Portal de Periódicos da CAPES

OCRL regulates lysosome positioning and mTORC1 activity through SSX2IP‐mediated microtubule anchoring

2021; Springer Nature; Volume: 22; Issue: 7 Linguagem: Inglês

10.15252/embr.202052173

1469-3178

Biao Wang, Wei He, Philipp P. Prosseda, Liang Li, Tia J. Kowal, Jorge A. Alvarado, Qing Wang, Yang Hu, Yang Sun,

Genetic and Kidney Cyst Diseases

Article13 May 2021free access Transparent process OCRL regulates lysosome positioning and mTORC1 activity through SSX2IP-mediated microtubule anchoring Biao Wang Biao Wang orcid.org/0000-0003-1049-8546 Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Wei He Wei He Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Philipp P Prosseda Philipp P Prosseda Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Liang Li Liang Li Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Tia J Kowal Tia J Kowal Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Jorge A Alvarado Jorge A Alvarado Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Qing Wang Qing Wang Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Yang Hu Yang Hu orcid.org/0000-0002-7980-1649 Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Yang Sun Corresponding Author Yang Sun [email protected] orcid.org/0000-0001-8735-4765 Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Palo Alto Veterans Administration, Palo Alto, CA, USA Search for more papers by this author Biao Wang Biao Wang orcid.org/0000-0003-1049-8546 Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Wei He Wei He Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Philipp P Prosseda Philipp P Prosseda Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Liang Li Liang Li Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Tia J Kowal Tia J Kowal Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Jorge A Alvarado Jorge A Alvarado Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Qing Wang Qing Wang Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Yang Hu Yang Hu orcid.org/0000-0002-7980-1649 Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Search for more papers by this author Yang Sun Corresponding Author Yang Sun [email protected] orcid.org/0000-0001-8735-4765 Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA Palo Alto Veterans Administration, Palo Alto, CA, USA Search for more papers by this author Author Information Biao Wang1, Wei He1, Philipp P Prosseda1, Liang Li1, Tia J Kowal1, Jorge A Alvarado1, Qing Wang1, Yang Hu1 and Yang Sun *,1,2 1Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, USA 2Palo Alto Veterans Administration, Palo Alto, CA, USA *Corresponding author. Tel: +1 650 724 3952; E-mail: [email protected] EMBO Reports (2021)22:e52173https://doi.org/10.15252/embr.202052173 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 Figures & Info Abstract Lysosomal positioning and mTOR (mammalian target of rapamycin) signaling coordinate cellular responses to nutrient levels. Inadequate nutrient sensing can result in growth delays, a hallmark of Lowe syndrome. OCRL mutations cause Lowe syndrome, but the role of OCRL in nutrient sensing is unknown. Here, we show that OCRL is localized to the centrosome by its ASH domain and that it recruits microtubule-anchoring factor SSX2IP to the centrosome, which is important in the formation of the microtubule-organizing center. Deficiency of OCRL in human and mouse cells results in loss of microtubule-organizing centers and impaired microtubule-based lysosome movement, which in turn leads to mTORC1 inactivation and abnormal nutrient sensing. Centrosome-targeted PACT-SSX2IP can restore microtubule anchoring and mTOR activity. Importantly, boosting the activity of mTORC1 restores the nutrient sensing ability of Lowe patients' cells. Our findings highlight mTORC1 as a novel therapeutic target for Lowe syndrome. SYNOPSIS OCRL is localized to the centrosome by its ASH domain and recruits microtubule anchoring factor SSX2IP to the centrosome, which is important in the formation of the microtubule organization center and regulation of lysosome positioning. Cells