A genome-wide RNA interference screen disentangles the Golgi tropism of LC3

ABSTRACT

Oleate, the most abundantly occurring cis-unsaturated fatty acid, has the particularity to induce the accumulation of MAP1LC3B/LC3 (microtubule associated protein 1 light chain 3 beta) at the trans-Golgi apparatus. A genome-wide RNA interference screen designed to identify the mechanisms of this LC3 redistribution led to the identification of a BECN1-PIK3C3-independent pathway that, however, requires the ATG12–ATG5 and ATG7-dependent conjugation system, and several genes/proteins involved in endoplasmic reticulum (ER)-to-Golgi anterograde protein transport, as well as the unfolded protein response, including the integrated stress response that results in the phosphorylation of EIF2A/eIF2α (eukaryotic translation initiation factor 2A). Functional experiments revealed that oleate blocks conventional protein secretion, stalling the process at the level of the trans-Golgi network. Oleate-induced blockade of protein secretion occurred even after depletion of ATG5, suggesting that it does not rely on the recruitment of LC3 to the Golgi apparatus (which does require ATG5). Rather, it appears that oleate and other pharmacological inhibitors of protein secretion with a similar mode of action provoke a perturbation of the trans-Golgi compartment that secondarily results in the local enrichment of LC3.

There are fundamental differences with respect to the effects of distinct classes of fatty acids on cellular and organismal physiology. Saturated fatty acids represented by palmitate, the most abundant endogenous or food-born molecule of this class, induce cellular stress including macroautophagy/autophagy through a canonical pathway that involves the ATG12–ATG5 and ATG7-dependent conjugation system, as well as the BECN1-PIK3C3-dependent pathway. Dietary cis-unsaturated fatty acids represented by oleate, the most abundant naturally occurring molecule of this category, induce signs of autophagy such as the redistribution of LC3 to cytoplasmic puncta. Intriguingly, such puncta arise in a BECN1-PIK3C3-independent (non-canonical) fashion, colocalize with the trans-Golgi network, and rarely involve two membranes. Trans-unsaturated fatty acids such as elaidate (the trans-isomer of oleate), which are particularly generated during industrial food processing, induce neither conventional autophagy nor non-conventional LC3 puncta, but rather inhibit the conventional pathway induced by saturated fatty acids, with no effect on non-conventional autophagy triggered by cis-unsaturated fatty acids.

Intrigued by these observations, which might be related to the health-improving effects of cis-unsaturated fatty acids (and the notorious toxicity of trans-unsaturated fatty acids), we decided to explore the mechanisms causing the oleate-induced relocation of LC3 to the Golgi apparatus [1]. For this, we transfected human osteosarcoma U2OS cells equipped with a green fluorescent protein (GFP)-LC3 biosensor with more than 18,000 distinct siRNAs targeting the majority of the known human mRNAs, and determined their effect on the formation of GFP-LC3 puncta elicited by either oleate or palmitate. The results of this genome-wide screen together with the post-deconvolutional validation, indicate that a series of proteins localized at the Golgi apparatus are indeed selectively required for the oleate-induced accumulation of LC3 at this organelle. These components include several structural Golgi proteins, proteins required for endoplasmic reticulum (ER)-to-Golgi anterograde protein transport (such as COPB1, COPB2 and STX5 [syntaxin 5]), as well as multiple genes associated with ER stress, including genes involved in the integrated stress response (IRS) such as EIF2A/eIF2α (eukaryotic translation initiation factor 2A) and several of its kinases. These findings were validated by using chemical inhibitors of the ER-to-Golgi protein transport (such a brefeldin A or golgicide A) and genetic systems to abolish the IRS (namely, a knockin mutation of EIF2A to render it non-phosphorylable, or the knockout of the four known EIF2A kinase genes).

We found that oleate affects the subcellular morphology of the Golgi apparatus, correlating with a blockade of conventional (Golgi-dependent) protein secretion that causes secretory cargo to be stalled at the level of the trans-Golgi network. This oleate-induced blockade of protein secretion was observed using several different experimental systems including (i) an assay involving a thermosensitive vesicular stomatitis virus G (VSVG) protein that is retained in the ER until the temperature is lowered, (ii) the lipopolysaccharide-inducible release of TNF/TNFα from activated macrophages, and (iii) a retention using selective hooks (RUSH) system involving, on the one hand, streptavidin (the hook) targeted to the ER lumen and, on the other hand, as a “bait”, GFP fused to a streptavidin-binding peptide (SBP) that can be released from streptavidin by adding biotin. This latter system was also introduced into mice after replacing GFP by Gaussia luciferase (Gluc). Hydrodynamic injection was used to simultaneously introduce the 2 plasmids into mouse hepatocytes, one coding for an ER-lumenal streptavidin (the hook, in molar excess) and the other coding for the SBP-luciferase fusion protein (the reporter). In this in vivo system, protein secretion is orchestrated by biotin (which liberates the SBP-luciferase chimera from its streptavidin hook), and the intraperitoneal administration of oleate (but not palmitate) ahead of biotin, significantly reduces protein secretion.

