Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 May 31.
Published in final edited form as: Cell Rep. 2016 May 12;15(8):1660–1672. doi: 10.1016/j.celrep.2016.04.065

Autophagy promotes focal adhesion disassembly and cell motility of metastatic tumor cells through the direct interaction of paxillin with LC3

Marina N Sharifi 2,3,4,1, Erin E Mowers 2,5,1, Lauren E Drake 2,6, Chris Collier 2, Hong Chen 2, Marta Zamora 7, Stephanie Mui 2,3, Kay F Macleod 2,3,8
PMCID: PMC4880529  NIHMSID: NIHMS781702  PMID: 27184837

Summary

Autophagy is a conserved catabolic process that plays a housekeeping role in eliminating protein aggregates and organelles and is activated during nutrient deprivation to generate metabolites and energy. Autophagy plays a significant role in tumorigenesis, although opposing context-dependent functions of autophagy in cancer have complicated efforts to target autophagy for therapeutic purposes. We demonstrate that autophagy inhibition reduces tumor cell migration and invasion in vitro and attenuates metastasis in vivo. Numerous abnormally large focal adhesions (FAs) accumulate in autophagy-deficient tumor cells, reflecting a role for autophagy in FA disassembly through targeted degradation of paxillin. We demonstrate that paxillin interacts with processed LC3 through a conserved LIR motif in the amino terminal end of paxillin and that this interaction is regulated by oncogenic SRC activity. Together, these data establish a function for autophagy in FA turnover, tumor cell motility and metastasis.

Keywords: autophagy, paxillin, LC3, invasion, metastasis

Graphical Abstract

graphic file with name nihms781702u1.jpg

Introduction

Macro-autophagy (hereafter autophagy) is a catabolic process important for the degradation of damaged organelles and protein aggregates as well as the intracellular recycling of metabolites (Kroemer et al., 2010; Mizushima and Komatsu, 2011) that are used by tumor cells to survive nutrient stress, hypoxia and cytotoxic therapies (Amaravadi et al., 2011; Kimmelman, 2011; White, 2012). Consequently, autophagy has emerged as a potential therapeutic target (Amaravadi et al., 2011). However, such efforts are complicated by the growing realization that autophagy plays opposing roles in tumorigenesis, likely influenced by the stage of progression, driving oncogene and tissue type (Galluzzi et al., 2015). Mouse models support a tumor-suppressive role for autophagy in the early stages of tumorigenesis by promoting genome stability and limiting necrosis and inflammation (Degenhardt et al., 2006; Karantza-Wadsworth et al., 2007; Mathew et al., 2009). Conversely, autophagy is required for malignant progression (Guo et al., 2013; Karsli-Uzunbas et al., 2014; Rao et al., 2014; Rosenfeldt et al., 2013; Strohecker et al., 2013). Consistently, clinical studies have associated increased staining for the autophagy marker LC3 with melanoma metastases and with node positivity and decreased overall survival in human breast cancer (Lazova et al., 2012), whereas over-expression of the key autophagy regulator Beclin1 is linked to reduced latency of melanoma metastasis (Giatromanolaki et al., 2011).

Here, we report that autophagy is required for the motility and invasion of highly metastatic tumor cells due to a function for autophagy in promoting focal adhesion (FA) turnover. The key FA protein paxillin is degraded by autophagy, and paxillin is targeted to the autophagosome through the Src-regulated interaction of LC3 with a conserved LC3-interacting region (LIR) in paxillin. These results broaden our understanding of the multi-faceted role of autophagy in tumor progression and indicate that inhibiting autophagy may be an effective approach to blocking metastatic dissemination in the clinic

Results

Knockdown of Atg5 or Atg7 in metastatic mammary tumor cells inhibits autophagy without inhibiting cell growth

To explore the role of autophagy in metastasis, we used the 4T1 orthotopic mouse mammary tumor model, in which 4T1 murine mammary carcinoma cells implanted into the mammary fat pad of syngeneic Balb/C mice form primary tumors that spontaneously metastasize to the lung and liver (Aslakson and Miller, 1992). We inhibited autophagy in 4T1 cells through stable shRNA-mediated knockdown of either Atg5 or Atg7, two autophagy-regulatory proteins required for autophagosome formation, in accordance with published guidelines (Staskiewicz et al., 2013).

4T1 clones stably expressing shAtg5 exhibited no detectable Atg5 protein in the presence or absence of the autophagy inhibitor bafilomycin A1 (bafA1) (Figure 1A). As expected, knockdown of Atg5 also resulted in the absence of the Atg5–12 conjugate (Figure S1A). Although parental cells exhibited robust autophagic flux with bafA1 treatment, as measured by increased autophagosome marker LC3B-II, shAtg5 clones failed to do so either in the presence of bafA1 (Figure 1C, 1D) or under hypoxia (Figure S1B). Similar results were obtained upon Atg7 knockdown (Figure 1B, 1C, 1E). Immunofluorescence for endogenous LC3B revealed numerous punctae in parental and control cells but only diffuse cytoplasmic staining in autophagy-deficient cells (Figure 1F), demonstrating the lack of autophagosome formation. Finally, electron microscopy revealed abundant autophagosome/autolysosome structures in control but not shAtg5 cells (Figure 1G, red arrows).

Figure 1. Autophagy-deficient 4T1 cells do not exhibit altered growth rate.

Figure 1

Open in a new tab

(A) Western blot for Atg5 in parental, scrambled shRNA and two Atg5 shRNA-expressing 4T1 clones in the presence or absence of 100 nM bafA1 (4 h). (B) Detection of Atg7 by IP followed by western blotting in parental, control and Atg7 shRNA cells. (C) Densitometric quantification of LC3B-II in parental and scrambled, Atg5 and Atg7 shRNA cells ±bafA1. Mean ± SEM, n=3. (D–E) Representative western blot for LC3B-I and II in Atg5- and Atg7-deficient cells ±bafA1. (F) Immunofluorescence for endogenous LC3. (G) Electron microscopy of autophagy-competent (left) and -deficient (right) 4T1 cells. (H–I) Cell growth in complete medium over 96 h. Mean ± SEM, n=3. (J) Cell viability after 24 h in complete medium lacking glucose or at 1% O2 assessed by propidium iodide exclusion. Mean ± SEM, n=2.

Despite this marked inhibition of autophagy in Atg5- and Atg7-deficient 4T1 cells, there was no significant decrease in cell growth (Figure 1H, 1I) or viability compared with parental and scrambled shRNA cells, even when deprived of glucose or oxygen (Figure 1J). Thus, although metastatic 4T1 tumor cells exhibit high levels of autophagic flux, inhibition of autophagy did not limit cell growth in vitro.

Autophagy is required for metastasis of tumor cells from the mammary gland to the lung and liver

Injection of engineered 4T1 tumor cells into the mammary fat pad of syngeneic Balb/C female mice produced palpable mammary tumors within 2 weeks and large tumors within 4 weeks. LC3B-positive puncta/autophagosomes were readily visible in control tumors at 3 weeks but were absent in Atg5- and Atg7-deficient tumors (Figure 2A), indicating that autophagy was inhibited. However, consistent with in vitro data (Figure 1I), autophagy-deficient tumors did not exhibit reduced growth in vivo (Figure 2B), and there was no significant difference in the Ki-67 index (figure 2C) or TUNEL staining (figure 2D) between the autophagy-deficient and control tumors. Taken together, these data indicate that 4T1 tumor cells do not depend on autophagy for tumor cell proliferation or survival in vivo.