from Lowe syndrome patients and mouse models have persistent perinuclear lysosome positioning. OCRL deficiency causes mTORC1 inactivation and nutrient sensing defects. OCRL deficiency leads to failed centrosomal microtubule nucleation. OCRL localizes to centrosome via its ASH domain and recruits SSX2IP through its RhoGAP domain to regulate microtubule nucleation. Introduction Lysosome positioning and motility are tightly linked to mTORC1 activity and nutrient sensing (Korolchuk et al, 2011; Lawrence & Zoncu, 2019). In a nutrient-rich environment, lysosomes migrate to the cell periphery, which activates mTORC1 via membrane-bound signaling molecules (Cai et al, 2006; Korolchuk et al, 2011). In a nutrient-poor environment, lysosomes move to the perinuclear region and cluster at the microtubule-organizing center (MTOC), which inhibits mTOR activity (Matteoni & Kreis, 1987; Korolchuk et al, 2011). Defects in lysosomal positioning and mTOR dysfunction contribute to developmental and degenerative neurological diseases (Parenti et al, 2015; Sabatini, 2017). Lowe syndrome is a rare X-linked developmental neurologic disease characterized by eye abnormalities, renal proximal tubular dysfunction, and intellectual and growth retardation (Bokenkamp & Ludwig, 2016). Mutations in OCRL1, which encodes an inositol 5-phosphatase that dephosphorylates PI(4,5)P2 to PI4P, cause Lowe syndrome (Zhang et al, 1995). OCRL is a multidomain protein with a characteristic phosphatase domain, ASH (ASPM-SPD-2-Hydin) domain, and a RhoGAP-like domain (Erdmann et al, 2007). The ASH and RhoGAP-like domains mediate interaction of OCRL with proteins involved in trafficking, including Rab proteins, clathrins, and proteins associated with cytoskeleton, such as Cdc42 and Rac1 (Lowe, 2005; Erdmann et al, 2007; Hagemann et al, 2012). The ASH domain is also involved in centrosome-related processes (Ponting, 2006; Schou et al, 2014). Disease-causing mutations have been shown in the phosphatase domain (Bokenkamp & Ludwig, 2016) and in non-catalytic ASH-RhoGAP domains (Erdmann et al, 2007; McCrea et al, 2008). However, the underlying mechanisms linking non-catalytic OCRL mutations to Lowe syndrome remain elusive. Lysosome has been increasingly recognized to play critical roles in degenerative diseases of the brain, eye, and kidneys (Parenti et al, 2015; Festa et al, 2018). Early studies of Lowe syndrome highlighted autophagosome–lysosome fusion dysfunction and elevated serum lysosomal enzymes in affected patients (Ungewickell & Majerus, 1999; De Leo et al, 2016). More recently, lysosome positioning and mTOR signaling defects have been linked to intellectual disability and growth retardation, two major features of Lowe syndrome (Crino, 2011; Parenti et al, 2015; Sabatini, 2017). Because developmental delay can occur in patients without renal dysfunction (Abdalla et al, 2018), the systemic growth retardation in this disease could be related to an inability to correctly respond to nutrient and growth factor signals. Clinical data show that induction of tubular low molecular weight proteinuria, one of the main kidney dysfunction phenotypes in Lowe syndrome, can occur in patients treated with the mTOR inhibitor rapamycin (Fantus et al, 2016). Whether mTOR signaling is involved in Lowe syndrome and how lysosome dysfunction contributes to disease pathogenesis remain unclear. In this study, we investigated the role of lysosome positioning and the mTOR pathway in Lowe syndrome. Results Persistent perinuclear lysosome positioning in OCRL-deficient cells We examined lysosome distribution in cells of OCRL-deficient patients (Lowe 1676 and 3265 fibroblasts), and in embryonic fibroblasts (MEFs) prepared from OCRL knockout mice and from a humanized Lowe syndrome mouse model (IOB) (Bothwell et al, 2011). LAMP1 immunostaining of control cells (normal human fibroblasts, NHF) and wild type MEFs showed peripheral localization of lysosomes in complete medium and perinuclear localization after 5 h of starvation (Figs 1A–E and EV1A and B). In contrast, OCRL-deficient cells exhibited persistent