Functional integrity of the SEL1L-HRD1 complex is critical for endoplasmic reticulum-associated degradation and organismal viability

Abstract

The SEL1L-HRD1 complex is the most conserved branch of endoplasmic reticulum-associated degradation (ERAD), yet whether SEL1L is strictly required for HRD1 function in mammals has remained unclear. Here, we show, using complementary in vivo and in vitro approaches, that direct SEL1L-HRD1 binding is essential for ERAD activity and neonatal survival. Three knock-in mouse models targeting this interface reveal a clear genotype-phenotype relationship: the L709P variant, which abolishes SEL1L-HRD1 association, causes complete neonatal lethality; the partially disruptive S658P variant results in partial lethality; and the P699T mutation preserves the interaction and yields normal viability. Mechanistically, our data show that the SEL1L-HRD1 interface is essential for ERAD complex formation and activity, enabling both substrate handoff and E2 enzyme recruitment, and that the L709P mutation effectively uncouples these core steps of the mammalian ERAD pathway. These findings establish SEL1L-HRD1 coupling as a core requirement for mammalian ERAD function and early postnatal viability.

Keywords: ER quality control; ERAD; SEL1L variants; SEL1L–HRD1 interaction; neonatal lethality.

Fig. 1.

Homozygous Sel1L^L709P KI mice exhibit neonatal lethality. (A) Schematic diagram of the human SEL1L protein domain structure, highlighting the position of three variants at SLR-C. SP, signal peptide; FNII, fibronectin type II domain; SLR-N/M/C, N-, middle-, and C-terminal Sel1-like repeats; TM, transmembrane domain. (B) ClustalW sequence alignment demonstrating evolutionary conservation of residues L709 (red) and P699 (green). (C) Number and percent of mice of each genotype at postnatal (P) days 0 and 21. (D) Kaplan–Meier survival curves for neonates over the first 30 h after birth. n, mouse numbers. ****P < 0.0001 (comparing L709 or S658P KI to WT or P699T KI mice) by Log-rank (Mantel–Cox) test. (E) Body weights of WT and KI neonates at P0. Data, mean ± SEM; n.s., not significant by one-way ANOVA with Dunnett’s multiple comparisons test. (F and G) Hematoxylin and eosin (H&E) staining of WT and Sel1L^L709P KI P0 pups, with livers and pancreas shown in G. Br, brain; S, spinal cord; H, heart; Lu, lung; Li, liver; I, intestine. n = 3 mice per group.

Fig. 2.

SEL1L L709P mutation impairs ERAD function in vivo and in vitro. (A and B) Immunoblots of core ERAD components (SEL1L, HRD1, OS9) and the endogenous substrate IRE1α in livers from P0 WT, Sel1L^L709P, and Sel1L^S658P KI mice, and 2-mo-old Sel1L^f/f and Sel1L^AlbCre mice (A), with quantification of protein levels shown in (B), normalized to the loading control HSP90. Asterisks, nonspecific bands. n = 4 mice per group. (C) qPCR analyses of ERAD-related gene expression in P0 livers, normalized to L32. n = 3 mice per group. (D) Reducing and nonreducing SDS-PAGE followed by immunoblotting to detect high-molecular-weight (HMW) aggregates of proAVP-G57S in SEL1L^−/− HEK293T cells transfected with the indicated SEL1L-FLAG constructs. Quantification of HMW proAVP normalized to the loading control HSP90 shown on the Right. n = 3 independent samples. Values are mean ± SEM. Statistical comparisons are made relative to SEL1L^L709P KI. n.s., not significant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 by one-way ANOVA with Dunnett’s multiple comparisons test.

Fig. 3.