Figure 2. Autophagy inhibition reduces metastasis to the lung and liver.

Figure 2

Open in a new tab

(A) Immunohistochemistry for LC3B in control (i), Atg5-deficient (ii) and Atg7-deficient (iii) 4T1 primary tumors at 3 w. (B) Orthotopic tumor volume of parental and scrambled, Atg5, and Atg7 shRNA tumors over time (p = 0.2106, repeated measures ANOVA). Mean ± SEM, n=8. (C–D) Quantification of nuclear proliferation marker Ki-67 (C) and TUNEL staining (D) of tumors after 4 w. Mean ± SEM, n=6. (E) Representative H&E-stained sections of lungs and livers of mice with parental and scrambled, Atg5, and Atg7 shRNA tumors 4-w post-orthotopic injection. (F) Mean number of lung metastases in mice with control or autophagy-deficient tumors at 4 w (p<0.0001, one-way ANOVA). Mean ± SEM, n=15 (parental), n=20 (scrambled shRNA), n=18 (Atg5 shRNA-1 and -2), n=8 Atg7 (shRNA-1 and -2). (G) Mean number of lung metastases at 4 w in mice with parental tumors that received PBS or 60 mg/kg HCQ i.p. every 3 d following tumor implantation (p<0.0001, one-way ANOVA). Mean ± SEM, n=8 per group.

Although primary tumor growth was unaffected, inhibition of autophagy markedly reduced lung and liver metastases (Figure 2E, 2F). Numerous macro-metastases were evident in lung and liver sections from mice with parental and scrambled control tumors but not mice with autophagy-deficient tumors (Figure 2E). Quantification of lung metastases demonstrated a statistically significant reduction (p<0.0001) in mice bearing autophagy-deficient tumors relative to parental or control tumors (Figure 2E). To confirm these effects pharmacologically, mice with parental tumors were treated with hydroxychloroquine (HCQ), an FDA-approved inhibitor of autophagy with the same mechanism of action as bafA1 (Klionsky et al., 2016). As with genetic inhibition of autophagy, HCQ treatment did not alter primary tumor growth (Figure S2A) but significantly reduced lung metastasis (p<0.0001) (Figure 2G). Together, these results demonstrate that autophagy is required for spontaneous metastasis in the 4T1 in vivo model.

The known functions of autophagy in promoting cell survival during extracellular matrix (ECM) detachment, growth factor withdrawal and nutrient deprivation (Fung et al., 2008; Kuma et al., 2004) are believed to promote progression following escape from the primary tumor. Indeed, autophagy is required for tumor cell survival in the bloodstream during hepatocellular carcinoma metastasis (Peng et al., 2013). To investigate whether autophagy is required at later stages of metastasis in the 4T1 model, engineered tumor cells were injected directly into the circulation via the tail vein, bypassing earlier steps in the metastatic cascade. After 2 weeks, autophagy-deficient tumor cells formed as many lung metastases as parental and control cells (Figure S2B, S2C), indicating that autophagy is not required in this model for tumor cell survival in the circulation or metastatic outgrowth at secondary sites. This is consistent with our finding that autophagy is not required for 4T1 tumor cell proliferation or survival in vitro (Figure 1G–H) or in primary tumors in vivo (Figure 2B–D) and indicates that reduced metastasis of autophagy-deficient tumors (Figure 2E–F) was due to failure to escape from the primary tumor.

Autophagy is required for tumor cell motility in vitro

Consistent with the observed defect in spontaneous metastasis in vivo, autophagy-deficient 4T1 cells exhibited markedly decreased migration and invasion through collagen compared with control cells (Figures 3A, 3B). Inhibition of autophagy with chloroquine (CQ) also inhibited the motility of 4T1 cells (Figure 3C). Inhibition of autophagy in the metastatic MDA-MB-231 human breast cancer (Figure S3A–C) and B16.F10 mouse melanoma (Figure S3D–E) cell lines produced similar effects. Furthermore, autophagy-deficient 4T1 cells exhibited markedly altered morphology in time-lapse differential interference contrast (DIC) imaging (Supplemental videos 1–3). Unlike control cells, autophagy-deficient cells did not spread out (Figure 3D) or form protrusions (Figure 3E). Together, these results indicate that autophagy is required for metastatic tumor cell migration, invasion and spreading in vitro.

Figure 3. Autophagy is required for escape from the primary tumor in vivo and cell migration in vitro.

Figure 3

Open in a new tab

(A–B) Quantification of migration and invasion through collagen for parental, scrambled shRNA and Atg5- (A) and Atg7-deficient (B) 4T1 clones (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001, one-way ANOVA). Mean ± SEM, n=3. (C) Invasion and migration of parental 4T1 cells ±50 μM CQ. Mean ± SEM, n=3. (D) Depth heat map of cell spreading of scrambled and Atg5-deficient 4T1 cells. (E) Representative DIC images of control and autophagy-deficient 4T1 cells with altered cell morphology

Given the change in the morphology of the autophagy-deficient cells, which could be consistent with a transition to a more epithelial cellular program, as well as recent work suggesting that the autophagy adapter protein p62/Sqstm1 regulates Twist degradation (Qiang et al., 2014), we assessed Twist and E-cadherin protein levels in autophagy-deficient cells but found no difference relative to autophagy-competent cells (Figure S3F). Furthermore, because cell motility is driven by ATP, the production of which can be supported by autophagy, we also examined whether the motility defects of autophagy-deficient cells were due to insufficient ATP levels. As expected, Atg5 knockdown reduced cellular ATP levels. However, although treatment with methyl pyruvate restored ATP levels (Figure S3G), it failed to rescue cell invasion (Figure S3H) or protrusion formation (Figure S3I). Thus, the defective motility of autophagy-deficient tumor cells was not due to reduced ATP levels.

Focal adhesion disassembly is impaired in autophagy-deficient tumor cells

Focal adhesions (FAs) are large macromolecular complexes that link the actin cytoskeleton to the ECM to provide traction, and properly regulated FA dynamics are critical for cell migration (Parsons et al., 2010; Ridley, 2011; Webb et al., 2002). Immunofluorescence for the FA proteins paxillin and zyxin revealed an increase in both FA size and number in autophagy-deficient cells (Figure 4A–C). Although paxillin is recruited early to nascent FAs (Laukaitis et al., 2001), zyxin is found only in mature FAs (Zaidel-Bar et al., 2003), indicating that the more numerous and larger FAs in the autophagy-deficient cells were mature. This phenomenon was replicated upon bafA1 treatment of both parental 4T1 and B16.F10 cells (Figure S4B, S4C). As with the morphologic changes (Figure 3E), the alteration in FAs was density-independent (Figure 4A and Figure S4A). The FA abnormalities were observed in autophagy-deficient cells grown on glass, fibronectin or collagen I (Figure S4D), indicating that defects in cell spreading and FA accumulation were not due to altered interaction with or modulation by the cellular substratum.

Figure 4. Autophagy-deficient 4T1 cells have defective FA disassembly.

Figure 4

Open in a new tab

(A) Immunofluorescence of FA proteins paxillin and zyxin in control and autophagy-deficient cells. (B–C) Quantification of FAs/nucleus (B) and average FA size (C) per 63x field (*p<0.05, ***p<0.001, one-way ANOVA). Mean ± SEM, n=10. (D) Timelapse TIRF imaging of EGFP-paxillin FAs in control and Atg5-deficient cells. Inset: representative disassembly of adhesions (arrows) over 15 min (control) or 30 min (Atg5-deficient). (F) Rate constant for increase in EGFP-paxillin fluorescence during FA formation (mean ± SEM, n=11) and for decrease in EGFP-paxillin fluorescence during FA turnover (****p<0.0001, mean ± SEM, n=13 (scrambled shRNA), n=15 (Atg5 shRNA)).

