In response to amino acid supply, mTORC1, a master regulator of cell growth, is recruited to the lysosome and activated by the small GTPase Rheb. However, the intracellular localization of Rheb is controversial. In this study, we showed that a significant portion of Rheb is localized on the Golgi but not on the lysosome. GFP–Rheb could activate mTORC1, even when forced to exclusively localize to the Golgi. Likewise, artificial recruitment of mTORC1 to the Golgi allowed its activation. Accordingly, the Golgi was in contact with the lysosome at an newly discovered area of the cell that we term the Golgi–lysosome contact site (GLCS). The number of GLCSs increased in response to amino acid supply, whereas GLCS perturbation suppressed mTORC1 activation. These results suggest that inter-organelle communication between the Golgi and lysosome is important for mTORC1 regulation and the Golgi-localized Rheb may activate mTORC1 at GLCSs.

The mammalian target of rapamycin complex 1 (mTORC1) constitutes a master regulator of cell growth, and responds to nutrients, growth factors and energy to regulate cellular anabolic and catabolic processes (Dibble and Manning, 2013; Laplante and Sabatini, 2012; Saxton and Sabatini, 2017; Shimobayashi and Hall, 2016). mTORC1 primarily consists of mTOR, regulatory-associated protein of mTOR (Raptor; also known as RPTOR), and mammalian lethal with SEC13 protein 8 (mLST8; also known as GβL) (Laplante and Sabatini, 2013; Shimobayashi and Hall, 2016; Zoncu et al., 2011). Active mTORC1 phosphorylates a series of substrates including ribosomal protein S6 kinase (S6K) and EIF4E-BP1, one of the eukaryotic translation initiation factor 4E-binding proteins (4E-BPs) to promote protein, lipid and nucleotide synthesis, leading to cell growth (Shimobayashi and Hall, 2016).

Amino acids activate mTORC1 through RAG proteins and the Ragulator complex (Sancak et al., 2008, 2010). The homologs RAGA or RAGB (also known as RRAGA and RRAGB; hereafter RAGA/B), form heterodimers with RAGC or RAGD (also known as RRAGC and RRAGD; hereafter RAGC/D) (Sancak et al., 2008). The RAGs localize to lysosomes by binding to the scaffold Ragulator complex, consisting of LAMTOR1 (also known as p18), LAMTOR2 (p14), LAMTOR3 (MP1), LAMTOR4 (C7orf59) and LAMTOR5 (HBXIP) (Bar-Peled et al., 2012; Sancak et al., 2010). During starvation, mTORC1 is mostly dispersed in the cytoplasm. Upon amino acid stimulation, the RAGA/B–RAGC/D complex is converted into the RAGA/B-GTP–RAGC/D-GDP form, binds to Raptor, and recruits mTORC1 to the lysosome surface (Sancak et al., 2008). Recent studies have revealed that cytosolic amino acids signal to mTORC1 through specific sensor molecules. The GATOR1 complex acts as a GTPase-activating protein (GAP) for RAGA/B, and the GATOR2 complex plays a positive role in mTORC1 regulation by binding to GATOR1 at the lysosome (Bar-Peled et al., 2013). Cytosolic leucine directly binds to the leucine sensor sestrin 2, and suppresses the sestrin-2–GATOR2 interaction to activate mTORC1 (Wolfson et al., 2015). Cytosolic arginine directly binds to the arginine sensor CASTOR1, which interacts with GATOR2 in the absence of arginine, and regulates the CASTOR1–GATOR2 interaction for the activation of mTORC1 (Chantranupong et al., 2016).

Ras homolog enriched in brain (Rheb), a small GTP-binding protein, is an mTORC1 activator that has been suggested to be localized on the lysosome (Aspuria and Tamanoi, 2004; Menon et al., 2014). In response to amino acid supply, recruitment of mTORC1 to the lysosomal membrane leads to an mTORC1–Rheb interaction, the critical mechanism for mTORC1 activation (Sancak et al., 2010). In response to insulin and growth hormones, the phosphoinositide 3-kinase (PI3K) and AKT pathway suppress the tuberous sclerosis complex (TSC), the GAP of Rheb (Dibble and Manning, 2013; Inoki et al., 2003). It has been also reported that growth factor signaling requires nutrients to activate mTORC1 (Nobukuni et al., 2005; Sancak et al., 2008). However, the lysosomal localization of Rheb remains controversial, and localization of Rheb on the plasma membrane, endoplasmic reticulum (ER), Golgi and peroxisome has also been reported (Buerger et al., 2006; Hanker et al., 2010; Jiang and Vogt, 2008; Manifava et al., 2016; Thomas et al., 2014; Zhang et al., 2013). Therefore, the cellular location where Rheb activates mTORC1 remains an enigma.

In this study, we investigated Rheb localization in the Golgi as well as the mTORC1 regulatory mechanism by which Rheb activates mTORC1 at an inter-organelle contact site between the Golgi and lysosome.

Rheb is partially localized on the Golgi

First, we investigated the localization of endogenous Rheb by performing immunofluorescence studies in HEK293T cells. In addition to dispersed signals throughout the cell, some signals were accumulated at perinuclear regions (Fig. 1A). These signals represent bona fide Rheb localization as they disappeared when Rheb expression was knocked down with siRNA (Fig. 1A). The perinuclear signals were colocalized with the Golgi marker protein GM130 (also known as GOLGA2) (Fig. 1A). These results indicate that at least a proportion of Rheb is localized on the Golgi (Manifava et al., 2016). Just like endogenous Rheb, transiently expressed GFP–Rheb showed Golgi localization, as indicated by colocalization with GM130, in HEK293T cells (Fig. 1B), and in mouse embryonic fibroblasts (MEFs) and HeLa cells (Fig. S1A,B). The trans-Golgi marker protein, adaptin γ, was also found to colocalize with GFP–Rheb (Fig. S1C). As membrane association of Rheb is dependent on its C-terminal farnesylation (Buerger et al., 2006), farnesyl-transferase inhibitor (FTI) treatment led to defective Golgi localization for both endogenous Rheb and GFP–Rheb; however, it did not affect Golgi integrity (Fig. S1D,E). These localization patterns did not change with or without the addition of amino acids and/or insulin (Fig. S1F). In contrast to Rheb, mTORC1 is thought to localize mainly to the lysosome (Manifava et al., 2016; Sancak et al., 2008), and we observed that mTOR colocalized with the lysosome marker Lamp1 under conditions of amino acid replenishment in HEK293T and MEF cells (Fig. 1C; Fig. S1G). However, we did not clearly observe colocalization of mTOR and GFP–Rheb (Fig. 1C; Fig. S1G). These results prompted us to investigate how these distinct localization patterns support their functional relationship.

Fig. 1.

Endogenous Rheb and GFP–Rheb are localized on the Golgi. (A) HEK293T cells were cultured in DMEM for 24 h and then fixed and stained for immunofluorescence with anti-Rheb antibody (1:200) and anti-GM130 (1:3200) diluted in blocking buffer. For Rheb knockdown, HEK293T cells were transfected with control siRNA (scramble) or Rheb siRNA. After 48 h of transfection, cells were fixed and stained with the indicated antibodies. The right row shows higher magnification views of the boxed area. (B) HEK293T cells were transiently transfected with plasmid pMRX-puro harboring GFP–Rheb for 24 h and then fixed and stained with anti-GM130 antibody (1:3200). The right row shows higher magnification views of the boxed area. (C) HEK293T cells were transiently transfected with GFP–Rheb for 24 h and then starved for 1 h in amino acid-free DMEM followed by re-stimulation using DMEM supplemented with dialyzed FBS for 10 min. The cells were fixed and stained with anti-mTOR antibody (1:400) and anti-Lamp1 antibody (1:1000). Scale bars: 5 μm.

Fig. 1.

Endogenous Rheb and GFP–Rheb are localized on the Golgi. (A) HEK293T cells were cultured in DMEM for 24 h and then fixed and stained for immunofluorescence with anti-Rheb antibody (1:200) and anti-GM130 (1:3200) diluted in blocking buffer. For Rheb knockdown, HEK293T cells were transfected with control siRNA (scramble) or Rheb siRNA. After 48 h of transfection, cells were fixed and stained with the indicated antibodies. The right row shows higher magnification views of the boxed area. (B) HEK293T cells were transiently transfected with plasmid pMRX-puro harboring GFP–Rheb for 24 h and then fixed and stained with anti-GM130 antibody (1:3200). The right row shows higher magnification views of the boxed area. (C) HEK293T cells were transiently transfected with GFP–Rheb for 24 h and then starved for 1 h in amino acid-free DMEM followed by re-stimulation using DMEM supplemented with dialyzed FBS for 10 min. The cells were fixed and stained with anti-mTOR antibody (1:400) and anti-Lamp1 antibody (1:1000). Scale bars: 5 μm.

