Exosomes Induce Fibroblast Differentiation into Cancer … · Tumor Microenvironment Exosomes...

Tumor Microenvironment

Exosomes Induce Fibroblast Differentiationinto Cancer-Associated Fibroblasts throughTGFb SignalingCassandra Ringuette Goulet1,2,3, Genevi�eve Bernard1,2, Sarah Tremblay1,St�ephane Chabaud1,2, St�ephane Bolduc1,2, and Fr�ed�eric Pouliot2,3

Abstract

A particularly important tumor microenvironment relation-ship exists between cancer cells and surrounding stromal cells.Fibroblasts, in response to cancer cells, become activated andexhibit myofibroblastic characteristics that favor invasive growthand metastasis. However, the mechanism by which cancer cellspromote activation of healthy fibroblasts into cancer-associatedfibroblasts (CAF) is still not well understood. Exosomes arenanometer-sized vesicles that shuttle proteins and nucleic acidsbetween cells to establish intercellular communication. Here,bladder cancer–derived exosomes were investigated to determinetheir role in the activation of healthy primary vesical fibroblasts.Exosomes released by bladder cancer cells are internalized byfibroblasts and promoted the proliferation and expression of CAFmarkers. In addition, cancer cell–derived exosomes contain TGFb

and in exosome-induced CAFs SMAD-dependent signaling isactivated. Furthermore, TGFb inhibitors attenuated CAF markerexpression in healthy fibroblasts. Therefore, these data demon-strate that bladder cancer cells trigger the differentiation of fibro-blasts to CAFs by exosomes-mediated TGFb transfer and SMADpathway activation. Finally, exosomal TGFb localized inside thevesicle and contributes 53.4% to 86.3% of the total TGFb presentin the cancer cell supernatant. This studyhighlights anew functionfor bladder cancer exosomes as novel modulators of stromal celldifferentiation.

Implication: This study identifies exosomal TGFb as newmolecular mechanism involved in cancer-associated fibroblastactivation. Mol Cancer Res; 16(7); 1196–204. �2018 AACR.

IntroductionThe interactionbetween themicroenvironment and cancer cells

has a critical role in tumordevelopment. Tumor cells start to shapetheir microenvironment at an early phase of the malignantprogression by cell–cell interaction and paracrine mechanisms.They produce a variety of growth factors, chemokines, andmatrix-degrading enzymes, which in turn promote diverse events, such astumor angiogenesis, extracellular matrix remodeling, cellularproliferation, invasion, and metastasis, as well as mediatemechanisms that are therapeutically resistant (1). The tumormicroenvironment is composedof diverse typesof cells, includinginfiltrating immune cells, bloodand lymphatic vascular networks,and carcinoma-associated fibroblasts (CAF; ref. 2). CAFs consti-tute the majority of noncancerous cells within a tumor and sharemany similarities withmyofibroblasts found during wound heal-

ing. They express, among other things, a-smooth muscle actin(aSMA), fibroblast-activated protein (FAP), and Galectin (3, 4).CAFs arise from nearby healthy fibroblasts (HF) that have beendifferentiated under the tumor microenvironment conditions,and they contribute to the formation of cancer-altered stroma(5, 6). They create a microenvironment that promotes cancergrowth, invasiveness, vascularization, establishment of the pre-metastatic niche,metastasis, and therapy resistance (7–10).More-over, the presence of CAFs in vivo correlates with poor prognosis ofseveral cancer types (11–14). Therapeutics targeting CAF activa-tion may therefore be attractive, but the mechanisms responsiblefor the transformation of HFs into CAFs are not well understood.

There is a growing interest in the cancer cell–microenvironmentcommunication mediated by secreted vesicles termed exosomes(15). Exosomes are nanometer-sized vesicles secreted from theendosomal membrane compartment by diverse cell types (16).Exosomes carry multiple biologically active cargo components,such as proteins, nucleic acids, and lipids, which strongly varywith the cellular origin and the physiologic state of the cell thatsecretes the exosomes. They mediate cell-to-cell communicationand activate signaling pathways by acting as natural vehicles fordelivering their contents fromdonor cells to target cells. Exosomesmediate their biological functions, which include immunomo-dulation (17), cancer progression (18), and epigenetic reprogram-ming (19), through receptor-mediated endocytosis, pinocytosis,or phagocytosis, or else through fusion with the cell membraneresulting in direct release of contents into the cytoplasm (20).With our growing understanding of the biology of exosomes, thetheory of passive exosome loading, trafficking, and release is

1Centre de recherche en organog�en�ese exp�erimentale/LOEX, RegenerativeMedicine Division, CHU de Qu�ebec Research Center, Quebec, Canada. 2UrologyDivision, Department of Surgery, LavalUniversity, Quebec, Canada. 3Laboratoired'Urologie-Oncologie Exp�erimentale, Cancer Axis, CHU de Qu�ebec ResearchCenter, Quebec, Canada.

Note: Supplementary data for this article are available at Molecular CancerResearch Online (http://mcr.aacrjournals.org/).

Corresponding Author: Fr�ed�eric Pouliot, CHU de Quebec and LavalUniversity, Quebec City, Quebec G1J 1Z4, Canada. Phone: 418-525-4444;Fax: 418-691-5562; E-mail: [email protected]

doi: 10.1158/1541-7786.MCR-17-0784

�2018 American Association for Cancer Research.