To determine whether the FA abnormalities in autophagy-deficient tumor cells were the result of altered FA dynamics, we assessed FA assembly and disassembly rates using timelapse imaging of EGFP-paxillin FAs in control and Atg5-deficient 4T1 cells (Figure 4D, supplemental videos 4–5). EGFP-paxillin incorporation into and disappearance from FAs was linear on a semilogarithmic plot of fluorescence intensity over time, as described previously (Webb et al., 2004), allowing us to calculate assembly and disassembly rate constants (Figure 4E). Although Atg5-deficient cells exhibited no difference in FA assembly rates relative to control cells, the disassembly rate was significantly reduced; FAs disassembled completely within 15 min in control cells (Figure 4D, top panel, arrows) but required ≥30 min in Atg5-deficient cells (Figure 4E, bottom panel, arrows). Therefore, inhibition of autophagy increases FA size and number due to impaired FA disassembly.

Aberrant FAs in autophagy-deficient tumor cells are associated with paxillin accumulation

Cells lacking FAK (Ilic et al., 1995), catalytically active Src (Fincham and Frame, 1998) or paxillin (Hagel et al., 2002) exhibit decreased spreading and migration associated with increased FA number and size and exhibit impaired FA disassembly (Webb et al., 2004), similar to our observations in autophagy-deficient 4T1s. Although we did not detect any changes in FAK or Src protein levels or phosphorylation (Figure S5A, S5B), we observed a marked increase in total paxillin protein levels in autophagy-deficient 4T1 cells (Figure 5A, 5B). BafA1 increased paxillin protein levels in parental but not autophagy-deficient 4T1 cells (Figure 5A–C) as well as in MDA-MB-231 and B16.F10 cells (Figure 5D, 5E). Paxillin levels were also elevated in lysates from Atg5- and Atg7-deficient tumors relative to control primary 4T1 tumors (Figure 5F). Immunohistochemistry confirmed these increased paxillin levels in autophagy-deficient tumors (Figure 5G). Thus, autophagy inhibition is associated with paxillin accumulation in vitro and in vivo.

Figure 5. Autophagy-deficient 4T1 cells have increased paxillin.

Figure 5

Open in a new tab

(A–B) Western blot for paxillin in control and autophagy-deficient 4T1 cells ±100 nM bafA1 (4 h) to inhibit autolysosomal degradation. (C) Densitometric quantification of paxillin fold increase with bafA1 in parental and scrambled, Atg5, and Atg7 shRNA 4T1 cells. Mean ± SEM, n=3. (D–E) Western blot for paxillin ±bafA1 (4 h) in MDA-MB-231 (D) and B16.F10 (E) cells. (F) Western blot for paxillin in lysates from parental and scrambled, Atg5 and Atg7 shRNA-expressing 4T1 primary tumors 3 w after orthotopic injection. (G) Paxillin immunohistochemistry in parental and scrambled, Atg5 and Atg7 shRNA-expressing 4T1 primary tumor sections 3 w after orthotopic injection. Scale bars = 50 μm. (H) Western blot confirming paxillin knockdown in autophagy-deficient 4T1 cells. (I) Paxillin and zyxin immunofluorescence in autophagy-deficient 4T1 cells treated with control or paxillin siRNA. (J–K) Quantification of the effect of siRNA-mediated paxillin knockdown in autophagy-deficient 4T1 cells on migration and invasion (**p<0.01) Mean ± SEM, n=3.

To determine whether increased paxillin levels underlie the cell motility and FA defects in autophagy-deficient 4T1s, we used siRNA to knock down paxillin expression in autophagy-deficient 4T1 cells to levels similar to those in parental cells (Figure 5H). Reducing paxillin levels in autophagy-deficient cells normalized FA morphology (Figure 5I). Importantly, knockdown of paxillin and normalization of FAs rescued the motility of autophagy-deficient cells (Figure 5J, 5K). Together, these results demonstrate that inhibition of autophagy in metastatic tumor cells results in the accumulation of paxillin, thereby inhibiting FA disassembly, cell spreading and cell motility.

Paxillin is stabilized in autophagy-deficient cells and colocalizes with LC3

Increased paxillin in autophagy-deficient cells was not due to increased paxillin mRNA levels (Figure S5C) but was associated with increased protein stability (Figure S5D), indicating that inhibition of autophagy reduces paxillin degradation. Paxillin can be degraded at the proteasome (Abou Zeid et al., 2006), and autophagy inhibition can reduce proteasomal activity (Korolchuk et al., 2009). However, proteasomal inhibition with MG132 or ALLN did not increase paxillin levels in control 4T1 cells (Figure S5E), and proteasomal degradation was not inhibited in autophagy-deficient 4T1 cells (Figure S5F). Thus, we investigated a direct role for autophagy in degrading paxillin.

By live cell confocal imaging, we detected mApple-paxillin in punctate cytosolic structures that co-localized with the autophagosome marker EGFP-LC3B (Figure 6A, Figure S6A), confirming that paxillin colocalizes with autophagosomes in autophagy-competent 4T1 cells. EGFP-LC3B also co-localized with mApple-paxillin-positive FA structures (Figure 6B, Figure S6A), indicating that autophagosomes co-localize with FAs in parental cells, consistent with previous reports (Sandilands et al., 2011). In combination with our data demonstrating that both transient and permanent inhibition of autophagic flux increase paxillin levels, the colocalization of paxillin with autophagosomes strongly suggests that paxillin is degraded by autophagy. Immunohistochemistry revealed punctate paxillin staining in 4T1 tumors expressing scrambled but not Atg5 shRNA (Figure S6B), consistent with the presence of paxillin at autophagosomes in vivo.

Figure 6. Paxillin co-localizes with autophagosomes.

Figure 6

Open in a new tab

(A–B) 4T1 cells expressing mApple-paxillin and EGFP-LC3B were treated with bafA1, and live cells were imaged by confocal microscopy. 3D reconstruction reveals that mApple-paxillin colocalizes with EGFP-LC3B (arrows) (A) and EGFP-LC3B colocalizes with FAs (arrows) (B). (C) Alignment of human paxillin residues 37–43 (bottom) with the −3 to +4 residues of known tryptophan- and tyrosine-containing LIRs. (D–E) Co-IP of mApple-paxillin and endogenous paxillin with GFP-LC3 in parental 4T1 (D) and B16.F10 cells (E). (F) Western blot for paxillin in control and LC3B-deficient 4T1 cells ±100 nM bafA1 (4 h). (G) Immunofluorescence of FA proteins paxillin and zyxin in control and LC3B-deficient cells. (H–I) Quantification of invasion and migration of autophagy-deficient 4T1 cells relative to parental and scrambled cells (**p<0.01, ***p<0.001). Mean ± SEM, n=3.

