SDF-1/CXCL12 induces directional cell migration and spontaneous metastasis via a CXCR4/Gαi/mTORC1 axis

The FASEB Journal • Research Communication

SDF-1/CXCL12 induces directional cellmigration and spontaneous metastasis viaa CXCR4/Gai/mTORC1 axis

Patricia Dillenburg-Pilla,* Vyomesh Patel,* Constantinos M. Mikelis,*Carlos Rodrigo Zarate-Blades,† Colleen L. Doçi,* Panomwat Amornphimoltham,*Zhiyong Wang,* Daniel Martin,* Kantima Leelahavanichkul,* Robert T. Dorsam,*Andrius Masedunskas,* Roberto Weigert,* Alfredo A. Molinolo,* and J. Silvio Gutkind*,1

*Oral and Pharyngeal Cancer Branch, National Institute of Dental and Craniofacial Research, and†Laboratory of Immunology, National Eye Institute, National Institutes of Health, Bethesda,Maryland, USA

ABSTRACT Multiple humanmalignancies rely onC-X-Cmotif chemokine receptor type 4 (CXCR4) and its ligand,SDF-1/CXCL12 (stroma cell–derived factor 1/C-X-Cmotifchemokine 12), to metastasize. CXCR4 inhibitors promotethe mobilization of bone marrow stem cells, limiting theirclinical application for metastasis prevention. We investi-gated the CXCR4-initiated signaling circuitry to identifynew potential therapeutic targets. We used HeLa humancancer cells expressing high levels of CXCR4 endoge-nously.We found thatCXCL12promotes theirmigration inBoyden chamber assays and single cell tracking. CXCL12activated mTOR (mechanistic target of rapamycin) po-tently in a pertussis-sensitive fashion. Inhibition of mTORcomplex 1 (mTORC1) by rapamycin [drug concentrationcausing 50% inhibition (IC50) = 5 nM] and mTORC1/mTORC2byTorin2 (IC50 =6nM),orbyknockingdownkeymTORC1/2 components, Raptor and Rictor, respectively,decreased directional cell migration toward CXCL12. Wedeveloped a CXCR4-mediated spontaneous metastasismodel by implanting HeLa cells in the tongue of SCID-NOD mice, in which 80% of the animals develop lymphnode metastasis. It is surprising that mTORC1 disruptionby Raptor knockdown was sufficient to reduce tumorgrowth by 60% and spontaneous metastasis by 72%, whichwere nearly abolished by rapamycin. In contrast, disruptingmTORC2 had no effect in tumor growth or metastasiscompared with control short hairpin RNAs. These datasuggest thatmTORC1may represent a suitable therapeutictarget in human malignancies using CXCR4 for their meta-static spread.—Dillenburg-Pilla, P., Patel, V, Mikelis, C. M.,Zarate-Blades, C. R., Doçi, C. L., Amornphimoltham, P.,Wang, Z., Martin, D., Leelahavanichkul, K., Dorsam, R. T.,Masedunskas, A., Weigert, R., Molinolo, A. A, and Gutkind,J. S. SDF-1/CXCL12 induces directional cell migration and

spontaneous metastasis via a CXCR4/Gai/mTORC1 axis.FASEB J. 29, 000–000 (2015). www.fasebj.org

Key Words: mTOR • rapamycin • cancer • lymphangiogenesis •

chemotaxis

CHEMOKINES ARE MASTER regulators of cell migration (1).Currently, there are 48 knownmolecules that belong to thiscytokine superfamily, which are divided in 2 groupsdepending on their function: inflammatory and homeo-static. Inflammatory chemokines, such as IL-8, C-X-Cmotifchemokine (CXCL) 1, and CXCL2, are induced duringinflammatory events, and multiple chemokines can bindand activate shared receptors (2). In contrast, homeostaticchemokines are constantly expressed and released asthey play crucial physiologic roles, such as tissue re-generation and stemcellmaintenance, which also results inamore restricted ligandusage by its receptors (3). TheC-X-C motif chemokine receptor type 4 (CXCR4) 7-trans-membrane G protein–coupled receptor binds to the ho-meostaticC-X-Cmotif chemokine12(CXCL12, alsoknownas stromal cell–derived factor 1, or SDF-1). CXCL12 is se-creted in normal, nonpathologic state bymany organs suchas liver, bone marrow, lung, and lymph nodes (2). CXCR4expression correlates with poor prognosis in many tumortypes, andmost of the sites that secrete its ligand, CXCL12,are frequently colonized by metastatic cells (4). Given thekey role that CXCR4plays in cancermetastasis, therapeutictargeting of CXCR4 has been tested in clinical trials (5, 6).Althougheffective indecreasingmetastaticdisease,CXCR4inhibitors promote the mobilization of bone marrow stemcells from their niche (5), a side effect that limits theirclinical application. In fact, CXCR4 antagonists are cur-rently approved by theU.S. Food andDrug Administrationfor hematopoietic stem and progenitor cell mobilization

Abbreviations: CXCL12, C-X-C motif chemokine 12;CXCR4, C-X-C motif chemokine receptor type 4; EGF, epi-dermal growth factor; FACS, flow cytometry; H2BGFP, histoneand GFP fusion protein; LPA, lysophosphatidic acid; mTOR,mechanistic target of rapamycin; mTORC, mTOR complex;PTX, pertussis toxin; SDF-1, stroma cell–derived factor 1;shRNA, short hairpin RNA; siRNA, small interfering RNA

1 Correspondence: National Institute of Dental and Cra-niofacial Research/National Institutes of Health, 30 ConventDr., Building 30, Room 320, Bethesda, MD 20892-4340, USA.E-mail: [email protected]: 10.1096/fj.14-260083This article includes supplemental data. Please visit http://

www.fasebj.org to obtain this information.

0892-6638/15/0029-0001 © FASEB 1

The FASEB Journal article fj.14-260083. Published online December 2, 2014.

for autologous transplantation in patients with non-Hodgkin lymphoma and multiple myeloma (6–8). To cir-cumvent this, we hypothesized that the study of the CXCR4downstream signaling circuitry may help identify targetsthat couldbeaffectedbydrugsand that couldpotentiallybeexplored as new therapeutic options for many humanmalignancies that depend on CXCR4 for their metastaticspread without provoking hematopoietic stem and pro-genitor cell mobilization.

The PI3K/Akt/mTOR (phosphoinsoitide 3-kinase/AKT/mechanistic target of rapamycin) signaling pathwayplays amajor role in regulating cell growth and survival andis one of the most frequently deregulated biochemicalroutes in cancer (9). It is noteworthy thatmany oncogenesand tumor suppressorgenesdirectlyor indirectly affect thispathway, and mutations of and genetic and epigeneticalterations in components of thePI3K/mTORpathway areamong the most prevalent alterations in human malig-nancies (10–13).

mTOR is a major effector of PI3K/Akt pathway, andCXCL12 signaling through CXCR4 has been shown toinduce Akt phosphorylation/activation (14, 15). However,mTOR exerts different functions depending on the com-plex in which it is engaged. Two accessory proteins, knownas Raptor (regulatory-associated protein of mTOR) andRictor (rapamycin-insensitive companion of mTOR), de-finemTOR complex 1 (mTORC1) andmTOR complex 2(mTORC2), respectively (16, 17). Both complexes havebeen associated with a variety of functions, including cellmetabolism, growth, proliferation, autophagy and proteinsynthesis for mTORC1, and survival, metabolism, and cy-toskeletal organization for mTORC2 (9, 18). The effec-tiveness of inhibitors of the mTOR pathway for cancertreatment is under current evaluation in multiple clinicalsettings, and some mTOR inhibitors have already beenapproved for clinical use (19–23).Nonetheless, the preciseroleofmTORand its complexes ineach tumor type is yet tobe fully understood, as is the possibility that mTOR maycontribute to cancer metastasis.

Although theabilityofCXCL12 to stimulatemTORiswellestablished, the functional contribution ofmTOR signalingto CXCR4-mediated migration and metastasis is poorly un-derstood. The latter may be of direct cancer relevance, asmTOR blockade is not known to cause bone marrow stemcell mobilization (24, 25), whereas it has been reported ingastric carcinomas that mTOR is required for CXCL12-mediatedmigration in vitro (26). By use of cells that expressCXCR4 endogenously, we show that CXCR4/Gai activatesmTORC1 andmTORC2, and that this activation is requiredfor chemotaxis. Moreover, we took advantage of a novel invivo system to monitor CXCR4-mediated spontaneous me-tastasis to the lymph nodes to investigate whether mTORrepresents a suitable antimetastatic target. It is surprisingthat we found that, although the 2 mTOR complexesplay a role in CXCR4-mediated migration in vitro, onlymTORC1 disruption decreases tumor growth and theability of tumor cells to spontaneouslymetastasize to lymphnodes. This suggests that rapamycin and its analogs, whichinhibit primarily mTORC1, may represent promising tar-geted agents preventing metastasis of many highly aggres-sive cancers that use CXCR4 for the guided migration ofcancer cells from their primary tumors to their secondarycolonization sites.