perinuclear localization in both fed and starved conditions (Figs 1A–E and EV1A and B). This persistent perinuclear lysosome localization in Lowe 1676 cells could be rescued by re-expression wild-type OCRL and the enzymatic deficient OCRL (D499A mutant) also showed a slight rescued effect (Fig EV1C and D), indicating that lysosomal abnormality was caused by OCRL loss. Next, we examined lysosomal biosynthesis by Western blot. We found no significant difference in the LAMP1 protein expression between normal and OCRL-deficient cells (Fig EV1E and F). Subsequently, we analyzed the dynamics of lysosome movement via transfection of a fluorescently labeled lysosome marker (LAMP1-GFP) in live cells. Lysosomes in Lowe 1676 cells were significantly larger than in NHF cells (Fig EV1G and H, Movies EV1 and EV2). To determine lysosome mobility, we used single particle tracking of NHF and Lowe 1676 cells transfected with LAMP1-GFP (Fig 1F). We analyzed lysosome mobility in the peripheral region of these cells; single particle tracking revealed that the lysosomes moved significantly slower in Lowe 1676 cells than in NHF cells (Fig 1G and H). To further confirm the lysosome mobility defects in OCRL knockout and IOB MEFs, we analyzed the formation of lysosome tubulation, which is dependent on lysosome anterograde movement (Li et al, 2016). During an 8-h starvation period, lysosome tubulation occurred in the wild-type MEFs, but not in the OCRL-deficient MEFs (Fig 1I and J, Movies EV3–EV5). We therefore concluded that OCRL deficiency prevents anterograde trafficking of lysosomes and consequently leads to a persistent perinuclear lysosome-positioning defect. Figure 1. OCRL deficiency leads to perinuclear lysosome positioning A. NHF, Lowe 1676, and Lowe 3265 cells were grown in complete media or 5 h serum starvation, then immunostained with LAMP1 antibody. Dashed lines in images outline cell boundaries. Scale bar, 10 μm. B. Wild-type (Ocrl+/y) and Ocrl−/− MEF cells were grown in complete media or 5 h serum starvation, then immunostained with indicated antibodies. Dashed lines in images outline cell boundaries. Scale bar, 10 μm. C. Quantitative analyses of lysosome perinuclear distribution in the experiments shown in A. The intracellular distribution of LAMP1-positive vesicles was quantified as described in the Materials and Methods. n = 3 per group in which 30 cells were counted per condition per experiment. D, E. Quantitative analyses of lysosome and mTOR perinuclear distribution in the experiments shown in B. The intracellular distribution of mTOR-positive vesicles was quantified as described in the Materials and Methods. n = 3 per group in which 30 cells were counted per condition per experiment. F, G. Representative images of the trajectory of GFP signals from NHF (n = 4) and Lowe 1676 (n = 4) cells transfected with lysosome-GFP. Trajectories were obtained as described (He et al, 2016) and labeled in different colors with the same colored circles showing the start of the track. G. Measured MSD (⟨∆r2 (t)⟩) as a function of delay time t for the mobile lysosome trajectories taken at F. H. Diffusion coefficient of lysosome in the experiments shown in G. Means ± SEM of n = 4 cells per group. I. Snap images of WT (Ocrl+/y), OCRL knockout (Ocrl−/−), and IOB MEFs transfected with lysosome-GFP. After transfection, WT, OCRL knockout, and IOB MEFs were left in serum starvation for 8 h. Scale bar, 5 μm. J. Quantitative analyses of tubulated lysosomes in the experiments shown in H. n = 3 per group in which 10 cells were counted per condition per experiment. Data information: In (C,D,E,J), data are presented as mean ± SEM. *P ≤ 0.05, **P ≤ 0.01, ***P≤ 0.001, ****P ≤ 0.0001, n.s., not significant, calculated by two-tailed Student's t-test. Download figure Download PowerPoint Click here to expand this figure. Figure EV1. Lysosome and mTOR positioning defects in OCRL-deficient cells A. WT (Ocrl+/y) and IOB MEFs cultured either in complete medium or serum starvation medium for 5 h were stained with LAMP1 antibody. Dashed lines in images outline cell boundaries. Scale