Paxillin interacts directly with LC3

Paxillin can be ubiquitinated (Didier et al., 2003), and cargo receptor proteins like p62/Sqstm1 and Nbr1 are known to target ubiquitinated proteins to autophagosomes for degradation (Rogov et al., 2014); both have been implicated in the regulation of cell migration (Kenific et al., 2016; Qiang et al., 2014), However, neither paxillin accumulation (Figure S6C, S6D) nor reduced cell migration or invasion (Figure S6E, S6F) were observed in 4T1 cells knocked down for p62/Sqstm1 or Nbr1, indicating that neither protein is involved in paxillin degradation. Cytosolic proteins and macromolecular structures can also be targeted for autophagy via direct interaction with processed LC3 (Rogov et al., 2014). We identified a putative LC3-interacting region (LIR), defined by the presence of the conserved sequence [W/F/Y]-xx-[L/I/V] (Birgisdottir et al., 2013) in paxillin at amino acids 40–43 (Figure 6C). The putative paxillin LIR resembles that of Nbr1, with the aromatic residue Y40 in position +1 and the hydrophobic residue I43 in position +4. The acidic residue in the +2 position and threonine in the −1 position have also been shown to contribute to interactions of validated LC3-interacting proteins with LC3 (Birgisdottir et al., 2013). Furthermore, the putative LIR in paxillin is highly conserved from Xenopus through humans (Figure S6G).

Given that paxillin colocalizes with LC3B in the cytosol and at FAs (Figure 6A, 6B), we tested for an interaction between LC3B and paxillin. We successfully co-immunoprecipitated mApple-paxillin and trace levels of endogenous paxillin with EGFP-LC3B in both 4T1 (Figure 6D) and B16.F10 cells (Figure 6E). Furthermore, an in vitro binding assay demonstrated that paxillin was pulled down with GST-LC3B but not GST, demonstrating that LC3B is able to directly bind paxillin in the absence of any adaptors (Figure S6H). Consistent with these results, shRNA-mediated knockdown of LC3B (Figure S6I) led to accumulation of paxillin (Figure 6F), enlarged FAs (Figure 6G) and reduced cell motility (Figure 6H, 6I), phenocopying the effects of Atg5 and Atg7 deficiency. These data illustrate the requirement for a direct interaction between paxillin and LC3B-II to promote targeted degradation of paxillin by autophagy and focal adhesion disassembly.

Defining a LIR motif in paxillin that is regulated by Src

To determine whether the interaction of paxillin with LC3 requires the putative LIR motif, we generated a paxillin mutant in which the critical tyrosine at the +1 position of the putative LIR was mutated to alanine (Y40A) as well as a mutant in which positions +2 through +4 were mutated to alanine (QEIAAA). The Y40A and QEIAAA mutants localized properly to focal adhesions (Figure S7E), but both mutations significantly reduced the colocalization of mApple-paxillin with EGFP-LC3 (Figure 7A, 7B, 7C) in 4T1 cells stably depleted of endogenous paxillin (Figure S7A). These mutations also abrogated the co-immunoprecipitation of mApple-paxillin with EGFP-LC3 (Figure 7D, lane 3 and lane 8), although the Y40A mutation exhibited a greater inhibitory effect on the paxillin-LC3 interaction than the QEIAAA mutation. Furthermore, cells expressing the mApple-paxillin mutants exhibited reduced motility relative to cells expressing wildtype mApple-paxillin (Figure S7B). These results validate the LIR motif in paxillin and highlight the key function of the Y40 residue in the interaction of paxillin with LC3.

Figure 7. The LIR motif of paxillin is critical for interaction with LC3 and is SRC-regulated.

Figure 7

Open in a new tab

(A–C) Images (A) and quantification (B–C) of shPaxillin 4T1 cells co-expressing EGFP-LC3B and mApple-paxillin WT, Y40A or QEIAAA. %Cells with colocalization (**p<0.01): Mean ± SEM, n=2, 15–20 cells/condition/experiment. Colocalization events per cell (***p<0.001): Mean ± SEM, n=30. (D) Co-IP of mApple-paxillin WT and Y40A or QEIAAA with GFP-LC3 in shPaxillin 4T1 cells in the presence or absence of active SrcY527F. (E) Quantification of migration and invasion of scrambled shRNA and shAtg5 cells in the presence or absence of SrcY527F (**p<0.01, ***p<0.001). Mean ± SEM, n=4. (F) Quantification of the SrcY527F-simulated fold change in the migration and invasion of cells expressing mApple-paxillin WT, Y40A, or QEIAAA (*p<0.05, **p<0.01, ****p<0.0001). Mean ± SEM, n=6.

A tyrosine at position +1 is less commonly found in LIR motifs than the canonical +1 tryptophan (Figure 6C) (Birgisdottir et al., 2013). Intriguingly, the Y40 residue of paxillin is a known phosphorylation target of Src tyrosine kinase, although its functional significance is not well established (Schaller and Schaefer, 2001; Webb et al., 2005). Co-immunoprecipitation of mApple-paxillin with EGFP-LC3 was strongly stimulated by constitutively active SrcY527F in both paxillin knockdown 4T1 cells (Figure 7D lane 2 and lane 4) and B16.F10 cells (Figure S7C). Significantly, the Y40A mutation blocked the effect of SrcY527F (Figure 7D, lane 5), while the QEIAAA mutation significantly blunted it (Figure 7D lane 10), indicating that Src modulates the interaction of paxillin with LC3 via the identified LIR motif. Interesting, the Y31 and Y118 residues in paxillin, which are the sites classically regulated by Src phosphorylation (Deakin and Turner, 2008) were not required for either the basal interaction of paxillin with LC3 or the response to Src (Figure S7C). Taken together, these results indicate that the LIR motif of paxillin and the Y40 residue in particular is critical to the interaction of paxillin with LC3B and that this interaction is regulated by oncogenic Src.

Autophagic degradation of paxillin is required for Src-stimulated motility

To determine whether the Src-regulated interaction of paxillin and LC3 underlies the motility defects of autophagy-deficient cells, we performed transwell migration assays. Although the expression of SrcY527 strongly stimulated the motility of control 4T1 cells, it failed to do so in autophagy-deficient shAtg5 cells (Figure 7E). Furthermore, the motility of cells expressing mApple-paxillin-WT was strongly stimulated by co-expression of SrcY527F (Figure 7E), but cells expressing either the Y40A or QEIAAA mutants were defective for Src-stimulated motility (Figure 7F). These data demonstrate that stimulation of the interaction between paxillin and LC3 by oncogenic Src and the subsequent autophagic degradation of paxillin promotes cellular migration.

In summary, our work identifies a role for autophagy in promoting metastatic tumor cell migration and invasion by degrading paxillin and promoting FA disassembly that is dependent on the Src-regulated interaction of paxillin with processed LC3B.

Discussion

We have demonstrated that autophagy promotes cell spreading, migration, and invasion of highly metastatic tumor cells in vitro and is required for early steps in the metastatic cascade in the 4T1 mouse mammary tumor model in vivo, with similar effects observed in the MDA-MB-231 human breast cancer and B16.F10 melanoma cell lines. Our data establish that autophagy is required to promote FA disassembly by degrading the FA protein paxillin, that paxillin interacts directly with LC3B, and that the conserved amino-terminal LIR motif is critical to the interaction. Consistent with the knowledge that Y40 in the paxillin LIR motif is an established Src phosphorylation target, expression of constitutively active SrcY527F dramatically increased both the interaction of paxillin with LC3B and tumor cell motility. Furthermore, Atg5-deficient cells were refractory to the migration-stimulatory effects of SrcY527F, indicating that autophagy plays a critical role downstream of oncogenic Src in promoting FA disassembly through targeted degradation of paxillin. Interestingly, Src phosphorylation of a critical tyrosine in the LIR motif of FUNDC1 has been reported to modulate the interaction of this protein also with LC3B (Liu et al., 2012).

