The TRAPP complex mediates secretion arrest induced by stress granule assembly
2019; Springer Nature; Volume: 38; Issue: 19 Linguagem: Inglês
10.15252/embj.2019101704
ISSN1460-2075
AutoresFrancesca Zappa, Cathal Wilson, Giuseppe Di Tullio, Michele Santoro, Piero Pucci, Maria Monti, Davide D’Amico, Sandra Pisonero‐Vaquero, Rossella De Cegli, Alessia Romano, Moin A. Saleem, Elena Polishchuk, Mario Failli, Laura Giaquinto, Maria Antonietta De Matteis,
Tópico(s)Endoplasmic Reticulum Stress and Disease
ResumoArticle20 August 2019free access Source DataTransparent process The TRAPP complex mediates secretion arrest induced by stress granule assembly Francesca Zappa Corresponding Author [email protected] orcid.org/0000-0003-1263-3828 Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Cathal Wilson Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Giuseppe Di Tullio Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Michele Santoro Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Piero Pucci Federico II University, Naples, Italy Search for more papers by this author Maria Monti Federico II University, Naples, Italy Search for more papers by this author Davide D'Amico Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Sandra Pisonero-Vaquero Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Rossella De Cegli Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Alessia Romano Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Moin A Saleem Bristol Renal, Bristol Medical School, University of Bristol, Bristol, UK Search for more papers by this author Elena Polishchuk Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Mario Failli Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Laura Giaquinto Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Maria Antonietta De Matteis Corresponding Author [email protected] orcid.org/0000-0003-0053-3061 Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Federico II University, Naples, Italy Search for more papers by this author Francesca Zappa Corresponding Author [email protected] orcid.org/0000-0003-1263-3828 Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Cathal Wilson Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Giuseppe Di Tullio Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Michele Santoro Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Piero Pucci Federico II University, Naples, Italy Search for more papers by this author Maria Monti Federico II University, Naples, Italy Search for more papers by this author Davide D'Amico Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Sandra Pisonero-Vaquero Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Rossella De Cegli Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Alessia Romano Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Moin A Saleem Bristol Renal, Bristol Medical School, University of Bristol, Bristol, UK Search for more papers by this author Elena Polishchuk Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Mario Failli Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Laura Giaquinto Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Search for more papers by this author Maria Antonietta De Matteis Corresponding Author [email protected] orcid.org/0000-0003-0053-3061 Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy Federico II University, Naples, Italy Search for more papers by this author Author Information Francesca Zappa *,1, Cathal Wilson1, Giuseppe Di Tullio1, Michele Santoro1, Piero Pucci2, Maria Monti2, Davide D'Amico1, Sandra Pisonero-Vaquero1, Rossella De Cegli1, Alessia Romano1, Moin A Saleem3, Elena Polishchuk1, Mario Failli1, Laura Giaquinto1 and Maria Antonietta De Matteis *,1,2 1Telethon Institute of Genetics and Medicine, Pozzuoli (Naples), Italy 2Federico II University, Naples, Italy 3Bristol Renal, Bristol Medical School, University of Bristol, Bristol, UK *Corresponding author. Tel: +1 8058 375229; E-mail: [email protected] *Corresponding author. Tel: +39 8119 230620; E-mail: [email protected] EMBO J (2019)38:e101704https://doi.org/10.15252/embj.2019101704 [The