bar, 5 μm. B. Quantitative analyses of lysosome perinuclear distribution in the experiments shown in A. n = 3 per group in which 30 cells were counted per condition per experiment. C. Lowe 1676 transfected with GFP-OCRL-WT, GFP-OCRL-D499A, GFP-OCRL-△ RhoGAP, and GFP-OCRL-△ASH+RhoGAP for 72 h were grown in complete media, then immunostained with LAMP1 antibody. Dashed lines in images outline cell boundaries. Scale bar, 10 μm. D. Quantitative analyses of lysosome peripheral distribution in the experiments shown in C. The intracellular distribution of LAMP1-positive vesicles was quantified as described in the Materials and Methods. n = 3 per group in which 30 cells were counted per condition per experiment. E, F. Immunoblot analysis of endogenous LAMP1 and GAPDH in NHF, Lowe 1676 and NHF cells treated with OCRL siRNA or control siRNA for 48 h. LAMP1 over GAPDH ratios normalized to NHF cells, three independent experiments. G. Snap images of NHF and Lowe 1676 cells transfected with lysosome-GFP are shown. Scale bar, 10 μm. H. Lysosome size in H was measured. n = 3 per group in which 20 cells were counted per condition per experiment. I. Quantitative analyses of mTOR perinuclear distribution in the experiments shown in A. n = 3 per group in which 20 cells were counted per condition per experiment. Data information: In (B,D,H,I), data are presented as mean ± SEM. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, n.s., not significant, calculated by two-tailed Student's t-test. Download figure Download PowerPoint mTORC1 inactivation and nutrient sensing abnormality in OCRL-deficient cells In the presence of nutrients, mTORC1 is recruited and activated at the surface of peripheral lysosomes (Cai et al, 2006; Korolchuk et al, 2011). Since we observed that numbers of peripheral lysosomes were reduced in OCRL-deficient cells, we hypothesized that mTOR localization might also be affected. In wild-type MEFs, mTOR was localized peripherally in fed conditions and in the perinuclear region in starved conditions (Fig 1B and E). However, OCRL-deficient cells exhibited persistent perinuclear mTOR localization in both fed and starved conditions (Fig 1B and E), indicating that OCRL depletion does not affect the recruitment of mTOR to the lysosomal membrane. mTOR staining of IOB MEFs confirmed this phenotype (Fig EV1A and I). Consistent with these results for mTOR localization, we found that in wild-type cells, serum starvation decreased mTORC1 activity compared with fed conditions, as assayed by the phosphorylation status of mTOR Ser2448 and p70 S6 Kinase (Thr389), whereas in OCRL-deficient cells mTOR activity was consistently lower than in wild-type cells in fed conditions, and also less sensitive to serum starvation (Figs 2A and B, and EV2A and B). Importantly, retina and brain tissues derived from OCRL knockout and Lowe syndrome mice (IOB) also showed decreased mTORC1 activity (Fig EV2C). In addition, we found that reintroducing OCRL in Lowe 1676 cells could significantly restore mTORC1 activity (Fig 2C). We further investigated whether OCRL enzyme activity contributes to the mTOR signaling pathway. However, OCRL inhibitor YU142670 did not affect the phosphorylation status of S6K in NHF cells (Fig EV2D) and reintroducing OCRL-D499A, a mutant with no phosphatase activity, in Lowe 1676 cells partially increased p70 S6 Kinase (Thr389) activity (Fig EV2E and F). These results show that mTORC1 activity is relatively low in OCRL-deficient cells and tissues, which could result in serum sensing defects. We also analyzed the effect of both serum and amino acid starvation, which is more physiologic than serum starvation, to further confirm whether decreased mTORC1 activity could affect nutrient sensing. We found that mTORC1 activity was restored significantly more slowly after nutrient replenishment in Lowe patient-derived fibroblasts than in controls, as assayed by the phosphorylation status of S6K and mTOR Ser2448 (Fig 2D–F). Figure 2. OCRL deficiency impairs lysosome positioning leading to mTOR inactivation and abnormal nutrient sensing