Combined regulation of mTORC1 and lysosomal acidification ... · rapamycin (mTOR) as a GSK-3 target.25–28 mTOR is a nutrient sensor that controls protein synthesis, cell size and

ORIGINAL ARTICLE

Combined regulation of mTORC1 and lysosomal acidificationby GSK-3 suppresses autophagy and contributes to cancercell growthI Azoulay-Alfaguter1, R Elya1, L Avrahami1, A Katz and H Eldar-Finkelman

There is controversy over the role of glycogen synthase kinase-3 (GSK-3) in cancer progression. Recent work has implicated GSK-3 inthe regulation of mammalian target of rapamycin (mTOR), a known player in malignant transformation. Autophagy, a self-degradation pathway, is inhibited by mTOR and is tightly associated with cell survival and tumor growth. Here we show that GSK-3suppresses autophagy via mTOR complex-1 (mTORC1) and lysosomal regulation. We show that overexpression of GSK-3 isoforms(GSK-3α and GSK-3β) activated mTORC1 and suppressed autophagy in MCF-7 human breast cancer cells as indicated by reducedbeclin-1 levels and upregulation of sequestosome 1 (p62/SQSTM1). Further, overexpression of GSK-3 increased the number ofautophagosomes and inhibited autophagic flux. This activity was directly related to reduced lysosomal acidification triggered byGSK-3 (in which GSK-3β has a stronger impact). We found that lysosomal acidification is reduced in MCF-7 cells that also exhibitincreased levels of autophagosomes and p62/SQSTM1 and increased activity of mTORC1. Subsequently, treating cells with GSK-3inhibitors restored lysosomal acidification, enhanced autophagic flux and inhibited mTORC1. Furthermore, GSK-3 inhibitorsinhibited cell proliferation. We provide evidence that GSK3-mediated mTORC1 activity and GSK-3-mediated lysosomal acidificationoccur via distinct pathways, yet both mTORC1 and lysosomes control cell growth. Finally, we show that GSK-3-reduced lysosomalacidification inhibits endocytic clearance as demonstrated by reduced endocytic degradation of the epidermal growth factorreceptor. Taken together, our study places GSK-3 as a key regulator coordinating cellular homeostasis. GSK-3 inhibitors may beuseful in targeting mTORC1 and lysosomal acidification for cancer therapy.

Oncogene advance online publication, 15 December 2014; doi:10.1038/onc.2014.390

INTRODUCTIONThe serine threonine glycogen synthase kinase-3 (GSK-3) is amulti-functional enzyme.1–5 In mammals, GSK-3 is expressed as twoisozymes, GSK-3α and GSK-3β.6 GSK-3 has emerged as a potentialtherapeutic target for treatment of various diseases includingdiabetes, Alzheimer’s disease and affective disorders.2,5,7,8 The roleof GSK-3 in cancer and tumor progression is still under debate.9–12

The initial link of GSK-3 with cancer was based on its involvement inthe canonical Wnt/β-catenin signaling pathway; GSK-3 targetsβ-catenin for proteosomal degradation.13–15 Inhibition of GSK-3 hasbeen thus considered a risk for cancer. Nevertheless, growingevidence supports the idea that GSK-3 enhances cancer cellproliferation. Treatment with GSK-3 inhibitors reduces cell pro-liferation and inhibits tumor growth in models of prostate, colon,pancreatic and ovarian cancers and glioblastoma andleukemia.16–19 The ability of GSK-3 to modulate pro-inflammatoryand anti-apoptotic processes involving nuclear factor-κB,20–22 thetumor suppressor protein p5323 and cyclin D1 levels24 also suggeststhat GSK-3 promotes cell growth.In previous work, we and others identified mammalian target of

rapamycin (mTOR) as a GSK-3 target.25–28 mTOR is a nutrientsensor that controls protein synthesis, cell size and cellgrowth.29–31 mTOR is found in two complexes, mTOR complex-1(mTORC1) and mTORC2, in which it associates with raptor and

rictor, respectively.32 Signaling through mTORC1 is involved in cellproliferation and tumor progression,29–31,33 and rapamycin and itsanalogs, which inhibit mTORC1 function, are used in cancertherapy.34–36 mTORC1 is a major suppressor of macroautophagy(herein referred to as autophagy), a highly regulated processresponsible for intracellular degradation. The autophagosomesequesters cellular materials and subsequently fuses with alysosome, which facilitates cargo degradation.37–40 Aberrantautophagic activity has a prominent impact on cell survival andtumor growth. Autophagy may be required for tumor suppressionas it preserves cellular homeostasis and limits genomedamage.41–46 On the other hand, autophagy may acceleratecancer at later stages by providing nutrients under stressedor starved conditions.41–46 Hence, maintaining ‘normal’ cellproliferation requires ‘correct’ balance between biosynthesis andautophagy. Whether GSK-3 maintains this balance via regulationof mTORC1 and autophagy is unclear.In the present study, we show that GSK-3 activates mTORC1 and

suppresses autophagy in breast cancer cells. We further identify anew role for GSK-3 in reducing lysosomal acidification that in turnsuppresses autophagic flux and inhibits endocytic clearance. Ourstudy places GSK-3 as a key player coordinating cellular home-ostasis and identified new mechanisms linking GSK-3 with cancercells growth.

Department of Human Molecular Genetics and Biochemistry, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel. Correspondence: Professor H Eldar-Finkelman,Department of Human Molecular Genetics and Biochemistry, Sackler School of Medicine, Tel Aviv University, Ramat Aviv, Tel Aviv 69978, Israel.E-mail: [email protected] authors contributed equally to this work.Received 31 August 2014; revised 10 October 2014; accepted 16 October 2014

Oncogene (2014), 1–11© 2014 Macmillan Publishers Limited All rights reserved 0950-9232/14

www.nature.com/onc

http://dx.doi.org/10.1038/onc.2014.390

mailto:[email protected]

http://www.nature.com/onc

RESULTSGSK-3 is a positive regulator of mTORC1We first examined whether GSK-3 regulates mTORC1 activity incancer cells. We used human breast adenocarcinoma MCF-7 cellsthat show growth sensitivity to rapamycin (Figure 1a). MCF-7 cellswere transiently transfected with green fluorescent protein (GFP)-tagged GSK-3α or GSK-3β constructs. (GFP-tagged constructs wereused because GPF tags make cell imaging experiments possible,similar results were obtained with non-tagged GSK-3.) mTORactivity was determined by evaluating the phosphorylation levelsof the mTOR target p70 ribosomal S6 kinase (S6K-1, Thr389) and itstarget ribosomal S6 protein (S6, Ser240/244). Phosphorylation levels

of S6K-1 significantly increased and those of S6 showed a clear butmodest increase in cells that overexpressed GFP-GSK-3α or GFP-GSK-3β (Figure 1b). To determine which of the mTOR complexeswere regulated by GSK-3, similar experiments were performed inthe presence of rapamycin. Rapamycin completely abolishedGSK-3-mediated phosphorylation of mTOR targets indicating thatGSK-3 activates mTORC1 (Figure 1c). GSK-3 elevated mTORC1activity even under starvation conditions such as low serum or noglucose (Figure 1d). It was difficult to assess GSK-3 activation ofmTORC1 in the absence of amino acids because the signal wasvery weak (Figure 1d). The TSC1 and TSC2 complexes that mediatemTORC1 activity mainly respond to stimulation by growth factors

a

24 48 72

0.5

1.0

1.5

2.0

Time (Hr)

Cel

l gro

wth

NTRap

β-actin

pS6K1

S6K1

pS6

S6

GSK-3GFP

GSK-3α/β

b c

GSK-3GFP

pS6K1

S6K1

- + + +- -Rap

Ctrl GSK3α GSK3β

GSK-3α/β

pS6/S6

**

0.0

0.5

1.0

1.5

2.0

d

0.1% FCS

pS6K1

S6K1

-Glucose -aa

e

GSK3α

pS6K1

β-actin

S6K1

GSK3β

f

g

TCS+/+ TCS-/-

pS6K1

β-actin

S6K1

pS6K1/S6K1*

*

0.0

0.5

1.0

1.5

2.0

Fol

d o

f Ctr

l

0.00

0.25

0.50

0.75

1.00

pS6K1/S6K1

**

Fol

d o

f Ctr

l

pS6/S6

0.0

0.25

0.50

0.75

1.00

*

**

Fol

d of

NT

*

pS6K1/S6K1

0.0

0.25

0.50

0.75

1.00

**

*

Fol

d of

NT

*

h

Cel

ls x

106

0

10

20

30

40

24 48 72

NTSB216783

AR-A01148L803-mtsCT99012

Time (Hr)