copyright line of this article was changed on 2 September 2019 after original online publication.] PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract The TRAnsport Protein Particle (TRAPP) complex controls multiple membrane trafficking steps and is strategically positioned to mediate cell adaptation to diverse environmental conditions, including acute stress. We have identified the TRAPP complex as a component of a branch of the integrated stress response that impinges on the early secretory pathway. The TRAPP complex associates with and drives the recruitment of the COPII coat to stress granules (SGs) leading to vesiculation of the Golgi complex and arrest of ER export. The relocation of the TRAPP complex and COPII to SGs only occurs in cycling cells and is CDK1/2-dependent, being driven by the interaction of TRAPP with hnRNPK, a CDK substrate that associates with SGs when phosphorylated. In addition, CDK1/2 inhibition impairs TRAPP complex/COPII relocation to SGs while stabilizing them at ER exit sites. Importantly, the TRAPP complex controls the maturation of SGs. SGs that assemble in TRAPP-depleted cells are smaller and are no longer able to recruit RACK1 and Raptor, two TRAPP-interactive signaling proteins, sensitizing cells to stress-induced apoptosis. Synopsis The TRAnsport Protein Particle (TRAPP) complex, which controls multiple membrane trafficking steps along the secretory pathway, mediates cellular adaptation to acute stress by associating with and driving the recruitment of the COPII coat to stress granules (SGs), thereby halting secretion and possibly limiting energy consumption. The TRAPP complex associates with and drives the recruitment of the COPII coat to SGs. The TRAPP complex promotes the maturation of SGs. Relocation of TRAPP complex and COPII to SGs depends on CDK1/2 and occurs only in cycling cells. Association of TRAPP complex and COPII with SGs halts secretion. Introduction The TRAPP (TRAnsport Protein Particle) complex is a conserved multimolecular complex intervening in multiple segments of membrane trafficking along the secretory, the endocytic, and the autophagy pathways (Kim et al, 2016). Originally identified in yeast as a tethering factor acting in ER-to-Golgi trafficking, it was subsequently discovered to act as a GEF for Ypt1 and Ypt31/32 in yeast and for Rab1 and possibly Rab11 in mammals (Yamasaki et al, 2009; Westlake et al, 2011; Zou et al, 2012). The TRAPP complex has a modular composition and is present as two forms in mammals: TRAPPII and TRAPPIII which share a common heptameric core (TRAPPC1, TRAPPC2, TRAPPC2L, TRAPPC3, TRAPPC4, TRAPPC5, TRAPPC6) and additional subunits specific for TRAPPII (TRAPPC9/TRAPPC10) or for TRAPPIII (TRAPPC8, TRAPPC11, TRAPPC12, TRAPPC13; Sacher et al, 2018). TRAPPII has been implicated in late Golgi trafficking while TRAPPIII has a conserved role in the early secretory pathway (ER-to-Golgi) and in autophagy (Yamasaki et al, 2009; Scrivens et al, 2011). Recent lines of evidence have expanded the range of activities of the TRAPP complex by showing that it takes part in a cell survival response triggered by agents that disrupt the Golgi complex (Ramírez-Peinado et al, 2017) and can drive the assembly of lipid droplets in response to lipid load (Li et al, 2017). The importance of the TRAPP complex in humans is testified by the deleterious consequences caused by mutations in genes encoding distinct TRAPP subunits. Mutations in TRAPPC2L, TRAPPC6A, TRAPPC6B, TRAPPC9, and TRAPPC12 cause neurodevelopmental disorders leading to intellectual disability and dysmorphic syndromes (Khattak & Mir, 2014; Milev et al, 2017, 2018; Harripaul et al, 2018; Mohamoud et al, 2018), mutations in TRAPPC11 lead to ataxia (Koehler et al, 2017), and mutations in TRAPPC2 lead to the spondyloepiphyseal dysplasia tarda (SEDT; Gedeon et al, 1999). Spondyloepiphyseal dysplasia tarda is characterized by short stature, platyspondyly, barrel chest, and premature osteoarthritis that manifest in late childhood/prepubertal age. We have shown that pathogenic mutations or deletion of TRAPPC2 