MolecularCancerResearch

Mol Cancer Res; 16(7) July 20181196

on April 9, 2021. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst April 10, 2018; DOI: 10.1158/1541-7786.MCR-17-0784

http://crossmark.crossref.org/dialog/?doi=10.1158/1541-7786.MCR-17-0784&domain=pdf&date_stamp=2018-6-16

http://mcr.aacrjournals.org/

beginning to be replaced by the idea that these processes areorchestrated anddeliberate (21, 22). In cancer, emerging evidencesuggests that exosomes play a role in facilitating tumorigenesis byregulating angiogenesis, immunity, and metastasis (23–25). Veryfew studies have investigated the function of exosomes in thecontext of cancer-stroma activation (26, 27), and there remainmany unanswered questions about the factors, shuttled by exo-somes, involved in stromal activation.

In this study, we hypothesize that bladder cancer exosomes canalter the function of HFs surrounding the tumor, leading toalterations that may promote tumor growth and progression.Therefore, we examined whether exosomes produced by bladdercancer cells could transfer information toHFs and trigger a cellularresponse. We found that bladder cancer cells communicated withHFs through release of exosomes. We showed that HFs acquired aCAF phenotype following exposure to cancer-derived exosomesand that the differentiation ofHFs toCAFswas associatedwith theactivation of the TGFb/SMAD signaling pathway. We also foundthat exosomal TGFb, localized inside the vesicle, contributes for53.4% to 86.3% of the total TGFb present in the cancer cellsupernatant. Thus, this study identifies a new molecular mecha-nism involved in CAF activation.

Materials and MethodsEthics statement

Bladder biopsies from paediatric patients undergoing nonon-cologic urologic surgery were obtained at the CHU de Qu�ebecResearch Center in accordance with recognized ethical guidelinesof Declaration ofHelsinki andwith the institutional review boardof Laval University (Quebec, Canada). All patients provided theirformal, informed, written consent, each agreeing to supply abiopsy for this study.

Cell isolation and cultureHFs were isolated from two different human bladder biopsies

as previously described (28). Briefly, the stroma was separatedfrom the urothelium after incubation overnight at 4�C in HEPESbuffer with 500 mg/mL thermolysin (Sigma-Aldrich). The fibro-blasts were enzymatically dissociated from the extracellularmatrix by treating the stroma with 0.125 U/mL collagenase H(Roche) for 30 minutes at 37�C under gentle agitation. Then,fibroblasts were cultured in DMEM supplemented with 10% FBS(Invitrogen) and antibiotics (100 U/mL penicillin and 25 mg/mLgentamicin; Sigma-Aldrich). RT4, T24, and SW1710 bladdercancer cell lines were obtained from the ATCC (HTB-2, HTB-4)or DSMZ (ACC 426) and cultured in DMEM containing 10% FBSand antibiotics. All cells were cultured fewer than five passagesafter purchasing or receiving them for all the experiments andtested for mycoplasma contamination. Cell supernatant wasaliquoted and stored at �80�C until the ELISA assay.

Exosomes production and isolationTo remove any residual bovine exosomes from the FBS,

serum was treated with the FBS Exosome Depletion Kit (NorgenBiotek Corp.) according to the manufacturer's instructions.Bladder cancer cells were cultured in DMEM containing 10%exosome-depleted FBS for 48 hours. Then, the conditionedmedium was centrifuged to 2,000 � g for 30 minutes to removecells and debris, and supernatant was mixed with 0.5 volumesof the Total Exosome Isolation Reagent (Invitrogen). Samples

were mixed by vortexing and incubated at 4�C overnight. Then,they were centrifuged at 10,000 � g for 60 minutes at 4�C.Exosomes, contained in the pellet, were resuspended in PBS.The protein concentration was measured with a BCA Kit(Pierce; Thermo Fisher Scientific).

Electron microscopyIsolated exosomes were loaded onto copper grids covered with

a carbon film and stained with 3% uranyl acetate (Sigma). Thegridwas air-dried, and electronmicrographswere acquiredusing aJEOL, JEM-1230 (Tokyo, Japan) transmission electron micro-scope at 80 kV excitation voltage.

Sucrose density gradientsIsolated exosomes were placed on top of nine successive layers

(from 1.10 to 1.18 g/mL) of sucrose density gradient and cen-trifuged at 200,000 � g for 12 hours at 4�C. After centrifugation,series of fractions (1.0 mL each � 9) were collected, the proteinconcentration was measured by using a BCA Kit (Pierce; ThermoFisher Scientific), and samples were lysed with RIPA buffer. Thesesamples (10 mg) were separated by SDS-PAGE using a 12%polyacrylamide gel. Proteins were visualized by silver stainingusing a Silver Stain kit (Pierce; Thermo Fisher Scientific) accordingto the manufacturer's instructions.

Exosome size distribution and concentration measurementTo determine exosome size distribution and concentration,

semiautomated nanoparticle tracking-based analyses were per-formed using a NanoSight (NS300) apparatus (Malvern Instru-ments Ltd.). Samples were diluted to provide counts within thelinear range of the instrument (i.e., 3� 108 to 109 per mL). Threevideos of 1-minute duration were recorded for each sample,with a frame rate of 30 frames per second. Particle movementwas analyzed by NTA software (NTA 2.3; NanoSight Ltd.)according to the manufacturer's protocol. The NTA softwarewas optimized to first identify and then track each particle on aframe-by-frame basis.

