Lipid biosynthesis perturbation impairs endoplasmic reticulum-associated degradation

Abstract

The relationship between lipid homeostasis and protein homeostasis (proteostasis) is complex and remains incompletely understood. We conducted a screen for genes required for efficient degradation of Deg1-Sec62, a model aberrant translocon-associated substrate of the endoplasmic reticulum (ER) ubiquitin ligase Hrd1, in Saccharomyces cerevisiae. This screen revealed that INO4 is required for efficient Deg1-Sec62 degradation. INO4 encodes one subunit of the Ino2/Ino4 heterodimeric transcription factor, which regulates expression of genes required for lipid biosynthesis. Deg1-Sec62 degradation was also impaired by mutation of genes encoding several enzymes mediating phospholipid and sterol biosynthesis. The degradation defect in ino4Δ yeast was rescued by supplementation with metabolites whose synthesis and uptake are mediated by Ino2/Ino4 targets. Stabilization of a panel of substrates of the Hrd1 and Doa10 ER ubiquitin ligases by INO4 deletion indicates ER protein quality control is generally sensitive to perturbed lipid homeostasis. Loss of INO4 sensitized yeast to proteotoxic stress, suggesting a broad requirement for lipid homeostasis in maintaining proteostasis. A better understanding of the dynamic relationship between lipid homeostasis and proteostasis may lead to improved understanding and treatment of several human diseases associated with altered lipid biosynthesis.

Keywords: Doa10; ER quality control; Hrd1; Saccharomyces cerevisiae; endoplasmic reticulum-associated degradation (ERAD); phospholipid metabolism; protein degradation; sterol; translocon quality control; yeast genetics.

Figure 1

Screen for genes required for degradation of a model translocon-associated protein.A, schematic of Deg1-Sec62-His3 following aberrant translocon engagement. Following the integration of the two transmembrane segments of Sec62, the N-terminal tail of the fusion protein loops into and persistently engages (i.e., clogs) the translocon (14). Upon clogging, Deg1-Sec62(±His3) undergoes N-linked glycosylation and is ubiquitylated by Hrd1 and Ubc7 (which is anchored at the ER membrane by Cue1). Deg1-Sec62-His3 possesses, in sequence, Deg1 (the N-terminal 67 amino acids from the yeast transcriptional repressor MATα2), a FLAG epitope (F), Sec62, two copies of Staphylococcus aureus Protein A (2xProtA), and the His3 enzyme. Ub, ubiquitin. B, yeast of the indicated genotypes transformed with an empty vector or a plasmid encoding Deg1-Sec62-His3 were spotted onto media containing or lacking histidine (His). C, DOA10 locus of the query strain used for the genome-wide screen. DOA10 was replaced with a cassette containing Deg1-Sec62-His3 and natMX4 as two independent genes, each with its own promoter and transcriptional terminator. D, overview of the genome-wide screen. See text and Table 1 for details.

Figure 2

Ino2 and Ino4 are required for Deg1∗-Sec62 degradation.A, WT yeast or yeast lacking either HRD1 or INO4 were transformed with a plasmid encoding Deg1∗-Sec62 or an empty vector and subjected to cycloheximide chase and western blot analysis to detect Deg1∗-Sec62 and Pgk1. B, as in (A), but with WT yeast or yeast lacking either HRD1, INO4, or INO2. Means of percent Deg1∗-Sec62 remaining for three to four biological replicates are plotted. Error bars represent the SEM. Means of percent Deg1∗-Sec62 remaining at 60 min were evaluated by one-way ANOVA followed by Holm-Šídák multiple comparison tests (only pairs relative to WT yeast were compared).