Stable expression of GFP–Rheb activates mTORC1

Under starvation conditions, mTORC1 is inactivated and its substrate, S6K, is dephosphorylated (Fig. 2A) (Burnett et al., 1998). However, stable expression of GFP–Rheb activated mTORC1 even under conditions of amino acid starvation (Fig. 2A). Addition of insulin further increased the GFP–Rheb-dependent activation (data not shown). To determine whether cytosolic or lysosomal mTORC1 is involved in this activation, we used MEF cells that had a knockouts in LAMTOR1, a Ragulator complex subunit (Nada et al., 2009). Without Ragulator, the scaffold complex that localizes Rag GTPase to the lysosome, mTORC1 cannot be recruited to the lysosome (Sancak et al., 2010). GFP–Rheb-dependent mTORC1 activation did not occur in LAMTOR1-knockout MEF cells but did occur in wild-type MEF cells (Fig. 2B). This indicates that mTORC1, which is localized at the lysosome via Ragulator, is involved in this activation, even though a large portion of mTORC1 would be dispersed, owing to starvation, in these conditions.

Fig. 2.

Stable expression of Rheb can activate mTORC1. (A) MEF cells with or without stable GFP or GFP–Rheb expression were starved in amino acid-free DMEM for 60 min and then were lysed and immunoblotted with anti-phosphorylated S6K (P-S6K) (1:1000), anti-S6K (1:1000), anti-4E-BP1 (1:1000), anti-α-tubulin (1:10,000) and anti-GFP (1:1000) antibodies. The ratio of P-S6K to S6K was quantified from three independent experiments. Bars represent the means±s.e.m. ***P<0.001 (two-tailed paired Student's t-test). (B) LAMTOR1-knockout (p18KO) MEF cells and wild-type (WT) MEF cells were infected for 48 h with the viruses transducing the indicated proteins and starved for amino acids as described in A. The cells were then lysed and immunoblotted with the indicated antibodies, as above. The ratio of P-S6K to S6K was quantified from three independent experiments. Bars represent the means±s.e.m. ***P<0.001 (two-tailed paired Student's t-test).

Fig. 2.

Stable expression of Rheb can activate mTORC1. (A) MEF cells with or without stable GFP or GFP–Rheb expression were starved in amino acid-free DMEM for 60 min and then were lysed and immunoblotted with anti-phosphorylated S6K (P-S6K) (1:1000), anti-S6K (1:1000), anti-4E-BP1 (1:1000), anti-α-tubulin (1:10,000) and anti-GFP (1:1000) antibodies. The ratio of P-S6K to S6K was quantified from three independent experiments. Bars represent the means±s.e.m. ***P<0.001 (two-tailed paired Student's t-test). (B) LAMTOR1-knockout (p18KO) MEF cells and wild-type (WT) MEF cells were infected for 48 h with the viruses transducing the indicated proteins and starved for amino acids as described in A. The cells were then lysed and immunoblotted with the indicated antibodies, as above. The ratio of P-S6K to S6K was quantified from three independent experiments. Bars represent the means±s.e.m. ***P<0.001 (two-tailed paired Student's t-test).

Golgi-localized GFP–Rheb can activate mTORC1

To explore how Rheb localization affects mTORC1 activity, we attempted to artificially target GFP–Rheb to the Golgi, lysosome, and ER. For this purpose, a GFP-binding protein (GBP) (Katoh et al., 2015; Kubala et al., 2010) was fused to proteins residing in each organelle [FLAG–GBP-tagged Rab1A and Rab33B for the Golgi, Rab7 and Rab9A for lysosomes, and cytochrome b5 (Cytb) for the ER]. These fusion proteins were then stably expressed in the GFP–Rheb-expressing MEF cells (Fig. 3A). Our results showed that the expression of FLAG–GBP–Rab1A efficiently targeted GFP–Rheb specifically to the Golgi, as confirmed by signal overlap with GM130, but not to lysosomes or ER (Fig. 3B,C; Fig. S2A). The specificity of organelle localization was also observed for all other constructs and GFP–Rheb was exclusively targeted to each organelle (Fig. 3B,C; Fig. S2A,B). We then assessed mTORC1 activity in each of these cells under conditions of amino acid starvation. When GFP–Rheb was targeted to the lysosome, either with FLAG–GBP–Rab7 or by FLAG–GBP–Rab9A, mTORC1 was activated even under conditions of amino acid starvation (Fig. 3D,E). This is expected, considering that cellular mTORC1 is localized to the lysosome. Moreover, when GFP–Rheb was localized to the Golgi through FLAG–GBP–Rab1A or FLAG–GBP–Rab33B expression, mTORC1 could be also activated (Fig. 3D,E). In contrast, when GFP–Rheb was targeted to the ER by means of FLAG–GBP–Cytb expression, mTORC1 activity was completely suppressed, which was also observed in the cells expressing empty GFP vector (Fig. 3D,E). We noted that co-expression of GFP–Rheb with FLAG–GBP suppressed mTORC1 activity compared with the activity with GFP–Rheb alone, probably because although GBP not only affects GFP–Rheb localization to the Golgi, but also because it might affect GFP–Rheb mobility in cells and partially suppress mTORC1 activity induced by GPF–Rheb under starvation conditions. To determine whether expression of the GBP constructs alone without GPF–Rheb affects mTORC1 activity, we observed mTORC1 activity when expressing GBP constructs without expressing GFP–Rheb and found that mTORC1 activity was not affected by GBP constructs under starvation or replenishment conditions (Fig. S2C). Although the mobility of Rheb was altered if the farnesylation did not occur (data not shown), the mobility of GFP-Rheb in each western blot was uniform (Fig. 3D), indicating that lipid modification is not altered in these targeting systems. These results clearly indicate that Rheb can activate mTORC1 even when they are exclusively localized at the Golgi, but not the ER.

Fig. 3.

Golgi-localized Rheb can activate mTORC1. (A) A schematic of the method of targeting of GFP–Rheb to the Golgi with FLAG–GBP-tagged Rab1A. (B) MEF cells stably expressing GFP–Rheb with FLAG–GBP–Rab1A or FLAG–GBP–Rab7 were fixed and stained with anti-GM130 (1:1000) or anti-Lamp1 (1:1000) antibodies. For GFP–Rheb and FLAG–GBP–Cytb co-expression, the cells were transfected with the RFP–Sec61β ER marker and then fixed and observed. The right row shows higher magnification views of the boxed area. Scale bars: 5 μm. (C) Colocalization quantified as Pearson's correlation coefficient (PCC) between GFP–Rheb and GM130, Lamp1 or RFP–Sec61β. The localization of GFP–Rheb with Lamp1 or RFP–Sec61β in FLAG–GBP–Rab1A-expressing cells, GFP–Rheb with GM130 or RFP–Sec61β in FLAG-GBP-Rab7-expressing cells, GFP–Rheb with GM130 or Lamp1 in FLAG–GBP–Cytb-expressing cells is shown in Fig. S2. More than 30 cells were quantified. Bars represent the means±s.e.m. (D) MEF cells stably expressing GFP, GFP–Rheb, or GFP–Rheb with FLAG–GBP, FLAG–GBP–Rab1A, FLAG–GBP–Rab33B, FLAG–GBP–Rab7, FLAG–GBP–Rab9A or FLAG–GBP–Cytb were starved for 60 min in amino acid-free DMEM. Cells were then lysed and immunoblotted with anti-FLAG antibody (1:1000) and the indicated antibodies as in Fig. 2. (E) Quantification of the ratio of P-S6 K to S6 K from blots as shown D (three independent experiments). Bars represent the means±s.e.m. *P<0.05 (two-tailed paired Student's t-test).

Fig. 3.

Golgi-localized Rheb can activate mTORC1. (A) A schematic of the method of targeting of GFP–Rheb to the Golgi with FLAG–GBP-tagged Rab1A. (B) MEF cells stably expressing GFP–Rheb with FLAG–GBP–Rab1A or FLAG–GBP–Rab7 were fixed and stained with anti-GM130 (1:1000) or anti-Lamp1 (1:1000) antibodies. For GFP–Rheb and FLAG–GBP–Cytb co-expression, the cells were transfected with the RFP–Sec61β ER marker and then fixed and observed. The right row shows higher magnification views of the boxed area. Scale bars: 5 μm. (C) Colocalization quantified as Pearson's correlation coefficient (PCC) between GFP–Rheb and GM130, Lamp1 or RFP–Sec61β. The localization of GFP–Rheb with Lamp1 or RFP–Sec61β in FLAG–GBP–Rab1A-expressing cells, GFP–Rheb with GM130 or RFP–Sec61β in FLAG-GBP-Rab7-expressing cells, GFP–Rheb with GM130 or Lamp1 in FLAG–GBP–Cytb-expressing cells is shown in Fig. S2. More than 30 cells were quantified. Bars represent the means±s.e.m. (D) MEF cells stably expressing GFP, GFP–Rheb, or GFP–Rheb with FLAG–GBP, FLAG–GBP–Rab1A, FLAG–GBP–Rab33B, FLAG–GBP–Rab7, FLAG–GBP–Rab9A or FLAG–GBP–Cytb were starved for 60 min in amino acid-free DMEM. Cells were then lysed and immunoblotted with anti-FLAG antibody (1:1000) and the indicated antibodies as in Fig. 2. (E) Quantification of the ratio of P-S6 K to S6 K from blots as shown D (three independent experiments). Bars represent the means±s.e.m. *P<0.05 (two-tailed paired Student's t-test).