SEL1L L709P mutationis associated with mild ER stress in vivo. (A and B) Phos-tag gel analyses of phosphorylated IRE1α in liver tissues from P0 WT, Sel1L^L709P, and Sel1L^S658P KI mice. Livers from mice injected with PBS (CON) or tunicamycin (TM) were included as controls (A). Lamda phosphatase (λPP) treatment of the lysates was used to show protein phosphorylation. Total IRE1α and BiP protein levels were assessed using the standard SDS-PAGE. Quantification of BiP protein level normalized to GAPDH and the ratio of p-IRE1α in total IRE1α shown in (B). n = 4 mice per group. (C and D) RT-PCR analysis of Xbp1 mRNA splicing in P0 and control livers (as in A) with quantification of percent of Xbp1s mRNA in total Xbp1 mRNA shown in (D). n = 4 mice per group. (E) Coimmunoprecipitation (Co-IP) of endogenous SEL1L from P0 livers, followed by immunoblotting. HSP90, loading control. n = 3 mice per group. (F and G) Immunoblot analyses of PERK, total and phosphorylated eIF2α (p-eIF2α) in P0 and control livers (as in A), with quantification of protein level of PERK normalized to HSP90 and the ratio of p-eIF2α in total eIF2α shown in (G). n = 3 mice per group. Values are shown as mean ± SEM. Statistical comparisons are made relative to SEL1L^L709P. n.s., not significant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 using one-way ANOVA with Dunnett’s multiple comparisons test.

Fig. 4.

SEL1L L709P mutation causes mild ER dilation in hepatocytes and pancreatic acinar cells. (A) Representative TEM images of hepatocytes from P0 WT, Sel1L^L709P, and Sel1L^S658P livers. n = 3 mice with 30 to 40 cells per mouse. (B) Representative TEM images of P0 pancreatic acinar cells. n = 3 mice with 20 to 30 cells per mouse. N, nucleus; M, mitochondria; ER, endoplasmic reticulum (arrows); ZG, zymogen granules; Glyc, glycogen. Asterisk, dilated ER.

Fig. 5.

SEL1L L709P mutation abolishes SEL1L–HRD1 interaction in vivo and in vitro. (A and B) Dimeric OS9–SEL1L–HRD1 complex structure. A close-up view highlights the SEL1L–HRD1 interface with the location of three variants shown (B). APH, amphipathic helix of SEL1L. (C–E) Coimmunoprecipitation (Co-IP) of endogenous SEL1L from MEFs derived from WT, Sel1L^S658P, Sel1L^L709P, and Sel1L^ERCremice, followed by immunoblotting for ERAD components (C and D), with quantification shown in (E). GAPDH, a loading control. n = 3 mice per group. (F and G) Co-IP of endogenous SEL1L from livers of P0 neonates (WT, Sel1L^S658P, and Sel1L^L709P), followed by Western blot analyses of ERAD components (F), with quantification shown in (G). HSP90, a loading control. n = 3 mice per group. Values, mean ± SEM. Statistical comparisons are made relative to SEL1L^L709P. n.s., not significant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 using one-way ANOVA with Dunnett’s multiple comparisons test.

Fig. 6.

The SEL1L–HRD1 interaction is required for substrate engagement and E2 recruitment to HRD1 under basal conditions. (A and B) CHX translation shut-off assay in KI and KO HEK293T, with quantification normalized to HSP90 shown in (B). n = 3 independent samples. (C and D) IP of endogenous SEL1L in the indicated KI and KO HEK293T cells, followed by immunoblotting for ERAD components (C), with quantification normalized to SEL1L shown in (D); HSP90, a loading control. n = 3 independent samples. (E and F) IP of proAVP (G57S)-HA in transfected KI and KO HEK293T cells with quantification normalized to proAVP shown in (F). HSP90, loading control. The asterisk indicates a nonspecific band. n = 3 independent samples. (G) Coimmunoprecipitation of HRD1 in KI and KO HEK293T cells. HSP90, loading control. The asterisk indicates a nonspecific band. n = 3 independent samples. (H) Immunoblot of poly-ubiquitinated proAVP (G57S)-HA in transfected KO or KI HEK293T cells treated with 10 μM MG132 for 4 h. The quantification of AVP ubiquitination shown below the gel from three independent samples. HSP90, loading control. Values, mean ± SEM. Statistical comparisons are made relative to SEL1L^L709P. n.s., not significant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 determined by simple linear regression (B) or by using one-way ANOVA with Dunnett’s multiple comparisons test (D and F).

Fig. 7.

Model illustrating the functional and physiological roles of SEL1L–HRD1 coupling. Our data reveal a coherent structure–function–physiology relationship: the extent of SEL1L–HRD1 disruption corresponds directly to ERAD impairment and stratifies organismal survival.