The ER membrane protein complex promotes biogenesis of sterol-related enzymes maintaining cholesterol homeostasis

ABSTRACT The eukaryotic endoplasmic reticulum (ER) membrane contains essential complexes that oversee protein biogenesis and lipid metabolism, impacting nearly all aspects of cell physiology. The ER membrane protein complex (EMC) is a newly described transmembrane domain (TMD) insertase linked with various phenotypes, but whose clients and cellular responsibilities remain incompletely understood. We report that EMC deficiency limits the cellular boundaries defining cholesterol tolerance, reflected by diminished viability with limiting or excessive extracellular cholesterol. Lipidomic and proteomic analyses revealed defective biogenesis and concomitant loss of the TMD-containing ER-resident enzymes sterol-O-acyltransferase 1 (SOAT1) and squalene synthase (SQS, also known as FDFT1), which serve strategic roles in the adaptation of cells to changes in cholesterol availability. Insertion of the weakly hydrophobic tail-anchor (TA) of SQS into the ER membrane by the EMC ensures sufficient flux through the sterol biosynthetic pathway while biogenesis of polytopic SOAT1 promoted by the EMC provides cells with the ability to store free cholesterol as inert cholesteryl esters. By facilitating insertion of TMDs that permit essential mammalian sterol-regulating enzymes to mature accurately, the EMC is an important biogenic determinant of cellular robustness to fluctuations in cholesterol availability. This article has an associated First Person interview with the first author of the paper.

(C) Comparison of EMC sedimentation as in (A) from LMNG lysates of WT and EMC2-depleted cells. The mature EMC (fractions 7-9, blue) and putative assembly intermediates (fractions 4-6, red) are indicated and reflected by changes to sedimentation of EMC1/5/6/10.
(D) Transcript levels of non-targeted EMC subunits determined by qRT-PCR from U2OS Flp-In TM TRex TM cells depleted of EMC6 relative to non-targeting control (NTC)-treated cells as described in Figure 1B. Means ± S.D. (n = 3) are shown and significance determined by Students t-test: *p ≤ 0.05, ***p ≤ 0.001.

(E -G) Knockout of EMC5 and EMC6 by CRISPR/Cas9.
Western blots of WCL from selected U2OS Flp-In TM TRex TM clones knocked out for EMC5 (E, ∆5) or EMC6 (F, ∆6) by CRISPR/Cas9 using different sgRNAs and probed with antibodies to the indicated EMC subunits and tubulin (TUB). Colored circles indicate clone identity throughout the study. (G) Genomic sequencing of individual ∆5, ∆6 and ∆SQS clones used in this study.
(N) Expression of GFP-P2A-RFP-3xFLAG-SOAT1 reporter in U2OS Flp-In TM TRex + EV cells. Cells were treated as described in Figure 4F and whole-cell lysates separated by SDS-PAGE.

Journal of Cell Science • Supplementary information
(D -F) Cholesterol auxotrophy results from loss of SQS activity.

(G -K) SQS expression is compromised in EMC-deficient cells.
(G) Immunofluorescence images of endogenous SQS in WT and EMC-deficient cells collected by confocal microscopy. Scale bar = 10 µm.  J. Cell Sci.: doi:10.1242/jcs.223453: Supplementary information Normalized H/L and M/L ratios were converted to a log2 scale and plotted against one another. Dashed lines demarcate the log2 = 0.5 threshold (corresponding to ≥30% change). Identified proteins are represented as single points, proteins decreased by ≥30% are annotated with gene names, and SQS is highlighted in red. Proteins not consistently reduced by ≥30% in all cell lines are shown in brackets. (E) WT cells treated with avasimibe (AVA, 1 µM) for the indicated times with western blots of WCLs probed with antibodies to SQS and SM.

Journal of Cell Science • Supplementary information
(F) WT cells grown in 5% FCS or 5% LPDS were treated with AVA (5 µM, 96 h) and visualized by crystal violet staining. Scale bar = 100 µm.
(I) Pulse-chase assay performed as in Figure 6D with cells expressing either HA:SQS opsin or an HA:SQS opsin whose TMD was replaced with the TMD of Sec61β (HA:SQS-Sec61βTMD opsin ).
(B) In silico prediction of SOAT1 TMD hydrophobicity using the Kyte & Doolittle (K-D) method and calculated for the TMD only (light green) or using a sliding window (WINDOW = 5 aa, dark green) by ProtScale (https://web.expasy.org/protscale/). The calculated K-D value of the SQS TMD is indicated (dashed line). SOAT1 contains several TMDs whose K-D vales are lower (TM4, 5) or comparable (TM2,7,9) to that of the SQS TMD.