Exosome uptake by fibroblastsExosomes were prelabeled with a PKH26 Cell Linker Kit

(Sigma-Aldrich) according to the manufacturer's instructions.Briefly, exosomes were resuspended in 1 mL Diluent C, and 4 mLPKH26 was mixed with 1 mL Diluent C separately. The exosomesuspension and the PKH26 solution were mixed and incubatedfor 4 minutes. The labeling reaction was stopped by adding anequal volume of 1% FBS. To remove the extra dye, samples wereultracentrifuged at 4,000 � g, washed with PBS, and centrifugedagain at 4,000 � g in Vivaspin 2 100 kDa MWCO spin columns(GE Healthcare). Exosomes were resuspended in 1 mL of PBS.PKH26-labeled exosomes (1mg/mL)were incubated for 24hourswith HFs grown on coverslips to 50% confluence in a 12-wellplate. The coverslips were then washed three times with PBSsolution and fixed with 4% PFA for 20 minutes. The HFs werelabeled with Vimentin (1/1,000; Abcam) primary antibody dilut-ed in PBS containing 1% (w/v) BSA (Sigma-Aldrich) for 60minutes at room temperature, rinsed in PBS, and then incubatedfor 45 minutes with Alexa Fluor 488–conjugated secondary anti-bodies (1:1,000; Invitrogen). The nuclei were stained by usingHoechst 33258 (5 mg/mL; Sigma-Aldrich), and samples wereobserved using a Zeiss Axio ImagerM2microscope equippedwithan Axiocam HR Rev3 camera.

Cancer Exosomes Induce Fibroblast Differentiation into CAFs

www.aacrjournals.org Mol Cancer Res; 16(7) July 2018 1197




CAF inductionHFswere cocultured with freshly isolated exosomes (1mg/mL)

for 48 hours inDMEM10%FBS. rhTGFb1 (4 ng/mL; Proteintech)was used as a positive control. Then, proteins were collected inRIPA buffer containing protease inhibitors cOmplete (Roche),quantified with a BCA Kit, and stored at �80�C until performingWestern blot assay. For neutralization of TGFb, exosomes werepreincubated with 10 mg/mL of human TGFb 1,2,3 antibody(R&D Systems) for 1 hour at room temperature before they wereadded to the culture medium.

ImmunocytochemistryHFswere cocultured with freshly isolated exosomes (1mg/mL)

for 48 hours in DMEM 10% FBS. rhTGFb1 (4 ng/mL) was used asa positive control. The cells were then washed with PBS andfixed with 4% (v/v) paraformaldehyde/PBS for 30 minutes. Forimmunocytochemistry, permeabilization was performed using0.25% (v/v) Triton X-100/PBS (Sigma-Aldrich) for 30 minutesfollowed by blocking using PBS 5% BSA for 1 hour. Incubationwith primary antibodyaSMA (1/1,000; Abcam)was performed atroom temperature for 1 hour in PBS 5% BSA. The cells wererinsed in PBS before incubation with anti-rabbit Alexa Fluor488–conjugated secondary antibody (1/1,000; Invitrogen) for1 hour. Cells were then mounted and counterstained usingVECTASHIELD mounting medium with 4',6-diamidino-2-phe-nyl-indole (DAPI; Vector Laboratories).

Cell proliferation assayTo assess the proliferation of HFs in the presence of bladder

cancer-derived exosomes, the CyQUANT NF assay (InvitrogenLtd.) was used according to the manufacturer's instructions.

HFs were plated at 5,000 cells per well in DMEM in 96-wellwhite walled clear bottomed plates (Corning Life Sciences) andincubated with 1 mg/mL of exosomes overnight at 37�C in 5%CO2. The following day, the cells were incubated with �1CyQUANT NF dye binding solution for 60minutes at 37�C,and the fluorescence intensity was measured at excitationwavelength of 485 nm and the emission detection at 530 nmusing Varioskan Flash microplate reader (Thermo ElectronCorporation).

ImmunoblotEqual amounts of proteins (10 mg) were loaded in 12%

polyacrylamide gels, resolved by SDS-PAGE, and transferred topolyvinylidene difluoride membranes. The membranes wereblocked for 30minuteswith 5%nonfatmilk and 0.05%Tween 20in PBS. The membranes were incubated with primary antibodiesovernight at 4�C followed by 45 minutes at room temperaturewith HRP-conjugated secondary antibodies (Jackson Immunor-esearch Laboratories). Protein expression was detected usingAmersham ECL Western Blotting Detection Reagent (GE Health-care). Bands were imaged by using Fusion Fx7 imager (VilberLourmat) and analyzed with ImageJ software (NIH, Bethesda,MD). The following antibodies were used: CD9 (1/1,000; Biole-gend), CD63 (1/1,000; Biolegend), CD81 (1/1,000; Biolegend),Flotilin (1/1,000; Proteintech), Alix (1/1,000; ProteinTech), Hsp70(1/1,000; Biolegend), GM130 (1/1,000; BD Biosciences),b-actin (1/5,000; Abcam), CK18 (1/50; Life Technologies),aSMA (1/5,000; Abcam), Galectin (1/1,000; R&D Systems),FAP (1/1,000; Novus Biologicals), pSMAD2 (1/500; Millipore),SMAD2/3 (1/1,000; Millipore), and Tubulin (1/1,000; NovusBiologicals).

A CRT4 kDa T24 SW1710 RT4 T24 SW1710 FBS

Exosomes Cell lysates

CD81 25

CD63

75

50

37

45 CK18

Flotillin 47

GM130130

Alix 95

Actin 42

CD9 25

Hsp70 70

BDensity (g/mL)

Fraction 9 8 6 5 4 3 2 1 7 9 6 5 4 3 2 1 9 8 6 5 4 3 2 1 7 8 7

SW1710 T24 RT4

34

43

5572

130

SW1710 T24 RT4

Figure 1.