1.0

2.03.0

0.0ps6k

1/s6

k1

1.0

2.0

3.0

0.0

βαCtrl βαCtrl

siβsiαCtrl

β-actin

Figure 1. GSK-3 regulates mTORC1 and cell growth. (a) MCF-7 cell growth is sensitive to rapamycin: MCF-7 cells were treated with 50 nM

rapamycin (Rap) for 3 days and were counted every 24 h in triplicate samples as described in Materials and methods. (b) GSK-3 activates mTORtargets: MCF-7 cells were transiently transfected with GFP-GSK-3α or GFP-GSK-3β constructs. Control cells (Ctrl) were transfected with aconstruct encoding only GFP. Phosphorylation and total expression levels of S6K-1 (Thr389) and S6 (Ser240/244) were determined byimmunoblot analyses. Levels of endogenous total GSK-3 or GFP-GSK-3 proteins were determined by staining with an anti-GSK-3 antibody.β-Actin was analyzed to ensure equal loading. (c) GSK-3 activates mTORC1: cells were treated and analyzed as describe in b, except that cellswere treated with rapamycin (50 nM, 4 h). (d) GSK-3 rescues mTORC1 activity under starvation conditions: cells were treated and analyzed asdescribe in b, except that cells expressing GFP-GSK-3 were incubated with low serum (0.1% fetal calf serum), medium lacking glucose ormedium lacking amino acids (aa) for 4 h. (e) Inhibition of expression of GSK-3 isozymes inhibits mTORC1: MCF-7 cells were transfected withsiRNA targeting GSK-3α or GSK-3β. Control cells (Ctrl) were transfected with appropriate control siRNA. Levels of endogenous GSK-3 andβ-actin are shown. (f) GSK-3 inhibitors inhibit mTORC1 activity: MCF-7 cells were treated with SB-216763 (SB, 10 μM), AR-A014418 (AR, 20 μM),CT99012 (CT, 10 μM) and L803-mts (L803, 40 μM) for 4 h. (g) GSK-3 activates mTORC1 via the TSC1/2 complex: ‘wild-type’ MEF or MEF-TSC− /−

cells were treated with GSK-3 inhibitors as described in f. (h) GSK-3 inhibitors inhibit cell growth: MCF-7 cells were treated with SB-216763, AR-A014418, CT99012 and L803-mts (as described in f for 3 days). Cells were counted every 24 h in triplicate samples. For all panels,representative gels are shown. Calculated ratios of indicated phosphorylated to total protein were evaluated by densitometry analysis and areshown in bar graphs that present means of three to five independent experiments ± s.e.m., *Po0.05. NT, non-treated.

GSK-3 promotes breast cancer cell growthI Azoulay-Alfaguter et al

2

Oncogene (2014), 1 – 11 © 2014 Macmillan Publishers Limited

or glucose, but not to amino acids;47–50 therefore, these datasuggest that GSK-3 mediates mTORC1 activity (at least in part) viathe TSC1/2 complex.We next determined whether inhibition of GSK-3 produces an

effect opposite to that of overexpression. Inhibition of eitherGSK-3α or GSK-3β expression with small interfering RNA (siRNA)reduced S6K-1 phosphorylation (Figure 1e). There was a smallreduction in the phosphorylation of S6 that was not statisticallysignificant (data not shown). It is possible that ablation of theexpression of one isozyme alone was not sufficient to inhibit thephosphorylation of S6. Activity of both isozymes was then reducedby treating cells with selective GSK-3 inhibitors: the ATP-competitive small-molecule inhibitors SB-216763, ARA0114418and CT99012 and a cell-permeable peptide substrate competitiveinhibitor L803-mts.26 The three GSK-3 inhibitors reduced thephosphorylation levels of S6K-1 and S6 (Figure 1f). To furtherexamine whether GSK-3 mediates this effect via TSC1/2, we usedMEF cells lacking TSC1/2 that express constitutively activemTORC1 (MEF-TSC− /−).51 mTORC1 activity was inhibited byGSK-3 inhibitors in ‘wild-type’ MEF-TSC+/+ cells, whereas noinhibition was observed in MEF-TSC− /− cells (Figure 1g). Theseresults supported our notion that GSK-3 mediates mTORC1 activityvia the TSC1/2 complex. Finally, MCF-7 cells were treated withSB-216763, ARA0114418, CT99012 or L803-mts and cell prolifera-tion was determined. Treatment of cells with each of these GSK-3inhibitors inhibited cell proliferation by about 50% as comparedwith non-treated cells (Figure 1h). Together, our data show thatGSK-3 is a positive regulator of mTORC1, and this ability correlateswith regulation of cell growth.

GSK-3 inhibits autophagy and triggers accumulation ofautophagosomesAs mTORC1 suppresses autophagy, we asked whether GSK-3regulates autophagy. Autophagy is initiated by de novo formationof double-membrane vesicles called autophagosomes thatsequester intracellular waste.37–39 Fusion of autophagosomes withlysosomes facilitates cargo degradation.37–39 We examined levelsof several markers known to correlate with the number of auto-phagosomes and with autophagic activity. Microtubule-associatedprotein 1 light chain 3-II (LC3-II) is a lipidated form of LC3-I that isprimarily associated with phagophore (that is, autophagosomeprecursor) and autophagosome membranes.52,53 LC3-II levels thusserve as an index for the number of autophagosomes. Levels ofLC3-II significantly increased in cells that overexpressed GFP-GSK-3α or GFP-GSK-3β (Figure 2a, upper panel). To further confirmthe formation of autophagic vesicles, cells were co-transfectedwith GFP-GSK-3 (green) and RFP-LC3 (red) plasmids. Consistently,GSK-3α or GSK-3β increased the number of LC3-II puncta per cell(Figure 2b, lower panel). Thus, unexpectedly, we found that GSK-3increases the amount of autophagosomes.To further analyze autophagy, we determined levels of

p62/SQSTM1, an autophagy receptor protein that interactswith ubiquitinated proteins and LC3-II.54 p62/SQSTM1 (sequesto-some 1) levels thus increase on autophagy inhibition, whereasreduced p62/SQSTM1 levels indicate increased autophagicresponse and flux.54 Levels of p62/SQSTM11 were significantlyincreased in cells overexpressing GFP-GSK-3α or GFP-GSK-3β(Figure 2c, upper panel). Consistent with this, cells that were co-transfected with Cherry-GSK-3α or β (red) and GFP-p62/SQSTM1(green) showed an increase in p62/SQSTM1 puncta that werebrighter and larger than those in control cells (Figure 2d, lowerpanel). GSK-3β consistently gave a more robust effect onaccumulation of p62/SQSTM1 puncta than did GSK-3α.To further examine alterations in autophagy, levels of beclin-1, a

key factor that initiates formation of autophagosomes aftersuppression of mTORC1,55 was determined. Overexpression ofGFP-GSK-3α or GFP-GSK-3β reduced beclin-1 levels further

indicating suppression autophagy, likely via activation of mTORC1by GSK-3 (Figure 2e).The accumulation in LC3-II/autophagosomes (Figures 2a and b)

could not result from activation of mTORC1 (as mTORC1 inhibits theformation of autophagosomes). Hence, elevation in LC3-II likelyreflected inhibition in autophagosomes’ turnover due to defects inlysosomes. To further verify this possibility, cells overexpressingGSK-3 were treated with chloroquine (CQ), a reagent that neutralizelysosomal pH. Although CQ increased levels of LC3-II (as expected),overexpression of either GSK-3 isozyme did not further elevatelevels of LC3-II in the presence of CQ, indicating that GSK-3/lysosome axis was responsible for the reduced flux (Figure 2f;detailed analysis of GSK-3-lysosomes is described in Figure 4). Wethen examined the role of mTORC1 in inhibiting autophagic flux bymanipulating p62/SQSTM1 with rapamycin. Treatment with rapa-mycin largely blocked GSK-3 elevation of p62/SQSTM1 (Figure 2g).Taken together, GSK-3 suppresses autophagy and flux via mTORC1and lysosomes. In addition, it is noteworthy that suppression ofautophagy and upregulation of p62/SQSTM1 are important playersin cancer progression,56,57 further supporting the link betweenGSK-3 and cancer.

Autophagy and lysosomal acidification are suppressedH in MCF-7cellsThe role of GSK-3 in inhibiting autophagic flux was an intriguingfinding and hinted at a mechanism controlling lysosomal activity.Lysosomes are acidic organelles and their ‘correct’ acidification(pHo4.5) is critical for proteolytic activity.58–60 Lysosomeacidification can be visualized in live cells by staining of acidicorganelles with LysoTracker Red dye. The LysoTracker Red signalin MCF-7 cells was very weak compared with the signal in the‘non-cancerous’ mammary epithelial MCF-10A cells (Figure 3a).(Treatment with CQ reduced the fluorescence of the red vesiclesand confirmed dye specificity26 (data not shown).) In addition,LC3-II and p62/SQSTM1 expression levels were significantly higherin MCF-7 cells than in MCF-10A cells (Figure 3b). The lysosomalhydrolase cathepsin D (CatD) serves as a marker for lysosomalactivity. CatD is proteolyzed in acidified lysosomes to produce anactive mature fragment (mCatD). Consistently, mCatD levels weresignificantly lower in MCF-7 cells than in MCF-10A cells, verifyingreduced lysosomal acidification in MCF-7 cells (Figure 3b). Lamp2,a marker for lysosomes, was abundant in both cell lines(Figure 3b). Finally, MCF-7 cells showed significantly higher levelsof mTORC1 activity than did MCF-10A cells (Figure 3c).Interestingly, overexpression of GSK-3 had a similar impact on

mTORC1 and autophagy in the non-cancer-derived MCF-10A cellsas in the MCF-7 cells. In MCF-10A cells, overexpression of GSK-3αor GSK-3β increased mTORC1 activity and reduced beclin-1 levels(Figure 3d). Overexpression of GSK-3 also increased levels of LC3-IIand p62/SQSTM1 and reduced levels of mCatD (Figure 3e). Takentogether, these data show that MCF-7 cells are deficient oflysosomal and autophagic activity. In addition, we conclude thatGSK-3 impacts on mTORC1 and autophagy is a likely a broaderphenomenon that is not restricted to MCF-7 cells.