mTORC1 can be activated when artificially localized at the Golgi

We postulated that mTORC1 might be activated even if artificially localized to the Golgi. To test this hypothesis, we targeted mTORC1 to the Golgi by engineering a MEF cell line stably expressing GFP-Raptor, an mTORC1-specific subunit (Hara et al., 2002; Kim et al., 2002). GFP–Raptor was mostly distributed throughout cytoplasm under conditions of amino acid starvation, accumulated at the lysosome upon amino acid replenishment (Fig. S3A) and colocalized with endogenous mTOR (Fig. S3B). Next, we stably expressed FLAG–GBP–Rab33B, which is localized at the Golgi (Zheng et al., 1998) in MEF cells stably expressing GFP–Raptor. As expected, GFP–Raptor was targeted to the Golgi and colocalized with GM130 (Fig. 4A–C). Furthermore, together with GFP–Raptor, endogenous mTOR was exclusively recruited to the Golgi whereas all other localization mostly disappeared (Fig. 4D,E). In contrast, endogenous mTOR was not colocalized with GM130 in the control FLAG–GBP-expressing cells (Fig. 4D,E). We also established cell lines stably expressing FLAG–GBP–Rab6A, FLAG–GBP–Rab9A, and FLAG–GBP–Cytb for Golgi, lysosome and ER targeting in GFP–Raptor-expressing MEF cells, respectively. In these cells, mTOR and GFP–Raptor were efficiently recruited to each organelle (Figs S4A,B). We then examined mTORC1 activity in these cells. mTORC1 targeted to the lysosome could be activated (Fig. 4F), which is to be expected given the bona fide lysosomal localization of mTORC1 (Fig. S4A,B). Notably, even when GFP–Raptor and endogenous mTOR were targeted to the Golgi by means of FLAG–GBP–Rab6A or FLAG–GBP–Rab33B, mTORC1 activity was maintained at the same level as that observed in control cells (Fig. 4F). This indicates that Golgi-targeted mTORC1 can be activated. In contrast, ER-targeted mTORC1 cannot be activated, which suggested that endogenous Rheb is not present in the ER (Fig. 4F). During starvation conditions, mTORC1 was not able to be activated in any cell lines, suggesting that unknown mechanisms of mTORC1 activation are coming into play in this experimental system (data not shown). We also determined mTORC1 activity by using Raptor–GFP to force targeting of mTORC1 to the above-mentioned organelles; however, we obtained a similar result to that found when mTORC1 was targeted to each organelle, as shown in Fig. 4F (data not shown). Taken together, these results indicate that mTORC1 can be activated efficiently even when it is targeted to the Golgi.

Fig. 4.

mTORC1 can be activated by amino acids when tethered at the Golgi. (A) Schematic of the method of targeting of GFP–Raptor to the Golgi with FLAG–GBP-tagged Rab33B. (B) MEF cells stably expressing GFP–Raptor with FLAG–GBP or FLAG–GBP–Rab33B were starved for 60 min in amino acid-free DMEM and then replenished for 10 min with amino acids in DMEM. The cells were fixed and stained with anti-GM130 antibodies (1:1000). Scale bars: 5 μm. (C) Colocalization quantified as the Pearson's correlation coefficient (PCC) between GFP–Raptor and GM130. More than 30 cells were quantified. Bars represent the means±s.e.m. (D) Cells described in B were fixed and stained with anti-mTOR antibody (1:400), anti-GM130 antibody or anti-Lamp1 antibody. Arrowheads highlight the colocalization of GFP–Raptor with mTOR and the lysosome marker Lamp1. Scale bars: 5 μm. (E) Colocalization quantified as the PCC between mTOR and either GM130 or Lamp1. More than 30 cells were quantified. Bars represent the means±s.e.m. (F) MEF cells or MEF cells stably expressing GFP–Raptor with FLAG–GBP, FLAG–GBP–Rab6A, FLAG–GBP–Rab33B, FLAG–GBP–Rab9A or FLAG–GBP–Cytb were starved for 60 min in amino acid-free DMEM and then replenished with amino acids in DMEM for 10 min. The cells were lysed and immunoblotted with indicated antibodies as in Fig. 3D.

Fig. 4.

mTORC1 can be activated by amino acids when tethered at the Golgi. (A) Schematic of the method of targeting of GFP–Raptor to the Golgi with FLAG–GBP-tagged Rab33B. (B) MEF cells stably expressing GFP–Raptor with FLAG–GBP or FLAG–GBP–Rab33B were starved for 60 min in amino acid-free DMEM and then replenished for 10 min with amino acids in DMEM. The cells were fixed and stained with anti-GM130 antibodies (1:1000). Scale bars: 5 μm. (C) Colocalization quantified as the Pearson's correlation coefficient (PCC) between GFP–Raptor and GM130. More than 30 cells were quantified. Bars represent the means±s.e.m. (D) Cells described in B were fixed and stained with anti-mTOR antibody (1:400), anti-GM130 antibody or anti-Lamp1 antibody. Arrowheads highlight the colocalization of GFP–Raptor with mTOR and the lysosome marker Lamp1. Scale bars: 5 μm. (E) Colocalization quantified as the PCC between mTOR and either GM130 or Lamp1. More than 30 cells were quantified. Bars represent the means±s.e.m. (F) MEF cells or MEF cells stably expressing GFP–Raptor with FLAG–GBP, FLAG–GBP–Rab6A, FLAG–GBP–Rab33B, FLAG–GBP–Rab9A or FLAG–GBP–Cytb were starved for 60 min in amino acid-free DMEM and then replenished with amino acids in DMEM for 10 min. The cells were lysed and immunoblotted with indicated antibodies as in Fig. 3D.

mTORC1 is activated at the site where the Golgi and lysosome are in contact

The above results prompted us to consider the possibility that lysosome-localized mTORC1 might be activated by Golgi-localized Rheb at the Golgi–lysosome contract site. First, to examine the existence of the contact site between Golgi and the lysosome, we employed a proximity ligation assay, which uses two antibodies and yields a spot fluorescence signal if the antigen proteins exist within 40 nm (Zatloukal et al., 2014). We found that when antibodies against the GM130 Golgi protein and the Lamp1 lysosome protein were simultaneously employed, several fluorescence spot signals were detected in each cell (Fig. 5A). However, when either antibody was used alone, these spot signals were not detected (Fig. 5A). Furthermore, these fluorescence signals overlapped with some of the GFP–Rheb fluorescence signals (Fig. 5A). In addition, even when using the antibodies against adaptin γ, a trans-Golgi protein, and Lamp1, the fluorescence spot signals could be detected (Fig. S5A). Upon FTI treatment, which prevents the Golgi localization of Rheb by inhibiting the farnesylation of its C-terminus (Fig. S1D,E), GFP-Rheb did colocalize with the signals from GM130 and Lamp1 proximity ligation assay and instead was present in the cytosol (Fig. S5B). As FTI treatment does not affect Golgi integrity, the formation of proximity ligation assay signals with anti-GM130 and anti-Lamp1 antibodies were not affected (Fig. S5B). Furthermore, after Brefeldin A (BFA)-mediated disruption of the Golgi structure (Chardin and McCormick, 1999), these signals became undetectable, highlighting that the Golgi structure is required for this signal formation (Fig. 5B). Next, we disrupted lysosome integrity with lysosomotropic compound L-Leucyl-L-leucine methyl ester (LLOMe) (Maejima et al., 2013; Thiele and Lipsky, 1990; Uchimoto et al., 1999). Damaged lysosomes could be marked by staining for galectin-3 (Gal3), a lectin that recognizes carbohydrate chains exposed to the cytosol owing to lysosome damage (Maejima et al., 2013; Paz et al., 2010). We established MEF cells stably expressing GFP–Gal3, and detected the GM130 and Lamp1 proximity ligation assay signal after LLOMe treatment. Following LLOMe treatment, the GM130 and Lamp1 proximity ligation assay signals were dramatically decreased only in cells in which GFP–Gal3 punctate signals existed (Fig. 5C,D). Furthermore, anti-mTOR and anti-GM130 yielded fluorescent spots in the proximity ligation assay (Fig. S5C). Likewise, the combination of anti-Rheb and an antibody against Lamp1 generated spot signals in the proximity ligation assay (Fig. S5D). As expected, fluorescent signals of the proximity ligation assay for anti-Rheb and anti-mTOR were also detected (Fig. S5E). TSC2 is recruited to lysosomes and plays a role as a GAP for Rheb, thereby inhibiting its signaling to mTOR (Carroll et al., 2016; Demetriades et al., 2014, 2016; Menon et al., 2014; Zhang et al., 2003). Therefore, we aimed to determine whether proximity ligation assay signals could be formed between and the Golgi protein GM130. As expected, signals were detected between GFP–TSC2 and GM130 (Fig. S6A). Furthermore, GFP–TSC2 partially localized to the lysosome under both starvation and nutrient-rich conditions in MEF cells (Fig. S6B).

Fig. 5.

The Golgi contacts the lysosome at the perinuclear region. (A) Duolink, using anti-GM130 (1:000) and anti-Lamp1 (1:000) antibodies, was used to detect the GLCS in MEF cells stably expressing GFP–Rheb. The right row shows higher magnification views of the boxed area. (B) The cells described in A were treated with 1 μg/ml BFA for 2 h and then fixed and subjected to Duolink for detection of the GLCS. The right row shows higher magnification views of the boxed area. (C) MEF cells stably expressing GFP–Gal3 were treated without or with 1000 μM LLOMe for 3 h. After washing LLOMe out for 2 h in DMEM, the cells were subjected to Duolink for detection of the GLCS. (D) Quantification of the number of GLCSs from three independent experiments as shown in C. Bars represent the means±s.e.m. ***P<0.001, two-tailed paired Student's t-test. Scale bars: 5 μm.

