Gp78 is a cell surface receptor that also functions as an E3 ubiquitin ligase in the endoplasmic reticulum (ER)-associated degradation (ERAD) pathway. The Gp78 ligand, the glycolytic enzyme phosphoglucose isomerase (PGI; also called autocrine motility factor, AMF), functions as a cytokine upon secretion by tumor cells. AMF is internalized through a PI3K- and dynamin-dependent raft endocytic pathway to the smooth ER; however, the relationship between AMF and Gp78 ubiquitin ligase activity remains unclear. AMF uptake to the smooth ER is inhibited by the dynamin inhibitor, dynasore, is reduced in Gp78 knockdown cells and induces the dynamin-dependent downregulation of its cell surface receptor. AMF uptake is Rac1-dependent and is inhibited by expression of dominant-negative Rac1 and the Rac1 inhibitor NSC23766, and is therefore distinct from Cdc42- and RhoA-dependent raft endocytic pathways. AMF stimulates Rac1 activation, but this is reduced by dynasore treatment and is absent in Gp78-knockdown cells; therefore, AMF activities require Gp78-mediated endocytosis. AMF also prevents Gp78-induced degradation of the mitochondrial fusion proteins, mitofusin 1 and 2 in a dynamin-, Rac1- and phosphoinositide 3-kinase (PI3K)-dependent manner. Gp78 induces mitochondrial clustering and fission in a manner dependent on GP78 ubiquitin ligase activity, and this is also reversed by uptake of AMF. The raft-dependent endocytosis of AMF, therefore, promotes Rac1-PI3K signaling that feeds back to promote AMF endocytosis and also inhibits the ability of Gp78 to target the mitofusins for degradation, thereby preventing Gp78-dependent mitochondrial fission. Through regulation of an ER-localized ubiquitin ligase, the raft-dependent endocytosis of AMF represents an extracellular regulator of mitochondrial fusion and dynamics.
Gp78 is a key E3 ubiquitin ligase in the endoplasmic reticulum (ER)-associated degradation (ERAD) pathway that targets misfolded and physiological substrates for ubiquitylation and subsequent degradation by the proteasome (Christianson et al., 2011; Fang et al., 2001). Known Gp78 substrates include CD3-delta, the T cell receptor, ApoB lipoprotein, HMG-CoA reductase, cystic fibrosis transmembrane conductance regulator and the metastasis suppressor KAI1 (Fang et al., 2001; Liang et al., 2003; Morito et al., 2008; Song et al., 2005; Zhong et al., 2004). Gp78 overexpression in mammary gland induces preneoplastic hyperplasia, ductal branching, dense alveolar lobule formation and also KAI1 downregulation (Joshi et al., 2010). In metastatic HT-1080 fibrosarcoma cells, Gp78 E3 ubiquitin ligase targeting of KAI1 promotes sarcoma metastasis (Tsai et al., 2007). Gp78 targets ER-localized HMG-CoA reductase and Insig-1, involved in cholesterol regulation (Lee et al., 2006; Liu et al., 2012; Song et al., 2005), and liver-specific Gp78 knockdown mice show reduced degradation of HMG-CoA reductase and Insig-1/2 and are protected from diet or age-induced obesity and glucose intolerance (Liu et al., 2012). We have recently identified the mitochondria fusion proteins (mitofusin 1 and 2; Mfn1/2) as Gp78 substrates and shown that Gp78 promotes mitochondrial fission and mitophagy of depolarized mitochondria (Fu et al., 2013). Via degradation of specific substrates Gp78 therefore regulates various physiological processes but also tumor progression and metastasis. However, how Gp78 ubiquitin ligase activity in ERAD is regulated is poorly understood.