MATERIALS AND METHODS

Reagents

All chemical and reagents were purchased from Sigma-Aldrich(Woodlands, TX, USA) and all antibodies were purchased fromCell Signaling Technology (Beverly, MA, USA) unless otherwisestated. mTOR inhibitors rapamycin and Torin2 were purchasedfromLCLaboratories (Woburn,MA,USA)andTocrisBioscience(Ellisville, MO, USA), respectively. CXCL12, epidermal growthfactor (EGF), and lysophosphatidic acid (LPA) were purchasedfrom R&D Systems (Minneapolis, MN, USA).

Cell culture, transfection, and lentivirus infection

HeLa cells were cultured inDMEM supplemented with 10% fetalbovine serum at 37°C in 95% air/5%CO2 (Invitrogen, Carslbad,CA, USA). Small interfering RNA (siRNA) transfection was per-formed using Lipofectamine RNAiMAX reagent and 50 nM ofSMARTpool siRNA for Raptor or Rictor (Thermo Fisher Scien-tific,Woburn,MA,USA).All analyseswereperformedbetween48and 72 h after transfection. Stable knockdown of Raptor, Rictor,and CXCR4, and H2B-GFP stable cell lines were achieved byinfectingHeLacellswith lentivirus expressing therespective shorthairpin RNA (shRNA) (Open Biosystems, Huntsville, AL, USA)or H2B-GFP (Addgene, Cambridge, MA, USA). Selection wasstarted 7 d after infection using puromycin (1 mg/ml). Experi-mentsusingknockdowncellswereperformed5 to7passages afterselection was done, always in the presence of puromycin.

Chemokine receptor expression profile

Gene expression analysis was performed by RNA sequencing.RNA was isolated from cell lines during exponential growth andthen submitted to RNA sequencing. Indexed RNA sequencinglibraries were prepared using Truseq RNA sample Prep Kit, ver-sion 2 (Illumina, San Diego, CA, USA) and sequenced in paired-end mode on a Illumina Hiseq2000 sequencer. Raw data weremapped to the human genome (hg19 version) using GSNAPsoftware in SNP-aware mode. Aligned reads were imported intothe AVADIS NGS v1.6 software for read filtering and transcriptquantification using the DESeq algorithm. Clustering and otherstatistical analysis was performed by Avadis NGS.

Immunoblot analysis

Proteins from subconfluent HeLa cells were extracted at 4°Cusing lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1% NonidetP-40) supplemented with proteases and phosphatase inhibitors(protease inhibitor cocktail, Sigma-Aldrich; 1 mM Na3VO4 and1mMNaF). Protein quantificationwas assessed usingDCproteinassay (Bio-Rad,Hercules, CA,USA) following themanufacturer’sprotocol. A total of 30 mg of protein was loaded in an SDS-PAGE,and proteins were transferred to nitrocellulose membranes. Allmembranes were blocked in milk and incubated with primaryantibody (1:1000) for 1 h at room temperature. The followingprimary antibodies were used: phospho-S6, S6, phospho-Akt 473and 308, glyceraldehyde phosphate dehydrogenase, anda-tubulin.To detect signal, we used secondary antibodies conjugated tofluorochrome: monkey anti-rabbit coupled to IRDye800CW andgoat anti-mouse coupled to IRDye 700CW (Li-Cor Bioscience,Lincoln, NE, USA). Images were acquired using Odyssey (Li-CorBioscience) and processed using Odyssey Application Software,v3 (Li-Cor Bioscience).

2 Vol. 29 March 2015 DILLENBURG-PILLA ET AL.The FASEB Journal x www.fasebj.org

Boyden chamber chemotaxis assay

Chemotaxis assaywasperformedusing a48-wellBoyden chamberwith an 8 mmpore size polyvinyl pyrrolidone–free polycarbonatemembrane (NeuroProbe, Gaithersburg, MD, USA). Cells wereserum starved, and 8 mmmembranes were coated with collagentype I (10mg/ml) at 4°C for 6 to 8h (BDBioscience, San Jose,CA,USA). Next, cells were added to the upper chamber and che-moattractant was added to the lower chamber in serum-freemedium, and cells were allowed to migrate for 16 h in humidchambers at 37°C. Membrane was then fixed, stained with he-matoxylin, and cells from the upper chamber removed witha cotton swab. Migrated cells (in the lower chamber) werecounted under a high-magnification field.

Chemotaxis assay using m-Slides

Chemotaxis assay was performed following manufacture’s in-struction. In brief, the observation area of the chemotaxischamber (IBIDI LLC, Verona, WI, USA) was coated with fibro-nectin (100mg/ml)or collagen type I (10mg/ml) for 1h followedby 1 wash with water, then let dry for 1 h. HeLa cells transfectedwith siControl, siRaptor, or siRictor 48 h before the assay wereseeded in serum-free condition and left to adhere for 3 to 4 h at37°C in a humid chamber. CXCL12 (90 ng/ml) was used asa chemoattractant and was added to the upper reservoir. Imageswere collected every hour for 24 h with a 103 objective usinga Zeiss LSM 700 confocal microscope equipped with a CO2- andtemperature-controlled chamber. Data were analyzed by theImageJ (http://rsb.info.nih.gov/ij/) plug-in Manual Tracking(http://rsbweb.nih.gov/ij/plugins/track/track.html), followed by theChemotaxis and Migration tool from IBIDI (http://www.ibidi.de/applications/ap_chemo.html).

Animal studies

All animal studies were carried out according to U.S. NationalInstitutes of Health (NIH)–approved protocols (ASP 07-442), incompliance with the NIH Guide for the Care and Use of LaboratoryAnimals. FemaleSCID-NOD(NCI, Frederick,MD,USA), 4 to6wkof age and weighing 18 to 20 g, were used in the study and werehoused in appropriate sterile filter-capped cages and permittedfood and water ad libitum. All handling and tumor implantationprocedures were conducted in a laminar-flow biosafety hood.

Establishment of spontaneous metastasis model inSCID-NOD mice

HeLa cells were cultured to 70–80% confluence, trypsinized,washed in PBS, and suspended in serum-free DMEM at 50,000cells/50ml. For implanting, animalswere anesthetizedwith 2%to5% isoflurane (Baxter Healthcare Corporation, Deerfield, IL,USA), the tonguewas exposed using tweezers, and 50ml ofmediacontaining the cells was injected submucosally into the posteriorareaof the tongue.Theanimalwas thenmonitoreduntil recovery(1 to 2 min). All animals were then assessed weekly for tumorformation and progression. All animals developed visible tumorsin the tongue after 3 wk. For the initial analysis of tumor metas-tasis, animals were euthanized at wk 5 after tumor implantation,and inguinal, abdominal, and cervical lymph nodes were re-trieved and analyzed histopathologically. For all further studies,only cervical lymph nodes were analyzed.

Flank injections were performed initially with different cellconcentration injected subcutaneously in SCID-NOD mice. Atotal of 2,000,000 cells/100ml was the conditionwhere all animals

developed visible tumors after 3 wk, thus resembling the tonguemodel. Animals were monitored weekly and euthanized at wk 5after tumor implantation. Inguinal, abdominal, and cervicallymph nodes were retrieved and analyzed histopathologically.

Immunofluorescence, immunohistochemistry, andhistopathologic analysis

Tongueswerecut into4 sectionsof approximately the same thickness,following its major axis. These sections were fixed and embedded ina single paraffin block for histopathologic analysis and immunohis-tochemistry or in optimal cutting temperature compound for immu-nofluorescence. Multiple 8 mm sections were cut and stained aspreviously described (27, 28). For immunohistochemistry, we usedrabbit polyclonal anti-LYVE1 1:200 (Abcam, Cambridge, MA, USA)and rat monoclonal anti-CD31 1:50 (BD Bioscience), and for immu-nofluorescence,weusedrabbitpolyclonalanti-LYVE11:250(Abcam).

Two-photon microscopy

Time-lapse and 3-dimensional acquisitions were performed usingan Olympus IX81 microscope equipped with FluoView 1000 scan-ninghead(OlympusAmerica Inc.,CenterValley,PA,USA) thatwascustomized for 2-photon microscopy as previously described (29).

Flow cytometry (FACS) analysis

Cells were collected and stained using anti-CXCR4 (clone 44717)or isotype control (clone 133303) antibody conjugated to phy-coerythrin for 1 h at 4°C (R&D Systems) in PBS 2% fetal bovineserum. After staining, cells were washed, and 100,000 events wereacquired inFACSCaliburusingCellQuest software(BDBioscience)and analyzed by FlowJo software version 9.4.11 (TreeStar,Ashland, OR, USA).

Population doubling assay

To calculate the population doubling, 30,000 cells were seededandkept in culture inDMEM10%fetal bovine serumfor24, 48, or72 h. The number of cells per well was counted at each of the endpoints and applied to the formula x = [log10(NH/N1)]/log10(2)](30), whereN1 is the inoculum cell number (30,000) andNH thenumber of collected cells. To yield the cumulated doublings, thepopulation doubling for each passage was calculated and thenadded to the population doubling of the previous day (28).