Fig. 5.

The Golgi contacts the lysosome at the perinuclear region. (A) Duolink, using anti-GM130 (1:000) and anti-Lamp1 (1:000) antibodies, was used to detect the GLCS in MEF cells stably expressing GFP–Rheb. The right row shows higher magnification views of the boxed area. (B) The cells described in A were treated with 1 μg/ml BFA for 2 h and then fixed and subjected to Duolink for detection of the GLCS. The right row shows higher magnification views of the boxed area. (C) MEF cells stably expressing GFP–Gal3 were treated without or with 1000 μM LLOMe for 3 h. After washing LLOMe out for 2 h in DMEM, the cells were subjected to Duolink for detection of the GLCS. (D) Quantification of the number of GLCSs from three independent experiments as shown in C. Bars represent the means±s.e.m. ***P<0.001, two-tailed paired Student's t-test. Scale bars: 5 μm.

Taken together, these results demonstrate the existence of a sub-organelle region where the Golgi and lysosome are in close contact, which we term the Golgi–lysosome contact site (GLCS).

GLCS formation is involved in mTORC1 regulation

We next examined whether the GLCS integrity is correlated with mTORC1 activity. To monitor mTORC1 activity, we observed the translocation of transcription factor EB (TFEB). TFEB is primarily localized to the cytoplasm under basal nutrient conditions, owing to mTORC1-dependent phosphorylation, and translocates to the nucleus when mTORC1 activity is suppressed upon starvation (Fig. S7A) (Roczniak-Ferguson et al., 2012). MEF cells stably expressing GFP–TFEB were treated with BFA, which disrupted Golgi structure and decreased the GLCS signals, which induced the nuclear translocation of GFP–TFEB under basal nutrient conditions, indicating the suppression of mTORC1 activity (Fig. 6A,B). Furthermore, GFP–TFEB was transiently expressed in stable mStrawberry (mSt)-Gal3-expressing MEF cells. When lysosome integrity and GLCS were disrupted by LLOMe treatment, GFP–TFEB translocated to the nucleus, indicating that mTORC1 activity was downregulated (Fig. 6C,D). To further determine mTORC1 activity upon BFA or LLOMe treatment, we measured the phosphorylation of ribosomal protein S6 at S235/236 (pS6) as a readout for mTORC1 activity (Pende et al., 2004). As shown in Fig. S7B,C, the pS6 staining was absent in BFA- or LLOMe-treated cells. Furthermore, when Golgi structure, together with the GLCS, was disturbed by BFA treatment, mTORC1 activity was suppressed (Fig. S7D). Furthermore, to determine whether perturbation of the lysosome affects mTORC1 through phosphorylation of S6K (a direct substrate of mTOR), Bafilomycin A (Baf A), a v-ATPase inhibitor that blocks lysosome acidification was employed. As shown in Fig. S7E, mTORC1 activity was suppressed when mTORC1 was targeted to the lysosome. These results suggest that an intact GLCS correlates with normal mTORC1 activity.

Fig. 6.

Golgi or lysosomal rupture suppresses mTORC1 activity and GLCS number increases in response to amino acids. (A) MEF cells stably expressing GFP–TFEB were treated with 1 μg/ml of BFA for 2 h and then fixed and incubated with anti-GM130 antibody (1:1000). Scale bars: 5 μm. (B) Quantification of the percentage of nuclear-localized GFP–TFEB from three independent experiments as shown in A. Bars represent the means±s.e.m. ***P<0.001 (two-tailed paired Student's t-test). (C) GFP–TFEB was transiently expressed in MEF cells stably expressing mSt–Gal3. The cells were treated without or with 1000 μM LLOMe for 3 h. After washing for 2 h in DMEM, the cells were fixed and observed. Scale bars: 5 μm. (D) Quantification of the percentage of nuclear-localized GFP–TFEB from three independent experiments as shown in C. Bars represent the means±s.e.m. ***P<0.001 (two-tailed paired Student's t-test). (E) MEF cells were cultured in DMEM (Nutrient), starved for 60 min in amino acid-free DMEM (S: 1 h), or replenished for 5 or 10 min with amino acids in DMEM (S: 1 h +AA 5min or 10min, respectively). Cells were then fixed and stained with DAPI (1:1000) and anti-Lamp1 antibody (1:1000). Scale bars: 10 μm. (F) Duolink, using anti-GM130 antibody (1:1000) and anti-Lamp1 antibody (1:1000), was used to detect the GLCS in MEF cells stably expressing GFP–Rheb after being starved for 60 min in amino acid-free DMEM or replenished for 10 min with amino acids in DMEM. Scale bars: 25 μm. (G) The number of GLCS per cell was counted in more than 100 cells treated as in E. Bars represent the means±s.e.m. from three independent experiments. *P<0.05 (two-tailed paired Student's t-test).

Fig. 6.

Golgi or lysosomal rupture suppresses mTORC1 activity and GLCS number increases in response to amino acids. (A) MEF cells stably expressing GFP–TFEB were treated with 1 μg/ml of BFA for 2 h and then fixed and incubated with anti-GM130 antibody (1:1000). Scale bars: 5 μm. (B) Quantification of the percentage of nuclear-localized GFP–TFEB from three independent experiments as shown in A. Bars represent the means±s.e.m. ***P<0.001 (two-tailed paired Student's t-test). (C) GFP–TFEB was transiently expressed in MEF cells stably expressing mSt–Gal3. The cells were treated without or with 1000 μM LLOMe for 3 h. After washing for 2 h in DMEM, the cells were fixed and observed. Scale bars: 5 μm. (D) Quantification of the percentage of nuclear-localized GFP–TFEB from three independent experiments as shown in C. Bars represent the means±s.e.m. ***P<0.001 (two-tailed paired Student's t-test). (E) MEF cells were cultured in DMEM (Nutrient), starved for 60 min in amino acid-free DMEM (S: 1 h), or replenished for 5 or 10 min with amino acids in DMEM (S: 1 h +AA 5min or 10min, respectively). Cells were then fixed and stained with DAPI (1:1000) and anti-Lamp1 antibody (1:1000). Scale bars: 10 μm. (F) Duolink, using anti-GM130 antibody (1:1000) and anti-Lamp1 antibody (1:1000), was used to detect the GLCS in MEF cells stably expressing GFP–Rheb after being starved for 60 min in amino acid-free DMEM or replenished for 10 min with amino acids in DMEM. Scale bars: 25 μm. (G) The number of GLCS per cell was counted in more than 100 cells treated as in E. Bars represent the means±s.e.m. from three independent experiments. *P<0.05 (two-tailed paired Student's t-test).

As lysosomal positioning is related to cellular nutrient responses (Korolchuk et al., 2011), we asked whether the GLCS is affected by amino acid availability. We first confirmed the dynamics of lysosomes in response to amino acids. As shown in Fig. 6E, lysosomes accumulated at the perinuclear region in nutrient-containing conditions and dispersed to the peripheral region of cells in starvation conditions. Re-addition of amino acids again induced perinuclear clustering of lysosomes. Next, we observed the dynamics of GLCS in response to the nutrient conditions. As shown in Fig. 6F,G, at 10 min after replenishment of amino acids, the number of GLCS (14.2±2.3/cell) was approximately double that observed during starvation conditions (6.7±0.8/cell). A similar result was also observed in MEF cells (Fig. S7F). Furthermore, a time course of amino acid replenishment showed that mTORC1 activity was gradually increased within 30 min of amino acid stimulation then decreased slowly after 30 min of amino acid replenishment (Fig. S7G). The formation of GLCS showed a similar tendency to that of the mTORC1 activity as shown in Fig. S7H, as indicated by the ratio of P-S6K1 to total S6K1. These results suggest that the GLCS is at least partly involved in the regulation of mTORC1 activity.

