Syntaxin 17 promotes lipid droplet formation by regulating the distribution of acyl-CoA synthetase 3

Lipid droplets (LDs) are ubiquitous organelles that contain neutral lipids and are surrounded by a phospholipid monolayer. How proteins specifically localize to the phospholipid monolayer of the LD surface has been a matter of extensive investigations. Here we show that syntaxin 17 participates in LD biogenesis by regulating the distribution of acyl-CoA synthetase 3 (ACSL3), a key enzyme for LD biogenesis that redistributes from the endoplasmic reticulum to LDs during LD formation. Time course experiments revealed that syntaxin 17 binds to ACSL3 in the initial stage of LD formation, and that ACSL3 is released as a consequence of competitive binding of SNAP23 to syntaxin 17 in the maturation stage. We propose a model in which ACSL3 redistributes from the endoplasmic reticulum to LDs through association with syntaxin 17 and SNAP23-mediated dissociation from syntaxin 17. We also provide evidence that lipid raft-like structures are important for LD formation and SNAREs-ACSL3 interactions.


Introduction 27
Lipid droplets (LD) are ubiquitous organelles that store neutral lipids such as triacylglycerol 28 (TAG) and sterol esters, and play central roles in energy and lipid metabolism (Walther and 29 Farese, 2012). LDs are dynamic and diverse organelles, their size and number depending on replication (Stordeur et al., 2014;Welte, 2015). 37 LDs are unique among cellular organelles in that they are surrounded by a phospholipid 38 monolayer. LD formation starts in the endoplasmic reticulum (ER) at pre-defined or random  Figure 3E), although the ∆GATE mutant as well as the wild-type 160 protein may be close to Stx17 in the absence of OA ( Figure 3F). These findings combined 161 with the fact that the SNARE domain of Stx17 is required for LD formation ( Figure 1E,F  Given that Stx17 regulates the redistribution of ACSL3 from the ER to LDs likely through 179 protein-protein interaction, we reasoned that other protein(s) might modulate this interaction. 180 We focused on SNAP23 because previous studies revealed the involvement of this protein in 181 lipid droplet formation and localization (Boström et al., 2007;Jägerström et al. 2009), and 182 our interactome analysis identified SNAP23 as an Stx17-interacting protein (data not shown). 183 We examined whether SNAP23, like Stx17, localizes to the MAM, in addition to the ER and 184 mitochondria. As shown in Figure 4A, SNAP23 was found to be highly enriched in the 185 MAM fraction. Of note is that a significant amount of ACSL3 was also recovered in the 186

MAM fraction. 187
Silencing of SNAP23 ( Figure 4B   Because both ACSL3 and SNAP23 interact with the SNARE domain of Stx17, we 211 sought to determine whether they bind to Stx17 in a synergic or competitive manner. When 212 SNAP23 was knocked down, the number of PLA dots representing the proximity between 213 Stx17 and ACSL3 was significantly increased ( Figure 5C), whereas ectopic expression of 214 SNAP23 reduced the number of the PLA dots ( Figure 5D). Furthermore, in cells expressing 215 GFP-ACSL3 wild-type but not mutants lacking the ability to bind Stx17, i.e., the ∆GATE 216 and ∆TMD mutants, the proximity of Stx17 to SNAP23 was disrupted ( Figure 5E). These 217 findings suggest that SNAP23 and ACSL3 compete for Stx17 binding. 218

219
The MAM, but not tethering between the MAM and mitochondria, is important for 220

LD formation 221
Next, we examined the effect of depletion of PACS-2, a multifunctional sorting protein that 222 is required for maintaining MAM integrity (Myhill et al., 2008;Simmen et al., 2005), and 223 mitofusin 2 (Mfn2), a key tether for ER-mitochondria (Naon et al., 2016). As shown in 224 Figure 6A, Mfn2 depletion did not affect OA-induced LD formation, whereas PACS-2 was 225 found to be required for LD formation. Consistent with these findings, the proximity signal 226 for FLAG-Stx17 and ACSL3 was reduced upon depletion of PACS-2, but not Mfn2, and OA 227 increased the signal ( Figure 6B). 228 We assessed the Stx17 milieu in cells depleted of PACS-2 or Mfn2 by means of 229 digitonin sensitivity. Although some punctate Stx17-positive structures were observed in 230 Mfn2-silenced cells, they were mostly abolished by digitonin treatment ( Figure 6C

Transient binding of Stx17 to ACSL3 during LD formation 238
Given that SNAP23 and ACSL3 compete for Stx17 binding, one attractive hypothesis for the 239 regulation of the redistribution of ACSL3 from the ER to nascent LDs due to Stx17 and 240 SNAP23 is that at the onset of LD formation ACSL3 first binds to Stx17, and then SNAP23 241 competitively binds to Stx17, releasing ACSL3 from Stx17 to allow its redistribution to the 242 LD surface. To test this hypothesis, we monitored the change of the binding partner during 243 LD maturation. Immunoprecipitation ( Figure 7A) and PLA ( Figure 7B) revealed that the 244 binding of ACSL3 to Stx17 was augmented at 1-hr incubation with OA and then decreased, 245 whereas the binding of Stx17 to SNAP23 increased up to 6 hr. suggesting that on site supply of acetyl-CoA is necessary for the growth of LDs. Therefore, 255 how proteins selectively bind to the LD surface, which is surrounded by a phospholipid 256 monolayer, in a maturation stage-dependent manner is one of the most key questions to be 257 addressed in LD biogenesis. In this study, we demonstrated that Stx17, a SNARE protein 258 localized in the MAM, in conjunction with another MAM protein, SNAP23, regulates the 259 translocation of ACSL3 and perhaps LPCAT1 from the ER to nascent LDs. On the other 260 hand, the distribution of some LD-localized proteins such as DGAT2 and Tip47/PLIN3, 261 which translocate from the ER and cytosol to LDs, respectively, was found to be independent 262 of Stx17. These findings suggest that Stx17 selectively interacts with and regulates the 263 localization of a subset of proteins for LD biogenesis.   (Figures 2C and 4D), immunofluorescence images 379 obtained were analyzed using ImageJ software (NIH).

Subcellular fractionation 420
Subcellular fractionation was performed as described previously (Arasaki et al., 2015). 421 Experiments were repeated two or three times with similar results. 422 423

Statistical analyses 429
The results were averaged, expressed as the mean ± SD or SEM, and analyzed using a paired