Gp78 is also the cell surface receptor for autocrine motility factor (AMF) (Nabi et al., 1990; Silletti et al., 1991). AMF is phosphoglucose isomerase (PGI), a glycolytic enzyme that plays an essential role in gluconeogenesis (Watanabe et al., 1996) and that also functions as a cytokine upon non-classical secretion by tumor cells (Fairbank et al., 2009; Funasaka and Raz, 2007; Kanbe et al., 1994; Silletti et al., 1996; Tsutsumi et al., 2002; Tsutsumi et al., 2004; Yanagawa et al., 2004). AMF stimulates tumor cell motility and proliferation through activation of protein kinase C, the small GTPases RhoA and Rac1 and p27Kip1 inducing reorganization of focal contacts and tumor cell migration (Kanbe et al., 1994; Silletti et al., 1996; Tsutsumi et al., 2002). AMF also exhibits anti-apoptotic activity and chemoprotective ability, promoting tumor cell survival via PI3K/Akt signaling (Haga et al., 2003; Kojic et al., 2008; Tsutsumi et al., 2003) and protecting against the ER stress response through both PI3K/Akt signaling and regulation of ER calcium release (Fu et al., 2011). Gp78 internalizes AMF via a non-caveolar, dynamin- and PI3K-dependent raft-mediated endocytic pathway to the smooth ER (Benlimame et al., 1998; Kojic et al., 2007; Kojic et al., 2008; Le et al., 2002; Le and Nabi, 2003). We recently localized the initiation of Gp78-dependent substrate ubiquitylation to the peripheral smooth ER domain (St-Pierre et al., 2012). However, the relationship between the raft-dependent endocytosis of AMF, its signaling and Gp78 function in ERAD remains unknown.
We show here that raft-dependent Gp78-mediated AMF endocytosis is Rac1-dependent and required for downstream AMF activation of Rac1, defining a distinct role for raft-dependent endocytosis in AMF/Gp78 signaling. We further show that the Rac1-dependent endocytosis of AMF to the smooth ER inhibits Gp78 ubiquitin ligase activity preventing Gp78-dependent mitofusin degradation and Gp78-mediated mitochondrial fission. AMF regulation of Gp78 defines novel extracellular mechanisms of regulation of a key component of ERAD.
Dynamin-dependent internalization of AMF downregulates its cell surface receptor Gp78
Raft-dependent pathways include endocytosis via caveolae as well as non-caveolar pathways such as dynamin-dependent RhoA and dynamin-independent Cdc42 endocytosis (Lajoie and Nabi, 2010; Sandvig et al., 2011). To test whether AMF follows dynamin-dependent or -independent pathways, we treated cells with the reversible small molecule dynamin inhibitor, dynasore (Macia et al., 2006). Uptake of FITC-conjugated AMF (AMF-FITC), transferrin (Tf-FITC) and cholera toxin b-subunit (Ctb-FITC) were analyzed by flow cytometry of protease-treated MDA-435 and HT-1080 cells treated with either dynasore or mβCD, a cholesterol disrupting agent that inhibits raft-dependent endocytosis. As shown in Fig. 1A, pretreatment of cells with dynasore prevented internalization of AMF and Tf, but not Ctb while mβCD prevented AMF and Ctb uptake, but not Tf. As reported previously (Benlimame et al., 1998; Le et al., 2002; Le and Nabi, 2003), AMF-FITC is internalized to the smooth ER labeled with the 3F3A anti-Gp78 mAb in HT-1080 fibrosarcoma cells (Fig. 1B). Following dynasore treatment, cell-associated AMF-FITC intensity was reduced by ∼75%, paralleling data obtained by flow cytometry, and AMF-FITC showed reduced association with the 3F3A-labeled smooth ER.
Doxycyclin inducible Gp78-targeted shRNA lentiviral vectors incorporating two shRNAs (sh2 and sh6) that effectively targeted Gp78 (St-Pierre et al., 2012) were used to generate Gp78 knockdown HT-1080 cells. Doxycyclin treatment resulted in reduced Gp78 expression by western blot that was generally more pronounced in sh6 transfected cells (Fig. 2A). Cell surface Gp78 expression assessed by flow cytometry analysis was significantly reduced in sh6, but not sh2, expressing HT-1080 cells and AMF-FITC uptake was significantly reduced in both sh2 and sh6 expressing cells (Fig. 2B). Furthermore, AMF-FITC internalization was associated with a reduction of cell surface Gp78 expression in both HT-1080 and MDA-435 tumor cells (Fig. 2C). Treatment with dynasore inhibited AMF uptake and also prevented AMF-induced loss of cell surface Gp78. Washout of dynasore prior to addition of AMF-FITC partially reversed the inhibition of AMF-FITC internalization and the loss of cell surface Gp78. AMF internalization is therefore associated with downregulation of its cell surface receptor.
