATF6 family.

VCP is a protein belonging to the AAA ATPase protein family and is associated with various cellular functions, including ERAD, lysosomal degradation, membrane fusion, DNA repair, and cell division.^36,37) With regard to ERAD, VCP forms a complex with SEL1L/Hrd1 on the ER membrane and retro-transports misfolded proteins from the ER to the cytoplasm.^3,4,45) Therefore, we attempted to establish CRISPR/Cas9-mediated VCP-deficient HEK293 cells, as with ERAD-related factors, but were unable to do so. This may be because VCP has diverse functions, making it difficult to establish constitutive deficient cells. On the other hand, VCP has been considered an attractive target for the treatment of cancer and neurological diseases.^36–42,46) VCP consists of an N-terminal domain (N-domain) followed by two ATPase domains (D1 and D2). The N-domain has been reported to be involved in interactions with several factors such as gp78,^35,40,42) and the two ATPase domains are target sites for the development of VCP inhibitors.^39,42) In this study, three different VCP inhibitors were used, but their effects differed not only on NFE2L1 but also on GADD153, CREB3L2, and LC3. This is thought to reflect the diversity of VCP functions and differences in the inhibition patterns of each compound. Among them, CB-5083 is the newest VCP inhibitor among three, and in addition to its effect on NFE2L1, it also significantly increased full-length CREB3L2, raising expectations for its future application in ER and Golgi apparatus stress analysis. On the other hand, mutations in the VCP gene have been found in cancer cells that show resistance to VCP inhibitors^47,48) and in rare inherited neurodegenerative diseases such as frontotemporal dementia (FTD) and amyotrophic lateral sclerosis (ALS).^31,46,49) In this study, three mutant VCPs, R155H, A232E, and E305Q/E578Q (DKO), were also tested, but only the DKO mutant showed a significant effect. The finding that the R155H and A232E VCP mutations did not affect NFE2L1 is consistent with previous reports indicating that both mutations are deeply implicated in autophagy deficiency.³¹⁾ In contrast, it is intriguing that the stabilization and processing of co-transfected V5-tagged NFE2L1 by bortezomib was remarkably disturbed by the DKO mutant VCP (Fig. 5B). However, the effect of DKO mutant VCP on NFE2L1 stabilization in the absence of bortezomib was negligible compared with SEL1L/Hrd1 deficiency or CB-5083 treatment, despite sufficient expression of DKO mutant VCP (Figs. 2, 5). At present, the reasons for these discrepancies remain unclear, but further research is needed to deepen our understanding of the ER protein quality control mechanism mediated by VCP.

We finally investigated two NFE2L1 proteins that appeared after being transported from the ER to the cytoplasm via VCP when proteasome dysfunction occurred. In this study, we observed that high molecular weight form induced by proteasome impairment disappeared more quickly with CHX treatment (Fig. 6A). To characterize this issue, we tried to use z-VAD-fmk (z-VAD) and nelfinavir (Nelf). z-VAD and Nelf are, respectively, developed as a pan-caspase inhibitor and an aspartyl protease inhibitor against HIV treatment, but they are recently reported to inhibit NGLY-1 and DDI2.^50,51) However, the effects of each compound on NFE2L1 were unclear, and Nelf alone induced CREB3L2 cleavage (Supplementary Fig. 4). Therefore, we constructed three NFE2L1 genes with mutations in the DDI2 cleavage and/or N-glycosylated sites and evaluated their processing and stabilities (Fig. 6B). In cells expressing NFE2L1 with mutated DDI2 cleavage (mDDI2), the high molecular weight form was markedly increased, but the reduction rate following CHX treatment was comparable to that of the wild-type. This suggests that DDI2-mediated cleavage of high molecular weight form does not contribute to the faster decrease observed in high molecular weight form. Recently, Chavarria et al. reported that the N-terminal structure of NFE2L1 plays an important role in ER transport, RAD23-dependent ubiquitination, and DDI2-mediated cleavage by transfecting various types of NFE2L1 constructs.⁴³⁾ It is possible that the approximately 100 amino acids in its N-terminal region may also be involved in the stability after retro-translocation. On the other hand, the high molecular weight form in cells expressing the hypoglycosylated NFE2L1 (Q4) tended to be more unstable regardless of the DDI2 cleavage site mutation. When testing the transcriptional activity of wild-type NFE2L1 and hypoglycosylated NFE2L1 (Q4) using a luciferase reporter with an antioxidant response element (ARE), the transcriptional activity by the Q4 mutant tended to be weaker than that of the wild-type (Supplementary Fig. 5). This result is consistent with the finding of Zhang et al., who reported a decrease in NFE2L1 activity when seven asparagine residues were replaced with glutamine.⁵²⁾ Tachida et al. previously reported that human NFE2L1 genes replaced nine asparagine residues with glutamine or aspartic acid changes in their processing and SDS-PAGE migration.²⁵⁾ However, the stability of each mutant NFE2L1 protein has not been fully investigated yet. Therefore, our finding using hypoglycosylated form (Q4) is expected to be useful for further analysis. In particular, it is considered important to determine whether the mutant forms abnormal structures within ER and is degraded, or whether it is unstable in the cytoplasm. In addition, it is necessary to analyze the changes in intracellular localization and transcriptional activity caused by proteasome deficiency in each mutant NFE2L1 variant.