Likely due to the aggressive nature of the metastatic tumor cells examined here, we did not observe significant reductions in cell growth or viability in vitro (Figure 1I, 1J) or in vivo (Figure 2B) following autophagy inhibition. Importantly, although autophagy-deficient 4T1 cells were unable to metastasize successfully from an orthotopic site, they were able to colonize the lung when injected directly into the circulation, indicating that survival in the circulation and outgrowth in the lung were not impaired by loss of autophagy in this model. Thus, we have been able to separate a specific requirement for autophagy in cell motility during metastasis that is independent of effects on proliferation or cell viability.

The link between autophagy and paxillin was first suggested based on genetic interactions between Atg1 and paxillin in D. melanogaster (Chen et al., 2008). However, the degradation of paxillin is not the only mechanism by which autophagy can modulate FA dynamics. Treatment of patient-derived glioblastoma cells with phosphatidylinositol 3-kinase, an upstream regulator of autophagy, increased tumor cell motility through mitochondrial redistribution to the cortical cytoskeleton to promote FA turnover (Caino et al., 2015). Additionally, FIP200, a protein that forms part of the Ulk1/Atg13 pre-initiation complex (Hara et al., 2008), inhibits FAK autophosphorylation and cellular migration when over-expressed (Abbi et al., 2002). Finally, AMPK activation induces autophagy but reduces cell migration in an Ulk1- and FIP200-dependent manner in glioblastoma, prostate cancer cells and normal fibroblasts (Caino et al., 2013). We did not observe any change in FAK autophosphorylation (Figure S5A) or FIP200 levels (data not shown) when autophagy was inhibited, but it is possible that alterations in autophagic flux could alter FIP200-FAK signaling in other systems not studied here.

Our work adds to a growing body of work linking autophagy to tumor cell motility, invasion and metastasis (Kenific et al., 2016; Lock et al., 2014; Mackintosh et al., 2012; Qiang et al., 2014). A recent study identified a critical role for autophagy in the secretion of MMPs and cytokines during RAS-driven invasion (Lock et al., 2014), and knockdown of Atg12 has been shown to inhibit invasion in a glioblastoma cell line (Mackintosh et al., 2012). Overexpression of autophagy genes has been associated with the more aggressive mesenchymal subtype of primary glioblastoma, and the substrate adaptor p62/Sqstm1 and autophagic flux were required for the invasion and migration of glioblastoma stem-like cell lines (Galavotti et al., 2013). However, an autophagy-independent role for p62 in promoting cell migration by binding Twist and preventing its degradation has also been identified (Qiang et al., 2014). We observed that paxillin turnover by autophagy was independent of p62/Sqstm1 (Figure S6C), indicating that autophagy can modulate cellular dynamics through both p62-independent and p62-dependent mechanisms. We also show that in the highly metastatic tumor cells examined here, NBR1, a different cargo receptor protein implicated in the migration of Ras-expressing epithelial cells (Kenific et al., 2016), does not stabilize paxillin (Figure S6D). However, we cannot rule out the possibility that adaptors other than p62/Sqstm1 and NBR1 may facilitate binding of paxillin to LC3B. Finally, the levels of Twist and E-cadherin were unchanged in autophagy-deficient 4T1 cells (Figure S3F), indicating that the motility defects of autophagy-deficient tumor cells here were not explained by the mechanisms reported in other studies (Kenific et al., 2016; Qiang et al., 2014). This suggests that autophagy regulates cell motility through different mechanisms in different cell types, and that the mechanism described here may be specific to highly metastatic and mesenchymal tumor cells. Finally, the requirement for autophagy in the survival of dormant tumor cells in GIST (Gupta et al., 2010) and ovarian cancer (Lu et al., 2008) illustrates an additional role for autophagy at a later step in the metastatic cascade. Together with our work identifying a critical role for autophagy in FA disassembly through paxillin degradation and its requirement for escape from the primary tumor, these studies highlight the potential utility of inhibiting autophagy to block tumor metastasis.

Experimental Procedures

Cell culture

Cells were grown in DMEM/10% FBS/ 1% Pen/Strep (1% MEM-NEAA was added to 4T1 media). Mouse Atg7 (TRCN0000375444), p62 (TRCN0000098619), Nbr1 (TRCN0000238310), MAP1LC3B/Atg8 (TRCN0000120800), paxillin (TRCN0000305025) and non-targeting control shRNA (SHC002) and custom shRNA to human/mouse ATG5/Atg5 (GCATTATCCAATTGGTTT seed sequence) were from Sigma. Non-targeting control (Santa Cruz sc-37007) and paxillin siRNA (Santa Cruz sc-36197) were used at 50 nM. ATP was measured with the ATP Bioluminescence Assay Kit CLS II (Roche).

Chemicals

BafA1 and MG132 were from Enzo. Chloroquine, ALLN and human plasma fibronectin were from Sigma, methyl pyruvate was from Fluka, and rat tail collagen I and matrigel were from BD.

Transwell assays

For migration assays, 5x104 (4T1), 4x104 (MDA-MB-231) or 9x104 (B16.F10) cells in serum-free medium were seeded per 8-μm pore cell culture insert (BD). Inserts were pre-coated with collagen I (BD; 4T1 and MDA-MB231) or matrigel (BD; B16.F10) for invasion assays, and 1x105 (4T1), 9x104 (MDA-MB-231) or 1.5x105 (B16.F10) cells were seeded per insert. Complete medium was added outside of the insert, and plates were incubated at 37°C for 16 h. Cells were fixed in 4%PFA and Giemsa-stained, non-migrated cells were scraped from the inner side of the insert, and migrated/invaded cells were counted in five 20x fields per insert. Assays were performed in triplicate.

Mice and tumor studies

For orthotopic injections, MAP-tested 4T1 cells (104 cells in 100 μl per mouse) were injected into the mammary fat pad of 8-week-old female Balb/C mice (Charles River). Tumor volume (calculated as L x W2) was measured weekly with electronic calipers. Hydroxychloroquine (60 mg/kg) was delivered i.p. every 3 days following tumor implantation. Tumors, lungs and livers were harvested at 4 weeks post-orthotopic injection, fixed in 10% NBF, embedded in paraffin, sectioned and stained as described below. For tail vein injections, 5 x 105 cells in 100 μl were injected per 8-week-old female Balb/C mouse (Charles River), and lungs were harvested at 2 weeks and processed as above. ImageJ was used to segment and count lung metastases on 8 H&E-stained 25-μm serial lung sections per mouse for both spontaneous and experimental metastasis assays.

Immunohistochemistry

Ki67, TUNEL staining and LC3B immunohistochemistry were performed as described (Rosenfeldt et al., 2012). Paxillin was detected in situ using a specific antibody (Santa Cruz H-114, 1:80 dilution) and standard immunohistochemistry. Stained sections were digitized and quantified using a ScanScope XT automated slide scanning system and Spectrum Plus image analysis software (Aperio). For Ki-67, a tuned nuclear quantification v9 algorithm was used to report DAB+ nuclei as a percent of total (hematoxylin-stained) nuclei on 10 random non-necrotic 0.25-mm2 sections per slide. For TUNEL, a tuned color deconvolution v9 algorithm was used to report DAB+ area as percent of total tissue area for each section.