AMF-Rac1 signaling and endocytosis regulation
To test the small Rho GTPase dependence of AMF endocytosis, HT-1080 cells were transfected with wild-type (WT), dominant active (DA) and dominant negative (DN) c-Myc-tagged RhoA, Rac1 and Cdc42 plasmids and then incubated with AMF-FITC for 30 minutes. Upon expression of DA-Rac1, AMF-FITC showed significantly elevated association with the 3F3A-labeled smooth ER compared to WT-Rac1 and DN-Rac1 transfected cells (Fig. 3A). Quantification of cell-associated AMF-FITC intensity showed that DA-Rac1 transfection increased while DN-Rac1 reduced AMF-FITC internalization (Fig. 3A). Expression of WT, DA or DN RhoA or Cdc42 did not impact on AMF-FITC uptake (Fig. 3A). AMF internalization was previously shown to be PI3K-dependent (Kojic et al., 2007) and AMF uptake was inhibited by dynasore, the Rac1-inhibitor NSC23766 and the PI3K inhibitor LY294002 without altering the surface expression of Gp78 (Fig. 3B). The raft-dependent endocytosis of AMF is therefore dynamin-dependent and requires PI3K-Rac1 signalling.
AMF was previously reported to induce activation of RhoA and Rac1, but not Cdc42, in A375 melanoma cells (Tsutsumi et al., 2002). We observed the robust, rapid and transient activation of Rac1 30–60 minutes after AMF treatment of HT-1080 cells but did not detect activation of RhoA (Fig. 4A). Importantly, Gp78 knockdown completely abrogated AMF induced Rac1 signaling (Fig. 4B) demonstrating that extracellular AMF signals through Gp78 to activate Rac1. Dynasore treatment alone did not affect basal Rac1 activation status but did reduce AMF activation of Rac1 in both MDA-435 and HT-1080 cells (Fig. 4C). Gp78-mediated, raft-dependent endocytosis of AMF therefore enhances its downstream signaling to Rac1.
AMF inhibits Gp78-dependent mitofusin degradation and mitochondrial fission
Gp78 ubiquitin ligase activity induces mitochondrial fission and degradation of the mitochondrial fusion proteins, Mitofusin 1 and 2 (Mfn1 and Mfn2) (Fu et al., 2013). Overexpression of full-length Gp78 in Cos7 cells results in degradation of both Mfn1 and Mfn2 and is inhibited by AMF treatment. Treatment of Gp78 overexpressing cells with dynasore, the Rac1-inhibitor NSC23766 and the PI3K inhibitor LY294002 does not affect degradation of Mfn1 and Mfn2 (Fig. 5A). However, these inhibitors of AMF endocytosis prevent AMF inhibition of Gp78 ubiquitin ligase activity. HT-1080 cells express elevated levels of Gp78 that efficiently targets KAI1 for degradation (St. Pierre et al., 2012). Similarly, Gp78 knockdown in sh6 transfected HT-1080 cells, although not sh2 transfected cells, results in elevated levels of Mfn1 and Mfn2 similar to those observed following AMF treatment of untransfected and control shRNA transfected HT-1080 cells (Fig. 5B). The basis for the retention of effective targeting of Mfn1/Mfn2 in sh2 Gp78 knockdown cells is not clear, however these cells do not generally exhibit as robust or consistent a Gp78 knockdown as sh6 knockdown cells. AMF treatment did not affect Mfn1 and Mfn2 levels in sh6 Gp78 knockdown cells. Interestingly, while we observed increased expression of the Gp78 substrate KAI1 in sh6 Gp78 knockdown cells, it was not reversed by