Here, we have revealed that, whereas Rheb is largely localized to the Golgi, it can activate mTORC1 that is localized to the lysosome. We have also identified a novel inter-organelle contact site, termed the Golgi–lysosome contact site (GLCS) (Fig. 7). Involvement of the Golgi in mTORC1 regulation has gradually come to light through other lines of study. Recently, Rheb has also been shown to localize to the Golgi (Manifava et al., 2016). Thomas et al. have shown that knockdown of Rab1A, a small GTPase involved in vesicle transport from the ER to the Golgi, affects the interaction between Rheb and mTORC1 (Thomas et al., 2014). We have also reported that MTMR3, which is localized to the Golgi, interacts with mTORC1 and suppresses its activity (Hao et al., 2016; Taguchi-Atarashi et al., 2010). Recently, it has become recognized that a variety of contact sites exist between organelles that are used to exchange their contents such as lipid molecules, in addition to classical vesicular transport systems (Hamasaki et al., 2013; Helle et al., 2013; Liu et al., 2016; Rocha et al., 2009; Valm et al., 2017). These membrane contact sites are maintained by tethering the organelles to keep the membranes in proximity, typically within 30 nm (Helle et al., 2013). In many contact sites, specific protein machinery is responsible for the tethering. In the case of ER–mitochondria contact sites, VAPB–PTPIP51 and ERMES are important tethering machineries (Kornmann and Walter, 2010; Stoica et al., 2014). In the case of GLCS, the tethering machinery is still to be determined. However, it remains possible that there is no such molecule or machinery. Both the Golgi and lysosome are transported towards centrosomes via microtubules (Kreis, 1990; Matteoni and Kreis, 1987), which could eventually allow an opportunity for them to encounter each other. The localization of lysosomes is dramatically altered from the perinuculear region to the cell peripherally in response to nutrient availability at least in some cell lines (Hong et al., 2017; Kim and Cunningham, 2015; Korolchuk et al., 2011; Li et al., 2016). Interestingly, our observation (Fig. 6E) contrasts with that of the aforementioned reports, which suggest that lysosomes are clustered in the perinuclear region under starvation conditions and disperse to the cell periphery in response to nutrients. Because those findings were obtained in HeLa cells, the different mechanisms underlying lysosome positioning in response to nutrients in different cell types warrants investigation in future studies. Indeed, our observations regarding lysosome positioning in response to nutrient alterations was in accordance with our subsequent result showing that the number of GLCS increased in response to amino acid supply (Fig. 6F,G; Fig. S7F,H). LLOMe-induced lysosome rupture suppressed TORC1 activity (Fig. 6C,D). Therefore, it is highly possible that the re-positioning of the lysosome is involved in TORC1 activation, in part, through the formation of the GLCS.

Fig. 7.

Proposed model – Rheb activates mTORC1 at the GLCS. Rheb at least partially localizes to the Golgi. When the amino acids signal to the cells, mTORC1 translocates to lysosomes. To respond to the amino acid signals, the lysosomes accumulate at the perinuclear region, where the Golgi exists, leading to the formation of more GLCSs to provide a platform for the activation of mTORC1 by Rheb.

Fig. 7.

Proposed model – Rheb activates mTORC1 at the GLCS. Rheb at least partially localizes to the Golgi. When the amino acids signal to the cells, mTORC1 translocates to lysosomes. To respond to the amino acid signals, the lysosomes accumulate at the perinuclear region, where the Golgi exists, leading to the formation of more GLCSs to provide a platform for the activation of mTORC1 by Rheb.

If so, what is the role of GLCS-dependent regulation in addition to the Rag/Ragulator-dependent system (Sancak et al., 2008, 2010)? At the very least, the Rag/Ragulator system-dependent lysosomal localization of mTORC1 appears to be a prerequisite for GLCS-dependent TORC1 activation (Fig. 2B). One possibility is that the two systems allow for the fine-tuning of regulation. The tuberous sclerosis complex (TSC), a GTPase-activating protein complex that negatively regulates Rheb activity, is also localized to the lysosome (Demetriades et al., 2014, 2016; Menon et al., 2014; Carroll et al., 2016). Therefore, it is highly possible that TSC–Rheb interaction also occurs at the GLCS. If Rheb were also localized to the lysosome, it would be suppressed by the TSC, leading to immediate inactivation of mTORC1. Therefore, compartmentalization of Rheb, mTORC1 and TSC would be required at the suborganelle level, and communication at the inter-organelle level may allow such regulation. How such fine-tuning is possible remains to be elucidated and needs to be investigated in the future, although Rab1A and/or MTMR3 may play roles in such regulation (Hao et al., 2016; Thomas et al., 2014).

In addition, our findings provide significant data showing that mTORC1 can be activated when artificially targeted to the Golgi (Fig. 4F), suggesting the possible role of the GLCS in mTORC1 activation; however, we cannot completely eliminate the possibility that mTORC1 persists at low levels on the lysosome and is responsible for mTORC1 activation. Furthermore, additional inter-organelle regulation between Rheb and mTORC1, other than that described in this study, may also be involved. Indeed, in addition to the lysosome, it has been shown that mTOR is present in the ER, Golgi, plasma membrane, cytosol, endosomes, mitochondria, and nucleus, although we did not detect such diverse localization in our study (Fig. 1C; Fig. S1G) (Drenan et al., 2004; Liu and Zheng, 2007; Sancak et al., 2008; Schieke et al., 2006; Zhou et al., 2015). Based on our findings, further detailed localization analyses will provide a complete picture of mTORC1 regulation.

Cell culture

Cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (Wako, Osaka, Japan) containing 10% fetal bovine serum (FBS; Gibco, Waltham, MA) and penicillin (50 U/ml)-streptomycin (50 μg/ml) in a 5% CO2 incubator at 37°C. For starvation treatment, cells were washed once with phosphate-buffered saline (PBS) and cultured in Eisen's balanced salt solution (EBSS) (Sigma-Aldrich, St. Louis, MO,) or DMEM (high glucose) with sodium pyruvate without amino acids (Wako) supplemented with or without dialyzed FBS, which was prepared using Slide-A-Lyzer Dialysis Cassettes (Thermo Fisher Scientific, Waltham, MA), for 60 min or for an indicated time. For nutrient replenishment, cells were cultured in DMEM with dialyzed FBS for 10 min or an indicated time after starvation treatment. Leu-Leu methyl ester hydrobromide (LLOMe; Sigma-Aldrich) (333 mM) dissolved in ethanol was used as a stock solution, and 1000 μM diluted from stock solution was administrated for 3 h and washed out with DMEM for 2 h. Brefeldin A (CST, Danvers, MA) (10 mg) was dissolved in DMSO to create a stock solution and cells were treated with 1 μg/ml or 4 μg/ml for the described time. FTI-277 (Sigma-Aldrich) (5 mM) was dissolved in Milli-Q water to create a stock solution and cells were treated with 10 μM for 24 h. Bafilomycin A (BioViotica, Dransfeld, Germany) was dissolved in DMSO and cells were treated with 1 μM for 1 h.

Cells, plasmids and virus preparation

LAMTOR1-knockout MEF cells were described previously (Nada et al., 2009). Wild-type Rheb, cloned from mouse cDNA, Raptor, cloned from Myc-Raptor (Hao et al., 2016), and TSC2, cloned from pcDNA3 Flag TSC2 obtained from Addgene (#14129, Cambridge, MA) (Manning et al., 2002), were inserted into the BamHI and NotI sites of pMRX-puro-EGFP (Itoh et al., 2011). Rab1A, Rab6A, Rab7, Rab9A and Rab33B cloned from pEF-BOS-FLAG-RabX (Itoh et al., 2008) or Cytochrome b5 cloned from mouse cDNA were inserted into the BamHI and NotI or BamHI and EcoRI sites of pMRX-bsr-FLAG-GBP. pMRX-bsr-FLAG-GBP was constructed by inserting GBP, cloned from pGEX-6P1-GFP Nanobody (Katoh et al., 2015), into pMRX-bsr-FLAG (Itoh et al., 2008) for this study. mRFP-Sec61β, GFP–Gal3, and mSt–Gal3 were kind gifts from Tamotsu Yoshimori (Department of Genetics, Graduate School of Medicine, Osaka University, Japan) (Hamasaki et al., 2013; Maejima et al., 2013). GFP–TFEB plasmid was obtained from Addgene (#38119) (Roczniak-Ferguson et al., 2012).

Transient transfection was performed using Lipofectamine 3000 (Invitrogen, Waltham, MA) according to the manufacturer's protocols. For retrovirus packaging, Plat-E cells were cultured in DMEM supplemented with 10% FBS (Morita et al., 2000) and related plasmids were transiently co-transfected with PLP-VSVG (Invitrogen). After 36 h of transfection, the retrovirus was harvested for infection. Stable transfection was established by retroviral infection and selection in growth medium containing 2 μg/ml puromycin or 10 μg/ml blasticidin (Invitrogen).

The control siRNA was obtained from Invitrogen and the Rheb siRNA (5′- CUAUGGAGUAUGUCUGAGGdTdT-3′, sense) was obtained from Dharmacon (Lafayette, CO). The siRNA (40 pmol per 3.5 cm dish) was diluted in 125 μl Opti-MEM (Invitrogen) and the Lipofectamine RNAiMAX (Invitrogen) (4 μl per 3.5 cm dish) was diluted in 125 μl Opti-MEM. The diluted siRNA was mixed with Lipofectamine RNAiMAX and incubated for 5 min at room temperature before being dropped onto cells.

Antibodies

The following antibodies were used: rabbit anti-mTOR (7C10), rabbit anti-p70 S6 kinase (49D7), rabbit anti-phospho-p70 S6 kinase (Thr389), rabbit anti-phospho-S6 ribosomal protein (Ser235/236), rabbit anti-4E-BP1, rabbit anti-GM130 (CST), mouse anti-Rheb (Abnova, Walnut, CA, USA), mouse anti-FLAG (M2), mouse anti-α-tubulin and mouse anti-β-tubulin (Sigma-Aldrich), mouse anti-GM130, mouse anti-Adaptin γ (BD Transduction Laboratories, Lexington, KY, USA), rat anti-LAMP1 (1D48) (Santa Cruz Biotechnology, Dallas, TX, USA), rabbit anti-LAMP1 (Abcam, Cambridge, UK), and rabbit anti-GFP antibody (MBL, Nagoya, Japan).