alter the ER export of procollagen (both type I and type II) and that TRAPPC2 interacts with the procollagen escort protein TANGO1 and regulates the cycle of the GTPase Sar1 at the ER exit sites (ERES; Venditti et al, 2012). The Sar1 cycle in turn drives the cycle of the COPII coat complex, which mediates the formation of carriers containing neo-synthesized cargo to be transported to the Golgi complex. While this role of TRAPPC2 in the ER export of PC might explain the altered ECM deposition observed in patient's cartilage (Tiller et al, 2001; Venditti et al, 2012), it leaves the late onset of the disease signs unexplained. We hypothesized that the latter could be due to an inability to maintain long-term cartilage tissue homeostasis, possibly due to an impaired capacity of chondrocytes to face the physiological stresses that underlie and guide their development and growth. These include mechanical stress that can induce oxidative stress leading to apoptosis (Henrotin et al, 2003; Zuscik et al, 2008). Here, by analyzing the cell response to different stresses, we show that TRAPPC2 and the entire TRAPP complex are components of stress granules (SGs), membrane-less organelles that assemble in response to stress (Protter & Parker, 2016). The recruitment of TRAPP to SGs has multiple impacts on the stress response as it induces the sequestration of Sec23/Sec24 (the inner layer of the COPII complex) onto SGs thus inhibiting trafficking along the early secretory pathway, leads to the inactivation of the small GTPase Rab1 with a consequent disorganization of the Golgi complex, and is required for the recruitment of signaling proteins, such as Raptor and RACK1, to SGs thus contributing to the anti-apoptotic role of SGs. Interestingly, TRAPP and COPII recruitment to SGs only occurs in actively proliferating cells and is under control of cyclin-dependent kinases (CDK 1 and 2). The TRAPP complex thus emerges as a key element in conferring an unanticipated plasticity to SGs, which adapt their composition and function to cell growth activity, reflecting the ability of cells to adjust their stress response to their proliferation state and energy demands. Results TRAPP redistributes to SGs in response to different stress stimuli To investigate the involvement of TRAPPC2 in the stress response, we exposed HeLa cells (Fig 1A), chondrocytes, fibroblasts or U2OS cells (Fig EV1A) to sodium arsenite (SA), a treatment that induces oxidative stress. SA treatment led to dissociation of TRAPPC2 from ERES, with which it associates under steady-state conditions (Venditti et al, 2012), with a complete relocation to roundish structures. Since oxidative stress is known to lead to the formation of stress granules (SGs; Anderson & Kedersha, 2002), we considered the possibility that these TRAPPC2-positive structures might be SGs. Indeed, co-labeling with an anti-eIF3 antibody, a canonical SG marker (Aulas et al, 2017), showed the co-localization of TRAPPC2 with eIF3 after SA treatment (Fig 1A). Interestingly, two different disease-associated mutant forms of TRAPPC2 (D47Y, R90X; Choi et al, 2009; Venditti et al, 2012) exhibited reduced or no association with SGs, respectively (Fig EV1B). Figure 1. The TRAPP complex is recruited to stress granules Localization of TRAPP complex components after sodium arsenite (SA) treatment in HeLa cells treated with 300 μM SA for 30 min. Fluorescence microscopy of fixed cells using antibodies against TRAPPC2 and eIF3 (to label SGs) at steady state and after SA treatment (top two rows). Other panels show localization of GFP-TRAPPC3 and endogenous TRAPPC1 to eIF3-labeled SGs and of endogenous TRAPPC9 and TRAPPC8 to G3BP-labeled SGs. DAPI (blue). Scale bar, 10 μm. Representative images of a time course analysis of TRAPPC2 redistribution to SGs. Cells were treated as in (A). The graph (B') shows quantification of TRAPPC2 localization at SGs over time as the ratio between TRAPPC2 (mean fluorescence intensity) in SG puncta and cytosolic TRAPPC2. Mean ± s.e.m. n = 50 cells per experiment, N = 3. Scale bar, 10 μm. Analysis of TRAPPC2 (green), TRAPPC8 (blue), and TRAPPC9 (orange) localization at SGs upon TRAPPC2, TRAPPC4, TRAPPC8, and TRAPPC9 depletion. Mean ± s.e.m. n > 100. *P < 0.05; ns: not significant, one-way ANOVA with Dunnett's multiple comparison test. Schematic representation of MS/MS analysis of TRAPPC2 interactors. In addition to the TRAPP components and known TRAPPC2 interactors, 53 RNA binding proteins known to associate with SGs were co-IP with TRAPPC2. Proteins that associate with TRAPPC2 only upon SA treatment are highlighted in red. Analysis of TRAPPC2 localization at SGs upon CLIC1, CRYAB, and ENO1 depletion. Mean ± s.e.m. n > 100. ****P < 0.0001; ns: not significant, one-way ANOVA with Dunnett's multiple comparison test. Download figure Download PowerPoint Click here to expand this figure. Figure EV1. Re-localization of TRAPP to SGs is independent of stress stimulus and cell type A rat chondrosarcoma cell line (RCS), human fibroblasts (HF), or U2OS cells were treated with SA (300 μM 60 min for RCS and U2OS, 500 μM 180 min for HF) and immunostained for TRAPPC2 and eIF3. HeLa cells transfected with 3XFLAG-tagged wild-type or mutant forms of TRAPPC2 treated with SA (300 μM 30 min) were immunostained with anti-FLAG and anti-G3BP antibodies. HeLa cells exposed to heat shock (44°C, 45 min) immunostained for TRAPPC2 and eIF3. Stress-induced localization of Bet3 (TRAPPC3) to SGs in yeast cells. Cells co-expressing Bet3-GFP and the stress granule marker RFP-PAB1 were exposed to heat shock (46°C, 10 min) and imaged. See also Movies EV1 and EV2. Association of TRAPP with SGs is reversible after SG disassembly. Cells were treated with SA for 30 min, the SA was washed out, and cells were left to recover for 120 min and processed for staining. Extracts from untreated and SA-treated HeLa cells were fractionated by size exclusion chromatography on a Superose6 column. Fractions were analyzed by Western blot using antibodies against TRAPPC2 and TRAPPC3 (components of both TRAPPII and TRAPPIII complexes), TRAPPC10 (a component of the TRAPPII complex), and TRAPPC12 (a component of the TRAPPIII complex). Arrows indicate protein standards: thyroglobulin (670 kDa), γ-globulin (158 kDa), ovalbumin(44 kDa). Volcano plots showing proteins IP with anti-FLAG Ab from TRAPPC2-FLAG or FLAG-transfected cells under control conditions (left) or upon SA treatment (right). The proteins with (P) < 0.05 and a log2 fold TRAPPC2-FLAG vs FLAG ratio > 0.6 (Student's t-test) were considered TRAPPC2 interactors and are labeled in red (TRAPP complex components), green (RNA binding proteins), and black (other interactors). Data information: (A–C, E) Scale bars, 10 μm. (D) Scale bar, 5 μm. Source data are available online for this figure. Download figure Download PowerPoint Other components of the TRAPP complex, such as TRAPPC1, TRAPPC3, TRAPPC9 (a TRAPPII-specific component) and TRAPPC8 (a TRAPPIII-specific component), were also recruited to SGs after SA treatment (Fig 1A), thus indicating that both TRAPPII and TRAPPIII complexes can associate with SGs. TRAPP recruitment to SGs also occurs in response to heat shock (Fig EV1C). In addition, TRAPP components associate with SGs in yeast cells exposed to heat stress (Fig EV1D, and Movies EV1 and EV2), thus suggesting that this is a conserved process in evolution. Of note, the association of TRAPP with SGs is fully reversible after stress removal (Fig EV1E). We studied the kinetics of TRAPPC2 redistribution to SGs. The appearance of eIF3-positive SGs occurred 7 min after SA treatment (Fig 1B), while TRAPPC2 re-localization to SGs began 15 min after exposure to stress and gradually increased over time, with massive recruitment occurring 60 min after treatment, thus lagging behind the initial assembly of SGs (Fig 1B and B'). We observed that oxidative stress did not affect the overall integrity of the TRAPP complexes (both TRAPPII and TRAPPIII), as analyzed by gel filtration (Fig EV1F). It appears that TRAPPC2 exerts a pivotal role since TRAPPC2 depletion significantly reduced the recruitment of both TRAPP complexes to SGs (Fig 1C), as indicated