Bladder cancer–derived exosomes characterization. A, TEM micrographs showing morphology of exosomes immunoprecipitated with anti-CD9 mAb fromRT4, T24, and SW1710 bladder cancer cells. Exosomes were stained with 2% uracyl acetate after being placed on carbon-coated TEM grid. B, Silver staining offractions resulted from sucrose density-gradient centrifugation of the crude exosome pellet. C, Protein blot of cell lysate, FBS, and exosomes samples forcommon exosome markers (Alix, HSP70, Flotillin, CD9, CD63, and CD81) and cell markers (GM130, Actin, CK18).

Ringuette Goulet et al.

Mol Cancer Res; 16(7) July 2018 Molecular Cancer Research1198




Proteinase K treatmentExosomes were permeabilized or not with 1% Triton X-100

and then incubated for 60 minutes at 37�C in the presence of100 mg/mL proteinase K (Invitrogen). Protease K was quenchedwith 2 mmol/L phenylmethylsulfonyl fluoride. Equivalentamounts of all samples were resolved with SDS-PAGE and sub-jected to Western blotting with a TGFb antibody (1/500; R&DSystems). Results were confirmed by using an ELISA according tothe manufacturer's protocol (Duoset; R&D Systems). The plateswere read at 450 nm using a SpectraMax Plus spectrometer withSoftmaxPro Ver 4.7.1.

Cycloheximide treatmentFor protein synthesis inhibition, HFs were incubated with

cycloheximide (100 mg/mL; Sigma-Aldrich) 1 hour before exo-some treatment. Then, HFs were cocultured with freshly isolatedexosomes (1mg/mL) for 48 hours in DMEM10% FBS containing500 mg/mL cycloheximide.

Statistical analysisFor graphical representation of data and statistical analysis,

GraphPad Prism was used. The results were expressed as mean �SE. Differences between the groups were considered significant atP < 0.05. Data were interpreted using one- or two-way ANOVA.

ResultsIsolation and characterization of bladder cancer cell–derivedexosomes

Exosomes were isolated from superficial RT4 or invasive T24and SW1710 bladder cancer cell supernatants. As shown in Fig.1A, the exosomes had a vesicle shape limited by a lipidmembranewith a diameter ranging from 30 to 100 nm. To examine theprotein heterogeneity, the exosomes derived from bladder cancercells were analyzed using sucrose density gradients (Fig. 1B). Theexosomes were purified by differential centrifugation, overlaid ona 0.2–1.8 g/mL sucrose density gradient, and centrifuged for 12hours at 200,000 � g. Gradient fractions were collected andanalyzed by silver staining. Then, the expression of exosomalmarkers CD9, CD63, CD81, Alix, Flotillin, and Hsp70, as well asthat of cellmarkers GM130, Actin, andCK18,was characterized inboth cell lysates and isolated exosomes by Western blotting (Fig.1C). The exosomes of each cancer cell line were quantified bynanoparticle tracking analysis (Fig. 2A–C). This revealed a typi-cally heterogeneous exosomal population with size rangingbetween 30 and 450 nm in diameter; themost enriched exosomeswere smaller than 200 nm (Fig. 2D). The isolation protocolretrieved 1.3 to 4.3 � 109 exosomes/mL (Fig. 2D). Collectively,these results suggest that our isolated exosomes had a high degree

A

C

B

SW1710-derivedexosomes

Mean 144.4 ± 1.0 nm

Mode 109.3 ± 9.3 nm

SD 65.8 ± 2.6 nm

D10 91.5 ± 1.0 nm

D50 122.0 ± 2.1 nm

D90 214.2 ± 2.3 nm

Particles/mL 4.3x108 ± 2.3x107

RT4-derivedexosomes

Mean 143.0 ± 2.0 nm

Mode 119.3 ± 12.9 nm

SD 56.7 ± 4.9 nm

D10 86.8 ± 2.6 nm

D50 131.4 ± 2.9 nm

D90 206.3 ± 4.7 nm


T24-derivedexosomes

Mean 182.5 ± 1.6 nm

Mode 151.1 ± 8.3 nm

SD 59.5 ± 1.1 nm

D10 120.0 ± 0.6 nm

D50 164.9 ± 1.7 nm

D90 253.1 ± 8.5 nm


Con

cent

ratio

n (p

artic

les/

mL)

Con

cent

ratio

n (p

artic

les/

mL)

Con

cent

ratio

n (p

artic

les/

mL)

Figure 2.

Particle size distributions and concentration of isolated exosomes. NanoSight analysis shows that RT4 (A), T24 (B), and SW1710-derived vesicles' isolation (C) delivergood yields of particles whose sizes are consistent with exosomes.






of purity and satisfied the size, density, structure, and molecularphenotype of exosome vesicles as described (29, 30).

Bladder cancer cell–derived exosomes are internalized by HFsand promote their proliferation

To investigate the internalization of exosomes by HFs, welabeled the exosomes with PKH26. As shown in Fig. 3A, bladdercancer–derived exosomes were internalized and accumulatedaround the nuclei inHFs after incubation for 24 hours. To analyzewhether the proliferation ability of HFs was affected by bladdercancer–derived exosomes, HFs were cultured with 1 mg/mL ofexosomes for 18 hours. As shown in Fig. 3B, cancer-derivedexosomes promoted the proliferation of HFs.