Statistical analysis

Data analysis was done using GraphPad Prism version 5.00 forWindows (GraphPad Software, La Jolla, CA, USA). One-wayANOVA followed byNewman-Keulsmultiple comparison tests wasused, andP valuesof,0.05were considered statistically significant.

RESULTS

CXCL12 induces HeLa cell migration andspontaneous metastasis through CXCR4

To investigate the underlying mechanisms by whichCXCL12 induces tumor cell migration and metastasis, wetook advantage of HeLa cells that express CXCR4 endog-enously. We first performed FACS analysis to confirmCXCR4 expression, and as seen in Fig. 1A, HeLa cells ex-press high levels of CXCR4.Moreover, when challenged ina chemotaxis assay, HeLa cells migration toward CXCL12

SDF-1/CXCL12 INDUCES DIRECTIONAL CELL MIGRATION 3

was remarkable, while control cells without chemo-attractant barely moved. Of note, treatment of HeLa cellswith pertussis toxin (PTX)completely preventedCXCL12-mediated migration without affecting EGF-induced mi-gration, demonstrating that HeLa cell migration toCXCL12 is mediated by coupling of the receptor to Gai(Fig. 1B). To further explore the role of CXCR4 in HeLacell migration andmetastasis, we generated stable CXCR4knockdown cells using lentivirus infection. Upon puro-mycin selection, a pool of shCXCR4+ cells was establishedand validated by FACS for decreased expression of the

chemokine receptor (Fig. 1C). As expected, cells with de-creased CXCR4 expression had defects on CXCL12-mediated chemotaxis (Fig. 1D). Although it has beenreported thatCXCL12/CXCR4 are critical for survival andcell proliferation in connective tissue sarcoma, gliomas,andkidney andprostate tumors (31–34),HeLacells donotrequire CXCR4 signaling for cell proliferation in vitro (Fig.1E). Given the high levels of CXCR4 expression in HeLacells and their propensity to spontaneous metastasis in anoral xenograft model (29), we hypothesized that theirdissemination to locoregional lymph nodes could be

A

0

50

100

150

200

250

***

VehiclePTX

Num

ber o

f cel

l/HM

FIsotypeshControlshCXCR4

CXCR4

LYVE 1

ventral tongue

dorsal tongue

LN Normal MetastaticLN

shControl

shCXCR4

CXCR4

IsotypeCXCR4

B C

Num

ber o

f cel

ls/H

MF

0

50

100

150

200

***

shCXCR4shControl

Vehicle CXCL12

shC shCXCR40

20

40

60

80

100

**

0 7 14 21 280

5

10

Time (days)

shCXCR4shControl

0 1 2 30

1

2

3

4

5shControlshCXCR4

Time (days)Tongue

0

20

40

60

80

100

Lym

ph n

odes

inva

ded

(%)

440

330

220

110

0100 101 102 103

Cou

nts

500μm

500μm 500μm 50μm

Tumor cells

Tumor cells / LYVE 1

Lym

ph n

odes

inva

ded

(%)

Tum

or s

ize

(mm

3 )

D

E F G H

I J K

Cum

ulat

ive

popu

latio

n do

ublin

g

100 101 102 103

Figure 1. CXCL12 induces HeLa cell migration and spontaneous metastasis through CXCR4. A) FACS data showing membraneexpression of CXCR4 in HeLa cells. B) Chemotaxis assay using Boyden chamber after 16 h exposure to vehicle, CXCL12 (50 ng/ml),or EGF (50 ng/ml) showing that CXCL12-mediated migration is PTX sensitive. C) Stable knockdown of CXCR4 was achievedusing shRNA, and FACS analysis was performed to confirm decreased expression of CXCR4 in HeLa cells stably expressingshCXCR4 compared to shControl. D) Chemotaxis assay measured after 16 h exposure to CXCL12 (50 ng/ml) in Boydenchamber, showing decreased cell migration to CXCL12 in shCXCR4 cells. E) Cell proliferation was assessed by populationdoubling, and it was not changed by CXCR4 knockdown. F–H) HeLa cells were injected in the tongue of SCID-NOD mice, andfull necropsy was performed at the experimental end point. F) Graph represents the percentage of invaded lymph nodes permouse. G) Histology from primary tumor and lymph nodes showing in the upper panel a poorly differentiated primary tumorand in the lower panels a normal and a metastatic lymph node. H) Live animal imaging of H2B-GFP expressing primary tumor(upper left) and lymph node metastasis (upper right) using 2-photon microscopy. LYVE1 immunohistochemistry showing highdensity of lymphatic vessel in the primary tumor (lower left). LYVE1 immunofluorescence analysis revealed H2B-GFP+ tumor cellsinside the LYVE1+ lymphatic vessel. I) CXCR4 knockdown did not affected primary tumor growth in tumor xenografts. J) Histologyfrom 4 representative primary tumors (dashed lines represent tumor limits). K) Percentage of invaded cervical lymph nodes permouse, showing that CXCR4 is required for lymph node metastasis. In each case, ANOVA was performed on 2 or 3 independentexperiments in vitro and n = 10 from 2 independent experiments in vivo. Data represent mean 6 SEM. **P , 0.01, ***P , 0.001.

4 Vol. 29 March 2015 DILLENBURG-PILLA ET AL.The FASEB Journal x www.fasebj.org

aCXCR4-mediated process. To test this, wefirst confirmedthe ability of HeLa cells to spontaneous metastasize byinjecting mice in the tongue and screening them for me-tastasis by performing full necropsies. The primary tumorswere large and aggressive, and the majority of the animalshad at least 1 invaded lymphnode.The secondary site in allcases was cervical lymph nodes, and the incidence of me-tastasis was between 70% and 100% (Fig. 1F). Histologicanalysis showed poor differentiated primary tumors andthe presence of tumor cells in cervical lymph nodes (Fig.1G). Because of the easy accessibility permitted by thistonguemodel system, live images fromprimary tumor andlymph node metastasis were acquired using a 2-photonmicroscope. In its upper panels, Fig. 1H shows snapshotsfrom Supplemental Movies 1 and 2, in which H2BGFP(histone and GFP fusion protein) tumor cells can be visu-alized in the tongue and within a cervical lymph node,respectively. Immunohistochemical analysis revealed thatHeLa xenografts are highly positive for the lymphaticmarker LYVE1, suggesting that those tumors have a com-plex lymphatic network. Moreover, immunofluorescenceusingH2BGFP tumors captured thepresenceof tumor cellsinside LYVE1+ vessels within the primary tumor. UsingshCXCR4+ and shControl+ cells, we induced tongue xeno-grafts to assess the contribution of CXCR4 in this experi-mental spontaneousmetastasismodel.Corroborating the invitro data, shCXCR4+ tumor cells showed no proliferationdefect in vivo, and primary tumors from shControl andshCXCR4 cells were similar regarding their growth and tu-mor size at the experimental endpoint (Fig. 1I, J ). Althougheffects in the primary tumor were absent, remarkably,CXCR4-deficientHeLacellswerenot capableof establishinglymph node metastasis (Fig. 1K). These data suggest thatspontaneous lymph node metastasis of HeLa tongue xeno-grafts rely onCXCR4 tometastasize to cervical lymphnodes;therefore, it may represent a novel, short-term, and suitableCXCR4-dependent spontaneous metastasis model.

CXCR4/Gai induces mTORC1 and mTORC2activation and the use of mTOR inhibitors abrogateCXCL12 mediated chemotaxis.

Steady-state HeLa cells in complete media were treatedwith a mTOR kinase inhibitor that blocks mTORC1 andmTORC2activity,Torin2(35); orwithmTORC1 inhibitorsrapamycin and Torin2. Long-term blockade of mTORdownstream targets were achieved with both pharmaco-logic approaches as judged by up to 72 h decrease inAktS473 phosphorylation upon Torin2 treatment and S6phosphorylation upon Torin2 and rapamycin treatment(Fig. 2A). HeLa cells were starved of serum overnightand treated with CXCL12. CXCR4 activation induced S6and AKTS473 phosphorylation, which was completely de-pendent of Gai, as it was abolishedby pretreatment of cellswith PTX (Fig. 2B). As expected, CXCL12-induced acti-vation of mTORC1 was blocked by rapamycin and Torin2;and activation of mTORC2 was blocked by Torin2, asjudged by decreased pAktS473 (35) (Fig. 2C). In order totest whether CXCR4-induced mTOR activation was re-quired for CXCL12 induced chemotaxis, we challengedHeLa cells in a Boyden chamber chemotaxis assay. In-terestingly, blockade of mTORC1 and -2 by Torin2 de-creased CXCR4-mediated migration but had no effect on