Taken together, under the unstimulated condition, NFE2L1 was found to be degraded as rapidly as CREB3/ATF6 family proteins. However, NFE2L1 was significantly stabilized when Hrd1 expression was abolished but not when EDEM2/TXNDC11 expression was abolished unexpectedly (Fig. 7). These results suggest that N-glycosylation in NFE2L1 might play other roles in protein maturation, transport, and degradation inside and outside the ER,⁵³⁾ but the details remain unclear. On the other hand, this study using cells deficient for gp78,²⁷⁾ RNF5,²⁸⁾ and USP19²⁹⁾ as Hrd1-related factors unexpectedly showed no effect at all on the ER protein expression we tested. It is already known that several other Ub ligases are localized in the ER membrane.^3,45) In addition, it has been demonstrated that UFMylation and ER-phagy are also involved in ER protein quality control.^54,55) Therefore, future studies on the recognition and degradation mechanisms of ER-resident transcription factors, including the NFE2L1 and CREB3/ATF6 family, will be helpful in elucidating the diverse protein quality control and signaling machineries in the ER under pathophysiological conditions.

Fig. 7. Schematic Representation of ER-Resident NFE2L1 Activation and Degradation

Newly synthesized NFE2L1(I*) undergoes glycosylation within the ER to become the mature form (I). Under normal conditions, matured NFE2L1 (I) is retro-translocated into cytosol, ubiquitinated by SEL1L/Hrd1, and degraded by the proteasomal pathway. Unlike the CREB3/ATF6 family, the degradation of NFE2L1 does not depend on EDEM2/TXNDC11. When proteasome dysfunction occurs, NFE2L1 also moves to the cytoplasm. NGLY-1 removes sugar chains from NFE2L1 and replaces its asparagine residue (N) with aspartic acid (D), after which DDI2 cleaves it. The deglycosylated form (II) tends to be less stable than the DDI2-mediated cleaved form (III). Activated NFE2L1 translocates to the nucleus and induces transcription of oxidative stress-related and proteasome subunit genes via antioxidant response elements (AREs) in the promoter region.

Acknowledgments

The authors are deeply grateful to Prof. Nico Dantuma and Prof. Michael Ristow for providing the EGFP-tagged VCP constructs and Luciferase reporter for ARE/NRF2, respectively.

DECLARATIONS

Funding

This study was partially supported by the Koshiyama Science and Technology foundation and Grants-in-Aid from the Japan Society for the Promotion of Science (JSPS, Japan, KAKENHI, Nos. 24K11444 to K.U.-Y. and 20K21751 and 25K03045 to K.O.). This work was supported by the Deanship of Scientific Research, Vice Presidency for Graduate Studies and Scientific Research, King Faisal University, Saudi Arabia (Grant No. KFU253878).

Author Contributions

M.Y., I.N., R.M., and K.O. performed the experiments; K.O. confirmed the results; T.T. prepared the NFE2L1 construct; R.M., M.Y., K.U.Y., T.T., and K.O. confirmed and discussed the results; and K.O., M.K., and M.Y. designed and prepared the manuscript.

Conflict of Interest

The authors declare no conflict of interest.