AMF treatment suggesting that AMF selectively impacts the ability of Gp78 to target the mitofusins for degradation (Fig. 5B). As observed for Gp78 transfected Cos7 cells, reduction of Mfn1 and Mfn2 degradation by AMF treatment of HT-1080 cells is prevented by inhibiting AMF endocytosis with dynasore, the Rac1-inhibitor NSC23766 and the PI3K inhibitor LY294002 (Fig. 5C). Overexpression of N-terminal Flag-tagged Gp78 (Flag-gp78-IRES-GFP) in Cos7 cells results in the fragmentation and perinuclear clustering of mitochondria labeled with co-transfected pOct-dsRed (Fu et al., 2013). Similarly, in serum-starved conditions, Gp78-dependent mitochondrial fission, reflected in the smaller size, reduced elongation and reduced mobility of mitochondria, requires Gp78 ubiquitin ligase activity as it is not observed in cells transfected with Gp78 mutated in the Ring finger domain (Flag-RINGmut-IRES-GFP) (Fig. 6A). Treatment of cells with AMF reversed Gp78-dependent mitochondrial fission but did not significantly affect mitochondrial size and mobility in the absence of Gp78 overexpression (Fig. 6B). To specifically determine whether Gp78 ubiquitin ligase activity affects mitochondrial dynamics, the rate of recovery of mitochondrial pOct-dsRed was assessed by fluorescence recovery after photobleaching (FRAP) (Fig. 7A). Flag-Gp78 overexpression reduced the mobile fraction and rate of recovery of pOct-dsRed labelled mitochondria. Conversely, overexpression of Flag-RINGmut increased the mobile fraction and rate of recovery of pOct-dsRed labeled mitochondria, suggesting that it is acting in a dominant-negative fashion to inhibit Gp78 regulation of mitochondrial dynamics. AMF treatment or co-transfection of Flag-Gp78 with HA-Ubmono, but not wild-type HA-Ub, similarly induced increased fluorescence recovery and mobile fraction of pOct-dsRed to levels observed in Flag-RINGmut transfected cells (Fig. 7B). HA-Ubmono did not affect recovery in the absence of Flag-Gp78 overexpression or in Flag-RINGmut transfected cells. This suggests that Gp78 ubiquitin ligase activity acts dominantly to promote mitochondrial fission. AMF treatment paralleled the effect of inhibition of Gp78 ubiquitin ligase, either upon expression of the Gp78 RING finger mutant or coexpression of wild-type Gp78 with HA-Ubmono, supporting a role for AMF in the regulation of Gp78 ubiquitin ligase activity and substrate degradation.
AMF did not affect mitochondrial mobility in untransfected Cos7 cells. We therefore undertook to determine whether increased mitofusin levels in AMF-treated HT-1080 cells and in sh6 Gp78 HT-1080 knockdown cells (Fig. 8) is associated with increased mitochondrial mobility. Stable, tet-inducible shCN and sh6 Gp78 cells were treated with doxyxyclin inducing both the shRNA and mRFP, and transfected with pOct-GFP. Mitochondrial mobility was assessed by fluorescence recovery after photobleaching (Fig. 8A). Representative confocal time lapse movies are presented in supplementary material Movies 1–4. Both AMF treatment and Gp78 knockdown enhanced mitochondria mobility in HT-1080 cells (Fig. 8B). Through regulation of mitofusin levels, AMF and Gp78 represent novel regulators of mitochondrial dynamics (Fig. 9).