EGF receptor– and LPA receptor–mediated migration(Fig. 2D). Moreover, treatment with rapamycin also de-creasedCXCR4-mediatedmigration inHeLa cells (Fig. 2),suggesting that blockade of mTORC1 is sufficient to ab-rogate CXCR4/Gai-mediated migration. Similar findingswere observed when analyzing the role of mTOR inCXCL12-initated migration in MCF-7 (Supplemental Fig.S1A) and T47D cells (Supplemental Fig. S1I), 2 represen-tative breast cancer cells in which CXCR4 uses Gai forchemotaxis (36). However, interestingly, rapamycin hadno effect onCXCL12-mediatedmigration in SUM-159 andMDA-MB-231 (Supplemental Fig. S1J, K), 2 triple-negativebreast cancer cells that have been previously reported toengageGa13 insteadofGai downstreamofCXCR4 for cellmigration (37), supporting amore specific role formTORdownstream of Gi signaling. None of themTOR inhibitorsinduced changes in theCXCR4 expression levels (Fig. 2F).These data indicate that CXCR4/Gai activates mTOR,which is required for CXCL12-mediated migration.

mTORC1 and mTORC2 are required for migrationdirectionality during CXCR4-mediated chemotaxisin vitro

Although rapamycin specifically blocks mTORC1, it hasbeen reported that long-term treatment with rapamycinhas also an inhibitory effect on mTORC2 (38). This in-direct effect of rapamycin onmTORC2makes it difficult todefine which of the mTOR complexes are involved in thedownstream pathway of CXCR4. To address this question,we used a loss-of-function approach using pools of siRNAsequences to knock down key components of each of thecomplexes: Raptor to affect mTORC1, and Rictor to affectmTORC2 (16, 17). Therefore, we generated cells that hadnonfunctional mTORC1 showing a decrease in S6 phos-phorylation, or that had nonfunctional mTORC2 withdecreased Akt phosphorylation at serine 473 (Fig. 3A andSupplemental Fig. S1B, C). Similar to the pharmacologicapproach, disruption of either mTOR complexes had noeffect onCXCR4 expression levels (Fig. 3B) anddecreasedCXCR4-mediated migration without affecting EGF- orLPA-mediated migration (Fig. 3C). To better understandthe involvement ofmTORcomplexes inCXCR4-mediatedmigration, we used a single-cell tracking-based strategy,which allowed us to study various aspects of cell migrationduring chemotaxis. Figure 3D and Supplemental Fig. S1Dshow dot-plot graphs representing the final position of in-dividual cells after 24 h chemotaxis in m-Slides. Black dotsrepresent cells that migrated forward, and red dots repre-sent cells that migrated backward to the chemoattractantgradient; the inner panel in each plot is the vector analysis,showed as a rose diagram.We observed that in the absenceof chemoattractant, these epithelial-derived tumor cellshad limited movement; however, when a gradient of che-moattractant (CXCL12) is applied, they migrate readily inthe direction of the gradient (Fig. 3D and SupplementalFig. S1D). In contrast, tumor cells with a nonfunctionalmTORC1 or mTORC2 were still able to move, but theylacked directionality. Indeed, analysis of statistical parame-ters of m-Slides chemotaxis revealed a minor, albeit signifi-cant, decrease in velocity (Fig. 3E and Supplemental Fig.S1E), as well as accumulated (Fig. 3F and Supplemental

SDF-1/CXCL12 INDUCES DIRECTIONAL CELL MIGRATION 5

Fig. S1F) and Euclidean (Fig. 3G and Supplemental Fig.S1G) distances; however, the major defect observed inRaptor and Rictor knockdown cells was in migration di-rectionality, as judged by a forwardmigration index similarto siControl cells in the absence of CXCL12 (Fig. 3H andSupplemental Fig. S1H). Taken together, these data showthatmTORactivationdownstreamCXCR4/Gai is requiredfor chemotaxis (directional migration) but not for che-mokinesis (random cell movement). Moreover, Raptorknockdown was sufficient to decrease CXCR4-mediateddirectional cell migration, suggesting that pharmacologicinhibition of mTORC1 could potentially be sufficient todecrease CXCR4-mediated migration/metastasis.

Pharmacologic blockade of mTORC1 by rapamycindecreases primary tumor growth and CXCR4-mediatedlymph node metastasis, and increases animal survival

We next used our CXCR4-dependent metastasis model totest whether our in vitro observation that blocking mTORC1is sufficient to decrease CXCR4-mediated migration wererelevant in the context of metastasis. Mice bearing tumorxenograftswere treatedwith vehicleor rapamycin (5mg/kg)daily. As shown in Fig. 4A–D, all animals developed primarytumors; however, tumor size in the rapamycin-treated

group was significantly smaller than in the vehicle-treatedgroup. Moreover, histologic analysis of cervical lymphnodes revealed that the incidence of invaded lymph nodesamong the rapamycin-treated group was dramatically de-creased (Fig. 4E), suggesting that pharmacologic blockadeof mTORC1 is sufficient and effective to decrease CXCR4-mediated metastasis. As a consequence of decreasingprimary tumor size and metastasis, animals from the rapa-mycin group had a major improvement in survival, with100% of animals alive 100 d after the tumor xenograft wasinduced, while all mice from the vehicle-treated group hadto undergo euthanasia due to tumor size to minimize ani-mal suffering before d 90 (Fig. 4F). These data suggest thatmTORC1blockade by rapamycin decreases primary tumorgrowth and CXCR4-mediated metastasis.

Stable disruption of mTORC1 by Raptor shRNA issufficient to decrease cell proliferation andCXCR4-mediated migration in vitro

In order to study the long-term inhibition of mTORC1and mTORC2, we generated stable Raptor and Rictorknockdown cells by infecting HeLa cells with lentivirusexpressing Raptor and Rictor shRNA-targeting sequences.After selection of cell pools resistant to puromycin, we

Figure 2. CXCL12 activates mTOR pathway downstream of CXCR4/Gai, and pharmacologic inhibition of mTOR abrogatesCXCR4-mediated migration. A) Rapamycin (100 nM) and Torin2 (100 nM) decreased the phosphorylation of mTOR downstreamtargets S6 (mTORC1) and AktS473 (mTORC2) in HeLa cells grown in complete media. B) CXCL12-induced mTORC1 andmTORC2 activation in serum-free conditions. S6 and AktS473 phosphorylation induced by CXCL12 is PTX sensitive, showing thatCXCL12 activates the mTOR pathway through a CXCR4/Gai axis. C) CXCL12-induced mTOR activation was blocked by mTORinhibitors rapamycin (mTORC1) and Torin2 (mTORC1 and mTORC2). D) Blockade of mTORC1 and mTORC2 by Torin2decreased CXCR4-mediated chemotaxis of HeLa cells. E) Inhibition of mTORC1 by rapamycin was sufficient to abrogate CXCR4-mediated migration in HeLa cells. F) FACS data showing membrane expression of CXCR4 in control, rapamycin-, and Torin2-treated HeLa cells. In each case, ANOVA was performed on 3 independent experiments. Data represent mean6 SEM. ***P, 0.001.

6 Vol. 29 March 2015 DILLENBURG-PILLA ET AL.The FASEB Journal x www.fasebj.org

analyzed Raptor and Rictor expression by Western blotanalysis to validate knockdowns, and mTORC2 andmTORC1 downstream signaling by analyzing S6 andAktS473 phosphorylation status. As expected, Rictor shRNA

abrogated AktS473 phosphorylation (Fig. 5A) and RaptorshRNA decreased S6 phosphorylation (Fig. 5E), indicatingthat these cells had stable nonfunctional mTORC2and mTORC1, respectively. Consistently, stable Rictor

10 10 10 100 1 2 30

Figure 3. Loss-of-function and single cell tracking approaches reveled that mTORC1 and mTORC2 are required for migrationdirectionality during CXCR4-mediated chemotaxis. A) Western blot analysis showing Raptor and Rictor knockdown and theireffect on the mTOR downstream targets S6 and AktS473 after transfection with control or with Raptor or Rictor siRNA (50 nM).HeLa cells with nonfunctional mTORC1 or mTORC2 upon Raptor and Rictor knockdown, respectively, were tested for CXCR4expression and challenged in chemotaxis assays. B) FACS data showing unchanged membrane expression of CXCR4 upon Raptoror Rictor knockdown. C) Chemotaxis was measured after 16 h exposure to CXCL12 (50 ng/ml), EGF (50 ng/ml), or LPA (1 mM)in Boyden chamber. Raptor and Rictor knockdown decreased CXCL12-mediated chemotaxis. D) Dot-plot graphs representing thefinal position of individual cells after 24 h chemotaxis in m-Slides as described in Materials and Methods. Black dots represent cellsthat migrated forward, and red dots represent cells that migrated backward to the chemoattractant gradient. Rose diagrams of eachgroup are shown in the inner panel of each dot-plot. E–H) Statistical analysis of m-Slides chemotaxis showing a small decrease invelocity (E), accumulated distance (F), and Euclidean distance (G) and a major defect in directionality in Raptor and Rictorknockdown cells, as judged by a forward migration index that was similar to siControl cells in the absence of CXCL12. In each case,ANOVA was performed on 3 independent experiments. Data represent mean 6 SEM. *P , 0.05, **P , 0.01, ***P , 0.001.