GSK-3 regulates lysosomal acidification and reduces autophagic fluxThe results presented in Figure 2 suggested that GSK-3 regulateslysosomal acidification. To further explore this possibility, MCF-7cells were transfected with GFP-GSK-3α or- GSK-3β and thenstained with LysoTracker Red to evaluate the acidic pool oflysosomes. The experiments were performed in starved cells (0.1%serum), as these conditions enabled a better detection of acidifiedlysosomes (which are present at low levels in MCF-7 cells asindicated). The numbers of lysosomal puncta and intensities weremarkedly reduced in cells overexpressing GSK-3β compared withcontrol cells (Figure 4a). The effect of GSK-3α was less pronouncedthan was the effect of GSK-3β: only in part of the cell population,


3

© 2014 Macmillan Publishers Limited Oncogene (2014), 1 – 11

we observed a reduction in numbers of puncta relative to controlcells (Figure 4a). It is possible that the effect of GSK-3α onlysosomes is dependent on additional cellular factors that are notknown at this point.As MCF-7 cells showed deficits in acidic lysosomes, we

examined whether inhibition of GSK-3 activity would restore thisdeficiency. The cells were treated with SB-216763, ARA0114418 orL803-mts and then stained with LysoTracker Red. Treatment withGSK-3 inhibitors increased the pool of acidified lysosomes(Figure 4b). Immunofluorescence analysis in fixed cells stainedwith anti-Lamp2 antibody indicated that GSK-3 inhibitors did notalter levels of Lamp2 (Figure 4c). Consistent results were obtained

in western blot (Figure 4d). Thus, GSK-3 inhibitors did not increasethe number of lysosomes, but rather altered acidification.We expected that improved lysosomal acidification achieved

with GSK-3 inhibitors would enhance autophagic flux. It wasdifficult to assess autophagic flux by measuring LC3-II becauseinhibition in mTORC1 increases LC3-II levels, although activation oflysosomes enhances LC3-II degradation resulting in a ‘zero’ neteffect. We thus used p62/SQSTM1 levels as a marker for flux:inhibition in mTORC1 activates autophagy and enhances p62/SQSTM1 degradation and activated lysosomes accelerates autop-hagosmses’ turnover resulting in p62/SQSTM1 degradation aswell. Changes in p62/SQSTM1 were determined by immuno-

LC3- II/LC3

0.0

0.5

1.0

1.5 * *p62/SQTM1

0.00.51.01.52.0 *

*

αGSK-3α/β

GSK-3GFP

LC3-I

β-actin

LC3-II

GFP LC3 MergeCtrl

GSK3α

GSK3�

Zoom Cherry p62/SQSTM1

GSK3α

Ctrl

Merge

GSK3�

GSK-3α/βGSK-3GFP

p62/SQTM1

β-actin

Beclin1

GSK-3α/βGSK-3GFP

10µm 5µm

β-actin

GSK-3GFP

GSK-3α/β

- + + +CQ

p62/ SQSTM1

- + + +- -Rap

β-actin

Fold

of

Ctrl

GSK-3GFP

GSK-3α/β

Fold

of

Ctrl

0.0

1.0

2.0

0.5

1.5

- + + +- -Rap

Ctrl α β

Ctrl β

Beclin1

0.250.5

0.751.0

0.0

* *

Fold

of C

trl

Ctrl α β

Fold

of C

trlLC3-IILC3-I

Figure 2. GSK-3 suppresses autophagy and triggers accumulation of autophagosomes. (a) GSK-3 overexpression increases levels of LC3-II:MCF-7 cells were transiently transfected with GFP-GSK-3α or GFP-GSK-3β constructs. Control cells (Ctrl) were transfected with a constructencoding only GFP. LC3-I and LC3-II levels were determined by immunoblot analysis. Calculated ratios of LC3-II/total LC3 (total LC3= LC3-I+LC3-II) are shown in the right panel. Levels of endogenous and GFP-GSK-3 proteins are also shown. β-Actin was analyzed as a control forequal loading. (b) GSK-3 upregulates numbers of autophagosomes per cell. MCF-7 cells were co-transfected with RFP-LC3 (red) and GFP-GSK-3α or GFP-GSK-3β (green) plasmids. Control cells (GFP) were transfected with a construct encoding GFP only. Live cells were imaged byconfocal microscopy. Arrows indicate representative cells expressing GSK-3 (green) and RFP-LC3 (red). Zoom of the LC3-II signal inautophagosomes is indicated by rectangle and yellow arrows (c). GSK-3 elevates p62/SQSTM1 levels: cells were treated as in a, except thatp62/SQSTM1 levels were determined by immunoblot analysis. (d) GSK-3 expression increases numbers of p62/SQSTM1 aggregates: MCF-7cells were co-transfected with GFP-p62/SQSTM1 and Cherry-GSK-3α or Cherry-GSK-3β expression plasmids. Control cells (Cherry) weretransfected with a construct encoding only Cherry. Live cells were imaged by confocal microscopy. Arrows indicate representative cellsexpressing GSK-3 (red) and GFP-p62/SQSTM1 (green). (e) GSK-3 expression reduces beclin-1 levels: cells were treated as in a, except thatcell lysates were blotted for beclin-1. (f) GSK-3 upregulates LC3-II via lysosomes: MCF-7 cells were transiently transfected with GSK-3 plasmidsas described in a. Cells were treated with CQ and levels of LC3-II were determined. No statistical differences were obtained in LC3-II levelsin cells expressing GSK-3 that were treated or non-treated with CQ. (g) GSK-3 inhibits autophagy via mTORC1: MCF-7 cells weretransiently transfected with GSK-3 plasmids as described in a. Cells were treated or not treated with rapamycin (50 nM, 4 h) and levels ofp62/SQSTM1 were determined. For all panels, representative gels are shown, and bar graphs present means of three to five independentexperiments ± s.e.m., *Po0.05.


4


fluorescence analyses in cells treated with GSK-3 inhibitors andstained with p62/SQSTM1 antibody. These experiments showedthat numbers of p62/SQSTM1 puncta were significantly reducedby the GSK-3 inhibitors (Figure 4e). Consistently, western blottingalso showed reduced p62/SQSTM1 levels on treatment of cellswith GSK-3 inhibitors (Figure 4f). Thus, GSK-3 inhibitors re-acidifylysosomes and enhance autophagic flux. As suppression inautophagy contributes to progression of early-stage cancers,41–45

GSK-3 inhibitors may be useful in treating these conditions.

GSK-3 controls mTORC1 and lysosomes via distinct mechanismsthat both act to regulate cell growthTo complete this part of our studies, we examined whethermTORC1 and lysosome pathways crosstalk61,62 via GSK-3 andwhether both mTORC1 and lysosomes underlie the impact ofGSK-3 inhibitors on cell growth. To determine whether GSK-3activates mTORC1 via lysosomes, we treated MCF-7 cells with CQ.This treatment inhibited the phosphorylation of mTORC1 target,S6, indicating that dysfunctional lysosomes may inhibit mTORC1(Figure 5a). However, CQ treatment did not prevent activation ofmTORC1 by GSK-3α and GSK-3β, indicating that GSK-3 activatesmTORC1 independently of lysosomal function (Figure 5a). Toexamine whether GSK-3 regulation of lysosomes occurs through

mTORC1, cells were treated with rapamycin and tested forlysosomal acidification. Rapamycin slightly enhanced lysosomalacidification; however, the effect of rapamycin was much weakerthan that observed with L803-mts (Figure 5b) or other GSK-3inhibitors (data not shown). From these results we conclude thatGSK-3/mTORC1 and GSK-3/ lysosomes axes represent indepen-dent processes.Inhibition of growth of MCF-7 cells when treated with GSK-3

inhibitors correlated well with the abilities of these compounds toinhibit mTORC1, the results with rapamycin and previous dataindicating that mTORC1 promote cell growth (see Figure 1 andLaplante and Sabatini,29 Cornu et al.,30 Proud31 andMamane et al.33). As GSK-3 inhibitors restored lysosomal acidifica-tion independently of mTORC1 as discussed above, we askedwhether lysosomal acidification also manipulates cell growth.There are no known drugs that directly enhance lysosomalacidification. We thus screened for conditions that resulted in re-acidification of lysosomes in MCF-7 cells. We found that starvation(either low serum or lack of amino acids) restored lysosomalacidification (Figure 5c). Starvation increased autophagic flux asjudged by reduced levels of LC3-II and p62/SQSTM1 (Figure 5d).As expected, starvation inhibited mTORC1 (Figure 5d). The factthat both rapamycin and starvation inhibited mTORC1, althoughonly starvation showed a strong impact on lysosomal acidification,