Data Availability

The data generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

1) Ellgaard L, Helenius A. Quality control in the endoplasmic reticulum. Nat. Rev. Mol. Cell Biol., 4, 181–191 (2003).
2) Vembar SS, Brodsky JL. One step at a time: endoplasmic reticulum-associated degradation. Nat. Rev. Mol. Cell Biol., 9, 944–957 (2008).
3) Ninagawa S, George G, Mori K. Mechanisms of productive folding and endoplasmic reticulum-associated degradation of glycoproteins and non-glycoproteins. Biochim. Biophys. Acta, Gen. Subj., 1865, 129812 (2021).
4) Hoseki J, Ushioda R, Nagata K. Mechanism and components of endoplasmic reticulum-associated degradation. J. Biochem., 147, 19–25 (2010).
5) Christianson JC, Ye Y. Cleaning up in the endoplasmic reticulum: ubiquitin in charge. Nat. Struct. Mol. Biol., 21, 325–335 (2014).
6) Fang S, Ferrone M, Yang C, Jensen JP, Tiwari S, Weissman AM. The tumor autocrine motility factor receptor, gp78, is a ubiquitin protein ligase implicated in degradation from the endoplasmic reticulum. Proc. Natl. Acad. Sci. U.S.A., 98, 14422–14427 (2001).
7) Hosokawa N, Wada I, Hasegawa K, Yorihuzi T, Tremblay LO, Herscovics A, Nagata K. A novel ER alpha-mannosidase-like protein accelerates ER-associated degradation. EMBO Rep., 2, 415–422 (2001).
8) Gnann A, Riordan JR, Wolf DH. Cystic fibrosis transmembrane conductance regulator degradation depends on the lectins Htm1p/EDEM and the Cdc48 protein complex in yeast. Mol. Biol. Cell, 15, 4125–4135 (2004).
9) Ninagawa S, Okada T, Takeda S, Mori K. SEL1L is required for endoplasmic reticulum-associated degradation of misfolded luminal proteins but not transmembrane proteins in chicken DT40 cell line. Cell Struct. Funct., 36, 187–195 (2011).
10) Jo Y, Sguigna PV, DeBose-Boyd RA. Membrane-associated ubiquitin ligase complex containing gp78 mediates sterol-accelerated degradation of 3-hydroxy-3-methylglutaryl-coenzyme A reductase. J. Biol. Chem., 286, 15022–15031 (2011).
11) Horimoto S, Ninagawa S, Okada T, Koba H, Sugimoto T, Kamiya Y, Kato K, Takeda S, Mori K. The unfolded protein response transducer ATF6 represents a novel transmembrane-type endoplasmic reticulum-associated degradation substrate requiring both mannose trimming and SEL1L protein. J. Biol. Chem., 288, 31517–31527 (2013).
12) Sun S, Shi G, Sha H, et al. IRE1α is an endogenous substrate of endoplasmic-reticulum-associated degradation. Nat. Cell Biol., 17, 1546–1555 (2015).
13) Murase R, Yamamoto A, Hirata Y, Oh-hashi K. Expression analysis and functional characterization of thioredoxin domain-containing protein 11. Mol. Biol. Rep., 49, 10541–10556 (2022).
14) Ruvkun G, Lehrbach N. Regulation and functions of the ER-associated Nrf1 transcription factor. Cold Spring Harb. Perspect. Biol., 15, a041266 (2023).
15) Oh-hashi K, Yamamoto A, Murase R, Hirata Y. Comparative analysis of CREB3 and CREB3L2 protein expression in HEK293 cells. Int. J. Mol. Sci., 22, 2767 (2021).
16) Murase R, Kato G, Oh-hashi K. Regulation of the ER-resident mannosidase EDEM2 in HEK293 cells. BPB Reports, 6, 193–199 (2023).
17) Asada R, Kanemoto S, Kondo S, Saito A, Imaizumi K. The signalling from endoplasmic reticulum-resident bZIP transcription factors involved in diverse cellular physiology. J. Biochem., 149, 507–518 (2011).
18) George G, Ninagawa S, Yagi H, Saito T, Ishikawa T, Sakuma T, Yamamoto T, Imami K, Ishihama Y, Kato K, Okada T, Mori K. EDEM2 stably disulfide-bonded to TXNDC11 catalyzes the first mannose trimming step in mammalian glycoprotein ERAD. eLife, 9, e53455 (2020).
19) Itoh K, Igarashi K, Hayashi N, Nishizawa M, Yamamoto M. Cloning and characterization of a novel erythroid cell-derived CNC family transcription factor heterodimerizing with the small Maf family proteins. Mol. Cell. Biol., 15, 4184–4193 (1995).