SDF-1/CXCL12 INDUCES DIRECTIONAL CELL MIGRATION 7

knockdown cells showed defects in CXCR4-mediated che-motaxis without affecting EGF-mediated migration (Fig.5B) or CXCR4 expression levels (Fig. 5C). Interestingly,cells with nonfunctional mTORC2 exhibited no changesin proliferation rate in vitro (Fig. 5D), as they had similarpopulation doubling rate compared to the pool of shCon-trol cells. Regarding cells with stable nonfunctionalmTORC1, defects on CXCR4-mediated chemotaxis wereobserved, corroborating the pharmacologic and transientknockdown approaches, with no effect on EGF-mediatedmigration (Fig. 5F) or CXCR4 membrane expression (Fig.5G). Nonetheless, Raptor ablation significantly decreasedthe proliferation rate of the cells. As shown in Fig. 5H,shRaptor+ cells required almost twice the time to doubletheir population compared with shControl cells. Thesedata suggest that in addition to be sufficient to decreaseCXCR4-mediated chemotaxis, targeting mTORC1 toimpair tumorprogression anddissemination also confersgrowth disadvantages to tumor cells, thus limiting theability to establishing primary and metastatic colonies.

Stable disruption of mTORC1 by Raptor shRNA issufficient to decrease tumor burden, angiogenesis,lymphangiogenesis, and lymph node metastasis

To rule out the possibility that an indirect effect of rapamy-cin long-term administration in mTORC2 could contribute

to the decrease in tumor burden and metastasis, we tookadvantage of our stable Raptor and Rictor knockdown cells.We induced tumor xenografts using the CXCR4-mediatedspontaneous metastasis model by injecting shControl,shRictor, or shRaptor cells and monitored animals weekly.At the experimental end point, primary tumor and cervicallymph nodes were collected for histologic analysis. Surpris-ingly, the disruption of mTORC2 by Rictor knockdown didnot affect tumor progression in vivo. As seen in Fig. 6A, B,shControl and shRictor cells generatedprimary tumorswithsimilar size. Moreover, despite affecting CXCR4-mediatedmigration in vitro (Fig. 6B), Rictor knockdown did not pre-vent CXCR4-mediated spontaneous metastasis in vivo (Fig.6C). Interestingly, cells with nonfunctional mTORC2 werestill able to induce changes in the tumor microenviromentand induce angiogenesis and lymphangiogenesis, asshControl and shRictor tumors showed similar levels of theendothelialmarker CD31 and the lymphaticmarker LYVE1(Fig. 6D–F). A completely different scenario was observedupon mTORC1 disruption in the tumor cells. Primarytumors from cells with nonfunctional mTORC1 due toRaptor knockdown were significantly smaller compared tothe shControl group (Fig. 6G, H). Interestingly, disruptionof mTORC1 function had a major effect on CXCR4-mediated metastasis, as Raptor knockdown tumors showeda significantdecrease in lymphnodemetastasis, withmost ofthe animals (16 of 20) free of metastases at the end of the

Figure 4. Rapamycin decreased primary tumor growth, decreased CXCR4-mediated lymph node metastasis, and increased animalsurvival. A–D) HeLa xenografts were induced, and vehicle and rapamycin (5 mg/kg per d) treatment started after the primarytumor was established (2 mm3). A) Primary tumor growth curve. B) Representative images at the experimental end point. C, D)Histologic analysis of the tumor area showing a decrease in primary tumor size upon rapamycin treatment. E) Number of invadedlymph nodes per mouse at the experimental end point. F) Rapamycin treatment increased animal survival. Statistical analysis wasperformed by ANOVA. n = 10 from 2 independent experiment. Data represent mean 6 SEM. ***P , 0.001.

8 Vol. 29 March 2015 DILLENBURG-PILLA ET AL.The FASEB Journal x www.fasebj.org

observationperiod(Fig.6I). Inaddition,adramatic impactofadysfunctionalmTORC1within thetumorcellswasobservedin the tumor microenvironment. Primary tumors from theshRaptor group had a significant decrease in the number ofCD31+ endothelial cells (Fig. 6J, K) as well as LYVE1+ lym-phatic vessels (Fig. 6J, L). Taken together, our data show thatmTOR is activated downstreamCXCR4/Gai axis and that itis essential to CXCR4-mediated directional migration andmetastasis.Ourdataalso suggest thatalthoughmTORC1andmTORC2 are activated uponCXCR4 stimulation,mTORC1seems to play amore important role in tumor progression invivo (Fig. 6M). While mTORC2 blockade affected chemo-taxis, disruption of mTORC1 significantly decreasedCXCR4-mediated migration and metastasis. In addition,mTORC1 disruption using pharmacologic or loss of func-tion approaches also impacted cell proliferation, which isrequired for primary tumor growth and the establishmentof secondary colonies. The ability of tumor cells to modifytheir microenviroment was also compromised by Raptorablation, thus suppressing angiogenesis and lymphangio-genesis, which limits the access of tumor cells to circulation.

DISCUSSION

In the present study, we used HeLa cells that express highlevels of CXCR4 endogenously as an experimental model

system to investigate the downstream signalingmechanisminvolved in CXCL12-mediated migration and metastasis.Our data show that binding of CXCL12 to CXCR4 inducesmTOR pathway activation in these epithelial-derived tu-mor cells. Moreover, mTOR activation downstream ofCXCR4 is strictly required for CXCL12-mediated migra-tion and is completely dependent of G proteins from theGai family. Similarfindingswereobserved inhumanbreastcancer cells exhibiting Gai-dependent migration and sig-naling downstream fromCXCR4. BecausemTORexists in2 different complexes, mTORC1 and mTORC2, and be-cause CXCL12 induces phosphorylation of downstreamtargets of both complexes, we used pharmacologic andloss-of-function approaches to address the contribution ofmTORC1 and mTORC2 in CXCR4-mediated directionalmigration. We demonstrated that disruption of eithermTORC1 or mTORC2 has a profound effect on di-rectionality during chemotaxis. To study the CXCR4/mTOR axis in vivo, we developed a model of CXCR4-dependent spontaneous metastasis in which 70–100% ofthe animals have lymph node metastasis at 1 month aftercell inoculation. Surprisingly, mTORC1 blockade was suf-ficient to decrease CXCR4-mediated metastasis as well astumor cell proliferation, angiogenesis, and lymphangio-genesis, whereas mTORC2 impairment had no de-monstrableeffect in tumorgrowthordissemination in vivo.

100 101 10 2 1030

10 10 10 100 1 2 30

Figure 5. Stable disruption of mTORC1 by Raptor shRNA was sufficient to decrease cell proliferation and CXCR4-mediatedmigration in vitro. A) Western blot analysis showing decreased expression of Rictor and down-regulation of mTORC2 downstreampathway upon stable knockdown of Rictor using shRNA. B) Boyden chamber chemotaxis assay showing decreased migration ofshRictor cells to CXCL12 (50 ng/ml) and unchanged migration to EGF (50 ng/ml). C) FACS data showing unaltered CXCR4membrane expression upon Rictor stable knockdown. D) Down-regulation of Rictor and mTORC2 downstream pathway had noeffect on cell proliferation in vitro. E) Decreased expression of Raptor and down-regulation of mTORC1 downstream pathwaydemonstrated by Western blot analysis in cells expressing Raptor shRNAs. F) Stable Raptor knockdown decreased CXCL12-mediated migration without affecting the ability of cells to migrate to EGF. G) Membrane expression of CXCR4 assessed by FACSwas not changed upon stable Raptor knockdown. H) Raptor stable knockdown was sufficient to decrease cell proliferation invitro. In each case, ANOVA was performed on 2 to 3 independent experiments. Data represent mean 6 SEM. **P , 0.01.

SDF-1/CXCL12 INDUCES DIRECTIONAL CELL MIGRATION 9

The PI3K/Akt/mTOR pathway is a major player inmany tumor types. It regulates cell metabolism, growth,survival, and proliferation as well as cytoskeleton organi-zation (9). Although many functions that are regulated bymTOR are easily associated with tumor progression, theprecise mechanisms by which the mTOR pathway

mediates tumor development and dissemination are notfully elucidated. Most of the effects of mTOR in cell mi-gration have been attributed to mTORC2 (17, 18, 39–42)due to its ability to activate small GTPases, such as Rac, andto control actin cytoskeleton organization (18). Recently,Liu et al. (40) reported that in neutrophilsmTORC2 is also

0

20

40

60

80

100

Inva

ded

Lym

ph n

odes

(%)

* **

0

5

10

15

20

25

Tum

or a

rea

(mm

2 )

** ***

shControl

shRaptor 1

shControl shRaptor0.000

0.002

0.004

0.006

0.008

**

shControl shRaptor0.0

0.1

0.2

0.3

**

LYVE 1

shControl

shRaptor

shControl

shRaptor

CD31

CD31

shControl

shRictor

LYVE 1

shControl

shRictor

shControl

shRictor 1 shRictor 2

shRaptor 2

0

5

10

15

20

25

0

20

40

60

80

100

Inva

ded

Lym

ph n

odes

(%)

shControl 1 2shRictor

shControl 1 2shRictor

shControl 1 2shRaptor

shControl 1 2shRaptor

ns

ns

shControl shRictor0.0000

0.0001

0.0002

0.0003

0.0004

0.0005

shControl shRictor0.0

0.1

0.2

0.3

0.4

Tum

or a

rea

(mm

2 )