Western blots and immunoprecipitation (IP) analyses

Primary antibodies were α–LC3, α-Atg5 (Novus); α–Atg7, α–β-actin (Sigma); α–Atg12, α-Nbr1, (Cell Signaling); α–paxillin (clone 177 (WB) and 349 (IF)), E-cadherin (BD); Twist (Abcam); α–GFP (Santa Cruz); and α–ubiquitin (Pierce). Cells were lysed in RIPA (1% sodium deoxycholate, 0.1% SDS, 1% Triton X-100, 10 mM Tris-HCl pH 8.0, 0.14 M NaCl) or NP40 buffer (LC3 western blots; 150 mM NaCl, 50 mM Tris-HCl pH 7.5, 1 mM EDTA, 1% NP40) with protease and phosphatase inhibitors. Lysates were centrifuged at 16300xg for 15 min at 4°C; supernatants were separated by SDS-PAGE. After transfer to nitrocellulose or PVDF (LC3), membranes were blocked in milk or BSA and incubated at 4°C overnight with primary antibody. Detection was performed by enhanced chemiluminescence after 2-h incubation with HRP-conjugated secondary antibodies (Dako).

For co-IPs, cells were lysed in IP lysis buffer (150 mM NaCl, 50mM Tris-HCl pH 8, 5 mM EDTA, 5% glycerol, 1% triton, 25 mM NaF) supplemented with phosphatase and protease inhibitors (Roche). Lysates were pre-cleared 1 h with binding control agarose beads and incubated overnight with GFP-Trap® agarose beads (Chromtek) at 4°C. Beads were washed in IP lysis buffer, and bound protein was eluted.

For in vitro binding assays, GFP-Trap® agarose beads (Chromtek) were incubated with 5 ug GST (Novus) or GST-LC3B (Enzo) in IP lysis buffer for 1 h at 4°C, washed and blocked with 5% BSA overnight. Subsequently, beads were washed and incubated 1 h with 0.5 ug paxillin (Origene) at 4°C. After final washes, bound protein was eluted.All densitometry was performed on scanned films using ImageJ (NIH). Band intensities were normalized to the loading control, and normalized values are reported as fold change relative to the first lane.

Immunofluorescence

For LC3B staining, cells were fixed in 4% PFA, permeabilized in 100% MeOH at −20°C and blocked in 2% FBS/1% goat serum/PBS. Otherwise, cells were fixed and permeabilized (3.7% PFA/PBS in 100 mM PIPES pH 6.8, 10 mM EGTA, 1 mM MgCl2, and 0.2% Triton-X), washed in TBS/0.1% triton and blocked in TBS/0.1% BSA. Primary antibodies were α-LC3 (Cell Signaling), α-Paxillin (BD clone 349), and α-Zyxin (B71, Mary Beckerle, University of Utah). Cells were imaged with an Axiovert200m widefield fluorescence microscope (Zeiss). Image deconvolution was performed with Openlab software (PerkinElmer); all other image analysis was performed with ImageJ.

Quantification of FA size and number

After image deconvolution, an ImageJ macro was written to perform the following steps: one round of background subtraction followed by thresholding on FAs with the renyi entropy algorithm. The analyze particles function was used to measure the number of FAs and total FA area. To confirm that this accurately represented the FAs in the original image, the particles analyzed were manually compared to the original. For each experiment, 10–12 images/120–150 cells per sample were analyzed.

Live cell microscopy and quantification of FA dynamics

Timelapse DIC was performed with an Olympus LCV110U VivaView microscope system. 3D cell depth imaging was performed using a Zeiss LSM780 microscope. Total internal reflection fluorescence (TIRF) imaging of EGFP-paxillin FAs was performed with a Leica TIRFM superresolution microscope system (evanescent wave depth 80 nm; images acquired every 60 s for 30 min). No significant photobleaching occurred. The fluorescent intensity of individual EGFP-paxillin FAs over time on background-subtracted images was determined using ImageJ. Semilogarithmic plots of fluorescent intensity as a function of time were linear for both assembly and disassembly of EGFP-paxillin FAs, and rate constants were calculated from the slopes. Average rate constants were determined from 10–15 FAs from 5 cells for each sample. Confocal imaging of mApple-paxillin and EGFP-LC3B was performed on an Olympus DSU Spinning Disk confocal microscope. Colocalization was analyzed on background-subtracted Z-stacks (0.5-μm slices) in ImageJ using the the JACoP object-based method (Bolte and Cordelieres, 2006) with 10–15 images (10–25 cells)/condition/experiment.

Statistics

Data were analyzed with GraphPad Prism. Significance was determined by unpaired Student’s t-test for 2-group comparisons and 1-way ANOVA for >2-group comparisons. A Tukey post-hoc test was used to identify significant pairwise differences when 1-way ANOVA identified a significant difference between means. Significance for tumor growth curves was determined by repeated-measures 2-way ANOVA.

Supplementary Material

1
2
3
Download video file (1.7MB, avi)
4
Download video file (2.9MB, avi)
5
Download video file (873.2KB, avi)
6
Download video file (327.1KB, avi)
7
Download video file (283.5KB, avi)

Acknowledgments

We thank Mary Beckerle (University of Utah) for the Zyxin antibody (B71); Rick Horwitz (University of Virginia) for the EGFP-paxillin expression plasmid; Sara Courtneidge (OHSU) for the pSG5-SRCY527F plasmid; Fred Miller (Wayne State University) for 4T1 tumor cells; Thomas Gajewski (University of Chicago) for B16.F10 melanoma cells; and Ana Pasapera (NIH NHLBI) for the paxillin IP protocol. This work is supported by NIH RO1 CA162405 (KFM), the University of Chicago Medical Scientist Training Program (MNS and EEM) and the Human Tissue Resource Center, Digital Light Microscopy and Flow Cytometry Core Facilities through the University of Chicago Cancer Center Support Grant (P30 CA014599).

Footnotes

Author contributions

Conceptualization, M.N.S, E.E.M, K.F.M.; Methodology, M.N.S, E.E.M, K.F.M.; Investigation, M.N.S, E.E.M, L.E.D., C.C., M.Z., S.M.; Resources, M.N.S, E.E.M, L.E.D, H.C.; Writing – Original Draft, M.N.S.; Writing – Review & Editing, M.N.S, E.E.M, K.F.M.; Funding Acquisition, K.F.M.; Supervision, K.F.M.;