a MCF-10AMCF-7

Lys

BF

P62/SQSTM1

β-actin

MCF-7 MCF-10A

LC3-I

LC3-II

b

Cat D

mCat D

p62/SQSTM1

*

*

0.0

0.5

1.0

1.5

2.0

mCat D

0.0

0.25

0.50

0.75

1.0

*

LC3-II

*

0.0

0.25

0.50

0.75

1.0 *

0.0

0.5

1.0

1.5

Fol

d of

MC

F-7

LC3-II/LC3*

*

0.0

0.5

1.0

1.5

2.0

Fol

d of

Ctr

l

mCatD

0.0

0.25

0.50

0.75

1.0

**

Fol

d of

Ctr

l

Lamp2

GSK-3GFP

GSK-3α/β

β-actin

p62/SQTM

LC3-I

e

LC3-II

MCF-10A

Cat D

mCat D

Lamp2

GSK-3GFP

GSK-3α/β

Beclin1

pS6K1

S6K1

d

β-actin

MCF-10A

0.0

0.25

0.50

0.75

1.0

Fol

d of

Ctr

l *

Beclin1

0.0

1.0

2.0

3.0

4.0

Fol

d of

Ctr

l

**pS6K1/S6K1

*

Ctrl α ββ

β

α

α

Ctrl

CtrlβαCtrl

βαCtrl

pS6K1

β-actin

S6K1

pS6

S6

MCF-7 MCF-10A

c

p62/SQSTM1

Figure 3. Autophagy and lysosomal acidification are suppressed in MCF-7 cells. (a) Lysosomal acidification is reduced in MCF-7 cells: MCF-7cells and MCF-10A cells were stained with LysoTracker Red (Lys) and live cells were imaged by confocal microscopy. White arrows indicateacidified lysosomal puncta (upper panel). Lower panel shows merged image of bright field with LysoTracker Red (BF). (b) Comparison betweenautophagy and lysosomes in MCF-7 and MCF-10A cells: expression levels of LC3-I and LC3-II, p62/SQSTM1, mCatD and Lamp2 weredetermined in MCF-7 cells and in MCF-10A cells by immunoblot analysis. Bar graphs show densitometry analysis of indicated proteinsexpressed in both cell lines. (c) mTORC1 activity is higher in MCF-7 cells: cells were treated as described in b, except that S6K-1 and S6 wereanalyzed by immunoblot analysis. (d) GSK-3 activates mTORC1 and inhibits autophagy in MCF-10A cells: MCF-10A cells were transientlytransfected with GFP-GSK-3α or GFP-GSK-3β constructs. Control cells (Ctrl) were transfected with a construct encoding GFP. Phosphorylationand total expression levels of S6K-1 (Thr389) and beclin-1 were determined by immunoblot analyses. Expression levels of endogenous GSK-3and GFP-GSK-3 proteins are also shown. β-Actin was analyzed as a control for equal loading (e). GSK-3 impacts autophagy and lysosomes inMCF-10A cells: cells were treated as in d, except that levels of LC3-II, p62/SQSTM1, CatD, mCatD and Lamp2 were determined. Bar graphs showdensitometry analyses of indicated proteins. For all panels, representative gels or images are shown, and densitometry analyses bar graphs ofindicated proteins present means of three to five independent experiments ± s.e.m., *Po0.05.


5


promoted us to compare growth rates of MCF-7 cells in thepresence of rapamycin or starvation medium. Rapamycin reducedcell growth by about 50%, however, starvation had an evenstronger impact on cell proliferation (Figure 5e). Hence, enhancedlysosomal acidification could be an additional player in inhibitingcell growth. (We acknowledge the possibility that additionalmechanisms activated or inhibited by starvation may beresponsible for this phenomenon.) Our results suggest that byinhibiting mTORC1 and ‘correcting’ lysosomal acidification GSK-3inhibitors inhibit cell growth.

GSK-3 impairs endocytic clearanceAs lysosomes are located at the crossroad of autophagic andendocytic degradation, it is possible that defects in lysosomal

acidification impair endocytic clearance. Epidermal growth factorreceptor (EGFR) undergoes endocytosis and degradation inlysosomes.63–65 We thus used EGFR as a marker for lysosomalfunction. To verify that EGFR clearance is mediated by lysosomes,MCF-10A cells (that express endogenous EGFR) were treated withCQ or leupeptin (which inhibits lysosomal proteases) and EGFRlevels were measured. Both CQ and leupeptin increased levels ofEGFR (Figure 6a). We then examined whether GSK-3 mediatesEGFR endocytic degradation. MCF-10A cells transfected with GFP-GSK-3 constructs were treated with EGF. EGFR downregulationwas observed after 60–240min treatment with EGF, but over-expression of GSK-3α or GSK-3β attenuated EGFR clearance(Figure 6b). Inhibition in EGFR clearance was accompanied withsustained activation of EGFR downstream targets ERK1 and ERK2(Erks) as indicated by increased ERKs phosphorylation (Figure 6b).

Lamp2

β-actin

NT L803mtsARSB

SB 216763

NT

L803mts

Lys BF

AR A0114418

Lamp2

SB 216763

L803mts

NT

AR A0114418

10µm

NT SB 216763 L803mts AR A0114418

BF

p62/

SQ

TM

p62/ SQTM1

β-actin

Ctrl GSK3α GSK3�G

FPLy

so

10µm

p62/SQTM1

NT SB AR L803mts0.00.20.40.60.81.0

NT L803mtsARSB

Figure 4. GSK-3 regulates lysosomal acidification and autophagic flux. (a) GSK-3 reduces lysosomal acidification: MCF-7 cells were transientlytransfected with GFP-GSK-3α or GFP-GSK-3β constructs. Control cells (Ctrl) were transfected with a construct encoding GFP only. Cells werestarved with 0.1% fetal calf serum (4 h) and stained with LysoTracker Red (Lys). Live cells were imaged by confocal microscopy. Arrows indicaterepresentative cells expressing GFP-GSK-3 (upper panel) and lysosomal puncta (lower panel). (b) GSK-3 inhibitors restore lysosomalacidification: MCF-7 cells were treated with 10 μM SB-216763, 20 μM AR-A014418 or 40 μM L803-mts for 4 h. Cells were stained with LysoTrackerRed (Lys). Left panel shows live cells imaged by confocal microscopy. Right panel shows merged image of bright field with LysoTracker Red(BF). (c) GSK-3 inhibitors do not alter Lamp2 expression: cells were treated as in b, except that cells were fixed with paraformaldehyde (PFA)and subjected to immunofluorescence analysis with anti-Lamp2 antibody. Images were taken by confocal microscopy. (d) Cells were treatedas in c, except that Lamp2 levels were determined by immunoblot analysis. (e) GSK-3 inhibitors decrease p62/SQSTM1 levels: MCF-7 cells weretreated with GSK-3 inhibitors as described in b. Cells were fixed with PFA and subjected to immunofluorescence analysis using anti-p62/SQSTM1 antibody. Images were taken by confocal microscopy. (f) Cells were treated as in d, except that p62/SQSTM1 levels were determinedby immunoblot analysis. For all panels, representative gels or images are shown and densitometry analyses bar graphs of indicated proteinspresent means of three to five independent experiments ± s.e.m. NT, non-treated.


6


These results indicated that GSK-3 slows down EGFR endocyticclearance.Lysosomal acidification restored by GSK-3 inhibitors observed in

MCF-7 cells could enhance endocytic degradation of EGFR. Weused MCF-7 cells that were transfected with EGFR plasmid (thesecells do not endogenously express EGFR). The expressed receptorwas functional as indicated by EGF-induced tyrosine phosphoryla-tion of EGFR and by activation of ERKs (Figure 6c). Treatment withthe GSK-3 inhibitor L803-mts reduced ‘basal’ levels of EGFR, EGFRtyrosine phosphorylation and ERK phosphorylation (Figure 6d). Tofurther verify that this reduction reflects EGFR endocytosis, weused MCF-7 cells that expressed GFP-EGFR, which were treatedwith EGF. At time 0, the EGFR signal was localized at the cellsurface (Figure 6e). Over time in the presence of EGF, the receptorwas internalized and gradually accumulated in endosomes as‘green’ cytoplasmic aggregates that remained at 4 h aftertreatment with EGF (Figure 6e, upper panel). In the presence ofthe GSK-3 inhibitor L803-mts, the EGFR signal disappeared after4 h treatment with EGF (Figure 6e, lower panel). To further verifythat enhanced EGFR clearance was mediated by lysosomes, thecells were treated with CQ. Treatment with CQ elevated EGFRaggregates and prevented their degradation by L803-mts(Figure 6g). Thus, enhanced degradation of EGFR by L803-mtswas mediated by lysosomes.