CD

31 p

ositi

ve c

ells

/μm

2

LYV

E1

posi

tive

cells

/μm

2

CD

31 p

ositi

ve c

ells

/μm

2

LYV

E 1

pos

itive

cel

ls/m

m2

A B C D

E F G H

I J K L

αi GDP

βγ

αiGTP

SDF-1CXCR4

mTORC1 •Directional migration•Growth•Angiogenesis•Lymphangiogenesis•Metastasis

•Directional migrationmTORC2βγ

M

Figure 6. Stable disruption of mTORC1 by Raptor shRNA was sufficient to decrease tumor burden, angiogenesis,lymphangiogenesis, and lymph node metastasis. A) Histology of 4 representative primary tumors from shControls and shRictor(dashed lines represents the tumor limits). B) Tumor area assessed at the experimental end point. C) Percentage of invadedlymph nodes per mouse showing that disruption of mTORC2 by Rictor knockdown did not affect CXCR4-dependent lymph nodemetastasis. Immunohistochemistry data revealed similar expression of CD31+ blood vessels (D, E) and LYVE1+ lymphatic vessels(D, F) within the primary tumor. G) Histology of 4 representative primary tumors from shControls and shRaptor (dashed linesrepresent tumor limits). H) Tumor area assessed at the experimental end point. I) Percentage of invaded lymph nodes permouse showing that disruption of mTORC1 by Raptor knockdown decreased CXCR4-dependent lymph node metastasis.Immunohistochemical data showing decreased expression of CD31+ blood vessels (J, K) and LYVE1+ lymphatic vessels (J, L)within the primary tumor. M) Schematic representation of CXCR4/Gai/mTOR pathway in CXCL12-mediated migration andmetastasis. In each case, ANOVA was performed. n = 10 from 2 independent experiments. Data represent mean6 SEM. *P, 0.05,**P , 0.01, ***P , 0.001.

10 Vol. 29 March 2015 DILLENBURG-PILLA ET AL.The FASEB Journal x www.fasebj.org

critical for cell migration through a mechanism that isdependent on cAMP and RhoA activation. ConcerningmTORC1, little is knownregarding its role in cellmigrationand metastasis. In particular, as a result of its importantfunction in protein synthesis (43, 44), mTORC1 has beendescribed to influence cell migrationmainly by regulatingexpressionof key components of themigratorymachinery,such as small GTPases and chemokine receptors (26, 45).By use of loss-of-function strategies, our study shows thatboth mTOR complexes are important for cell migrationdownstream of CXCR4. Of note, no changes in CXCR4 orin small GTPases RhoA, Rac, or CDC42 expression wereobserved after rapamycin and Torin2 treatment orRaptor/Rictor knockdown(datanot shown).Moreover, byuse of a single-cell tracking strategy, we showed thatblockade ofmTORC1 ormTORC2 decreases the ability ofcells to migrate directionally toward CXCL12, but it doesnot abrogate the ability of the cells tomove, indicating thatmTORC1 and mTORC2 are essential for chemotaxis butnot for chemokinesis. The underlying mechanism of suchan effect is not yet known, but it is tempting to speculatethat it might be related to the ability of the cells to polarizeand sense the chemoattractant gradient, as cells lose di-rectionality and not motility upon mTOR inhibition. Sup-porting this hypothesis, the connection between mTORpathway and cell polarity has already been reported, as anmTOR-regulating kinase, LKB1, is the human ortholog ofDrosophila melanogaster’s PAR-4 (46, 47), and it has beenshown that once activated, LKB1 can fully polarize a singlecell in the absence of cell–cell contact (48). However,similar to 20% of all cervical carcinomas (49), HeLa cellsdonot expressLKB1, suggesting theexistenceof a yet-to-beidentified mechanism linking mTOR and cell polarity,which warrants further investigation.

Although tumor cell migration is a key and rate-limitingevent in metastasis, it does not fully recapitulate the com-plex and multistep biologic process of metastasis. There-fore, we aimed to study the role of CXCR4/Gai/mTORaxis in vivo. Unfortunately, the limited number of sponta-neous metastasis models, together with the long-term du-rationand low incidenceof advanceddisease inmost of thecurrent protocols, represent substantial challenges in thestudy of metastasis (50). To overcome these difficulties,experimental approaches such as injection of tumor cellsin theheart or in the tail vein of experimental animals havebeen established. Even though these models have beenlargely used in cancer research, injecting cells directly intothe bloodstream does not fully recapitulate the completeprocess of tumor spread. In fact, how tumor cells gain ac-cess to blood and lymphatic vessels represents a key regu-latory step for tumor metastasis (50).

Considering that HeLa cells express high levels ofCXCR4, and considering our observation that they spon-taneously metastasize to cervical lymph nodes after in-jection in the tongue (29), we hypothesized that CXCR4could be the driver of HeLa cells spontaneous metastasis.Of interest, the rationale behind the tongue xenograftmodel relies in 3 important elements: 1) the tongue isa highly blood and lymphatic vascularized organ, 2) theneck area is fairly close to the primary tumor site and holdsabout one third of the body’s lymph nodes (51), and 3)carcinoma from uterus cervix metastasize to lymph nodessimilarly to most carcinomas, including those from the

head and neck, suggesting that themechanismunderlyinglymph node metastasis could be shared by different carcino-mas (52). Combining this model with a loss-of-functionapproach, we observed that HeLa require CXCR4 to me-tastasize. Indeed, shCXCR4+ cells showed severe defects tomigrate to CXCL12 in vitro and were not capable of in-vading lymph nodes in vivo. Of note, the requirement ofCXCR4 formetastasis, togetherwith the complete absenceof distant metastasis, especially considering lung and liver,which are common target organs when tumor cells areinjecteddirectly into thebloodstream, strongly suggest thatthis is an active and chemokine-regulated process, whichrequires tumor cells to gain access and follow a gradient ofchemokine to reach secondary organs. Furthermore, asa result of the easy access, live imaging from primarytongue tumor and cervical lymph node metastasis arepossible using 2-photonmicroscope, which allowsmultipleand sequential real-time images from primary and meta-static sites. Taken together, we can expect that this CXCR4-dependent spontaneousmetastasismodel, which results ina large fraction of mice exhibiting lymph node invasion,may provide a suitable experimental model for the futureevaluation of antimetastatic agents. Certainly, carefulstandardization will be required for the future de-velopment of similar spontaneous metastasis models foreach human cancer cell line of interest. In this context, wetook advantage of the CXCR4-dependent spontaneousmetastasis model to investigate whether mTOR activationwas required to tumor spontaneous metastasis in vivo,andmore importantly, what the contribution of mTORC1and mTORC2 are to the process. Indeed, our data showthat although both complexes are essential for CXCR4-mediated directional migration in vitro, mTORC1 mayplay a more important role in vivo. In this regard, disrup-tion of mTORC1 function by either rapamycin or byRaptor knockdown decreased several features of aggres-siveness, such as cell proliferation and chemotaxis in vitro,and tumor burden, vascularization, and ultimately lymphnode metastasis in vivo. Although it is not possible to dis-sociate at this stage the effects of mTORC1 disruption onthe primary tumor growth from tumor dissemination, it isimportant to consider that this may resemble the clinicalscenario. In fact, standard cancer therapies successfullydecrease primary tumor size, but often they have a morelimited impact on treating or preventing metastatic dis-ease, and consequently, metastasis remains the cause ofmore than 90% of all cancer-related deaths (50).

More studies areneeded tounderstand themechanismsby which mTORC1 contributes to CXCR4-mediated mi-gration and metastasis; however, one possibility is that inaddition to the loss of directionality in cell migration,tumorswithnonfunctionalmTORC1are also inefficient ininducingchanges in their tumormicroenvironmentand ineliciting autocrine/paracrine cytokine-initiated feedbackloops that stimulate tumor progression and dissemination(33, 53, 54). Indeed, the remarkable effect of mTORC1blockade in the primary tumor, and possibly cytokine-induced feedback loops, could help to explain the factthat only mTORC1 and not mTORC2 disruption havemajor consequences in CXCR4-mediated metastasis.Consistently, we have observed a significant decrease inangiogenesis and lymphangiogenesis in tumors with non-functional mTORC1 but no impact in endothelial or

SDF-1/CXCL12 INDUCES DIRECTIONAL CELL MIGRATION 11

lymphatic vascular network in response to Rictor knock-down in cancer cells.

Although CXCR4 itself does not currently representa drug-directed target for cancer therapy, identifyingdownstream targets involved in CXCR4-mediated migra-tion and metastasis may afford a new window of opportu-nity for pharmacologic intervention for aggressive tumors.Taken together, our data suggest that mTORC1 blockadeis sufficient to decrease primary tumor growth andCXCR4/Gai/mTORC1-dependent migration and metas-tasis in vivo.Therefore, drugs that blockmTORC1, such asthe U.S. Food and Drug Administration–approved rapa-mycin, RAD001, and Torisel, then could be explored as anoption for highly aggressive andmetastatic tumors that relyon CXCR4 to spread. Of note, rapamycin has not beenreported to promote expulsion of bone marrow stem cellsfrom their niche (24, 25), a side effect observed when tar-geting CXCR4 itself (5–8). Indeed, if these findings can betranslated into the clinic, such an approach would providean opportunity to target both primary and metastatictumors, thereby potentially improving patient outcome inthe caseofmultiplehighly aggressivemalignancies that useCXCR4 for their metastatic spread.