Western blotting

Cells were washed twice with ice-cold PBS, collected by centrifugation (2300 g for 5 min), and lysed in lysis buffer [50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% Triton X-100, complete protease inhibitor cocktail (Roche, Basel, Switzerland)]. Cell lysates were cleared by centrifugation at 20,400 g (MX201, TOMY, Tokyo, Japan) for 10 min at 4°C. Supernatants were collected and mixed with sample buffer (1×; 2% SDS, 100 mM DTT, 60 mM Tris-HCl pH 6.8, 10% glycerol, 0.001% Bromophenol Blue). After boiling for 5 min, the samples were subjected to SDS/polyacrylamide gel electrophoresis (PAGE). Then, the proteins were transferred to PVDF membranes (GE Healthcare, Bucks, UK) using transfer buffer (25 mM Tris base, 190 mM glycine, 20% methanol). The transferred membrane was blocked for 1 h at room temperature in 5% skimmed milk in TBS-T (25 mM Tris base, 137 mM NaCl, 2.7 mM KCl, 0.1% Tween 20, adjusting pH to 7.4). After blocking, the membrane was incubated overnight with appropriate dilutions of primary antibody in blocking buffer at 4°C. The membrane was washed three times in TBS-T and incubated at room temperature for 30 min with a 1:5000 dilution of HRP-conjugated secondary antibody (CST) in blocking buffer. The membrane was washed three times and visualized using the ECL Select western blotting detection reagent (GE Healthcare) on a Gene Gnome-5 chemiluminescence detector (Syngene, Cambridge, UK).

Immunofluorescence microscopy

To observe GFP–Rheb, mTOR, GFP–Raptor, GFP–TFEB, GFP–Gal3, mSt–Gal3, Golgi (GM130), lysosomes (Lamp1) and ER (RFP–Sec61β), cells were cultured on coverslips and fixed for 15 min with 4% paraformaldehyde in PBS. The fixed cells were permeabilized for 10 min with 50 μg/ml digitonin in PBS, and blocked with blocking buffer (0.2% gelatin-PBS), and then incubated with the indicated primary antibodies diluted in blocking buffer for 1 h at room temperature. Cells were washed three times in blocking buffer, incubated for 40 min with Alexa Fluor 488-, 568- or 633-conjugated secondary antibodies (Invitrogen) diluted in blocking buffer, and washed three times in blocking buffer. The samples were mounted using Slow Fade Gold (Invitrogen) and observed on a confocal laser-scanning microscope (TCS SP8; HC PL APO 63×/1.4 oil objective, Leica, Wetzlar, Germany). Colocalization was quantified using Pearson's correlation coefficient with Coloc 2 Fiji plugin (a distribution of ImageJ; https://imagej.net/Fiji/Downloads).

For endogenous Rheb observation, fixed cells were permeabilized for 10 min with permeabilization buffer (150 mM NaCl, 5 mM EDTA, 40 mM Tris-HCl pH 7.4, 0.25% gelatin, 0.25% NP40, 0.02% azide), blocked in blocking buffer (150 mM NaCl, 5 mM EDTA, 40 mM Tris-HCl pH 7.4, 0.25% gelatin, 0.05% NP40, 0.02% azide) for 30 min, and then incubated with anti-Rheb and anti-GM130 antibodies for 3 h at room temperature, washed three times in blocking buffer, incubated for 1 h with Alexa Fluor 488- and 568-conjugated secondary antibodies (Invitrogen) diluted in blocking buffer, and washed three times in blocking buffer. Then, the samples were mounted and observed as described above.

Proximity ligation assay

The Duolink in situ fluorescence kit was purchased from Sigma-Aldrich. The cells were fixed, permeabilized and blocked as described above. Then, the samples were incubated with primary antibodies including mouse anti-GM130 (1:1000) and rabbit anti-Lamp1 (1:1000), rabbit anti-mTOR (1:400) and mouse anti-GM130 (1:1000), or mouse anti-Rheb (1:200) and rabbit anti-Lamp1 (1:1000) diluted in blocking buffer (PBS with 0.2% gelatin) for 1 h. Cells were then washed two times in blocking buffer, incubated for 1 h at 37°C with proximity ligation assay (PLA) probes diluted in blocking buffer, washed twice for 5 min with PLA wash buffer A, and incubated for 30 min at 37°C with PLA ligation-ligase solution. Cells were then washed twice for 2 min with PLA wash buffer A, incubated with PLA amplification-polymerase solution for 100 min at 37°C, washed twice for 10 min with PLA wash buffer B, and then washed for 1 min with 0.01× PLA wash buffer B. The samples were mounted and observed as described above.

We appreciate technical support from the CFOS, Graduate School of Dentistry of Osaka University and all members of Noda laboratory for helpful discussion.

Author contributions

Conceptualization: F.H., T.I., T.N.; Methodology: T.I., T.N.; Formal analysis: K.K., T.I., T.N.; Investigation: K.K., T.I., S.I.; Resources: S.N., M.O.; Data curation: K.K., S.N., M.O., T.N.; Writing - original draft: F.H., T.N.; Writing - review & editing: F.H., S.N., M.O., T.N.; Supervision: T.I., S.N., M.O., T.N.; Project administration: T.N.; Funding acquisition: T.N.

Funding

This work was supported by Japan Society for Promotion of Science (grant number 16K07347) and the Ministry of Education, Culture, Sports, Science and Technology of Japan to T.N. (grant number 16H01202).