20) Kim HM, Han JW, Chan JY. Nuclear Factor Erythroid-2 Like 1 (NFE2L1): Structure, function and regulation. Gene, 584, 17–25 (2016).
21) Radhakrishnan SK, Lee CS, Young P, Beskow A, Chan JY, Deshaies RJ. Transcription factor Nrf1 mediates the proteasome recovery pathway after proteasome inhibition in mammalian cells. Mol. Cell, 38, 17–28 (2010).
22) Radhakrishnan SK, den Besten W, Deshaies RJ. p97-dependent retrotranslocation and proteolytic processing govern formation of active Nrf1 upon proteasome inhibition. eLife, 3, e01856 (2014).
23) Steffen J, Seeger M, Koch A, Krüger E. Proteasomal degradation is transcriptionally controlled by TCF11 via an ERAD-dependent feedback loop. Mol. Cell, 40, 147–158 (2010).
24) Lehrbach NJ, Breen PC, Ruvkun G. Protein sequence editing of SKN-1A/Nrf1 by peptide: N-glycanase controls proteasome gene expression. Cell, 177, 737–750.e15 (2019).
25) Tachida Y, Hirayama H, Suzuki T. Amino acid editing of NFE2L1 by PNGase causes abnormal mobility on SDS-PAGE. Biochim. Biophys. Acta, Gen. Subj., 1867, 130494 (2023).
26) Koizumi S, Irie T, Hirayama S, Sakurai Y, Yashiroda H, Naguro I, Ichijo H, Hamazaki J, Murata S. The aspartyl protease DDI2 activates Nrf1 to compensate for proteasome dysfunction. eLife, 5, e18357 (2016).
27) Zhang T, Xu Y, Liu Y, Ye Y. gp78 functions downstream of Hrd1 to promote degradation of misfolded proteins of the endoplasmic reticulum. Mol. Biol. Cell., 26, 4438–4450 (2015).
28) Yang Q, Chen X, Zhang Y, Hu S, Hu F, Huang Y, Ma T, Hu H, Tian H, Tian S, Ji YX, She ZG, Zhang P, Zhang XJ, Hu Y, Yang H, Yuan Y, Li H. The E3 ubiquitin ligase ring finger protein 5 ameliorates NASH through ubiquitin-mediated degradation of 3-hydroxy-3-methylglutaryl CoA reductase degradation protein 1. Hepatology, 74, 3018–3036 (2021).
29) Harada K, Kato M, Nakamura N. USP19-Mediated deubiquitination facilitates the stabilization of HRD1 ubiquitin ligase. Int. J. Mol. Sci., 17, 1829 (2016).
30) Mali P, Yang L, Esvelt KM, Aach J, Guell M, DiCarlo JE, Norville JE, Church GM. RNA-guided human genome engineering via Cas9. Science, 339, 823–826 (2013).
31) Tresse E, Salomons FA, Vesa J, Bott LC, Kimonis V, Yao TP, Dantuma NP, Taylor JP. VCP/p97 is essential for maturation of ubiquitin-containing autophagosomes and this function is impaired by mutations that cause IBMPFD. Autophagy, 6, 217–227 (2010).
32) Oh-hashi K, Kunieda R, Hirata Y, Kiuchi K. Biosynthesis and secretion of mouse cysteine-rich with EGF-like domains 2. FEBS Lett., 585, 2481–2487 (2011).
33) Kondo S, Hino SI, Saito A, Kanemoto S, Kawasaki N, Asada R, Izumi S, Iwamoto H, Oki M, Miyagi H, Kaneko M, Nomura Y, Urano F, Imaizumi K. Activation of OASIS family, ER stress transducers, is dependent on its stabilization. Cell Death Differ., 19, 1939–1949 (2012).
34) Hu S, Xiang Y, Qiu L, Wang M, Zhang Y. Activation of the membrane-bound Nrf1 transcription factor by USP19, a ubiquitin-specific protease C-terminally anchored in the endoplasmic reticulum. Biochim. Biophys. Acta Mol. Cell Res., 1869, 119299 (2022).
35) Zhong X, Shen Y, Ballar P, Apostolou A, Agami R, Fang S. AAA ATPase p97/valosin-containing protein interacts with gp78, a ubiquitin ligase for endoplasmic reticulum-associated degradation. J. Biol. Chem., 279, 45676–45684 (2004).
36) DeLaBarre B, Christianson JC, Kopito RR, Brunger AT. Central pore residues mediate the p97/VCP activity required for ERAD. Mol. Cell, 22, 451–462 (2006).
37) Pontifex CS, Zaman M, Fanganiello RD, Shutt TE, Pfeffer G. 2005 Valosin-Containing Protein (VCP): a review of its diverse molecular functions and clinical phenotypes. Int. J. Mol. Sci., 25, 5633 (2024).
38) Hanson PI, Whiteheart SW. AAA+ proteins: have engine, will work. Nat. Rev. Mol. Cell Biol., 6, 519–529 (2005).