by reduced TRAPPC8 and TRAPPC9 in SGs. By contrast, depletion of TRAPPC4, TRAPPC8, or TRAPPC9 did not significantly affect the recruitment of TRAPPC2 (Fig 1C). To investigate the mechanisms underlying the recruitment of TRAPPC2 to SGs, we performed a proteomics analysis of the TRAPPC2 interactors both under steady-state conditions and after SA treatment (see Materials and Methods). This analysis (Figs 1D and EV1G, and Dataset EV1) confirmed known TRAPPC2 interactors (e.g., TRAPP complex components, CLIC1, and ENO1) and revealed the presence of many RNA binding proteins (RBPs), including those with a central role in the assembly of SGs (Jain et al, 2016). Interestingly, while some of these RBPs were found among the TRAPPC2 interactors at steady state, most of them were significantly enriched among TRAPPC2 interactors upon SA exposure. These findings suggested that, as described for other components of SGs, TRAPPC2 may be recruited to growing SGs by a piggyback mechanism (Anderson & Kedersha, 2008), i.e., via the interaction with RBPs that are components of SGs. Finally, as some of the known interactors of TRAPPC2, such as CLIC1, ENO1, and CRYAB, have been described as associating with SGs (Jain et al, 2016), we investigated whether they were involved in TRAPPC2 recruitment and found that CLIC1, but not ENO1 or CRYAB, is involved in TRAPPC2 recruitment to SGs (Fig 1E). Altogether, these results led us to conclude that following stress the TRAPP complexes, through interaction of TRAPPC2 with multiple SG components, associate with SGs. TRAPPC2 is required for recruiting COPII to SGs The translocation of TRAPP to SGs prompted us to investigate whether other components of membrane trafficking machineries behaved similarly. We screened different coat complex components (COPI, COPII, clathrin adaptors, clathrin), and other cytosolic proteins associated with the exocytic and endocytic pathways. Out of the 28 proteins tested, only components of the inner layer of the COPII coat, Sec24 (in its four isoforms) and Sec23, but none of the others (Figs 2A and B, and EV2A–D) associated with SGs. Notably, the recruitment of Sec24 to SGs occurred with kinetics similar to those of the TRAPP complex (Fig 2B) and induced a marked reduction of Sec24 associated with ERES (Fig 2C). Figure 2. TRAPPC2 is required for Sec23/24 re-localization to SGs HeLa cells, untreated or treated with SA, were fixed and visualized by fluorescence microscopy using anti-Sec24C Ab, anti-eIF3 Ab, and DAPI (blue). Quantification of Sec24C redistribution to SGs over time after SA treatment [the ratio between Sec24C (mean fluorescence intensity) in SG puncta and cytosolic Sec24C]. Mean ± s.e.m. n = 50 cells per experiment, N = 3. Quantification of residual Sec24C in ERES after SA treatment. The data are expressed as percentage of steady-state values (CTRL). Mean ± s.e.m. n = 100 cells per experiment, N = 3. Representative images of TRAPPC2 localization in Sec23AB-KD and Sec24ABCD-KD cells treated with SA. Cells were fixed and visualized by fluorescence microscopy using anti-TRAPPC2 Ab, anti-G3BP Ab (to label SGs), and DAPI (blue). Quantification of TRAPPC2 redistribution to SGs after KD of the indicated Sec23 and Sec24 combinations, calculated as the ratio between TRAPPC2 (mean intensity) in SG puncta and cytosolic TRAPPC2 and expressed as % of CTRL. Mean ± s.e.m., n = 40–60 cells per experiment, N = 3. ns: not significant, One-way ANOVA with Dunnett's multiple comparison test. Representative images of Sec24C localization at SGs (stained for G3BP) in TRAPPC3-KD and TRAPPC2-KD cells treated with SA. Depletion of the entire TRAPP complex (via TRAPPC3 depletion) or of only TRAPPC2 reduces Sec24C recruitment. Graph, quantification of Sec24C at SGs, calculated as in (E). Mean ± s.e.m., n = 40–60 cells per experiment, N = 3. ****P < 0.0001, One-way ANOVA with Dunnett's multiple comparison test. Representative images of Sec24C localization at SGs (stained for G3BP) in TRAPPC2-KO cells treated with SA. Data information: (A, D, F, G) Scale bars, 10 μm. Download figure Download