Aspuria
,
P.-J.
and
Tamanoi
,
F.
(
2004
).
The Rheb family of GTP-binding proteins
.
Cell. Signal.
16
,
1105
-
1112
.
Bar-Peled
,
L.
,
Schweitzer
,
L. D.
,
Zoncu
,
R.
and
Sabatini
,
D. M.
(
2012
).
Ragulator is a GEF for the rag GTPases that signal amino acid levels to mTORC1
.
Cell
150
,
1196
-
1208
.
Bar-Peled
,
L.
,
Chantranupong
,
L.
,
Cherniack
,
A. D.
,
Chen
,
W. W.
,
Ottina
,
K. A.
,
Grabiner
,
B. C.
,
Spear
,
E. D.
,
Carter
,
S. L.
,
Meyerson
,
M.
and
Sabatini
,
D. M.
(
2013
).
A Tumor suppressor complex with GAP activity for the Rag GTPases that signal amino acid sufficiency to mTORC1
.
Science
340
,
1100
-
1106
.
Buerger
,
C.
,
DeVries
,
B.
and
Stambolic
,
V.
(
2006
).
Localization of Rheb to the endomembrane is critical for its signaling function
.
Biochem. Biophys. Res. Commun.
344
,
869
-
880
.
Burnett
,
P. E.
,
Barrow
,
R. K.
,
Cohen
,
N. A.
,
Snyder
,
S. H.
and
Sabatini
,
D. M.
(
1998
).
RAFT1 phosphorylation of the translational regulators p70 S6 kinase and 4E-BP1
.
Proc. Natl. Acad. Sci. USA
95
,
1432
-
1437
.
Carroll
,
B.
,
Maetzel
,
D.
,
Maddocks
,
O. D.
,
Otten
,
G.
,
Ratcliff
,
M.
,
Smith
,
G. R.
,
Dunlop
,
E. A.
,
Passos
,
J. F.
,
Davies
,
O. R.
,
Jaenisch
,
R.
, et al. 
(
2016
).
Control of TSC2-Rheb signaling axis by arginine regulates mTORC1 activity
.
Elife
5
,
1
-
23
.
Chantranupong
,
L.
,
Scaria
,
S. M.
,
Saxton
,
R. A.
,
Gygi
,
M. P.
,
Shen
,
K.
,
Wyant
,
G. A.
,
Wang
,
T.
,
Harper
,
J. W.
,
Gygi
,
S. P.
and
Sabatini
,
D. M.
(
2016
).
The CASTOR proteins are arginine sensors for the mTORC1 pathway
.
Cell
165
,
153
-
164
.
Chardin
,
P.
and
McCormick
,
F.
(
1999
).
Brefeldin A: the advantage of being uncompetitive
.
Cell
97
,
153
-
155
.
Demetriades
,
C.
,
Doumpas
,
N.
and
Teleman
,
A. A.
(
2014
).
Regulation of TORC1 in response to amino acid starvation via lysosomal recruitment of TSC2
.
Cell
156
,
786
-
799
.
Demetriades
,
C.
,
Plescher
,
M.
and
Teleman
,
A. A.
(
2016
).
Lysosomal recruitment of TSC2 is a universal response to cellular stress
.
Nat. Commun.
7
,
10662
.
Dibble
,
C. C.
and
Manning
,
B. D.
(
2013
).
Signal integration by mTORC1 coordinates nutrient input with biosynthetic output
.
Nat. Cell Biol.
15
,
555
-
564
.
Drenan
,
R. M.
,
Liu
,
X.
,
Bertram
,
P. G.
and
Zheng
,
X. F. S.
(
2004
).
FKBP12-rapamycin-associated protein or mammalian target of rapamycin (FRAP/mTOR) localization in the endoplasmic reticulum and the golgi apparatus
.
J. Biol. Chem.
279
,
772
-
778
.
Hamasaki
,
M.
,
Furuta
,
N.
,
Matsuda
,
A.
,
Nezu
,
A.
,
Yamamoto
,
A.
,
Fujita
,
N.
,
Oomori
,
H.
,
Noda
,
T.
,
Haraguchi
,
T.
,
Hiraoka
,
Y.
, et al. 
(
2013
).
Autophagosomes form at ER-mitochondria contact sites
.
Nature
495
,
389
-
393
.
Hanker
,
A. B.
,
Mitin
,
N.
,
Wilder
,
R. S.
,
Henske
,
E. P.
,
Tamanoi
,
F.
,
Cox
,
A. D.
and
Der
,
C. J.
(
2010
).
Differential requirement of CAAX-mediated posttranslational processing for Rheb localization and signaling
.
Oncogene
29
,
380
-
391
.
Hao
,
F.
,
Itoh
,
T.
,
Morita
,
E.
,
Shirahama-Noda
,
K.
,
Yoshimori
,
T.
and
Noda
,
T.
(
2016
).
The PtdIns3-phosphatase MTMR3 interacts with mTORC1 and suppresses its activity
.
FEBS Lett.
590
,
161
-
173
.
Hara
,
K.
,
Maruki
,
Y.
,
Long
,
X.
,
Yoshino
,
K.-I.
,
Oshiro
,
N.
,
Hidayat
,
S.
,
Tokunaga
,
C.
,
Avruch
,
J.
and
Yonezawa
,
K.
(
2002
).
Raptor, a binding partner of target of rapamycin (TOR), mediates TOR action
.
Cell
110
,
177
-
189
.
Helle
,
S. C. J.
,
Kanfer
,
G.
,
Kolar
,
K.
,
Lang
,
A.
,
Michel
,
A. H.
and
Kornmann
,
B.
(
2013
).
Organization and function of membrane contact sites
.
Biochim. Biophys. Acta-Mol. Cell Res.
1833
,
2526
-
2541
.
Hong
,
Z.
,
Pedersen
,
N. M.
,
Wang
,
L.
,
Torgersen
,
M. L.
,
Stenmark
,
H.
and
Raiborg
,
C.
(
2017
).
PtdIns3P controls mTORC1 signaling through lysosomal positioning
.
J. Cell Biol
.
216
,
4217
-
4233
.
Inoki
,
K.
,
Li
,
Y.
,
Xu
,
T.
and
Guan
,
K.
(
2003
).
Rheb GTPase is a direct target of TSC2 GAP activity and regulates mTOR signaling
.
Genes Dev.
17
,
1829
-
1834
.
Itoh
,
T.
,
Fujita
,
N.
,
Kanno
,
E.
,
Yamamoto
,
A.
,
Yoshimori
,
T.
and
Fukuda
,
M.
(
2008
).
Golgi-resident small GTPase Rab33B interacts with Atg16L and modulates autophagosome formation
.
Mol. Biol. Cell
19
,
2916
-
2925
.
Itoh
,
T.
,
Kanno
,
E.
,
Uemura
,
T.
,
Waguri
,
S.
and
Fukuda
,
M.
(
2011
).
OATL1, a novel autophagosome-resident Rab33B-GAP, regulates autophagosomal maturation
.
J. Cell Biol.
192
,
839
-
853
.
Jiang
,
H.
and
Vogt
,
P. K.
(
2008
).
Constitutively active Rheb induces oncogenic transformation
.
Oncogene
27
,
5729
-
5740
.
Katoh
,
Y.
,
Nozaki
,
S.
,
Hartanto
,
D.
,
Miyano
,
R.
and
Nakayama
,
K.
(
2015
).
Architectures of multisubunit complexes revealed by a visible immunoprecipitation assay using fluorescent fusion proteins
.
J. Cell Sci.
128
,
2351
-
2362
.
Kim
,
A.
and
Cunningham
,
K. W.
(
2015
).
A LAPF/phafin1-like protein regulates TORC1 and lysosomal membrane permeabilization in response to endoplasmic reticulum membrane stress
.
Mol. Biol. Cell
26
,
4631
-
4645
.
Kim
,
D.-H.
,
Sarbassov
,
D. D.
,
Ali
,
S. M.
,
King
,
J. E.
,
Latek
,
R. R.
,
Erdjument-Bromage
,
H.
,
Tempst
,
P.
and
Sabatini
,
D. M.
(
2002
).
mTOR interacts with raptor to form a nutrient-sensitive complex that signals to the cell growth machinery
.
Cell
110
,
163
-
175
.
Kornmann
,
B.
and
Walter
,
P.
(
2010
).
ERMES-mediated ER-mitochondria contacts: molecular hubs for the regulation of mitochondrial biology
.
J. Cell Sci.
123
,
1389
-
1393
.
Korolchuk
,
V. I.
,
Saiki
,
S.
,
Lichtenberg
,
M.
,
Siddiqi
,
F. H.
,
Roberts
,
E. A.
,
Imarisio
,
S.
,
Jahreiss
,
L.
,
Sarkar
,
S.
,
Futter
,
M.
,
Menzies
,
F. M.
, et al. 
(
2011
).
Lysosomal positioning coordinates cellular nutrient responses
.
Nat. Cell Biol.
13
,
453
-
460
.
Kreis
,
T. E.
(
1990
).
Role of microtubules in the organisation of the golgi apparatus
.
Cell Motil. Cytoskeleton
15
,
67
-
70
.
Kubala
,
M. H.
,
Kovtun
,
O.
,
Alexandrov
,
K.
and
Collins
,
B. M.
(
2010
).
Structural and thermodynamic analysis of the GFP:GFP-nanobody complex
.
Protein Sci.
19
,
2389
-
2401
.
Laplante
,
M.
and
Sabatini
,
D. M.
(
2012
).
MTOR signaling in growth control and disease
.
Cell
149
,
274
-
293
.
Laplante
,
M.
and
Sabatini
,
D. M.
(
2013
).
Regulation of mTORC1 and its impact on gene expression at a glance
.
J. Cell Sci.
126
,
1713
-
1719
.
Li
,
X.
,
Rydzewski
,
N.
,
Hider
,
A.
,
Zhang
,
X.
,
Yang
,
J.
,
Wang
,
W.
,
Gao
,
Q.
,
Cheng
,
X.
and
Xu
,
H.
(
2016
).
A molecular mechanism to regulate lysosome motility for lysosome positioning and tubulation
.
Nat. Cell Biol.
18
,
404
-
417
.
Liu
,
X.
and
Zheng
,
X. F. S.
(
2007
).
Endoplasmic reticulum and golgi localization sequences for mammalian target of rapamycin
.
Mol. Biol. Cell
18
,
1073
-
1082
.
Liu
,
L.-K.
,
Choudhary
,
V.
,
Toulmay
,
A.
and
Prinz
,
W. A.
(
2016
).
An inducible ER-Golgi tether facilitates ceramide transport to alleviate lipotoxicity
.
J. Cell Biol.
216
.
Maejima
,
I.
,
Takahashi
,
A.
,
Omori
,
H.
,
Kimura
,
T.
,
Takabatake
,
Y.
,
Saitoh
,
T.
,
Yamamoto
,
A.
,
Hamasaki
,
M.
,
Noda
,
T.
,
Isaka
,
Y.
, et al. 
(
2013
).
Autophagy sequesters damaged lysosomes to control lysosomal biogenesis and kidney injury