ERAD is a fundamental cellular process that targets proteins for degradation via a ubiquitin- and proteasome-dependent process (Hebert et al., 2010; Smith et al., 2011; Vembar and Brodsky, 2008). However, while much is known about the ERAD process per se, extracellular control of ERAD remains poorly defined. Using a combinatorial analysis of protein–protein interactions based on known ERAD substrates, Gp78 was identified, along with Hrd1, to be a key E3 ubiquitin ligase in ERAD (Christianson et al., 2011). Gp78 was originally identified as a cell surface 78 kDa glycoprotein on metastatic melanoma cells and subsequently as the receptor for AMF, and is also called AMFR (Nabi and Raz, 1987; Nabi and Raz, 1988; Nabi et al., 1990; Silletti et al., 1991). Full length sequencing of Gp78 cDNA identified cytoplasmic domain RING finger and Cue domains and Gp78 was subsequently shown to be a p97 binding protein that targeted the ERAD substrate CD3-delta for proteasomal degradation (Fang et al., 2001; Shimizu et al., 1999). Gp78 ubiquitin ligase activity has been shown to play a critical role in tumor progression and metastasis via targeting of the KAI1 metastasis suppressor (Tsai et al., 2007). The demonstration here that AMF regulates Gp78 ubiquitin ligase targeting of the mitofusin substrates for degradation defines a novel functional linkage between extracellular AMF cytokine function and the ER localized function of Gp78 in ERAD.
How extracellular AMF regulates the activity of ER-localized Gp78 remains unclear. Indeed, the fact that AMF does not prevent Gp78-dependent degradation of the KAI1 metastasis suppressor, known to be a Gp78 substrate (Joshi et al., 2010; Tsai et al., 2007), suggests that AMF does not generally inhibit Gp78 ubiquitin ligase activity. Gp78 is localized to a mitochondria-associated ER domain (Goetz et al., 2007; Wang et al., 2000) to which AMF is internalized. Interaction between the Gp78-labeled ER domain and mitochondria is calcium-sensitive (Goetz et al., 2007; Wang et al., 2000) and AMF impacts ER calcium release (Fu et al., 2011). By altering ER-mitochondria interaction, AMF could therefore selectively prevent Gp78 targeting of mitochondrial substrates, including the mitofusins, and thereby promote mitochondrial fusion. Indeed, consistent with Gp78 promotion of mitochondrial fission, the ER is recruited to sites of mitochondrial fission (Friedman et al., 2011). Inhibition of Gp78 ubiquitin ligase activity through expression of a Gp78 RING finger mutant or coexpression with HA-Ubmono, that prevents polyubiquitin chain elongation, increases mitochondrial fusion and may function as dominant negatives inhibiting Gp78 targeting of the mitofusins. The ability of internalized AMF to inhibit Gp78 degradation of the mitofusins defines a novel role for raft-dependent endocytosis of its extracellular ligand in the regulation of a key component of ERAD (Fig. 9).
The Cdc42-mediated, dynamin-independent CLIC/GEEC pathway is the major route of entry for GPI-anchored proteins and glycosphingolipid binding SV40 virus, Shiga toxin and Ctb to endosomes (Ewers et al., 2010; Römer et al., 2007; Sabharanjak et al., 2002). AMF is internalized via a tumor-cell-specific dynamin- and PI3K-dependent raft-endocytic pathway to the Gp78-positive endoplasmic reticulum that is negatively regulated by caveolin-1 and therefore non-caveolar (Kojic et al., 2007; Kojic et al., 2008; Lajoie and Nabi, 2010; Le et al., 2002). The ability of dynasore to inhibit AMF endocytosis but not that of Ctb shows, as we reported previously (Le and Nabi, 2003), that these two raft ligands follow different endocytic routes. Dynamin and RhoA-dependent raft endocytosis has been reported for IL2 internalization in T cells and the compensatory endocytosis of bladder umbrella cells (Khandelwal et al., 2010; Lamaze et al., 2001). Existence of a non-clathrin, RhoA-dependent endocytic route in yeast suggests that this pathway may represent a conserved endocytic pathway (Prosser et al., 2011). The demonstration here that AMF endocytosis is Rac1-, and not RhoA-, dependent defines the dynamin-dependent, Rac1-PI3K raft endocytic route followed by AMF to the ER as a distinct raft endocytosis pathway.