Differentiation of HFs to CAFs induced by bladder cancer cell–derived exosomes

The role of cancer exosomes in the differentiation of HFsto CAFs was investigated. To do so, an HF monolayer was

cultured in DMEM containing rhTGFb (4 ng/mL) or RT4, T24,or SW1710 cell–derived exosomes (1 mg/mL) for 48 hours. Itis well established that CAFs share key phenotypic markerswith myofibroblasts (6). Using immunocytochemistry, wevisualized the expression of aSMA in cells; as expected, rhTGFband exosome treatments strongly induced aSMA expression(Fig. 3C). Moreover, compared with untreated HFs, rhTGFband exosome treatments induced morphologic changes wherecells exhibited the classical stress fiber structures of aSMA andthe mechanical tensions typical of myofibroblasts. As CAFswere defined as the increased expression of not only aSMA butalso FAP and Galectin proteins, we quantified their expressionwith a Western blot analysis. Analyses showed that, comparedwith the untreated group, treatment with cancer exosomesresulted in increases in aSMA, FAP, and Galectin protein levels(Fig. 3D–E). Taken together, these results indicate that HFsundergo differentiation to CAFs in response to cancer exosomeexposure.

Vimentin PKH26 A

D

B

C

PKH26 Vimentin

HF

αSMA

HF + TGFβ 4 ng/mL

HF + Exosomes 1 mg/mL

E

αSMA

Tubulin

FAP

Galectin

25

50

75

100

125

αSM

A/T

ubul

in ****

* ****

HF

HF + TGFβ 4 ng/mL

HF + RT4-derived exosomes

HF + T24-derived exosomes

HF + SW1710-derived exosomes

0

50

100

150

200

250

Gal

ectin

/Tub

ulin

* * *

0

10

20

30

40R

FU

*

****

*

HF




0

50

100

150

200

FAP

/Tub

ulin

****

n.s.

n.s.n.s.

Figure 3.

Cancer-derived exosomes penetrate into HF, drive their proliferation and induce CAF markers expression. A, Fluorescence microscopy images demonstratedcolocalization of PKH26-labeled cancer exosomes (red; left) with primary vesical fibroblasts cells labeled with vimentin antibody (green; middle). Cellnuclei are labeled with Hoechst nuclear stain (blue). Exosomes was localized around the nucleus and in the lamellipodia of the fibroblasts. Scale bar ¼ 100 nm.B, HF were treated with 1 mg/mL of RT4, T24, or SW1710 cell–derived exosomes for 48 hours. Cell proliferation was evaluated with cyQUANT Kit (C) and HF wereexamined by immunocytochemistry for the expression of aSMA (scale bar ¼ 100 nm; D) or by Western blot analysis for the expression of aSMA, FAP, andGalectin proteins. E, Density analysis of Western blotting bands. rhTGFb1 (4 ng/mL) served as a positive control. Graphs show mean � SD. The difference betweengroups was analyzed by one-way ANOVA followed by post hoc analysis using Dunnett multiple comparison tests. � , P < 0.05; �� , P < 0.01; �� , P < 0.0001compared with the HF control group (n ¼ 6–9).






The role of TGFb pathways in the differentiation of HFs to CAFsTGFb has been shown to be involved in CAF differentiation in

breast and prostate cancer (31, 32). Therefore, we investigatedwhether cancer exosomes can trigger TGFb signaling pathwaysand thereafter initiate a process of differentiation of HFs toward aCAF phenotype. We first demonstrated the presence of TGFb inbladder cancer exosomes (Fig. 4A). Surprisingly, T24 cell line–derived exosomes showed significantly higher levels of TGFbthan RT4 and SW1710 cell line–derived exosomes. TGFb issynthesized in a latent complex with its prodomain, the laten-cy-associated peptide (LAP). To exert biological activity, TGFbrequires activation, which involves dissociation of LAP and TGFb.To determine the bioactive exosomal TGFb levels, we measuredby ELISA active (not acid-treated) and total (acid-treated) TGFb inour exosome preparations. As expected, we find that approxi-mately 2%–3% of the exosomally associated TGFb is active (Fig.4A). We next determine the proportion of secreted TGFb associ-ated with exosomes by comparing TGFb levels found in bladdercancer cell supernatant and supernatant depleted in exosomes.Using ELISA, data revealed that from 53.4% to 86.3% of super-natant TGFb is associated with exosomes (Fig. 4B).

TGFb binds sequentially to type II receptor and then type Ireceptor, which triggers a signaling cascade, involving phosphor-ylation of SMAD2 and 3, which subsequently mediate binding toSMAD4. This SMAD complex translocates to the nucleus toinitiate transcription of a host of genes, including aSMA (33).

Therefore, to assess whether the differentiation of HF to CAF isassociated with TGFb pathway activation, we examined the levelof phosphorylated SMAD2 1, 4, and 48 hours after bladder cancerexosomes treatment (Supplementary Fig. S1). Typical kineticfor TGFb1-mediated Smad2 phosphorylation is on the order of30 to 60 minutes consistent with ligand-mediated activation of areceptor kinase. Thus, the absence of SMAD2 activation in ourexosome-treated HF suggests that cells have to process cancer-derived exosomes to release exosomal TGFb and then induceSMAD2 signalization. To eliminate the TGFb that could beproduced by HFs, we blocked protein synthesis in HFs by addingcycloheximide. The results showed that SMAD2 phosphorylationwas increased by cancer exosomes in rhTGFb and RT4, T24, andSW1710 cell–derived exosomes treatments even in cycloheximidecondition (Fig. 4C).