39) Bastola P, Neums L, Schoenen FJ, Chien J. VCP inhibitors induce endoplasmic reticulum stress, cause cell cycle arrest, trigger caspase-mediated cell death and synergistically kill ovarian cancer cells in combination with Salubrinal. Mol. Oncol., 10, 1559–1574 (2016).
40) Anderson DJ, Le Moigne R, Djakovic S, et al. Targeting the AAA ATPase p97 as an approach to treat cancer through disruption of protein homeostasis. Cancer Cell, 28, 653–665 (2015).
41) Phan JM, Creekmore BC, Nguyen AT, Bershadskaya DD, Darwich NF, Mann CN, Lee EB. VCP activator reverses nuclear proteostasis defects and enhances TDP-43 aggregate clearance in multisystem proteinopathy models. J. Clin. Invest., 134, e169039 (2024).
42) Carrera Espinoza MJ, Tucker SK, Sureshkumar S, Gamble ME, Hakim NL, Orrantia S, Espitia CM, Cruickshank-Taylor AB, Wang W, Kelly KR, Carew JS, Nawrocki ST. Harnessing p97/VCP:a transformative AAA+ ATPase target for next-generation cancer therapeutics. Cancers, 17, 2945 (2025).
43) Chavarria C, Zaffalon L, Ribeiro ST, Op M, Quadroni M, Iatrou MS, Chapuis C, Martinon F. ER-trafficking triggers NRF1 ubiquitination to promote its proteolytic activation. iScience, 26, 107777 (2023).
44) Shmueli A, Tsai YC, Yang M, Braun MA, Weissman AM. Targeting of gp78 for ubiquitin-mediated proteasomal degradation by Hrd1: cross-talk between E3s in the endoplasmic reticulum. Biochem. Biophys. Res. Commun., 390, 758–762 (2009).
45) Krshnan L, van de Weijer ML, Carvalho P. Endoplasmic reticulum-associated protein degradation. Cold Spring Harb. Perspect. Biol., 14, a041247 (2022).
46) Kimonis VE, Fulchiero E, Vesa J, Watts G. VCP disease associated with myopathy, Paget disease of bone and frontotemporal dementia: review of a unique disorder. Biochim. Biophys. Acta, 1782, 744–748 (2008).
47) Bastola P, Wang F, Schaich MA, Gan T, Freudenthal BD, Chou TF, Chien J. Specific mutations in the D1-D2 linker region of VCP/p97 enhance ATPase activity and confer resistance to VCP inhibitors. Cell Death Discov., 3, 17065 (2017).
48) Wang F, Li S, Rosencrans WM, Cheng KW, Stott GM, Mroczkowski B, Chou TF. Sulforaphane is synergistic with CB-5083 and inhibits colony formation of CB-5083-resistant HCT116 cells. ChemMedChem, 17, e202200030 (2022).
49) Shaw CE. Capturing VCP: another molecular piece in the ALS jigsaw puzzle. Neuron, 68, 812–814 (2010).
50) Tomlin FM, Gerling-Driessen UIM, Liu YC, Flynn RA, Vangala JR, Lentz CS, Clauder-Muenster S, Jakob P, Mueller WF, Ordoñez-Rueda D, Paulsen M, Matsui N, Foley D, Rafalko A, Suzuki T, Bogyo M, Steinmetz LM, Radhakrishnan SK, Bertozzi CR. Inhibition of NGLY1 inactivates the transcription factor Nrf1 and potentiates proteasome inhibitor cytotoxicity. ACS Cent. Sci., 3, 1143–1155 (2017).
51) Fassmannová D, Sedlák F, Sedláček J, Špička I, Grantz Šašková K. Nelfinavir inhibits the TCF11/Nrf1-mediated proteasome recovery pathway in multiple myeloma. Cancers, 12, 1065 (2020).
52) Zhang Y, Lucocq JM, Hayes JD. The Nrf1 CNC/bZIP protein is a nuclear envelope-bound transcription factor that is activated by t-butyl hydroquinone but not by endoplasmic reticulum stressors. Biochem. J., 418, 293–310 (2009).
53) Yoshida Y, Takahashi T, Ishii N, Matsuo I, Takahashi S, Inoue H, Endo A, Tsuchiya H, Okada M, Ando C, Suzuki T, Dohmae N, Saeki Y, Tanaka K, Suzuki T. Sugar-mediated non-canonical ubiquitination impairs Nrf1/NFE2L1 activation. Mol. Cell, 84, 3115–3127.e11 (2024).
54) Liang JR, Lingeman E, Luong T, Ahmed S, Muhar M, Nguyen T, Olzmann JA, Corn JE. A genome-wide ER-phagy screen highlights key roles of mitochondrial metabolism and ER-resident UFMylation. Cell, 180, 1160–1177.e20 (2020).
55) Chino H, Mizushima N. ER-phagy: quality control and turnover of endoplasmic reticulum. Trends Cell Biol., 30, 384–398 (2020).

責任著者(Corresponding author)

J-STAGEへの登録はこちら（無料）