DISCUSSIONHere we show that GSK-3 networks with mTORC1, autophagy andlysosomes represent mechanisms in which GSK-3 coordinate

cellular homeostasis and regulate cell proliferation. We firstdetermined that GSK-3 is a positive regulator of mTORC1. Thispathway is known to promote cell proliferation and cancerdevelopment: elevated mTORC1 activity is detected in many typesof cancers and its inhibition by the drug rapamycin provokesanticancer effects.66–68 Inhibition of cell proliferation observedwith GSK-3 inhibitors is likely due to their ability to inhibitmTORC1, which mimics the anticancer activity of rapamycin;31,34,35

these data have important implications in breast cancer for whichmTOR is considered a therapeutic target.69,70 Our studies are inagreement with previous work that positioned GSK-3 as anactivator of the mTORC1 pathway.25,27 In contrast, results fromother analyses indicate that GSK-3 inhibits mTORC1.71–73 Onestudy revealed that a coordinated phosphorylation of TSC2 byGSK-3 and AMPK suppresses mTORC1 activity.71 A more recentstudy reported that GSK-3 interacts with AMPK and inhibits AMPKactivity, thus preventing the suppressive effects of AMPK onmTORC1.74 Thus, the role of GSK-3 in regulation of mTORC1 maydepend on cell type or cell context. This issue needs furtherclarification.We then broadened our focus to analyze the effect of GSK-3 on

regulation of autophagy. We found that GSK-3 inhibits autophagy—as would be expected from the fact that GSK-3 activatesmTORC1. However, we also observed an increase in the numbersof autophagosomes in cells that overexpressed GSK-3, suggestinginhibition in autophagic turnover due to a deficiency in lysosomalactivity. Indeed, further experiments showed that GSK-3 impairslysosomal acidification, which in turn reduces the turnover ofautophagosomes. We further indicated that GSK-3 regulates

GSK3α GSK3β

GSK-3GFP

GSK-3α/β

Lys

BF

NT Rap L803mts Rap +L803mtsb

β-actin

pS6

S6

- + + +- -CQ

Ctrla

cCtrl Strv

Lys

BF

Ctrl Strv

LC3-ILC3-II

p62/SQSTM1

pS6K-1

S6K-1

β-actin

d

24 48 720.0

0.5

1.0

1.5

2.0

96Time (Hr)

CtrlRapStrv

Cel

l gro

wth

e

Figure 5. GSK-3 regulates mTORC1 and lysosomes via distinct pathways and both act to regulate cell growth. (a) GSK-3 activates mTORC1independent of lysosomes: MCF-7 cells were transiently transfected with GFP-GSK-3α or GFP-GSK-3β constructs. Cells were treated or non-treated with CQ and phosphorylation and total expression levels of S6 (Ser240/244) were determined by immunoblot analyses. Levels ofendogenous total GSK-3 or GFP-GSK-3 proteins and β-actin are shown. No statistical differences were obtained in pS6/S6 in cells expressingGSK-3 that were treated or non-treated with CQ. (b) MCF-7 cells were treated with rapamycin L803-mts or combination of both reagents. Cellswere stained with LysoTracker Red (Lys) and imaged in confocal microscopy (upper panel). Lower panel shows merged image of cells at brightfield with LysoTracker Red (BF). (c) Starvation enhances lysosomal acidification: MCF-7 cells were incubated with starved medium (Strv, 0.1%fetal calf serum with no amino acids) for 4 h. Live cells were stained with LysoTracker Red (Lys) and imaged in confocal microscopy. (d) Celllysates prepared from starved cells were subjected to immunoblot analysis using indicated antibodies. (e) Cell proliferation was determined instarved cells (Strv) or cells treated with rapamycin (Rap, 50 nM). Number of cells was determined by XTT cell proliferation assay. For all panels,representative gels or images are shown of three independent experiments.


7


mTORC1 and lysosomes via independent pathways that culminateinto suppressed autophagy/autophagic turnover. This paradigm isillustrated in Figure 7.Autophagy is considered a tumor-suppressive mechanism.

Many human tumors are deficient in autophagy, and insufficientautophagy promotes tumorigenesis in mouse models.75–77

Autophagy constrains cancer development by inhibiting pro-oncogenic pathways such as oxidative stress, genome instabilityand improper protein clearance.37,38,44,45,78 Our results are thus inagreement with the paradigm that GSK-3 promotes cell growth bysuppressing autophagy. Furthermore, GSK-3-mediated elevationof p62/SQSTM11 supports this view. p62/SQSTM11 acts as anoncoprotein by activating oxidative stress, preventing repair ofgenome damage and by enhancing Ras-induced celltransformation.56,57 It should be noted that the role of autophagyis complex and may have opposing effects: autophagy isprotective in early stages of pathogenesis but appears to supportmalignancies with increased metabolic demand in advancedstages of cancer progression.41–45 It is noteworthy that MCF-7 cellsrepresent an early stage of cancer (as compared with other breastcancer cell lines), a feature that is well correlated with thesuppressed autophagy observed in these cells (Figure 3). Hence,

GSK-3 inhibitors may be particularly useful in treating breastcancers at early stage. However, the ‘pro-cancer’ or ‘anticancer’activity of GSK-3 inhibitors as a function of stage and tumoraggressiveness should be taken into consideration.Previous studies have linked GSK-3 with autophagy. Impaired

autophagy was found in hearts of GSK-3α knockout mice,79 andinhibition of GSK-3 blocked serum deprivation-induced autophagyvia activation of the acetyltransferase TIP60.80 It is possible thatunder serum starvation, conditions in which mTORC1 activity issuppressed, GSK-3 inhibition has no further effect. Alternatively,under certain conditions, TIP60 may not be required for activatedautophagy as argued previously.79 An additional study showedthat GSK-3 target, β-catenin, is a negative regulator of autophagyby regulating TCF4.81 Accordingly, GSK-3 may enhance autophagyby its ability to reduce β-catenin levels. However, in our cellsystem, GSK-3 suppressed autophagy suggesting that combinedactivity of GSK-3 toward its multiple targets likely dictates the ‘net’effect on autophagy.The role of GSK-3 in regulation of lysosome function is an

important finding of this work. Of particular significance is theability of GSK-3 inhibitors to restore lysosomal acidification inMCF-7 cells that show low levels of lysosomal acidification.

LeupNT CQ

EGFR

β-actin

MCF10A

MCF-7

EGFR

pEGFR

pERK1/2

ERK

EGF - +

β-actin

pERK1/2

ERK

Ctrl GSK3α GSK3β

EGFR

EGF 0 10 60 240 0 10 60 240 0 10 60 240

MCF-10A (min)

EGF

+L80

3mts

MCF-7

0 5’ 240’60’

10μm

10μm

(min)

NT

EGFR

pERK1/2

ERK

β-actin

pEGFR

MCF-7

**

EGFR

pERK2

L803 - +mts

L803 - +

L803 -

1.0

0.5

0.0

1.0

0.5

0.0

+

**

mts

mts

NT L803mts L803mts+CQ CQ

EGF 240’

GFP

BF

MCF-7

Fold

of

NT

Fold

of

NT

GSK-3GFP

GSK-3α/β

Figure 6. GSK-3 inhibits EGFR endocytic clearance. (a) Lysosomes regulate EGFR levels: MCF-10A cells were treated with 30 μM CQ or 40 μMleupeptin (Leup) for 4 h. EGFR levels were determined by western blot analysis. (b) GSK-3 attenuates downregulation of EGFR: MCF-10A cellswere transiently transfected with GFP-GSK-3α or GFP-GSK-3β constructs. Control cells (Ctrl) were transfected with a GFP-coding construct.Cells were treated with 50 ng/ml EGF, and EGFR levels or levels of phosphorylated ERK1/2 (Thr202/Tyr204) were determined at each time pointby western blot analysis. β-Actin levels were analyzed as a control equal loading. (c) Expression of EGFR in MCF-7 cells: cells were transientlytransfected with EGFR and treated with EGF for 10min. Increased tyrosine phosphorylation of EGFR and phosphorylated levels of ERK1/2relative to control cells indicated that EGFR was functional in these cells. (d) GSK-3 inhibitor reduces EGFR levels. Cells were treated as in c,except that cells were treated with L803-mts (40 μM) for 4 h. Levels of EGFR and total and phosphorylated ERK1/2 were determined. (e) GSK-3inhibitor accelerates EGFR endocytic clearance: MCF-7 cells were transiently transfected with of GFP-EGFR expression plasmid. Cells weretreated or not with 40 μM L803-mts and then treated with 50 ng/ml EGF for indicated times. Cells were fixed with paraformaldehyde, and GFP-EGFR signal was imaged by confocal microscopy. (f) As in e, except that cells were treated with L803-mts in the presence of CQ and imaged4 h after treatment with EGF. For all panels, representative gels or images are shown, and densitometry analyses bar graphs of indicatedproteins are means of three to five independent experiments ± s.e.m., **Po0.01.