To demonstrate that TGFb signaling shuttled by cancer exo-somes is directly involved in the differentiationofHFs toCAFs,weblocked TGFb signaling with TGFb1,2,3 neutralization antibody(10 ng/mL) followed by cancer exosome treatment for 48 hours.As shown in Fig. 4D, TGFb1,2,3 neutralization reversed theexpression of the aSMA CAF marker in HFs treated with rhTGFband cancer exosomes.

Localization of exosomal TGFbAn issue that remains unknown is whether TGFb is associated

with exosomes. Proteins may be bound to the surface membrane

SW1710T24RT40

5,000

10,000

15,000

20,000

SW1710T24RT40

2,000

4,000

6,000

Supernatant

Exosome

Exosome-depleted supernatant

0

50

100

150

αSM

A/T

ubul

in

DC

53.4%

86.3%

65.3%

Tubulin

αSMA

HF

HF + TGFβ 4 ng/mL




TGFβ1,2,3 neutralizing antibody (10 ng/mL) -- ++++ ---

A

SW1710T24RT4012345

50

75

100

125

Active

Total

B

pSMAD2

SMAD 2/3

Tubulin

+ CycloheximideUntreated(500 μg/mL)

Exos

omal

TG

Fβ1

(pg/

μg o

f exo

som

es)

% o

f TG

Fβ1

TGFβ

1 (p

g/m

L)

Figure 4.

Bladder cancer cell–derived exosomes contain TGFb1, which induce SMAD2 phosphorylation in HF. A, ELISA quantification of TGFb1 expression in bladdercancer-derived exosomes. Percent of active TGFb1 was measured by ELISA, without (active) and with (total) prior activation with acid (HCL 1N), showing thatexosomal TGFb is mainly in a latent form. B, TGFb ELISA was performed on supernatant, exosomes, and exosome-depleted supernatant, revealing theproportion of TGFb1 in supernatant associated to exosomes. C, HF were treated with cancer-derived exosomes (1 mg/mL) in presence or absence of 500 mg/mLcycloheximide for 48 hours. rhTGFb (4 ng/mL) served as a positive control. The levels of pSMAD2 were analyzed by Western blotting. D, HF were treatedwith rhTGFb (4 ng/mL) or bladder cancer cell–derived exosomes (1 mg/mL) in the presence or absence of TGFb1,2,3 neutralizing antibody (10 ng/mL) for48 hours. The levels of aSMA were analyzed by Western blotting.






of the exosome or located inside the vesicle. Because proteinase Kcannot penetrate through the lipid bilayer, it is most likely thatproteins susceptible to proteinase K degradation are localized onthe surface of vesicle. To clarify the localization of TGFb, cancer-derived exosomes were treated with proteinase K (100 mg/mL) inthe presence or absence of 1% Triton X-100, which was used topermeabilize the exosomal membranes. Western blot analysisdata revealed that the proteinase K treatment alone did not resultin the degradation of TGFb. However, we observed a shift ofTGFb molecular weight, suggesting that TGFb is associated withother proteins. With the addition of 1% Triton X-100, theexosomal membrane was disrupted, resulting in the completedigestion of the TGFb (Fig. 5A). To validate this result, TGFblevels were quantified using ELISA (Fig. 5B). The exosomalTGFb level was not significantly reduced by treatment withproteinase K, whereas the addition of Triton X-100 in combi-nation with proteinase K degraded over 90% of the TGFb. Todetermine if the small portion of external exosomal TGFb issufficient to induce CAF differentiation, we assessed the capac-ity of proteinase K treated exosomes to trigger differentiation ofHFs. As expected, proteinase K–treated exosomes maintained

their ability to differentiate HFs to CAFs (Fig. 5C). Collectively,these data indicate that the majority of TGFb proteins arelocalized inside exosomes.

DiscussionInteractions between cancer exosomes and the tumor micro-

environment have been the subject of renewed interest in recentyears. It is becoming increasingly evident that exosomes derivedfrom cancer cells impact the local stroma, driving production of adisease-associated microenvironment. Although currently suchdata are lacking for bladder cancer, studies have shown that gastricand colon cancer exosomes trigger the differentiation of stromalcells into CAFs (26, 34, 35). The main goal of this study was toinvestigate the effect of exosomes isolated from three differentbladder cancer cell lines on primary vesical HFs.

Studies on extracellular vesicles require special attention toexperimental design to ensure the specific isolation of exosomesand avoid cellular contaminants. In this work, we used an isola-tion method based on sedimentation reagents. Globally, wedemonstrate that vesicles isolated in RT4, T24, and SW1710

Figure 5.

Exosomal localization of TGFb. Exosomes were treated with proteinase K (100 mg/mL) or proteinase K-1% Triton X-100. A, Equivalent amounts of all sampleswere resolved with SDS-PAGE and subjected to Western blotting with TGFb antibody. B, Samples were also quantified by ELISA. Proteinase K treatmentdid not result in the degradation of TGFb unless Triton X-100 was added to permeabilize the exosome membrane. C, HF were treated with rhTGFb (4 ng/mL)or 1 mg/mL of exosomes treated or not with proteinase K (100 mg/mL) for 48 hours. The expression of aSMA were analyzed by Western blotting. Thedifference between groups was analyzed by two-way ANOVA followed by post hoc analysis using Tukey multiple comparison tests. Graphs show mean � SD.� , P < 0.05; �� , P < 0.01; ��, P < 0.0001 (n ¼ 3).






bladder cancer cell lines possess similar characteristics to exo-somes, including density, protein expression, and size distri-bution. NanoSight analyses reveal that our exosomes are quitepure, because only a low concentration of the isolated vesiclespresents sizes greater than 200 nm. Obviously, we cannot avoidpossible cellular contaminants, but these larger vesicles mayalso be due to exosome aggregates, as seen in electronicmicroscopy. These results are broadly consistent with similarstudies of exosomes from other cellular sources (29, 30).Interestingly, we noticed a great heterogeneity regarding theexpression level of exosomal protein. Indeed, althoughSW1710-derived exosomes express CD9, CD63, and Alix pro-teins, the expression level is much lower than that of RT4-derived exosomes. Moreover, the absence of Golgi apparatus(GM130) or cytoskeleton (Actin, CK18) markers reveals thatour method allows specific and effective exosome isolation.