PowerPoint Click here to expand this figure. Figure EV2. COPII proteins, but not components of other coat complexes or other cytosolic proteins associated with the exocytic and endocytic pathways, associate with SGs HeLa cells transfected with Sec23A-GFP, Sec24A-HA, Sec24B-HA, or Sec24D-YFP were treated with SA (300 μM 30 min) and processed for imaging. Proteins tested in this study that were found to associate (green tick) or not (red X) with SGs after SA treatment. Cells treated as in (A) stained with an antibody recognizing the coatomer I (COPI), which does not re-localize to SGs. Cells exposed to heat shock (44°C, 45 min) were stained for Sec24C or COPI, and G3BP. Data information: (A, C, D) Blue, DAPI, scale bar 10 μm. Download figure Download PowerPoint Since TRAPP and COPII are known to interact and TRAPP is recruited to ERES in a Sar1- and COPII-dependent manner (Lord et al, 2011; Venditti et al, 2012), we asked whether COPII and TRAPP recruitment to SGs is interdependent. The depletion of the COPII inner layer proteins Sec23 and Sec24 had no impact on the recruitment of the TRAPP complex to SGs (Fig 2D and E, Appendix Fig S1) while the depletion of the entire TRAPP complex (by KD of TRAPPC3, which destabilizes the entire complex) or of the TRAPPC2 subunit (by KD of TRAPPC2 or by TRAPPC2 KO via CRISPR-CAS9) abrogated the recruitment of COPII components to SGs (Fig 2F and G). These results indicated that the TRAPP complex, through its component TRAPPC2, drives the recruitment of COPII to SGs formed in response to acute oxidative stress. It is worth mentioning that COPII components have been reported to partition in membrane-less bodies called "Sec bodies" in Drosophila S2 cells (Zacharogianni et al, 2014). Sec bodies, which are distinct from SGs as they are devoid of RBPs, form exclusively in response to prolonged amino acid starvation and not in response to acute oxidative stress (Zacharogianni et al, 2014). The formation of Sec bodies in mammalian cells has remained unclear (Zacharogianni et al, 2014). In fact, we found that COPII components (Sec24C and Sec31A) do not form Sec bodies as they remain even more associated with ERES in amino acid-starved mammalian cells (Fig EV3A and B). We also found that Sec16, which is required for the formation of Sec bodies and SGs in response to amino acid starvation in Drosophila S2 cells (Zacharogianni et al, 2014; Aguilera-Gomez et al, 2017), neither significantly associates with SGs formed in response to stress in mammalian cells (Fig EV3C and D) nor is it required for the recruitment of COPII to these SGs (Fig EV3E and F). Click here to expand this figure. Figure EV3. Sec bodies do not form in amino acid-starved mammalian cells and Sec16 does not significantly associate with SGs nor is it required for the recruitment of COPII to SGs HeLa cells at steady state or starved for amino acids in HBSS for 8 h were stained for cTAGE5 (red) to visualize ERES and the indicated COPII components (Sec24C and Sec31) or Sec16A (green). Right panels show magnifications of the boxed areas. Human fibroblasts were treated as in (A). Cells were treated with SA (300 μM) and co-stained for endogenous Sec16 and G3BP as an SG marker. Cells exposed to heat shock (44°C, 45 min) were processed as described in (C). Representative images of Sec24C localization in Sec16-KD cells treated with SA. Cells were fixed and visualized by fluorescence microscopy using anti-Sec24C Ab, anti-G3BP Ab, and DAPI (blue). Quantification of Sec24C redistribution to SGs after KD calculated as the ratio between Sec24C (mean intensity) in SG puncta and cytosolic Sec24C. Mean ± s.e.m., n = 40–60 cells per experiment, N = 3. ns: not significant, Student's unpaired two-tailed t-test Sec16A and B mRNA levels in Sec16A+B KD cells were evaluated by qRT–PCR and expressed as fold change with respect to the MOCK. Dashed red line: MOCK. Data information: (A–E) Scale bars, 10 μm. Download figure Download PowerPoint Altogether these results indicate that COPII components can undergo two distinct phase separation events: They can associate in a