Immunoblot analysis of endogenous p-mTOR (Ser2448), total mTOR, p-p70 S6K (Thr389), and total p70 S6K in NHF, Lowe 1676 and Lowe 3265 cells grown in complete media or 5 h serum starvation. p-p70 S6K (Thr389) over total p70 S6K ratios normalized to NHF cells, three independent experiments. Immunoblot analysis of endogenous p-mTOR (Ser2448), total mTOR, p-p70 S6K (Thr389), and total p70 S6K in WT (Ocrl+/y) and IOB MEFs grown in complete medium or 5 h serum starvation. p-p70 S6K (Thr389) over total p70 S6K ratios normalized to NHF cells, three independent experiments. Immunoblot analysis of GFP, p-mTOR (Ser2448), total mTOR, p-p70 S6K (Thr389), and total p70 S6K in Lowe 1676 transfected with or without GFP-OCRL. NHF and Lowe 1676 cells were grown in complete media, amino acid (aa)/FBS starved for 5 h, or starved and then recovered in amino acid/FBS-containing medium, then immunoblotted using antibodies against endogenous p-mTOR (Ser2448), total mTOR, p-p70 S6K (Thr389), and total p70 S6K. Quantification of p-mTOR (Ser2448) over total mTOR ratios normalized to NHF cells as shown in D, n = 3. Quantification of p-p70 S6K (Thr389) over total p70 S6K ratios normalized to NHF cells as shown in D, n = 3. WT and IOB MEFs were grown in 1% FBS, 15% FBS media with or without 2 μmol MHY1485 for 4 days. Relative cell number is the ratio of cells on day 4 normalized to day 1, three independent experiments. ****, ####P < 0.0001, calculated by two-way ANOVA. NHF and Lowe 1676 cells were grown in 1% FBS, 15% FBS media with or without 2 μmol MHY1485 for 4 days. Relative cell number is the ratio of cells on day 4 normalized to day 1, three independent experiments. ****, ####P < 0.0001, calculated by two-way ANOVA. Data information: In (A–F), data are presented as mean ± SEM. *P ≤ 0.05, **P ≤ 0.01, ***P≤ 0.001, ****P ≤ 0.0001, n.s., not significant, calculated by two-tailed Student's t-test unless otherwise stated. Download figure Download PowerPoint Click here to expand this figure. Figure EV2. Disturbed nutrient sensing ability in Lowe patient's cells can be rescued by mTOR activator MHY1485 Quantification of p-mTOR over total mTOR ratios normalized to NHF cells in the experiments shown in Fig 2A, three independent experiments. Quantification of p-mTOR over total mTOR ratios normalized to NHF cells in the experiments shown in Fig 2B, three independent experiments. Immunoblot analysis of endogenous p-mTOR(Ser2448) and total mTOR in retina and brain tissues from WT (Ocrl+/y) and IOB mouse. p-mTOR over total mTOR ratios normalized to WT mouse tissues, three independent experiments. NHF cells were either treated with 0.1% (v/v) DMSO or 50 μM YU142670 for indicated time, then immunoblotted using antibodies for endogenousp-p70 S6K(Thr389) and total p70 S6K. Immunoblot analysis of GFP, p-mTOR (Ser2448), total mTOR, p-p70 S6K (Thr389), and total p70 S6K in Lowe 1676 cells transfected with GFP-OCRL-WT, GFP-OCRL-D499A, and GFP-OCRL-△RhoGAP. Quantification of p-p70 S6K (Thr389) over total p70 S6K ratios normalized to NHF cells in the experiments shown in E, n = 3. Lowe 1676 and Lowe 3265 cells were either treated with 0.1% (v/v) DMSO or 10 μM MHY1485 in DMSO treated for 24 h, then immunoblotted using antibodies for endogenous p-mTOR(Ser2448), total mTOR, p-p70 S6K(Thr389), and total p70 S6K. p-mTOR over total mTOR ratios normalized to NHF cells, three independent experiments. Quantification of p-p70 S6K(Thr389) over total p70 S6K ratios normalized to NHF cells in the experiments shown in G, n = 3. Immunoblot analysis of TSC2 and GAPDH in NHF cells treated with siRNA control or TSC2 siRNA (50 nM). Quantification of TSC2 protein expression level was normalized to GAPDH. Three independent experiments. NHF cells treated with siRNA control or TSC2 siRNA (50 nM) were grown in 15% FBS media for 5 days. Relative cell number is the ratio of cells on indicated day normalized to day 1, three independent experiments. Data information: In (A,B,C,F,G,H,I), data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, calculated by two-tailed Student's t-test. In (J), data are