The uptake of these exosomes by vesical HFs was observed byimmunofluorescence staining, confirming direct interactionbetween cancer exosomes and HFs. Cancer exosome treatmentbrings about sustained changes in the morphology, proteinexpression, and metabolism of HFs. Indeed, we observed anaugmentation of the proliferation rate and a dramatic upregu-lation of aSMA stress fibers, resulting in the characteristicspindle-shaped CAF morphology. Based on Western blot anal-ysis data, several CAF markers were overexpressed after exo-some treatment, including aSMA, FAP, and Galectin, suggestingthat bladder cancer cell-derived exosomes trigger differentia-tion of HFs to CAFs.

TGFb signaling is one of the major pathways controllingtumorigenesis not only in cancer cells but also in CAFs. Previousstudies showed evidence that TGFb is involved in the alteration oftumor stroma, notably by contributing to the acquisition of a CAFphenotype in stromal cells (32, 36, 37). Our data are consistentwith such findings, as exosomal TGFb, contributing for 53.4% to86.3% of the total TGFb present in the cancer cell supernatant,mediates SMAD pathway activation and drives the differentiationof HFs to CAFs. Moreover, TGFb neutralization renders cancerexosomes unable to trigger fibroblast differentiation, suggestingthat TGFb is a key regulator in HF-CAF transition.

A large proportion of exosome surface-associated proteins aretransmembrane proteins, but some of those may be non-cova-lently bound to the vesicle membrane and could thus be travelingwith the exosome between cells (38). Although the precise mech-anism of how exosomal TGFb is delivered to the recipient cell isnot known, its associationwith the outer exosomemembrane canbe evaluated by its susceptibility to digestion by an extracellularprotease. Interestingly, a Western blot analysis of proteinase K-treated exosomes revealed a shift in TGFbmolecular weight from�40 to �15 kDa. This finding led us to hypothesize that intra-vesicular TGFb is associated with transmembrane protein and isliberated when this protein is digested by proteinase K. We alsofound that TGFb levels were significantly affected by proteinase Ktreatment only in the presence of Triton X-100, demonstratingthat TGFb is localized inside exosomes. These results raise impor-tant questions of whether exosomal TGFb reaches HF TGFb

receptors and activates Smad signaling. Because TGFb neutraliz-ing antibodies reduced the expression levels ofaSMA, the effect ofexosomes should bemediated by TGFb released into themedium.We confirmedwith cycloheximide-treatedHF that Smad signalingactivation involve only exosomal TGFb and not cell-secretedTGFb. This result suggests that exosomal TGFb, which is localizedinside exosome, is released from exosomes prior to bind TGFbreceptors at the surface of HF. We could hypothesize that theexosomes burst in the extracellular media prior their internaliza-tion, releasing their TGFb, or even that the content of internalizedexosomes is recycled by endosomes and is then secreted byexocytosis, releasing TGFb in the extracellular media.

Taken together, our data show exosomes to be an additionalmechanism contributing to microenvironment modulation bybladder cancer cells. Our results give a new insight into the role ofexosomes in the differentiation of fibroblasts into CAFs. Thisstudy provides a better understanding of whether exosomescontribute to the establishment and persistence of cancer-alteredstroma, and it may offer a novel therapeutic modality to targetcancer growth and progression.

Disclosure of Potential Conflicts of InterestFrederic Pouliot has received speakers bureau honoraria from Sanofi,

Genzyme, Amgen, Astellas, and Janssen; is a consultant/advisory boardmemberof Sanofi, Abbvie, Astellas, Janssen, Genzyme, and Roche. No potential conflictsof interest were disclosed by the other authors.

Authors' ContributionsConception and design: C. Ringuette Goulet, S. Chabaud, S. Bolduc, F. PouliotDevelopment of methodology: C. Ringuette Goulet, S. Bolduc, F. PouliotAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): C. Ringuette Goulet, G. Bernard,Analysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): C. Ringuette Goulet, S. Chabaud,Writing, review, and/or revision of the manuscript: C. Ringuette Goulet,S. Chabaud, S. Bolduc, F. PouliotAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases):Study supervision: C. Ringuette Goulet, S. Bolduc

AcknowledgmentsThis work was supported by a CUOG/Bladder Cancer Canada Grant to

F. Pouliot and C. Ringuette Goulet and a Ferring Innovation Grant to S. Bolduc.C. Ringuette Goulet is the recipient of an FRQS Doctoral Research Award. FP isthe recipient of a FRQSClinician-Scientist Scholarship. S. Bolduc is the recipientof a Canadian Urological Association Scholarship and a Canadian Institute forHealth Research Grant (#258229).