TRAPP-dependent fashion with SGs in response to acute stress in mammalian cells (this report) and they can assemble in a Sec16-dependent manner in "Sec bodies" in Drosophila S2 cells in response to amino acid starvation (Zacharogianni et al, 2014). Interestingly, these two distinct phase separation events lead to different functional consequences (see below). We then investigated the nature of the association of COPII with SGs since the COPII coat cycles in a very dynamic fashion at ERES (D'Arcangelo et al, 2013). To this end, we permeabilized living cells (Kapetanovich et al, 2005) at steady state and upon stress induction. Treatment with digitonin forms pores at the plasma membrane that allows the exit of the cytosolic content into the extracellular space. Under these conditions, we found that the COPII pool recruited to SGs is less dynamic than the pool cycling at the ERES, since COPII remained associated with SGs in permeabilized cells exposed to stress while the COPII pool cycling at the ERES was completely lost upon permeabilization of unstressed cells (Fig 3A). We then assessed the possible interplay between the COPII fraction at ERES and the pool that eventually relocates to SGs. We asked whether changes in the association of COPII at the ERES could have an impact on the recruitment of COPII to SGs. To this end, we stabilized COPII and TRAPP at ERES by expressing either a constitutively active GTP-bound Sar1 mutant (Sar1-H79G) or decreasing the rate of GTP hydrolysis on Sar1 by depleting Sec31, which is a co-GAP that potentiates by an order of magnitude the Sar1 GAP activity of Sec23-24 (Bi et al, 2007). Under both conditions, COPII and TRAPP were more tightly associated with ERES, while their translocation to SGs upon exposure to stress was significantly reduced (Fig 3B–D). Thus, COPII/TRAPP recruitment to SGs is commensurate with the rate of their association at the ERES indicating that it is the cytosolic pool of COPII generated through the fast cycling at ERES that is recruited to SGs upon stress. Figure 3. Comparison and relationships between the ERES-associated and the SG-associated pools of TRAPP and COPII TRAPP and COPII associate more stably with SGs than with ERES. The membrane association of Sec24C was evaluated in non-permeabilized or permeabilized cells with or without SA treatment, as indicated. G3BP was used as an SG marker. Left panel insets, G3BP staining in non-permeabilized cells. Dashed white lines show the outline of permeabilized SA-untreated cells. Blue, DAPI. Stabilizing TRAPP and COPII at the ERES prevents their re-localization to SGs. HeLa cells overexpressing GFP-Sar1H79G were treated with SA and immunostained for Sec24C, TRAPPC2, and eIF3, as indicated. Blue, DAPI. Graphs show quantification of Sec24C or TRAPPC2 at SGs in GFP-Sar1H79G-expressing cells. Data are the ratio between Sec24C or TRAPPC2 mean intensity in SG puncta and Sec24C or TRAPPC2 mean intensity at ERES, expressed as a percentage of the non-transfected (NT) cells. Mean ± s.e.m. n = 60–70, three independent experiments. ****P < 0.0001, Student's unpaired two-tailed t-test. Effect of Sec31A depletion on TRAPPC2 recruitment to SGs. Cells were mock-treated or KD for Sec31, treated with SA, and then immunostained for TRAPPC2 and eIF3 as indicated. Graphs, quantification of TRAPPC2 at ERES and SGs after SA treatment. Mean ± s.e.m., one representative experiment; n = 60–80. **P < 0.001, ****P < 0.0001, Student's unpaired two-tailed t-test. Effect of Sec31A depletion on Sec24C recruitment to SGs. Cells were mock-treated or KD for Sec31 and then immunostained for Sec24C and cTAGE5 (to stain ERES, top panels) and for eIF3 (bottom panels). Insets show eIF3. Graphs, quantification of Sec24C at ERES and in SGs after SA treatment. n = 80. ****P < 0.0001, Student's unpaired two-tailed t-test. Data information; (A–D) Scale bars, 10 μm. Download figure Download PowerPoint The association of TRAPP/COPII with SGs is under the control of CDK1/2 To dissect the regulation of TRAPP/COPII recruitment to SGs, we screened a library of kinas
Referência(s)