This research was supported by the Intramural ResearchProgram of U.S. National Institutes of Health, NationalInstitute of Dental and Craniofacial Research. The authorsapologize to all of their colleagues for not citing some of theiroriginal studies as a result of space limitations.

REFERENCES

1. Moser, B., Wolf, M., Walz, A., and Loetscher, P. (2004)Chemokines: multiple levels of leukocyte migration control.Trends Immunol. 25, 75–84

2. Zlotnik, A., and Yoshie, O. (2012) The chemokine superfamilyrevisited. Immunity 36, 705–716

3. Zlotnik, A., Burkhardt, A. M., and Homey, B. (2011)Homeostatic chemokine receptors and organ-specific metasta-sis. Nat. Rev. Immunol. 11, 597–606

4. Muller, A., Homey, B., Soto, H., Ge, N., Catron, D., Buchanan,M. E., McClanahan, T., Murphy, E., Yuan, W., Wagner, S. N.,Barrera, J. L., Mohar, A., Verastegui, E., and Zlotnik, A. (2001)Involvement of chemokine receptors in breast cancer metastasis.Nature 410, 50–56

5. Broxmeyer, H. E., Orschell, C. M., Clapp, D. W., Hangoc, G.,Cooper, S., Plett, P. A., Liles, W. C., Li, X., Graham-Evans, B.,Campbell, T. B., Calandra, G., Bridger, G., Dale, D. C., andSrour, E. F. (2005) Rapid mobilization of murine and humanhematopoietic stem and progenitor cells with AMD3100,a CXCR4 antagonist. J. Exp. Med. 201, 1307–1318

6. Devine, S. M., Flomenberg, N., Vesole, D. H., Liesveld, J.,Weisdorf, D., Badel, K., Calandra, G., and DiPersio, J. F. (2004)Rapid mobilization of CD34+ cells following administration of theCXCR4 antagonist AMD3100 to patients with multiple myelomaand non-Hodgkin’s lymphoma. J. Clin. Oncol. 22, 1095–1102

7. Pusic, I., and DiPersio, J. F. (2010) Update on clinical experiencewith AMD3100, an SDF-1/CXCL12-CXCR4 inhibitor, in mobili-zation of hematopoietic stem and progenitor cells. Curr. Opin.Hematol. 17, 319–326

8. Devine, S. M., Vij, R., Rettig, M., Todt, L., McGlauchlen, K., Fisher, N.,Devine, H., Link, D. C., Calandra, G., Bridger, G., Westervelt, P., andDipersio, J. F. (2008) Rapid mobilization of functional donorhematopoietic cells without G-CSF using AMD3100, an antago-nist of the CXCR4/SDF-1 interaction. Blood 112, 990–998

9. Laplante, M., and Sabatini, D. M. (2012) mTOR signaling ingrowth control and disease. Cell 149, 274–293

10. Zoncu, R., Efeyan, A., and Sabatini, D. M. (2011) mTOR: fromgrowth signal integration to cancer, diabetes and ageing. Nat.Rev. Mol. Cell Biol. 12, 21–35

11. Yuan, T. L., and Cantley, L. C. (2008) PI3K pathway alterations incancer: variations on a theme. Oncogene 27, 5497–5510

12. Rosenbluth, J. M., Mays, D. J., Jiang, A., Shyr, Y., and Pietenpol,J. A. (2011) Differential regulation of the p73 cistrome bymammalian target of Rapamycin reveals transcriptionalprograms of mesenchymal differentiation and tumorigenesis.Proc. Natl. Acad. Sci. USA 108, 2076–2081

13. Posch, C., Moslehi, H., Feeney, L., Green, G. A., Ebaee, A.,Feichtenschlager, V., Chong, K., Peng, L., Dimon, M. T., Phillips,T., Daud, A. I., McCalmont, T. H., LeBoit, P. E., and Ortiz-Urda, S.(2013) Combined targeting of MEK and PI3K/mTOR effectorpathways is necessary to effectively inhibit NRAS mutant mela-noma in vitro and in vivo. Proc. Natl. Acad. Sci. USA 110, 4015–4020

14. Dutt, P., Wang, J. F., and Groopman, J. E. (1998) Stromalcell–derived factor-1 alpha and stem cell factor/kit ligand sharesignaling pathways in hemopoietic progenitors: a potentialmechanism for cooperative induction of chemotaxis. J. Immunol.161, 3652–3658

15. Sotsios, Y., Whittaker, G. C., Westwick, J., and Ward, S. G. (1999)The CXC chemokine stromal cell–derived factor activates a Gi-coupled phosphoinositide 3-kinase in T lymphocytes. J. Immunol.163, 5954–5963

16. Kim, D. H., Sarbassov, D. D., Ali, S. M., King, J. E., Latek, R. R.,Erdjument-Bromage, H., Tempst, P., and Sabatini, D. M. (2002)mTOR interacts with raptor to form a nutrient-sensitive complexthat signals to the cell growth machinery. Cell 110, 163–175

17. Sarbassov, D. D., Ali, S. M., Kim, D. H., Guertin, D. A., Latek,R. R., Erdjument-Bromage, H., Tempst, P., and Sabatini, D. M.(2004) Rictor, a novel binding partner of mTOR, definesa Rapamycin-insensitive and Raptor-independent pathway thatregulates the cytoskeleton. Curr. Biol. 14, 1296–1302

18. Jacinto, E., Loewith, R., Schmidt, A., Lin, S., Ruegg, M. A., Hall,A., and Hall, M. N. (2004) Mammalian TOR complex 2 controlsthe actin cytoskeleton and is Rapamycin insensitive. Nat. Cell Biol.6, 1122–1128

19. Stallone, G., Infante, B., Grandaliano, G., Schena, F. P., andGesualdo, L. (2008) Kaposi’s sarcoma and mTOR: a crossroadbetween viral infection neoangiogenesis and immunosuppression.Transpl. Int. 21, 825–832

20. Dancey, J. (2010) mTOR signaling and drug development incancer. Nat Rev Clin Oncol 7, 209–219

21. Atkins, M. B., Hidalgo, M., Stadler, W. M., Logan, T. F., Dutcher,J. P., Hudes, G. R., Park, Y., Liou, S. H., Marshall, B., Boni, J. P.,Dukart, G., and Sherman, M. L. (2004) Randomized phase IIstudy of multiple dose levels of CCI-779, a novel mammaliantarget of Rapamycin kinase inhibitor, in patients with advancedrefractory renal cell carcinoma. J. Clin. Oncol. 22, 909–918

22. Stallone, G., Schena, A., Infante, B., Di Paolo, S., Loverre, A.,Maggio, G., Ranieri, E., Gesualdo, L., Schena, F. P., andGrandaliano, G. (2005) Sirolimus for Kaposi’s sarcoma inrenal-transplant recipients. N. Engl. J. Med. 352, 1317–1323

23. Plas, D. R., and Thomas, G. (2009) Tubers and tumors:Rapamycin therapy for benign and malignant tumors. Curr.Opin. Cell Biol. 21, 230–236

24. Garcıa-Garcıa, C., Ibrahim, Y. H., Serra, V., Calvo, M. T.,Guzman, M., Grueso, J., Aura, C., Perez, J., Jessen, K., Liu, Y.,Rommel, C., Tabernero, J., Baselga, J., and Scaltriti, M. (2012)Dual mTORC1/2 and HER2 blockade results in antitumor ac-tivity in preclinical models of breast cancer resistant to anti-HER2therapy. Clin. Cancer Res. 18, 2603–2612

25. Janes, M. R., Vu, C., Mallya, S., Shieh, M. P., Limon, J. J., Li, L. S.,Jessen, K. A., Martin, M. B., Ren, P., Lilly, M. B., Sender, L. S., Liu,Y., Rommel, C., and Fruman, D. A. (2013) Efficacy of theinvestigational mTOR kinase inhibitor MLN0128/INK128 in mod-els of B-cell acute lymphoblastic leukemia. Leukemia 27, 586–594

26. Chen, G., Chen, S. M., Wang, X., Ding, X. F., Ding, J., and Meng,L. H. (2012) Inhibition of chemokine (CXC motif) ligand 12/chemokine (CXC motif) receptor 4 axis (CXCL12/CXCR4)-mediated cell migration by targeting mammalian target ofRapamycin (mTOR) pathway in human gastric carcinoma cells.J. Biol. Chem. 287, 12132–12141

27. Molinolo, A. A., Marsh, C., El Dinali, M., Gangane, N., Jennison,K., Hewitt, S., Patel, V., Seiwert, T. Y., and Gutkind, J. S. (2012)mTOR as a molecular target in HPV-associated oral and cervicalsquamous carcinomas. Clin. Cancer Res. 18, 2558–2568