Figure 3

Deg1∗-Sec62 degradation is sensitive to perturbation of lipid biosynthesis. Yeast of the indicated genotypes were transformed with a plasmid encoding Deg1∗-Sec62 or an empty vector and subjected to cycloheximide chase and western blot analysis to detect Deg1∗-Sec62 and Pgk1. Means of percent Deg1∗-Sec62 remaining for 3 to 5 biological replicates are plotted. Error bars represent the standard error of the mean. For the experiment depicted in (A), the final three lanes represent yeast supplemented with 500 μM inositol, 2 mM ethanolamine, and 2 mM choline from inoculation through cell harvest and cycloheximide chase. Means of percent Deg1∗-Sec62 remaining 60 min in (A) were evaluated by one-way ANOVA followed by Tukey’s multiple comparison test. Means of percent remaining at indicated times in (B–D) were evaluated by one-way ANOVA followed by Holm-Šídák multiple comparison tests (only pairs relative to wild-type yeast were compared). Means of percent remaining at 60 min in (E) were evaluated by an unpaired, two-tailed t test.

Figure 4

INO4 deletion impairs ERAD of Hrd1 and Doa10 substrates.A, ERAD substrates of Hrd1 and Doa10 analyzed in this figure. B–E, yeast of the indicated genotypes were transformed with a plasmid encoding indicated ERAD substrates or an empty vector and subjected to cycloheximide chase and western blot analysis to detect the ERAD substrate and Pgk1. Means of percent ERAD substrate remaining for 3 to 6 biological replicates are plotted. Error bars represent standard error of the mean. Means of percentage of ERAD substrate remaining at 60 min were evaluated by one-way ANOVA followed by Holm-Šídák multiple comparison tests (only pairs relative to wild-type yeast were compared). F, Flag; ProtA, Protein A; Ub, ubiquitin.

Figure 5

INO4 deletion does not impair degradation of a soluble, nucleoplasmic substrate of soluble ubiquitin ligase Slx5/Slx8. Yeast of the indicated genotypes were transformed with a plasmid encoding α2∗-Ura3-3HA (α2∗-UH) or an empty vector and subjected to cycloheximide chase and western blot analysis of α2∗-UH and Pgk1. Means of percent α2∗-UH remaining for four biological replicates are plotted. Error bars represent standard error of the mean. Means of percent α2∗-UH remaining at 60 min were evaluated by one-way ANOVA followed by Holm-Šídák multiple comparison tests (only pairs relative to wild-type yeast were compared).

Figure 6

Loss of INO4 reduces Ubc7-2HA abundance.Left, Yeast of the indicated genotypes were transformed with a plasmid encoding Hrd1-3HA, Ubc7-2HA, Cue1-HA, or an empty vector, harvested, lysed, and subjected to anti-HA and anti-Pgk1 western blotting. Means of steady state abundance for 3 to 4 biological replicates are plotted. Error bars represent the standard error of the mean. Means in (A and C) were evaluated by an unpaired, two-tailed t test. Means in (B) were evaluated by one-way ANOVA followed by a Holm-Šídák multiple comparison test (only pairs relative to wild-type yeast were compared).

Figure 7

INO4 deletion does not broadly impair translocation.A, electrophoretic migration of plasmid-encoded Deg1∗-Sec62 in yeast of the indicated genotypes was assessed by western blotting. B, top, Electrophoretic migration of plasmid-encoded CPY (a model post-translationally translocated protein) or OPY (a model co-translationally translocated protein) in yeast of the indicated genotypes was assessed by western blotting. Bottom, Means of the proportion of CPY or OPY that is unglycosylated (i.e., untranslocated) for three biological replicates are plotted. Error bars represent standard error of the mean. Means were evaluated by one-way ANOVA followed by Tukey’s multiple comparison test. C, lysates from ino4Δ yeast expressing Deg1∗-Sec62, CPY, or OPY were incubated in the presence or absence of Endoglycosidase H (Endo H) prior to western blotting.

Figure 8

Genetic perturbation of lipid biosynthesis sensitizes yeast to hygromycin B. Yeast of the indicated genotypes were serially diluted and spotted onto rich yeast agar medium (YPD) with or without hygromycin B. Plates were incubated at 30 °C and imaged after 1 to 2 days. Experiments were performed in triplicate (i.e., three biological replicates).