8


Deficiency in lysosomal acidification has been described in variouspathological disorders,59,60 although its role in cancer has notbeen clear.82 A previous study indicated reduced lysosomalacidification in cancer cells,83 and a recent study using proteomictools correlated reduced lysosomal acidification in breast cancercells with the degree of cell transformation and aggressiveness.84

These studies thus support our notions that, by ‘correcting’lysosomal acidification, GSK-3 inhibitors restore cellular home-ostasis and reverse, at least in part, uncontrolled cell proliferation.This was further indicated in our experiments with starved cellsthat showed enhanced lysosomal acidification and cells’ growthinhibition that exceeded the growth inhibition achieved withrapamycin (Figure 5).GSK-3 inhibitors improved lysosomal-mediated endocytic clear-

ance (Figure 6). Disruption of the endocytic machinery is observedin many types of cancers.63,85 Impaired deactivation of receptortyrosine kinases, such as EGFR, is a well-recognized cause ofneoplastic growth.85–88 Receptor trafficking from early endosomesto lysosomes is critical for downregulation of the EGFR mitogenicsignal. Hence, poor lysosomal activity resulting from defects inlysosomal acidification or maturation had an inhibitory impact onendocytic clearance. Here we showed that inhibition of GSK-3facilitated lysosome-mediated EGFR endocytic clearance. It isnoteworthy that increased expression of EGFR is detected in 16%of breast cancer tumors,89 and enhancement of EGFR clearance isconsidered a promising therapeutic approach in treating breastcancer as well as other cancer types.88 Furthermore, a combina-tion of EGFR and mTORC1 inhibitors has been successfully used asan anticancer treatment.90,91 Hence, GSK-3 inhibition may be auseful approach in cancers associated with elevated mTORC1activity and limited lysosomal acidification.

MATERIALS AND METHODSMaterialsAR-A014418, SB-216763 and CT99012 were purchased from Sigma(Rehovot, Israel). L803-mts peptide was synthesized by GenemedSynthesis, Inc. (San Antonio, TX, USA). Antibodies against phopsho-S6K-1(Thr389), S6K-1, phospho-S6 (Ser240/244), S6 protein, phospho-ERK1/2(Thr202, Tyr204) and LC3-I/II were from Cell Signaling Technologies (Beverly,MA, USA). Antibodies against GSK-3α/β, β-actin, cathepsin D, Lamp2 andEGFR were from Santa Cruz Biotechnology (Dallas, TX, USA). Anti-p62/SQSTM1 was from Medical and Biological Laboratories (Nagoya, Japan). Allother reagents were from Sigma.

PlasmidsGFP- or Cherry-tagged GSK-3α and GSK-3β in which GFP or Cherry werefused to the C-terminus of the enzymes were generated as describedpreviously.92 RFP-LC3-II was a kind gift from Dr Ehud Cohen (HebrewUniversity of Jerusalem). GFP-p62SQSTP was obtained from Dr ZvulunElazar from the Weizmann institute of Science, Rehovot. Human EGFRplasmids were a kind gift from Dr Ronit Pinkas-Kamaraski at Tel AvivUniversity and Dr Yosef Yarden at the Weizmann institute of Science,Rehovot.

Cell culture and transfectionsMCF-7 cells were maintained in Dulbecco's modified Eagle's medium (DMEM)supplemented with 10% fetal calf serum, 5mg/ml glutamine and 1%penicillin and streptomycin. MCF-10A cells were grown in DMEM-F12medium (1:1) supplemented with 5% fetal calf serum, 5mg/ml L-glutamine,0.02 μg/ml EGF, 10 μg/ml insulin, 0.5 μg/ml hydrocortisol, 0.1 g/ml choler-atoxin and 1% penicillin and streptomycin. Mouse embryonic fibroblast cellsdeficient in TSC1/2 were generously provided by Dr Kwiatkowski (HarvardMedical School) and were maintained in DMEM supplemented with 10%fetal calf serum as described for MCF-7 cells. MCF-7 or MCF-10A cells weretransiently transfected with the indicated constructs (3–7 μg) usingLipofectamine 2000 (Invitrogen, Camarillo, CA, USA) for MCF-7 cells andusing X-tremeGENE HP DNA Transfection Reagent (Roche, Basel, Switzerland)for MCF-10A cells. For experiments in which GSK-3 expression was silenced,cells were transfected with 50 nM GSK-3α or GSK-3β siRNA or with a negativecontrol siRNA (Thermo Scientific/Dharmacon, Waltham, MA, USA) using thetransfection reagent Dharmafect (Thermo Scientific/Dharmacon) accordingto the manufacturer's instructions. In some experiments, cells were treatedwith GSK-3 inhibitors 40 μM L803-mts, 20 μM AR-A014418, 10 μM SB-216763,30 μM chloroquine or 50 nM rapamycin for 4 h or as indicated. To monitor cellgrowth, exponential growth phase was harvested and seeded at 20 000 cellsper six-well plates. Cells were incubated with GSK-3 inhibitors as indicated.Fresh medium and inhibitors were added every 24 h to the cell cultures. Cellswere trypsinized and counted in Cellometer Auto T4 (Nexcelom Biosciences,Lawrence, MA, USA).

Gel electrophoresis and immunoblottingCells were collected and lysed in an ice-cold buffer G (20mM Tris-HCl, 10%glycerol, 1 mM EDTA, 1 mM EGTA, 0.5% Triton X-100, 0.5 mM orthovanadate,10mM β-glycerophosphate, 5 mM sodium pyrophosphate, 50 mM NaF, 1 mM

benzaminidine and protease inhibitors aprotenin, leupeptin and pepstatinA). Cell extracts were centrifuged at 14 000 g for 20min and supernatantswere collected. Protein concentrations were determined by Bradfordanalysis, and equal amounts of protein (20 μg) were subjected to gelelectrophoresis and western blot analysis using indicated antibodies.Analysis of β-actin levels demonstrated equal protein loading.

Live-cell imagingTo stain acidified lysosomes, cells grown on coverslips were incubated with50 nM LysoTracker Red (Molecular Probes, Eugene, OR, USA) for 15min at37 °C. The cells were washed with growing medium and were immediatelytaken for live-cell microscopy. For other experiments, the cells were growthon coverslips and treated as indicated. Live-cell images were taken using a63.0 × 1.40 OIL UV objective lens on a laser scanning confocal microscope(Leica, Solms, Germany, TCS-SP5 II) with spatial resolution of 50–70 nM.Images were generated using LAS-AF Lite software (Leica microsystems).

Immunofluorescent analysesCells were grown on coverslips and fixed with 4% paraformaldehyde for15min at room temperature. After fixation, cells were permabilized (0.2%Triton X-100), blocked with 3% bovine serum albumin and incubated withindicated antibodies (0.1% bovine serum albumin, 0.05% Triton X-100 inphosphate-buffered saline) overnight at 4 °C. Cells were then washed threetimes in phosphate-buffered saline, and secondary Alexa488-conjugatedantibody (Invitrogen) was applied for 1 h at room temperature. Confocalimages were acquired on the Leica TCS-SP5 II confocal microscope.

Statistical analysesAll experiments were repeated at least three times. The data are expressedas means ± s.d. (s.e.m.). Statistical analysis was performed using Student'st-test (two-tailed). The criterion for statistical significance was Po0.05.

Beclin1

Endocyticdegradation

lysosomes

p62/SQSTM

LC3IImTORC1

GSK-3

AutophagyAutophagic flux

Figure 7. GSK-3 suppresses autophagy via mTORC1 and lysosomes.GSK-3 activation of mTORC1 results in suppressed autophagy asindicated by reduced beclin-1 and elevated levels of p62/SQSTM1(marked by dashed lines). Note that LC3-II levels are decreased bymTORC1. GSK-3 reduces lysosomal acidification that in turn inhibitsautophagic flux (lysosomes enhance autophagic activity and flux).This is pronounced by elevation in LC3-II and p62/SQSTM1 (markedby dashed lines). Reduced lysosomal acidification may inhibitendocytic clearance as was demonstrated by reduced levels ofEGFR. Arrows indicate activation, blocked lines indicate inhibition.


9


CONFLICT OF INTERESTThe authors declare no conflict of interest.

ACKNOWLEDGEMENTSThis research was supported by the Israel Science Foundation grant No. 341/10 andthe Finghort Fund for Cancer Research at Tel Aviv University. We thank Dr DanKlionsky and Dr Zvulun Elazar for their constructive comments.

REFERENCES1 Grimes CA, Jope RS. The multifaceted roles of glycogen synthase kinase 3beta in

cellular signaling. Prog Neurobiol 2001; 65: 391–426.2 Eldar-Finkelman H. Glycogen synthase kinase-3: an emerging therapeutic target.

Trend Mol Med 2002; 8: 126–132.3 Doble BW, Woodgett JR. GSK-3: tricks of the trade for a multi-tasking kinase. J Cell

Sci 2003; 116: 1175–1186.4 Hur EM, Zhou FQ. GSK3 signalling in neural development. Nat Rev Neurosci 2010;

11: 539–551.5 Jope RS, Yuskaitis CJ, Beurel E. Glycogen synthase kinase-3 (GSK3): inflammation,

diseases, and therapeutics. Neurochem Res 2007; 32: 577–595.6 Woodgett JR. Molecular cloning and expression of glycogen synthase kinase-3/

factorA. EMBO J 1990; 9: 2431–2438.7 Eldar-Finkelman H, Martinez A. GSK-3 inhibitors: preclinical and clinical focus on

CNS. Front Mol Neurosci 2011; 4: 32.8 Hernandez F, Lucas JJ, Avila J. GSK3 and tau: two convergence points in Alzhei-

mer's disease. J Alzheimers Dis 2013; 33: S141–S144.9 Ougolkov AV, Billadeau DD. Targeting GSK-3: a promising approach for cancer

therapy? Future Oncol 2006; 2: 91–100.10 Patel S, Woodgett J. Glycogen synthase kinase-3 and cancer: good cop, bad cop?