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Abbi S, Ueda H, Zheng C, Cooper LA, Zhao J, Christopher R, Guan JL. Regulation of focal adhesion kinase by a novel protein inhibitor FIP200. Mol Biol Cell. 2002;13:3178–3191. doi: 10.1091/mbc.E02-05-0295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Abou Zeid N, Vallés AM, Boyer B. Serine phosphorylation regulates paxillin turnover during cell migration. Cell Commun Signal. 2006;4:8. doi: 10.1186/1478-811X-4-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Amaravadi RK, Lippincott-Schwartz J, Yin XM, Weiss WA, Takebe N, Timmer W, DiPaola RS, Lotze MT, White E. Principles and current strategies for targeting autophagy for cancer treatment. Clin Cancer Res. 2011;17:654–666. doi: 10.1158/1078-0432.CCR-10-2634. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Aslakson CJ, Miller FR. Selective events in the metastatic process defined by analaysis of the sequential dissemination of subpopulations of a mouse mammary tumor. Cancer Res. 1992;52:1399–1405. [PubMed] [Google Scholar]
  5. Birgisdottir AB, Lamark T, Johansen T. The LIR motif - crucial for selective autophagy. J Cell Sci. 2013;126:3237–3247. doi: 10.1242/jcs.126128. [DOI] [PubMed] [Google Scholar]
  6. Bolte S, Cordelieres FP. A guided tour into subcellular colocalization analysis in light microscopy. Journal of microscopy. 2006;224:213–232. doi: 10.1111/j.1365-2818.2006.01706.x. [DOI] [PubMed] [Google Scholar]
  7. Caino MC, Chae YC, Vaira V, Ferrero S, Nosotti M, Martin NM, Weeraratna A, O’Connell M, Jernigan D, Fatatis A, et al. Metabolic stress regulates cytoskeletal dynamics and metastasis of cancer cells. J Clin Invest. 2013;123:2907–2920. doi: 10.1172/JCI67841. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Caino MC, Ghosh JC, Chae YC, Vaira V, Rivadeneira DB, Faversani A, Rampini P, Kossenkov AV, Aird KM, Zhang R, et al. PI3K therapy reprograms mitochondrial trafficking to fuel tumor cell invasion. Proc Natl Acad Sci U S A. 2015;112:8638–8643. doi: 10.1073/pnas.1500722112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen GC, Lee JY, Tang HW, Debnath J, Thomas SM, Settleman J. Genetic interactions between Drosophila Atg1 and paxillin reveal a role for paxillin in autophagosome formation. Autophagy. 2008;4:37–45. doi: 10.4161/auto.5141. [DOI] [PubMed] [Google Scholar]
  10. Deakin NO, Turner CE. Paxillin comes of age. Journal of cell science. 2008;121:2435–2444. doi: 10.1242/jcs.018044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Degenhardt K, Mathew R, Beaudoin B, Bray K, Anderson KL, Chen G, Mukherjee C, Shi Y, Gelinas C, Fan Y, et al. Autophagy promotes tumor cell survival and restricts necrosis, inflammation and tumorigenesis. Cancer Cell. 2006;10:51–64. doi: 10.1016/j.ccr.2006.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Didier C, Broday L, Bhoumik A, Israeli S, Takahashi S, Nakayama K, Thomas SM, Turner CE, Henderson S, Sabe S, Ronai ZA. RNF5, a RING finger protein that regulates cell motility by targeting paxillin ubiquitination and altered localization. Mol Cell Biol. 2003;23:5331–5345. doi: 10.1128/MCB.23.15.5331-5345.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fincham VJ, Frame MC. The catalytic activity of Src is dispensable for translocation to focal adhesions but controls the turnover of these structures during cell motility. The EMBO J. 1998;17:81–92. doi: 10.1093/emboj/17.1.81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fung C, Lock R, Gao S, Salas E, Debnath J. Induction of autophagy during extracellular matrix detachment promotes cell survival. Mol Biol Cell. 2008;19:797–806. doi: 10.1091/mbc.E07-10-1092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Galavotti S, Bartesaghi S, Faccenda D, Shaked-Rabi M, Sanzone S, McEvoy A, Dinsdale D, Condorelli F, Brandner S, Campanella M, et al. The autophagy-associated factors DRAM1 and p62 regulate cell migration and invasion in glioblastoma stem cells. Oncogene. 2013;32:699–712. doi: 10.1038/onc.2012.111. [DOI] [PubMed] [Google Scholar]
  16. Galluzzi L, Pietrocola F, Bravo-San Pedro JM, Amaravadi RK, Baehrecke EH, Cecconi F, Codogno P, Debnath J, Gewirtz DA, Karantza V, et al. Autophagy in malignant transformation and cancer progression. The EMBO J. 2015;34:856–880. doi: 10.15252/embj.201490784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Giatromanolaki AN, Charitoudis GS, Bechrakis NE, Kozobolis VP, Koukourakis MI, Foerster MH, Sivridis EL. Autophagy patterns and prognosis in uveal melanomas. Mod Pathol. 2011;24:1036–1045. doi: 10.1038/modpathol.2011.63. [DOI] [PubMed] [Google Scholar]
  18. Guo JY, Karsli-Uzunbas G, Mathew R, Aisner SC, Kamphorst JJ, Strohecker AM, Chen G, Price S, Lu W, Teng X, et al. Autophagy suppresses progression of K-ras-induced lung tumors to oncocytomas and maintains lipid homeostasis. Genes & Dev. 2013;27:1447–1461. doi: 10.1101/gad.219642.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Gupta A, Roy S, Lazar AJF, Wang WL, McAuliffe JC, Reynoso D, McMahon J, Taguchi T, Floris G, Debiec-Rychter M, et al. Autophagy inhibition and antimalarials promote cell death in gastrointestinal stromal tumor (GIST) Proc Natl Acad Sci U S A. 2010;107:14333–14338. doi: 10.1073/pnas.1000248107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hagel M, George EL, Kim A, Tamimi R, Opitz SL, Turner CE, Imamoto A, Thomas SM. The adaptor protein paxillin is essential for normal development in the mouse and is a critical transducer of fibronectin signaling. Mol Cell Biol. 2002;22:901–915. doi: 10.1128/MCB.22.3.901-915.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hara T, Takamura A, Kishi C, Iemura S, Natsume T, Guan JL, Mizushima N. FIP200, a ULK-interacting protein, is required for autophagosome formation in mammalian cells. J Cell Biol. 2008;181:497–510. doi: 10.1083/jcb.200712064. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Ilic D, Furuta Y, Kanazawa S, Takeda N, Sobue K, Nakatsuji N, MNomura S, Fujimoto J, Okada M, Yamamoto T. Reduced cell motility and enhanced focal adhesion contact formation in cells from FAK-deficient mice. Nature. 1995;377:539–544. doi: 10.1038/377539a0. [DOI] [PubMed] [Google Scholar]
  23. Karantza-Wadsworth V, Patel S, Kravchuk O, Chen G, Mathew R, Jin S, White E. Autophagy mitigates metabolic stress and genome damage in mammary tumorigenesis. Genes & Dev. 2007;21:1621–1635. doi: 10.1101/gad.1565707. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Karsli-Uzunbas G, Guo JY, Price S, Teng X, Laddha SV, Khor S, Kalaany NY, Jacks T, Chan CS, Rabinowitz JD, White E. Autophagy Is Required for Glucose Homeostasis and Lung Tumor Maintenance. Cancer Discov. 2014;4:914–927. doi: 10.1158/2159-8290.CD-14-0363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kenific CM, Stehbens SJ, Goldsmith J, Leidal AM, Faure N, Ye J, Wittmann T, Debnath J. NBR1 enables autophagy-dependent focal adhesion turnover. J Cell Biol. 2016 doi: 10.1083/jcb.201503075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kimmelman AC. The dynamic nature of autophagy in cancer. Genes & Dev. 2011;25:1999–2010. doi: 10.1101/gad.17558811. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Klionsky DJ, Abdelmohsen K, Abe A, Abedin MJ, Abeliovich H, Acevedo Arozena A, Adachi H, Adams CM, Adams PD, Adeli K, et al. Guidelines for the use and interpretation of assays for monitoring autophagy (3rd edition) Autophagy. 2016;12:1–222. doi: 10.1080/15548627.2015.1100356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Korolchuk VI, Mansilla A, Menzies FM, Rubinsztein DC. Autophagy Inhibition Compromises Degradation of Ubiquitin-Proteasome Pathway Substrates. Mol Cell. 2009;33:517–527. doi: 10.1016/j.molcel.2009.01.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kroemer G, Marino G, Levine B. Autophagy and the integrated stress response. Mol Cell. 2010;40:280–293. doi: 10.1016/j.molcel.2010.09.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kuma A, Hatano M, Matsui M, Yamamoto A, Nakaya H, Yoshimori T, Ohsumi Y, Tokuhisa T, Mizushima N. The role of autophagy during the early neonatal starvation period. Nature. 2004;432:1032–1036. doi: 10.1038/nature03029. [DOI] [PubMed] [Google Scholar]