Gp78-mediated raft-dependent endocytosis enhances AMF activation of Rac1 and Rac1 activation promotes AMF endocytosis, suggesting that AMF signaling feeds back positively to promote its endocytosis. Similarly, AMF stimulates PI3K/Akt signaling that is also required for its endocytosis (Kojic et al., 2007). PI3K-Rac1 signaling has been implicated in various internalization processes including adenovirus endocytosis, macropinocytosis, toll-like receptor induced phagocytosis and apoptotic cell removal (Li et al., 1998; Mondal et al., 2011; Shen et al., 2010; Yoshida et al., 2009). Rac1 also controls the dynamin-independent, raft-dependent (CLIC/GEEC) uptake of adeno-associated virus and the syndecan-4 proteoglycan (Nonnenmacher and Weber, 2011; Sanlioglu et al., 2000). Rac1/PI3K signaling is therefore a general regulator of multiple endocytic processes, including that of AMF.
Numerous studies have described the relationship between clathrin-dependent receptor endocytosis and receptor signaling (Le Roy and Wrana, 2005; Polo and Di Fiore, 2006) however the role of non-clathrin endocytosis in receptor signaling is poorly understood. The requirement for raft endocytosis in AMF-Gp78 signaling shown here indicates that lipid raft signaling is based not only on the formation of plasma membrane signaling platforms (Lingwood and Simons, 2010) but can also require raft internalization. Endocytosis-dependent receptor-mediated activation of Rac1 has recently been linked to stimulation of endothelial and tumor cell motility. EphrinB regulation of the internalization of vascular endothelial growth factor receptor 3 (VEGFR3) is required for Rac1 activation and stimulation of angiogenesis (Wang et al., 2010). Met receptor endocytosis has recently been linked to oncogenicity and Rac1 activation (Joffre et al., 2011). AMF stimulates cell motility and metastasis (Funasaka et al., 2009; Tsutsumi et al., 2003) and both AMF and Gp78 are closely associated with tumor progression (Chiu et al., 2008; Fairbank et al., 2009). As a critical regulator of AMF-Gp78 signaling to Rac1 and of Gp78 ubiquitin ligase targeting of the mitofusins, raft-dependent endocytosis of AMF is therefore an important component of the role of this ligand-receptor pair in cell motility and tumor metastasis.
Materials and Methods
Antibodies and reagents
Monoclonal rat IgM anti-AMFR (3F3A) was used as ascites fluid (Nabi et al., 1990). Mouse c-Myc, Mfn1, Mfn2, KAI1 and RhoA mAbs were from Santa Cruz Biotechnology (Santa Cruz, CA) and Alexa647-conjugated and Alexa568-conjugated anti-rat IgM anti-mouse secondary antibodies from Molecular Probes (Eugene, OR). Rabbit phosphoglucose isomerase (referred to as AMF) was purchased from Sigma (P-9544, St. Louis, MO) and conjugated to FITC using Fluorescein-EX protein labeling kit (Molecular Probes). Effectene transfection reagent was purchased from Qiagen (Germany). RhoA, Rac1 and Cdc42 plasmids were a kind gift from Nathalie Lamarche (McGill University) and pOct-GFP and pOct-dsRed plasmids a kind gift from Heidi McBride (McGill University). RhoA and Rac1 beads, MLB lysis buffer and Rac1 mAb were from Millipore (Upstate). All other reagents were from Sigma (St. Louis, Mo).
MDA-435 human breast carcinoma and HT-1080 human fibrosarcoma cell lines (ATCC, Manassas, VA) were maintained in RPMI supplemented with 10% FBS, 100 IU/ml penicillin, 100 µg/ml streptomycin, 2 mM l-glutamine and 25 mM HEPES buffer at 37°C in a humid atmosphere (5% CO2/95% air). Selected Gp78-specific shRNA mir sequences (#2, #3, #6 and #8) (St. Pierre et al., 2012) were cloned from pGIPZ to pTRIPZ (OpenBiosystem) to generate doxycyclin-inducible Gp78 knockdown cell lines. Lentivirus particles, produced in HEK293T host cells, were used to infect HT-1080 cells. Infected cells were selected against puromycin (∼one week) and stable HT-1080 clones expressing control shCN and Gp78-targeted sh2 and sh6 maintained in medium supplemented with 2 mg/ml puromycin. To induce Gp78 knockdown, cells were cultured with doxycyclin (1 mg/ml) for 48 hours (refreshed after 24 hours).