Cancer Cell 2008; 14: 351–353.11 Birch NW, Zeleznik-Le NJ. Glycogen synthase kinase-3 and leukemia: restoring the

balance. Cancer Cell 2010; 17: 529–531.12 McCubrey JA, Davis NM, Abrams SL, Montalto G, Cervello M, Basecke J et al.

Diverse roles of GSK-3: tumor promoter-tumor suppressor, target in cancertherapy. Adv Biol Regul 2014; 54: 176–196.

13 Manoukian AS, Woodgett JR. Role of glycogen synthase kinase-3 in cancer:regulation by Wnts and other signaling pathways. Adv Cancer Res 2002; 84:203–229.

14 Wu D, Pan W. GSK3: a multifaceted kinase in Wnt signaling. Trends Biochem Sci2011; 35: 161–168.

15 Polakis P. The many ways of Wnt in cancer. Curr Opin Genet Dev 2007; 17: 45–51.16 Kotliarova S, Pastorino S, Kovell LC, Kotliarov Y, Song H, Zhang W et al. Glycogen

synthase kinase-3 inhibition induces glioma cell death through c-MYC, nuclearfactor-kappaB, and glucose regulation. Cancer Res 2008; 68: 6643–6651.

17 Zhu Q, Yang J, Han S, Liu J, Holzbeierlein J, Thrasher JB et al. Suppression ofglycogen synthase kinase 3 activity reduces tumor growth of prostate cancerin vivo. Prostate 2011; 71: 835–845.

18 Aguilar-Morante D, Morales-Garcia JA, Sanz-SanCristobal M, Garcia-Cabezas MA,Santos A, Perez-Castillo A. Inhibition of glioblastoma growth by the thiadiazoli-dinone compound TDZD-8. PLoS ONE 2010; 5: e13879.

19 Remsing Rix LL, Kuenzi BM, Luo Y, Remily-Wood E, Kinose F, Wright G et al. GSK3alpha and beta are new functionally relevant targets of tivantinib in lungcancer cells. ACS Chem Biol 2013; 9: 353–358.

20 Ougolkov AV, Fernandez-Zapico ME, Savoy DN, Urrutia RA, Billadeau DD. Glyco-gen synthase kinase-3beta participates in nuclear factor kappaB-mediated genetranscription and cell survival in pancreatic cancer cells. Cancer Res 2005; 65:2076–2081.

21 Hoeflich KP, Luo J, Rubie EA, Tsao MS, Jin O, Woodgett JR. Requirement forglycogen synthase kinase-3beta in cell survival and NF-kappaB activation. Nature2000; 406: 86–90.

22 Fishman P, Bar-Yehuda S, Ohana G, Barer F, Ochaion A, Erlanger A et al.An agonist to the A3 adenosine receptor inhibits colon carcinoma growth inmice via modulation of GSK-3 beta and NF-kappa B. Oncogene 2004; 23:2465–2471.

23 Pluquet O, Qu LK, Baltzis D, Koromilas AE. Endoplasmic reticulum stressaccelerates p53 degradation by the cooperative actions of Hdm2 and glycogensynthase kinase 3beta. Mol Cell Biol 2005; 25: 9392–9405.

24 Kim HM, Kim CS, Lee JH, Jang SJ, Hwang JJ, Ro S et al. CG0009, a novel glycogensynthase kinase 3 inhibitor, induces cell death through cyclin D1 depletion inbreast cancer cells. PLoS ONE 2013; 8: e60383.

25 Karyo R, Eskira Y, Pinhasov A, Belmaker R, Agam G, Eldar-Finkelman H.Identification of eukaryotic elongation factor-2 as a novel cellular target

of lithium and glycogen synthase kinase-3. Mol Cell Neurosci 2010; 45:449–455.

26 Avrahami L, Farfara D, Shaham-Kol M, Vassar R, Frenkel D, Eldar-Finkelman H.Inhibition of GSK-3 ameliorates beta-amyloid(A-beta) pathology and restoreslysosomal acidification and mTOR activity in the alzheimer's disease mousemodel: in vivo and in vitro studies. J Biol Chem 2013; 288: 1295–1306.

27 Shin S, Wolgamott L, Yu Y, Blenis J, Yoon SO. Glycogen synthase kinase (GSK)-3promotes p70 ribosomal protein S6 kinase (p70S6K) activity and cell proliferation.Proc Natl Acad Sci USA 2011; 108: E1204–E1213.

28 Shin S, Wolgamott L, Tcherkezian J, Vallabhapurapu S, Yu Y, Roux PP et al.Glycogen synthase kinase-3beta positively regulates protein synthesis and cellproliferation through the regulation of translation initiation factor 4E-bindingprotein 1. Oncogene 2013; 33: 1690–1699.

29 Laplante M, Sabatini DM. mTOR signaling in growth control and disease. Cell 2012;149: 274–293.

30 Cornu M, Albert V, Hall MN. mTOR in aging, metabolism, and cancer. Curr OpinGenet Dev 2013; 23: 53–62.

31 Proud CG. mTOR Signalling in Health and Disease. Biochem Soc Trans 2013; 39:431–436.

32 Kwiatkowski DJ, Manning BD. Tuberous sclerosis: a GAP at the crossroads ofmultiple signaling pathways. Hum Mol Genet 2005, 14 Spec No. 2: R251–R258.

33 Mamane Y, Petroulakis E, LeBacquer O, Sonenberg N. mTOR translation initiationand cancer. Oncogene 2006; 25: 6416–6422.

34 Wander SA, Hennessy BT, Slingerland JM. Next-generation mTOR inhibitors inclinical oncology: how pathway complexity informs therapeutic strategy. J ClinInvest 2011; 121: 1231–1241.

35 Chiang GG, Abraham RT. Targeting the mTOR signaling network in cancer. TrendsMol Med 2007; 13: 433–442.

36 Voss MH, Molina AM, Motzer RJ. mTOR inhibitors in advanced renal cellcarcinoma. Hematol Oncol Clin North Am 2011; 25: 835–852.

37 Choi AM, Ryter SW, Levine B. Autophagy in human health and disease. N. Engl JMed 2013; 368: 1845–1846.

38 Ravikumar B, Sarkar S, Davies JE, Futter M, Garcia-Arencibia M, Green-Thompson ZWet al. Regulation of mammalian autophagy in physiology and pathophysiology.Physiol Rev 2010; 90: 1383–1435.

39 Klionsky D. An overview of autophagy: Morphology, mechanism and regulation.Antioxid Redox Signal 2013; 20: 460–473.

40 Schneider JL, Cuervo AM. Autophagy and human disease: emerging themes. CurrOpin Genet Dev 2014; 26C: 16–23.

41 Kimmelman AC. The dynamic nature of autophagy in cancer. Genes Dev 2011; 25:1999–2010.

42 Chen N, Karantza V. Autophagy as a therapeutic target in cancer. Cancer Biol Ther2011; 11: 157–168.

43 Mathew R, White E. Autophagy in tumorigenesis and energy metabolism: friendby day, foe by night. Curr Opin Genet Dev 2011; 21: 113–119.

44 Eisenberg-Lerner A, Kimchi A. The paradox of autophagy and its implication incancer etiology and therapy. Apoptosis 2009; 14: 376–391.

45 Wu WK, Coffelt SB, Cho CH, Wang XJ, Lee CW, Chan FK et al. The autophagicparadox in cancer therapy. Oncogene 2011; 31: 939–953.

46 Mizushima N, Levine B, Cuervo AM, Klionsky DJ. Autophagy fights diseasethrough cellular self-digestion. Nature 2008; 451: 1069–1075.

47 Inoki K, Zhu T, Guan KL. TSC2 mediates cellular energy response to control cellgrowth and survival. Cell 2003; 115: 577–590.

48 Sancak Y, Bar-Peled L, Zoncu R, Markhard AL, Nada S, Sabatini DM. Ragulator-Ragcomplex targets mTORC1 to the lysosomal surface and is necessary for itsactivation by amino acids. Cell 2010; 141: 290–303.

49 Brugarolas J, Lei K, Hurley RL, Manning BD, Reiling JH, Hafen E et al. Regulation ofmTOR function in response to hypoxia by REDD1 and the TSC1/TSC2 tumorsuppressor complex. Genes Dev 2004; 18: 2893–2904.

50 Smith EM, Finn SG, Tee AR, Browne GJ, Proud CG. The tuberous sclerosis proteinTSC2 is not required for the regulation of the mammalian target of rapamycin byamino acids and certain cellular stresses. J Biol Chem 2005; 280: 18717–18727.