  31. Laukaitis CM, Webb DJ, Donais K, Horwitz AF. Differential dynamics of alpha 5 integrin, paxillin and alpha-actinin during formation and disassembly of adhesions in migrating cells. J Cell Biol. 2001;153:1427–1440. doi: 10.1083/jcb.153.7.1427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Lazova R, Camp RL, Klump V, Siddiqui SF, Amaravadi RK, Pawelek JM. Punctate LC3B expression is a common feature of solid tumors and associated with proliferation, metastasis, and poor outcome. Clin Cancer Res. 2012;18:370–379. doi: 10.1158/1078-0432.CCR-11-1282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Liu L, Feng D, Chen G, Chen M, Zheng Q, Song P, Ma Q, Zhu C, Wang R, Qi W, et al. Mitochondrial outer-membrane protein FUNDC1 mediates hypoxia-induced mitophagy in mammalian cells. Nat Cell Biol. 2012;14:177–185. doi: 10.1038/ncb2422. [DOI] [PubMed] [Google Scholar]
  34. Lock R, Kenific CM, Leidal AM, Salas E, Debnath J. Autophagy dependent production of secreted factors facilitates oncogenic RAS-driven invasion. Cancer Discov. 2014;4:466–479. doi: 10.1158/2159-8290.CD-13-0841. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Lu Z, Luo RZ, Lu Y, Zhang X, Yu Q, Khare S, Kondo S, Kondo Y, Yu Y, Mills GB, et al. The tumor suppressor gene ARHI regulates autophagy and tumor dormancy in human ovarian cells. J Clin Invest. 2008;118:3917–3929. doi: 10.1172/JCI35512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Mackintosh RL, Timpson P, Thorburn J, Anderson KI, Thornburn A, Ryan KM. Inhibition of autophagy impairs tumor cell invasion in an organotypic model. Cell Cycle. 2012;11:2022–2029. doi: 10.4161/cc.20424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mathew R, Karp CM, Beaudoin B, Vuoong N, Chen G, Chen HY, Bray K, Reddy A, Bhanot G, Gelinas C, et al. Autophagy suppresses tumorigenesis through elimination of p62. Cell. 2009;137:1062–1075. doi: 10.1016/j.cell.2009.03.048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mizushima N, Komatsu M. Autophagy: Renovation of cells and tissues. Cell. 2011;147:728–741. doi: 10.1016/j.cell.2011.10.026. [DOI] [PubMed] [Google Scholar]
  39. Parsons JT, Horwitz AF, Schwartz MA. Cell adhesion: integrating cytoskeletal dynamics and cellular tension. Nat Rev Mol Biol. 2010;11:633–643. doi: 10.1038/nrm2957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Peng YF, Shi YH, Shen YH, Ding ZB, Ke AW, Zhou J, Qiu SJ, Fan J. Promoting colonization in metastatic HCC cells by modulation of autophagy. PLoS ONE. 2013;8:e74407. doi: 10.1371/journal.pone.0074407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Qiang L, Zhao B, Ming M, Wang NJ, He TC, Hwang S, Thorburn A, He YY. Proc Natl Acad Sci U S A. AOP; 2014. Regulation of cell proliferation and migration by p62 through stabilization of Twist1. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
  42. Rao S, Tortola L, Perlot T, Wirnsberger G, Novatchkova M, Nitsch R, Sykacek P, Frank L, Schramek D, Komnenovic V, et al. A dual role for autophagy in a murine model of lung cancer. Nat Comm. 2014;5:3056–3070. doi: 10.1038/ncomms4056. [DOI] [PubMed] [Google Scholar]
  43. Ridley AJ. Life at the leading edge. Cell. 2011;145:1012–1022. doi: 10.1016/j.cell.2011.06.010. [DOI] [PubMed] [Google Scholar]
  44. Rogov V, Dotsch V, Johansen T, Kirkin V. Interactions between autophagy receptors and ubiquitin-like proteins form the molecular basis for selective autophagy. Mol Cell. 2014;53:167–178. doi: 10.1016/j.molcel.2013.12.014. [DOI] [PubMed] [Google Scholar]
  45. Rosenfeldt MT, Nixon C, Liu E, Mah LY, Ryan KM. Analysis of macroautophagy by immunohistochemistry. Autophagy. 2012;8:963–969. doi: 10.4161/auto.20186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Rosenfeldt MT, O’Prey J, Morton JP, Nixon C, MacKay G, Mrowinska A, Au A, Rai TS, Zheng L, Ridgway R, et al. p53 status determines the role of autophagy in pancreatic tumour development. Nature. 2013;504:296–300. doi: 10.1038/nature12865. [DOI] [PubMed] [Google Scholar]
  47. Sandilands E, Serrels B, McEwan DG, Morton JP, Macagno JP, McLeod K, Stevens C, Brunton VG, Langdon WY, Vidal M, et al. Autophagic targeting of Src promotes cancer cell survival following reduced FAK signalling. Nat Cell Biol. 2011;14:51–60. doi: 10.1038/ncb2386. [DOI] [PubMed] [Google Scholar]
  48. Schaller MD, Schaefer EM. Multiple stimuli induce tyrosine phosphorylation of the Crk-binding sites of paxillin. Biochem. 2001;360:57–66. doi: 10.1042/0264-6021:3600057. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Staskiewicz L, Thorburn J, Morgan MJ, Thorburn A. Inhibiting autophagy by shRNA knockdown. Autophagy. 2013;9:1449–1450. doi: 10.4161/auto.24895. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Strohecker AM, Guo JY, Karsli-Uzunbas G, Price SM, Chen GJ, Mathew R, McMahon M, White E. Autophagy sustains mitochondrial glutamine metabolism and growth of BrafV600E-driven lung tumors. Cancer Discov. 2013;3:1272–1285. doi: 10.1158/2159-8290.CD-13-0397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Webb DJ, Donais K, Whitmore LA, Thomas SM, Turner CE, Parsons JT, Horwitz AF. FAK-Src signaling through paxillin, ERK and MLCK regulates adhesion disassembly. Nat Rev Cancer. 2004;6:154–161. doi: 10.1038/ncb1094. [DOI] [PubMed] [Google Scholar]
  52. Webb DJ, Parsons JT, Horwitz AF. Adhesion assembly, disassembly and turnover in migrating cells - over and over and over again. Nat Cell Biol. 2002;4:E97–100. doi: 10.1038/ncb0402-e97. [DOI] [PubMed] [Google Scholar]
  53. Webb DJ, Schroder MJ, Brame CJ, Whitmore LA, Shabanowitz J, Hunt DF, Horwitz AF. Paxillin phosphorylation sites mapped by mass spectrometry. J Cell Sci. 2005;118:4925–4929. doi: 10.1242/jcs.02563. [DOI] [PubMed] [Google Scholar]
  54. White E. Deconvoluting the context-dependent role for autophagy in cancer. Nat Rev Cancer. 2012;12:401–410. doi: 10.1038/nrc3262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zaidel-Bar R, Ballestrem C, Kam Z, Geiger B. Early molecular events in the assembly of matrix adhesions at the leading edge of migrating cells. J Cell Sci. 2003;116:4605–4613. doi: 10.1242/jcs.00792. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS781702-supplement-1.pdf (35.5KB, pdf)
2
NIHMS781702-supplement-2.pdf (3.9MB, pdf)
3
Download video file (1.7MB, avi)
4
Download video file (2.9MB, avi)
5
Download video file (873.2KB, avi)
6
Download video file (327.1KB, avi)
7
Download video file (283.5KB, avi)

RESOURCES