For AMF treatment, cells were starved in serum-free medium overnight and then incubated with fresh medium containing AMF (25 µg/ml) for 0.5–6 hours, as indicated. Cells were transfected (48 hours) with RhoA, Rac1 and Cdc42 plasmids using Effectene transfection reagent (Qiagen). Gp78 knockdown HT-1080 cells were incubated with doxycyclin 48 hours before AMF treatment. Where indicated, cells were pretreated for 60 minutes with 100 µM dynasore, 100 µM PI3K inhibitor LY294002 or 24 hours with 100 µM the Rac1-inhibitor NSC23766 or 30 minutes with 5 mM mβCD, and then incubated with AMF in the presence of the drug. For dynasore washout experiments, dynasore treated cells were washed with fresh medium and incubated with AMF in the absence of dynasore.
Immunofluorescence labelling, FACS, western blotting, Rho GTPase activity assays and FRAP
Fluorescence-activated cell sorting (FACS) analysis of cell surface Gp78 expression and AMF-FITC internalization was performed as previously described (Kojic et al., 2007). Cells were incubated with 25 µg/ml AMF-FITC, 15 µg/ml Tf-FITC, or 10 µg/ml Ctb-FITC, for 30 minutes at 37°C. Cell surface-bound conjugate was removed with pronase (400 µg/ml; 5 minutes) except when cell surface Gp78 was also assessed. Confocal images were obtained with the 100×(NA 1.35) planapochromat objective of an Olympus FV1000 confocal microscope. Fluorescence intensities were determined using ImagePro image analysis software (Media Cybernetics, USA). Cells were harvested and lysates were western blotted for Gp78 and β-actin as described (Kojic et al., 2007). RhoA and Rac1-GTP pull-downs were performed according to the manufacturer's protocol (Millipore) and levels of RhoA GTP, total RhoA, Rac1-GTP and total Rac, were detected by western blotting using specific RhoA (Santa Cruz Biotechnology) and Rac1 (Millipore) antibodies, respectively. FRAP analysis of pOct-dsRed or pOct-GFP (in pTRIPZ HT-1080 cells) was performed with the 60× (NA 1.35) UPlanApo objective of an Olympus FV1000 with open pinhole (800 nm). Images were acquired every 3.3 seconds over 200 seconds and bleaching done with a 500 millisecond pulse of a 543 nm laser at 50% for pOct-dsRed or with a 250 millisecond pulse of a 488 nm laser at 85% for pOct-GFP. 4D analysis and tracking of pOCT-dsRed mitochondria was performed using a III-Zeiss spinning disk confocal and Sharpstack image analysis software (Intelligent Imaging Innovations). Stacks were acquired every 5 seconds over 2 minutes and mitochondria movement tracked over at least 5 frames. Tracking of pOct-dsRed mitochondria was performed using a III-Zeiss spinning disk confocal and Sharpstack image analysis software (Intelligent Imaging Innovations). Stacks were acquired every 5 seconds over 2 minutes and mitochondria movement tracked over at least five frames. pOct-GFP movies in pTRIPZ shRNA expressing HT-1080 cells were acquired on an Olympus FV1000 confocal microscope (488 nm laser; 60× objective) at 40 frames every 5 seconds and deconvolved (2D blind; Sharpstack, ImagePro).
J.S., L.D.K., P. St-P., P.T.C.W., M.F. and B.J. performed and analyzed the experiments; I.R.N. provided expertise and helped in designing and analyzing the experiments; J.S., L.D.K., and P. St-P. designed the majority of the experiments and analyzed all the experiments with I.R.N.; J.S., L.D.K. and I.R.N. wrote the manuscript and all coauthors corrected the manuscript.
This work was supported by Canadian Institutes for Health Research [grant number MT-15132 to I.R.N.]. J.S. is the recipient of a MITACS post-doctoral fellowship.