51 Zhang H, Cicchetti G, Onda H, Koon HB, Asrican K, Bajraszewski N et al. Lossof Tsc1/Tsc2 activates mTOR and disrupts PI3K-Akt signaling throughdownregulation of PDGFR. J Clin Invest 2003; 112: 1223–1233.

52 Mizushima N, Yoshimori T. How to interpret LC3 immunoblotting. Autophagy2007; 3: 542–545.

53 Kabeya Y, Mizushima N, Ueno T, Yamamoto A, Kirisako T, Noda T et al. LC3, amammalian homologue of yeast Apg8p, is localized in autophagosome mem-branes after processing. EMBO J 2000; 19: 5720–5728.

54 Bjorkoy G, Lamark T, Brech A, Outzen H, Perander M, Overvatn A et al.p62/SQSTM1 forms protein aggregates degraded by autophagy and has aprotective effect on huntingtin-induced cell death. J Cell Biol 2005; 171: 603–614.

55 He C, Levine B. The Beclin 1 interactome. Curr Opin Cell Biol 2012; 22: 140–149.


10


56 Duran A, Linares JF, Galvez AS, Wikenheiser K, Flores JM, Diaz-Meco MT et al. Thesignaling adaptor p62 is an important NF-kappaB mediator in tumorigenesis.Cancer Cell 2008; 13: 343–354.

57 Mathew R, Karp CM, Beaudoin B, Vuong N, Chen G, Chen HY et al.Autophagy suppresses tumorigenesis through elimination of p62. Cell 2009; 137:1062–1075.

58 Saftig P, Klumperman J. Lysosome biogenesis and lysosomal membrane proteins:trafficking meets function. Nat Rev Mol Cell Biol 2009; 10: 623–635.

59 Luzio JP, Pryor PR, Bright NA. Lysosomes: fusion and function. Nat Rev Mol Cell Biol2007; 8: 622–632.

60 Boya P. Lysosomal function and dysfunction: mechanism and disease. AntioxidRedox Signal 2012; 17: 766–774.

61 Korolchuk VI, Saiki S, Lichtenberg M, Siddiqi FH, Roberts EA, Imarisio S et al.Lysosomal positioning coordinates cellular nutrient responses. Nat Cell Biol 2011;13: 453–460.

62 Zoncu R, Bar-Peled L, Efeyan A, Wang S, Sancak Y, Sabatini DM. mTORC1 senseslysosomal amino acids through an inside-out mechanism that requires thevacuolar H(+)-ATPase. Science 2011; 334: 678–683.

63 Haglund K, Dikic I. The role of ubiquitylation in receptor endocytosis andendosomal sorting. J Cell Sci 2012; 125: 265–275.

64 Sorkin A, Goh LK. Endocytosis and intracellular trafficking of ErbBs. Exp Cell Res2009; 315: 683–696.

65 Avraham R, Yarden Y. Feedback regulation of EGFR signalling: decision making byearly and delayed loops. Nat Rev Mol Cell Biol 2011; 12: 104–117.

66 Zoncu R, Efeyan A, Sabatini DM. mTOR: from growth signal integration to cancer,diabetes and ageing. Nat Rev Mol Cell Biol 2010; 12: 21–35.

67 Harris TE, Lawrence JC Jr. TOR signaling. Sci STKE 2003; 2003: re15.68 Wullschleger S, Loewith R, Hall MN. TOR signaling in growth and metabolism.

Cell 2006; 124: 471–484.69 Wazir U, Wazir A, Khanzada ZS, Jiang WG, Sharma AK, Mokbel K. Current State of

mTOR Targeting in Human Breast Cancer. Cancer Genomics Proteomics 2014; 11:167–174.

70 Lauring J, Park BH, Wolff AC. The phosphoinositide-3-kinase-Akt-mTOR pathwayas a therapeutic target in breast cancer. J Natl Compr Canc Netw 2013; 11:670–678.

71 Inoki K, Ouyang H, Zhu T, Lindvall C, Wang Y, Zhang X et al. TSC2 integrates Wntand energy signals via a coordinated phosphorylation by AMPK and GSK3 toregulate cell growth. Cell 2006; 126: 955–968.

72 Gulen MF, Bulek K, Xiao H, Yu M, Gao J, Sun L et al. Inactivation of the enzymeGSK3alpha by the kinase IKKi promotes AKT-mTOR signaling pathway thatmediates interleukin-1-induced Th17 cell maintenance. Immunity 2012; 37:800–812.

73 Huang J, Zhang Y, Bersenev A, O'Brien WT, Tong W, Emerson SG et al. Pivotal rolefor glycogen synthase kinase-3 in hematopoietic stem cell homeostasis in mice.J Clin Invest 2009; 119: 3519–3529.

74 Suzuki T, Bridges D, Nakada D, Skiniotis G, Morrison SJ, Lin JD et al. Inhibition ofAMPK catabolic action by GSK3. Mol Cell 2013; 50: 407–419.

75 Mathew R, Kongara S, Beaudoin B, Karp CM, Bray K, Degenhardt K et al. Autop-hagy suppresses tumor progression by limiting chromosomal instability. GenesDev 2007; 21: 1367–1381.

76 Liang XH, Jackson S, Seaman M, Brown K, Kempkes B, Hibshoosh H et al. Inductionof autophagy and inhibition of tumorigenesis by beclin 1. Nature 1999; 402:672–676.

77 Takamura A, Komatsu M, Hara T, Sakamoto A, Kishi C, Waguri S et al. Autophagy-deficient mice develop multiple liver tumors. Genes Dev 2011; 25: 795–800.

78 Yang Z, Klionsky DJ. Mammalian autophagy: core molecular machinery and sig-naling regulation. Curr Opin Cell Biol 2010; 22: 124–131.

79 Zhou J, Freeman TA, Ahmad F, Shang X, Mangano E, Gao E et al. GSK-3alpha is acentral regulator of age-related pathologies in mice. J Clin Invest 2013; 123:1821–1832.

80 Lin SY, Li TY, Liu Q, Zhang C, Li X, Chen Y et al. GSK3-TIP60-ULK1 signaling pathwaylinks growth factor deprivation to autophagy. Science 2012; 336: 477–481.

81 Petherick KJ, Williams AC, Lane JD, Ordonez-Moran P, Huelsken J, Collard TJ et al.Autolysosomal beta-catenin degradation regulates Wnt-autophagy-p62 crosstalk.EMBO J 2013; 32: 1903–1916.

82 Kallunki T, Olsen OD, Jaattela M. Cancer-associated lysosomal changes: friendsor foes? Oncogene 2013; 32: 1995–2004.

83 Altan N, Chen Y, Schindler M, Simon SM. Defective acidification in human breasttumor cells and implications for chemotherapy. J Exp Med 1998; 187: 1583–1598.

84 Harris RA, Yang A, Stein RC, Lucy K, Brusten L, Herath A et al. Cluster analysis of anextensive human breast cancer cell line protein expression map database.Proteomics 2002; 2: 212–223.

85 Bache KG, Slagsvold T, Stenmark H. Defective downregulation of receptor tyrosinekinases in cancer. EMBO J 2004; 23: 2707–2712.

86 Yarden Y. The EGFR family and its ligands in human cancer. signalling mechan-isms and therapeutic opportunities. Eur J Cancer 2001; 37: S3–S8.

87 Arteaga CL. ErbB-targeted therapeutic approaches in human cancer. Exp Cell Res2003; 284: 122–130.

88 Davis NM, Sokolosky M, Stadelman K, Abrams SL, Libra M, Candido S et al.Deregulation of the EGFR/PI3K/PTEN/Akt/mTORC1 pathway in breast cancer:possibilities for therapeutic intervention. Oncotarget 2014; 5: 4603–4650.

89 Witton CJ, Reeves JR, Going JJ, Cooke TG, Bartlett JM. Expression of the HER1-4family of receptor tyrosine kinases in breast cancer. J Pathol 2003; 200: 290–297.

90 Doherty L, Gigas DC, Kesari S, Drappatz J, Kim R, Zimmerman J et al. Pilot study ofthe combination of EGFR and mTOR inhibitors in recurrent malignant gliomas.Neurology 2006; 67: 156–158.

91 Fan QW, Cheng CK, Nicolaides TP, Hackett CS, Knight ZA, Shokat KM et al. A dualphosphoinositide-3-kinase alpha/mTOR inhibitor cooperates with blockade ofepidermal growth factor receptor in PTEN-mutant glioma. Cancer Res 2007; 67:7960–7965.

92 Azoulay-Alfaguter I, Yaffe Y, Licht-Murava A, Urbanska M, Jaworski J, PietrokovskiS et al. Distinct molecular regulation of glycogen synthase kinase-3alpha isozymecontrolled by its N-terminal region: functional role in calcium/calpain signaling.J Biol Chem 2011; 286: 13470–13480.


11


Documents

Combined regulation of mTORC1 and lysosomal acidification ... · rapamycin (mTOR) as a GSK-3 target.25–28 mTOR is a nutrient sensor that controls protein synthesis, cell size and