presented as mean ± SEM. ****P < 0.0001, calculated by two-way ANOVA. Download figure Download PowerPoint Since mTORC1 functions as a central hub of nutrient signaling and cell growth (Sabatini, 2017; Lawrence & Zoncu, 2019), we next investigated the growth rate of OCRL-deficient cells. We detected a significantly slower growing rate of serum-stimulated IOB MEFs and Lowe patient cells compared with serum-stimulated control NHF cells (Fig 2G and H). Based on the mTORC1 activity phenotype that we observed, we proposed that boosting the activity of mTOR in OCRL-depleted cells might rescue the proliferation defect. For this scope, we used MHY1485, a previously reported cell-permeable mTOR activator which could stimulate mTOR, S6K1, and rpS6 phosphorylation (Cheng et al, 2015) and found that it could successfully increase mTORC1 activity in Lowe patient cells (Fig EV2G and H). Treatment with MHY1485 also restored the proliferation rate of Lowe model cells (Fig 2G and H). In support of this, knockdown of TSC2 protein, a well-known negative regulator of mTORC1, also increased the proliferation rate of Lowe model cells (Fig EV2I and J). These results suggested that mTORC1 inactivation was responsible for the difference in proliferation. Taken together, these results lead us to conclude that the low proliferation rate observed in OCRL-deficient cells was caused by mTORC1 inactivation. OCRL loss leads to defective centrosomal microtubule nucleation that impairs microtubule-based vesicle transport Lysosome transport is a microtubule-dependent process (Matteoni & Kreis, 1987). Perinuclear movement of lysosomes, which are typically near the microtubule-organizing center, requires lysosomal calcium channel TRPML1 activity (Li et al, 2016). Previous studies have shown that depleting OCRL leads to an accumulation of PI(4,5)P2 on lysosome membrane, which inhibits lysosomal TRPML1 activity (De Leo et al, 2016). Decreased TRPML1 activity should result in peripheral localization of lysosomes (Li et al, 2016). However, we and others (De Leo et al, 2016) observed lysosomes in the perinuclear region of OCRL-deficient cells (Fig 1A), suggesting that regulation of lysosome positioning occurs by an alternative pathway in this condition. Because in normal conditions lysosomes traffic along microtubules, we hypothesized that there was a microtubule defect in these OCRL-deficient cells. To examine microtubule dynamics in OCRL-deficient cells, we performed immunofluorescence staining of microtubules (α-, β-tubulins) and a centrosome protein (γ-tubulins). We found that OCRL-deficient cells were characterized by non-radial microtubule arrays that failed to anchor to the centrosome, indicating impaired MTOC function (Fig 3A–D). To confirm a possible MTOC abnormality in OCRL-deficient cells, we used siRNA to deplete OCRL in RPE-1 cells (Fig EV3-EV5). We verified OCRL depletion with a previously published OCRL antibody (Luo et al, 2012) and confirmed the specificity of this antibody using Lowe patient cells (Fig EV3B). We also found that exogenous expression of OCRL could restore microtubule nucleation in Lowe patient cells (Fig EV3E). Proper formation of radial microtubule arrays requires microtubule nucleation. To further assess whether OCRL-deficient cells exhibited defects in MTOC microtubule nucleation, we performed a microtubule regrowth assay. Microtubules depolymerize upon exposure to nocodazole and ice, depolymerization is reversed when cells are restored to 37°C. After 10 min of microtubule regrowth, NHF cells displayed efficient initial nucleation in the form of robust radial microtubule arrays. Lowe patient cells, however, exhibited inefficient initial nucleation with less distinct focal structures (Fig 3E and F). Taken together, these results led us to conclude that OCRL plays a novel role as a microtubule nucleation factor. Figure 3. OCRL deficiency causes failed microtubule nucleation Representative images of NHF, Lowe 1676, and Lowe 3265 cells immunostained with anti-α-tubulin (red) and anti-γ-tubulin (green) antibodies. Scale bar, 10 