28. Iglesias-Bartolome, R., Patel, V., Cotrim, A., Leelahavanichkul,K., Molinolo, A. A., Mitchell, J. B., and Gutkind, J. S. (2012)

12 Vol. 29 March 2015 DILLENBURG-PILLA ET AL.The FASEB Journal x www.fasebj.org

mTOR inhibition prevents epithelial stem cell senescence andprotects from radiation-induced mucositis. Cell Stem Cell 11,401–414

29. Amornphimoltham, P., Rechache, K., Thompson, J., Masedunskas,A., Leelahavanichkul, K., Patel, V., Molinolo, A., Gutkind,J. S., and Weigert, R. (2013) Rab25 regulates invasion andmetastasis in head and neck cancer. Clin. Cancer Res. 19,1375–1388

30. Cristofalo, V. J., Allen, R. G., Pignolo, R. J., Martin, B. G., andBeck, J. C. (1998) Relationship between donor age and thereplicative lifespan of human cells in culture: a reevaluation.Proc. Natl. Acad. Sci. USA 95, 10614–10619

31. Zhou, Y., Larsen, P. H., Hao, C., and Yong, V. W. (2002) CXCR4is a major chemokine receptor on glioma cells and mediatestheir survival. J. Biol. Chem. 277, 49481–49487

32. Schrader, A. J., Lechner, O., Templin, M., Dittmar, K. E.,Machtens, S., Mengel, M., Probst-Kepper, M., Franzke, A.,Wollensak, T., Gatzlaff, P., Atzpodien, J., Buer, J., and Lauber,J. (2002) CXCR4/CXCL12 expression and signalling in kidneycancer. Br. J. Cancer 86, 1250–1256

33. Mo, W., Chen, J., Patel, A., Zhang, L., Chau, V., Li, Y., Cho, W.,Lim, K., Xu, J., Lazar, A. J., Creighton, C. J., Bolshakov, S., McKay,R. M., Lev, D., Le, L. Q., and Parada, L. F. (2013) CXCR4/CXCL12 mediate autocrine cell-cycle progression in NF1-associated malignant peripheral nerve sheath tumors. Cell 152,1077–1090

34. Darash-Yahana, M., Pikarsky, E., Abramovitch, R., Zeira, E., Pal,B., Karplus, R., Beider, K., Avniel, S., Kasem, S., Galun, E., andPeled, A. (2004) Role of high expression levels of CXCR4 intumor growth, vascularization, and metastasis. FASEB J. 18,1240–1242

35. Liu, Q., Wang, J., Kang, S. A., Thoreen, C. C., Hur, W., Ahmed,T., Sabatini, D. M., and Gray, N. S. (2011) Discovery of 9-(6-aminopyridin-3-yl)-1-(3-(trifluoromethyl)phenyl)benzo[h][1,6]naphthyridin-2(1H)-one (Torin2) as a potent, selective, andorally available mammalian target of Rapamycin (mTOR) in-hibitor for treatment of cancer. J. Med. Chem. 54, 1473–1480

36. Sosa, M. S., Lopez-Haber, C., Yang, C., Wang, H., Lemmon,M. A., Busillo, J. M., Luo, J., Benovic, J. L., Klein-Szanto, A., Yagi,H., Gutkind, J. S., Parsons, R. E., and Kazanietz, M. G. (2010)Identification of the Rac-GEF P-Rex1 as an essential mediator ofErbB signaling in breast cancer. Mol. Cell 40, 877–892

37. Yagi, H., Tan, W., Dillenburg-Pilla, P., Armando, S.,Amornphimoltham, P., Simaan, M., Weigert, R., Molinolo, A. A.,Bouvier, M., and Gutkind, J. S. (2011) A synthetic biology approachreveals a CXCR4-G13-Rho signaling axis driving transendothelialmigration of metastatic breast cancer cells. Sci. Signal. 4, ra60

38. Sarbassov, D. D., Ali, S. M., Sengupta, S., Sheen, J. H., Hsu, P. P.,Bagley, A. F., Markhard, A. L., and Sabatini, D. M. (2006)Prolonged Rapamycin treatment inhibits mTORC2 assembly andAkt/PKB. Mol. Cell 22, 159–168

39. Hernandez-Negrete, I., Carretero-Ortega, J., Rosenfeldt, H.,Hernandez-Garcıa, R., Calderon-Salinas, J. V., Reyes-Cruz, G.,Gutkind, J. S., and Vazquez-Prado, J. (2007) P-Rex1 links mam-malian target of Rapamycin signaling to Rac activation and cellmigration. J. Biol. Chem. 282, 23708–23715

40. Liu, L., Das, S., Losert, W., and Parent, C. A. (2010) mTORC2regulates neutrophil chemotaxis in a cAMP- and RhoA-dependent fashion. Dev. Cell 19, 845–857

41. Kim, E. K., Yun, S. J., Ha, J. M., Kim, Y. W., Jin, I. H., Yun, J., Shin,H. K., Song, S. H., Kim, J. H., and Lee, J. S., Kim, C. D., Bae, S. S.(2011) Selective activation of Akt1 by mammalian target ofRapamycin complex 2 regulates cancer cell migration, invasion,and metastasis. Oncogene 30, 2954–2963

42. Gan, X., Wang, J., Wang, C., Sommer, E., Kozasa, T., Srinivasula,S., Alessi, D., Offermanns, S., Simon, M. I., and Wu, D. (2012)PRR5L degradation promotes mTORC2-mediated PKC-d phos-phorylation and cell migration downstream of Ga12. Nat. CellBiol. 14, 686–696

43. Dann, S. G., Selvaraj, A., and Thomas, G. (2007) mTORComplex1-S6K1 signaling: at the crossroads of obesity, diabetesand cancer. Trends Mol. Med. 13, 252–259

44. Yanagiya, A., Suyama, E., Adachi, H., Svitkin, Y. V., Aza-Blanc, P.,Imataka, H., Mikami, S., Martineau, Y., Ronai, Z. A., andSonenberg, N. (2012) Translational homeostasis via the mRNAcap-binding protein, eIF4E. Mol. Cell 46, 847–858

45. Liu, L., Luo, Y., Chen, L., Shen, T., Xu, B., Chen, W., Zhou, H.,Han, X., and Huang, S. (2010) Rapamycin inhibits cytoskeletonreorganization and cell motility by suppressing RhoA expressionand activity. J. Biol. Chem. 285, 38362–38373

46. Kemphues, K. J., Priess, J. R., Morton, D. G., and Cheng, N. S.(1988) Identification of genes required for cytoplasmiclocalization in early C. elegans embryos. Cell 52, 311–320

47. Martin, S. G., and St Johnston, D. (2003) A role for DrosophilaLKB1 in anterior–posterior axis formation and epithelialpolarity. Nature 421, 379–384

48. Baas, A. F., Kuipers, J., van der Wel, N. N., Batlle, E., Koerten,H. K., Peters, P. J., and Clevers, H. C. (2004) Completepolarization of single intestinal epithelial cells upon activationof LKB1 by STRAD. Cell 116, 457–466

49. Wingo, S. N., Gallardo, T. D., Akbay, E. A., Liang, M. C., Contreras,C. M., Boren, T., Shimamura, T., Miller, D. S., Sharpless, N. E.,Bardeesy, N., Kwiatkowski, D. J., Schorge, J. O., Wong, K. K., andCastrillon, D. H. (2009) Somatic LKB1 mutations promote cervicalcancer progression. PLoS ONE 4, e5137

50. Eckhardt, B. L., Francis, P. A., Parker, B. S., and Anderson, R. L.(2012) Strategies for the discovery and development of therapiesfor metastatic breast cancer. Nat. Rev. Drug Discov. 11, 479–497

51. Kumar, V., Abbas, A. K., Aster, J. C. (2014) Robbins and CotranPathologic Basis of Disease, 9th ed. Saunders, Philadelphia

52. Duenne, A. A., Werner, J. A. (2000) Functional anatomy oflymphatic vessels under the aspect of tumor invasion. In Lym-phatic Metastasis and Sentinel Lymphonodectomy (Schlag, P. M.,Veronesi, U., eds.), pp. 82–89. Springer, Berlin

53. Narita, M., Young, A. R., Arakawa, S., Samarajiwa, S. A.,Nakashima, T., Yoshida, S., Hong, S., Berry, L. S., Reichelt, S.,Ferreira, M., Tavare, S., Inoki, K., Shimizu, S., and Narita, M.(2011) Spatial coupling of mTOR and autophagy augmentssecretory phenotypes. Science 332, 966–970

54. Patel, V., Marsh, C. A., Dorsam, R. T., Mikelis, C. M.,Masedunskas, A., Amornphimoltham, P., Nathan, C. A., Singh,B., Weigert, R., Molinolo, A. A., and Gutkind, J. S. (2011)Decreased lymphangiogenesis and lymph node metastasis bymTOR inhibition in head and neck cancer. Cancer Res. 71,7103–7112

Received for publication July 8, 2014.Accepted for publication November 3, 2014.

SDF-1/CXCL12 INDUCES DIRECTIONAL CELL MIGRATION 13