.
EMBO J.
32
,
2336
-
2347
.
Manifava
,
M.
,
Smith
,
M.
,
Rotondo
,
S.
,
Walker
,
S.
,
Niewczas
,
I.
,
Zoncu
,
R.
,
Clark
,
J.
and
Ktistakis
,
N. T.
(
2016
).
Dynamics of mTORC1 activation in response to amino acids
.
Elife
5
,
1
-
21
.
Manning
,
B. D.
,
Tee
,
A. R.
,
Logsdon
,
M. N.
,
Blenis
,
J.
and
Cantley
,
L. C.
(
2002
).
Identification of the tuberous sclerosis complex-2 tumor suppressor gene product tuberin as a target of the phosphoinositide 3-kinase/Akt pathway
.
Cell
10
,
151
-
162
.
Matteoni
,
R.
and
Kreis
,
T. E.
(
1987
).
Translocation and clustering of endosomes and lysosomes depends on microtubules
.
J. Cell Biol.
105
,
1253
-
1265
.
Menon
,
S.
,
Dibble
,
C. C.
,
Talbott
,
G.
,
Hoxhaj
,
G.
,
Valvezan
,
A. J.
,
Takahashi
,
H.
,
Cantley
,
L. C.
and
Manning
,
B. D.
(
2014
).
Spatial control of the TSC complex integrates insulin and nutrient regulation of mtorc1 at the lysosome
.
Cell
156
,
1771
-
1785
.
Morita
,
S.
,
Kojima
,
T.
and
Kitamura
,
T.
(
2000
).
Plat-E: an efficient and stable system for transient packaging of retroviruses
.
Gene Ther.
7
,
1063
-
1066
.
Nada
,
S.
,
Hondo
,
A.
,
Kasai
,
A.
,
Koike
,
M.
,
Saito
,
K.
,
Uchiyama
,
Y.
and
Okada
,
M.
(
2009
).
The novel lipid raft adaptor p18 controls endosome dynamics by anchoring the MEK-ERK pathway to late endosomes
.
EMBO J.
28
,
477
-
489
.
Nobukuni
,
T.
,
Joaquin
,
M.
,
Roccio
,
M.
,
Dann
,
S. G.
,
Kim
,
S. Y.
,
Gulati
,
P.
,
Byfield
,
M. P.
,
Backer
,
J. M.
,
Natt
,
F.
,
Bos
,
J. L.
, et al. 
(
2005
).
Amino acids mediate mTOR/raptor signaling through activation of class 3 phosphatidylinositol 3OH-kinase
.
Proc. Natl. Acad. Sci. USA
102
,
14238
-
14243
.
Paz
,
I.
,
Sachse
,
M.
,
Dupont
,
N.
,
Mounier
,
J.
,
Cederfur
,
C.
,
Enninga
,
J.
,
Leffler
,
H.
,
Poirier
,
F.
,
Prevost
,
M.-C.
,
Lafont
,
F.
, et al. 
(
2010
).
Galectin-3, a marker for vacuole lysis by invasive pathogens
.
Cell. Microbiol.
12
,
530
-
544
.
Pende
,
M.
,
Um
,
S. H.
,
Mieulet
,
V.
,
Goss
,
V. L.
,
Mestan
,
J.
,
Mueller
,
M.
,
Fumagalli
,
S.
,
Kozma
,
S. C.
,
Thomas
,
G.
and
Sticker
,
M.
(
2004
).
S6K1-/-/S6K2-/- mice exibit preinatal lethality and rapamycin-sensitive 5′ -terminal oligopyrimidine mRNA translation and reveal a mitogen-activated protein kinase-dependent S6 kinase pathway
.
Mol. Cell. Biol.
24
,
3112
-
3124
.
Rocha
,
N.
,
Kuijl
,
C.
,
Van Der Kant
,
R.
,
Janssen
,
L.
,
Houben
,
D.
,
Janssen
,
H.
,
Zwart
,
W.
and
Neefjes
,
J.
(
2009
).
Cholesterol sensor ORP1L contacts the ER protein VAP to control Rab7-RILP-p150Glued and late endosome positioning
.
J. Cell Biol.
185
,
1209
-
1225
.
Roczniak-Ferguson
,
A.
,
Petit
,
C. S.
,
Froehlich
,
F.
,
Qian
,
S.
,
Ky
,
J.
,
Angarola
,
B.
,
Walther
,
T. C.
and
Ferguson
,
S. M.
(
2012
).
The transcription factor TFEB links mTORC1 signaling to transcriptional control of lysosome homeostasis
.
Sci. Signal.
5
,
ra42
.
Sancak
,
Y.
,
Peterson
,
T. R.
,
Shaul
,
Y. D.
,
Lindquist
,
R. A.
,
Thoreen
,
C. C.
,
Bar-Peled
,
L.
and
Sabatini
,
D. M.
(
2008
).
The Rag GTPases bind raptor and mediate amino acid signaling to mTORC1
.
Science
320
,
1496
-
1501
.
Sancak
,
Y.
,
Bar-Peled
,
L.
,
Zoncu
,
R.
,
Markhard
,
A. L.
,
Nada
,
S.
and
Sabatini
,
D. M.
(
2010
).
Ragulator-rag complex targets mTORC1 to the lysosomal surface and is necessary for its activation by amino acids
.
Cell
141
,
290
-
303
.
Saxton
,
R. A.
and
Sabatini
,
D. M.
(
2017
).
mTOR signaling in growth, metabolism, and disease
.
Cell
168
,
960
-
976
.
Schieke
,
S. M.
,
Phillips
,
D.
,
McCoy
,
J. P.
,
Aponte
,
A. M.
,
Shen
,
R.-F.
,
Balaban
,
R. S.
and
Finkel
,
T.
(
2006
).
The mammalian target of rapamycin (mTOR) pathway regulates mitochondrial oxygen consumption and oxidative capacity
.
J. Biol. Chem.
281
,
27643
-
27652
.
Shimobayashi
,
M.
and
Hall
,
M. N.
(
2016
).
Multiple amino acid sensing inputs to mTORC1
.
Cell Res.
26
,
7
-
20
.
Stoica
,
R.
,
De Vos
,
K. J.
,
Paillusson
,
S.
,
Mueller
,
S.
,
Sancho
,
R. M.
,
Lau
,
K.-F.
,
Vizcay-Barrena
,
G.
,
Lin
,
W.-L.
,
Xu
,
Y.-F.
,
Lewis
,
J.
, et al. 
(
2014
).
ER-mitochondria associations are regulated by the VAPB-PTPIP51 interaction and are disrupted by ALS/FTD-associated TDP-43
.
Nat. Commun.
5
,
3996
.
Taguchi-Atarashi
,
N.
,
Hamasaki
,
M.
,
Matsunaga
,
K.
,
Omori
,
H.
,
Ktistakis
,
N. T.
,
Yoshimori
,
T.
and
Noda
,
T.
(
2010
).
Modulation of local Ptdins3P levels by the PI phosphatase MTMR3 regulates constitutive autophagy
.
Traffic
11
,
468
-
478
.
Thiele
,
D. L.
and
Lipsky
,
P. E.
(
1990
).
Mechanism of L-leucyl-L-leucine methyl ester-mediated killing of cytotoxic lymphocytes: dependence on a lysosomal thiol protease, dipeptidyl peptidase I, that is enriched in these cells
.
Proc. Natl. Acad. Sci. USA
87
,
83
-
87
.
Thomas
,
J. D.
,
Zhang
,
Y.-J.
,
Wei
,
Y.-H.
,
Cho
,
J.-H.
,
Morris
,
L. E.
,
Wang
,
H.-Y.
and
Zheng
,
X. F. S.
(
2014
).
Rab1A is an mTORC1 activator and a colorectal oncogene
.
Cancer Cell
26
,
754
-
769
.
Uchimoto
,
T.
,
Nohara
,
H.
,
Kamehara
,
R.
,
Iwamura
,
M.
,
Watanabe
,
N.
and
Kobayashi
,
Y.
(
1999
).
Mechanism of apoptosis induced by a lysosomotropic agent, L-Leucyl-L-Leucine methyl ester
.
Apoptosis
4
,
357
-
362
.
Valm
,
A. M.
,
Cohen
,
S.
,
Legant
,
W. R.
,
Melunis
,
J.
,
Hershberg
,
U.
,
Wait
,
E.
,
Cohen
,
A. R.
,
Davidson
,
M. W.
,
Betzig
,
E.
and
Lippincott-Schwartz
,
J.
(
2017
).
Applying systems-level spectral imaging and analysis to reveal the organelle interactome
.
Nature
546
,
162
-
167
.
Wolfson
,
R. L.
,
Chantranupong
,
L.
,
Saxton
,
R. A.
,
Shen
,
K.
,
Scaria
,
S. M.
,
Cantor
,
J. R.
and
Sabatini
,
D. M.
(
2015
).
Sestrin2 is a leucine sensor for the mTORC1 pathway
.
Science (80-.)
351
,
43
-
48
.
Zatloukal
,
B.
,
Kufferath
,
I.
,
Thueringer
,
A.
,
Landegren
,
U.
,
Zatloukal
,
K.
and
Haybaeck
,
J.
(
2014
).
Sensitivity and specificity of in situ proximity ligation for protein interaction analysis in a model of steatohepatitis with mallory-denk bodies
.
PLoS ONE
9
,
e96690
.
Zhang
,
Y.
,
Gao
,
X.
,
Saucedo
,
L. J.
,
Ru
,
B.
,
Edgar
,
B. A.
and
Pan
,
D.
(
2003
).
Rheb is a direct target of the tuberous sclerosis tumour suppressor proteins
.
Nat. Cell Biol.
5
,
578
-
581
.
Zhang
,
J.
,
Kim
,
J.
,
Alexander
,
A.
,
Cai
,
S.
,
Tripathi
,
D. N.
,
Dere
,
R.
,
Tee
,
A. R.
,
Tait-Mulder
,
J.
,
Di Nardo
,
A.
,
Han
,
J. M.
, et al. 
(
2013
).
A tuberous sclerosis complex signalling node at the peroxisome regulates mTORC1 and autophagy in response to ROS
.
Nat. Cell Biol.
15
,
1186
-
1196
.
Zheng
,
J. Y.
,
Koda
,
T.
,
Fujiwara
,
T.
,
Kishi
,
M.
,
Ikehara
,
Y.
and
Kakinuma
,
M.
(
1998
).
A novel Rab GTPase, Rab33B, is ubiquitously expressed and localized to the medial Golgi cisternae
.
J. Cell Sci.
111
,
1061
-
1069
.
Zhou
,
X.
,
Clister
,
T. L.
,
Lowry
,
P. R.
,
Seldin
,
M. M.
,
Wong
,
G. W.
and
Zhang
,
J.
(
2015
).
Dynamic visualization of mTORC1 activity in living cells
.
Cell Rep.
10
,
1767
-
1777
.
Zoncu
,
R.
,
Efeyan
,
A.
and
Sabatini
,
D. M.
(
2011
).
mTOR: from growth signal integration to cancer, diabetes and ageing
.
Nat. Rev. Mol. Cell Biol.
12
,
21
-
35
.

Competing interests

The authors declare no competing or financial interests.

Supplementary information