μm. Quantitative analyses of focused microtubule arrays numbers at γ-tubulin (microtubule-organizing center MTOC) measured in the experiments shown in A. n = 3 cells per group, three independent experiments. WT (Ocrl+/y), OCRL knockout (Ocrl−/−), and IOB MEFs derived from the corresponding mice were costained with anti-α-tubulin and anti- γ-tubulin antibodies. Nuclei were stained with DAPI (blue). Scale bar, 10 μm. Quantitative analyses of focused microtubule arrays numbers at γ-tubulin (microtubule-organizing center MTOC) measured in the experiments shown in C. n = 3 per group in which 30 cells were counted per condition per experiment. Microtubule regrowth assay was performed on NHF, Lowe 1676, and Lowe 3265 cells with 20 μM nocodazole for 2 h on ice, followed by rinse with PBS, and addition of pre-warmed medium to induce microtubule regrowth. Cells were then immunostained with anti-α-tubulin and anti-γ-tubulin antibodies at the indicated time point after addition of pre-warmed medium. The magnified view shows the cytoplasmic microtubules organization center. Scale bar, 10 μm. Quantitative analyses of focused microtubule arrays numbers at γ-tubulin (microtubule-organizing center, MTOC) measured in the experiments shown in E. n = 3 per group in which 30 cells were counted per condition per experiment. Data information: In (B,D,F), data are presented as mean ± SEM. *P ≤ 0.05, **P ≤ 0.01, calculated by two-way ANOVA, multiple comparisons. Download figure Download PowerPoint Click here to expand this figure. Figure EV3. Perinuclear positioned mitochondria in OCRL-deficient cells Western blot verification of OCRL deficient in RPE-1cells treated with 48-h OCRL siRNA using characterized OCRL antibody (25). OCRL over GAPDH ratios normalized to siControl cells, n = 3. Western blot verification of OCRL deficiency in Lowe patient cells characterized antibody(25). OCRL over GAPDH ratios normalized to NHF cells, n = 3. RPE-1 cells transfected with OCRL siRNA for 48 h, then immunostained using anti- β-tubulin and anti- γ-tubulin antibodies. Nuclei were stained with DAPI (blue). Mismatched siRNA were used as controls. Scale bar, 10 μm. Quantitative analyses of focused microtubule asters at γ-tubulin (microtubule-organizing center, MTOC) was measured in the experiments shown in C. n = 3 per group in which 20 cells were counted per condition per experiment. Lowe 1676 cells were transfected with GFP-OCRL for 48 h, then immunostained with anti- β-tubulin antibody. The centrosomal microtubule nucleation in transfected cells was indicated by white arrowhead. Scale bars, 10 µm. NHF and Lowe 1676 cells were grown in complete medium and immunostained with anti- EEA1 antibody. Scale bars, 10 µm. Quantitative analyses of EEA1 perinuclear distribution in the experiments shown in F. n = 3 per group in which 20 cells were counted per condition per experiment. NHF and Lowe 1676 cells were grown in complete media, amino acid (aa)/FBS starved for 1 h, or starved and then recovered in complete media medium with 6 ng/ml insulin for 5 mins, then immunoblotted using antibodies against endogenous p-Akt (Ser473) and total Akt. p-Akt (Ser473) over total Akt ratios normalized to NHF cells are shown under each lane. Data information: In (A,B,D,F), data are presented as mean ± SEM. *P ≤ 0.05, ***P ≤ 0.001, n.s., not significant, calculated by two-tailed Student's t-test. Download figure Download PowerPoint Click here to expand this figure. Figure EV4. Rab8 does not mediate the centrosomal OCRL localization Representative images of RPE-1 cells immunostained with anti-OCRL (red) and anti-centrin3 (green) antibodies. Nuclei (blue) were stained with DAPI. RPE-1 cells were cultured in complete medium or serum starved medium for 24 h. Scale bar, 10 μm. Representative images of RPE-1 cells treated with 50 nm OCRL siRNA for 48 h, then immunostained with anti-OCRL (green) and anti-γ-tubulin (red) antibodies. Nuclei (blue) were stained with DAPI. Scale bar, 10 μm. Representative images of RPE-1 cells were treated with mismatched control siRNA or OCRL siRNA for 48 h, then immunostai