We next asked whether the blockade of protein secretion by oleate requires the presence of LC3 at the trans-Golgi network. However, knockout of Atg5 abolishes the oleate-induced relocation of LC3 to the Golgi apparatus without restoring protein secretion. This result suggests that an oleate-induced perturbation of the Golgi apparatus causes the relocation of LC3 to this organelle, whereas LC3 is not required for this disturbance to occur. In line with this interpretation, we have found in the past that enforced recruitment of LC3 or SQSTM1 (sequestosome 1) to the Golgi apparatus (induced by means of a chemogenetic system) does not affect the integrity of this organelle. However, we could use the phenotypic alterations induced by oleate with respect to the redistribution of LC3 (fused to red fluorescent protein, RFP) and the Golgi-associated B4GALT1 (beta-1,4-galactosyltransferase 1; fused to GFP) to screen a compound library and to identify pharmacological agents that mimic the cellular alterations induced by oleate. This screen led to the identification of several compounds including two antiprotozoal/antimalarial agents (mefloquine and quinacrine), several serotonin reuptake inhibitors (paroxetine, sertraline, as well as indatraline that also acts on dopamine and norepinephrine reuptake) and the nonselective calcium channel blocker fendiline as “oleate neighbors”. These “oleate mimetics” also inhibited conventional protein secretion, supporting the notion that this pathway of Golgi perturbation is indeed of pharmacological relevance (Figure 1).

Figure 1.

Oleate stalls Golgi-mediated protein secretion. (A) Oleate and its functional analogs, the “oleate mimetics”, perturb the Golgi morphology and function downstream of the integrated stress response, secondarily causing the redistribution of LC3 to this organelle. (B-E) Different experimental systems employed to monitor protein secretion. (B) The thermosensitive vesicular stomatitis virus G (VSVG) protein is retained in the endoplasmic reticulum (ER) at 40°C and traffics toward the plasma membrane upon temperature shift to 32°C. (C) TNF (tumor necrosis factor) secretion by bone marrow-derived macrophages (BMDM) is induced by lipopolysaccharide (LPS). (D) The retention using selective hooks (RUSH) technique is employed to sequester in the ER a GFP reporter that can be released with biotin in vitro, (E) or a Gaussia luciferase (Gluc)-reporter that is expressed in mouse hepatocytes through hydrodynamic injection and can be systemically released with biotin in vivo.

Funding Statement

GC is supported by a scholarship of the Fondation pour la Recherche Médicale. GK is supported by the Ligue contre le Cancer (équipe labellisée); Agence National de la Recherche (ANR) – Projets blancs; ANR under the frame of E-Rare-2, the ERA-Net for Research on Rare Diseases; AMMICa US23/CNRS UMS3655; Association pour la recherche sur le cancer (ARC); Association “Le Cancer du Sein, Parlons-en!”; Cancéropôle Ile-de-France; Chancelerie des universités de Paris (Legs Poix), Fondation pour la Recherche Médicale (FRM); a donation by Elior; European Research Area Network on Cardiovascular Diseases (ERA-CVD, MINOTAUR); Gustave Roussy Odyssea, the European Union Horizon 2020 Project Oncobiome; Fondation Carrefour; High-end Foreign Expert Program in China [GDW20171100085 and GDW20181100051], Institut National du Cancer (INCa); Inserm (HTE); Institut Universitaire de France; LeDucq Foundation; the LabEx Immuno-Oncology [ANR-18-IDEX-0001]; the RHU Torino Lumière; the Seerave Foundation; the SIRIC Stratified Oncology Cell DNA Repair and Tumor Immune Elimination (SOCRATE); and the SIRIC Cancer Research and Personalized Medicine (CARPEM).

Disclosure statement

OK and GK are scientific cofounders of Samsara Therapeutics. GK is a scientific cofounder of Therafast Bio.