The authors thank Bastien Par�e and Isabelle Lorthois for their criticalcomments and experimental input, Richard Janvier (IBIS) for help with electronmicroscopy, Dr. Denis Boudreau and J�er�emie Asselin (COPL) for their Nano-Sight technical assistance.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received December 20, 2017; revised February 19, 2018; accepted March 26,2018; published first April 10, 2018.

References1. Catalano V, TurdoA,Di Franco S,Dieli F, TodaroM, Stassi G. Tumor and its

microenvironment: a synergistic interplay. Semin Cancer Biol 2013;23:522–32.

2. BhomeR, BullockMD,Al SaihatiHA,GohRW,Primrose JN, SayanAE, et al.A top-down view of the tumor microenvironment: structure, cells andsignaling. Front Cell Dev Biol 2015;3:33.






3. Tang D, Gao J, Wang S, Ye N, Chong Y, Huang Y, et al. Cancer-associatedfibroblasts promote angiogenesis in gastric cancer through galectin-1expression. Tumor Biol 2015:1–11.

4. Huang Y, Zhou S, Huang Y, Zheng D, Mao Q, He J, et al. Isolation offibroblast-activation protein-specific cancer-associated fibroblasts. BioMedRes Int 2017;2017:4825108.

5. Gascard P, Tlsty TD. Carcinoma-associated fibroblasts: orchestrating thecomposition of malignancy. Genes Dev 2016;30:1002–19.

6. Kalluri R. The biology and function of fibroblasts in cancer. Nat Rev Cancer2016;16:582–98.

7. Zhuang J, Lu Q, Shen B, Huang X, Shen L, Zheng X, et al. TGFb1 secretedby cancer-associated fibroblasts induces epithelial-mesenchymal tran-sition of bladder cancer cells through lncRNA-ZEB2NAT. Sci Rep 2015;5:11924–13.

8. Attieh Y, Clark AG, Grass C, Richon S, Pocard M, Mariani P, et al. Cancer-associated fibroblasts lead tumor invasion through integrin-b3–dependentfibronectin assembly. J Cell Biol 2017;216:3509–20.

9. Chan JSK, SngMK, TeoZQ,ChongHC, Twang JS, TanNS. Targeting nuclearreceptors in cancer-associated fibroblasts as concurrent therapy to inhibitdevelopment of chemoresistant tumors. Oncogene 2018;37:160–73.

10. Liu J, Chen S, Wang W, Ning BF, Chen F, Shen W, et al. Cancer-associatedfibroblasts promote hepatocellular carcinoma metastasis through chemo-kine-activated hedgehog and TGF-b pathways. Cancer Lett 2016;379:49–59.

11. Zhi K, Shen X, Zhang H, Bi J. Cancer-associated fibroblasts are positivelycorrelated with metastatic potential of human gastric cancers. J Exp ClinCancer Res 2010;29:66–8.

12. Ha SY, Yeo SY, Xuan YH, Kim SH. The prognostic significance of cancer-associated fibroblasts in esophageal squamous cell carcinoma. PLoS One2014;9:e99955.

13. Cheng Y, Wang K, MaW, Zhang X, Song Y, Wang J, et al. Cancer-associatedbroblasts are associated with poor prognosis in esophageal squamous cellcarcinoma after surgery. Int J Clin Exp Med 2015;8:1896–903.

14. Lau EY, Lo J, Cheng BYL, Ma MK, Lee JM, Ng JK, et al. Cancer-associatedfibroblasts regulate tumor-initiating cell plasticity in hepatocellular carci-noma through c-Met/FRA1/HEY1 signaling. Cell Rep 2016;15:1175–89.

15. Milane L, Singh A, Mattheolabakis G, Suresh M, Amiji MM. Exosomemediated communication within the tumor microenvironment. J ControlRel 2015;219:278–94.

16. Cocucci E, Meldolesi J. Ectosomes and exosomes: shedding the confusionbetween extracellular vesicles. Trends Cell Biol 2015;7:42996.

17. Wei G, Jie Y, Haibo L, Chaoneng W, Dong H, Jianbing Z, et al. Dendriticcells derived exosomesmigration to spleen and induction of inflammationare regulated by CCR7. Sci Rep 2017;7:42996.

18. Boelens MC, Wu TJ, Nabet BY, Xu B, Qiu Y, Yoon T, et al. Exosome transferfrom stromal to breast cancer cells regulates therapy resistance pathways.Cell 2014;159:499–513.

19. Chen L, Charrier A, Zhou Y, Chen R, Yu B, Agarwal K, et al. Epigeneticregulation of connective tissue growth factor byMicroRNA-214 delivery inexosomes from mouse or human hepatic stellate cells. Hepatology2014;59:1118–29.

20. Minciacchi VR, Freeman MR, Di Vizio D. Extracellular vesicles in cancer:exosomes, microvesicles and the emerging role of large oncosomes. SeminCell Develop Biol 2016;40:41–51.

21. Colombo M, Moita C, van Niel G, Kowal J, Vigneron J, Benaroch P, et al.Analysis of ESCRT functions in exosome biogenesis, composition and

secretion highlights the heterogeneity of extracellular vesicles. J Cell Sci2013;126:5553–65.

22. Villarroya-Beltri C, Baixauli F, Gutierrez-Vazquez C, Sanchez-Madrid F,Mittelbrunn M. Sorting it out: regulation of exosome loading. SeminCancer Biol 2014;28:3–13.

23. Umezu T, Ohyashiki K, Kuroda M, Ohyashiki JH. Leukemia cell toendothelial cell communication via exosomal miRNAs. Oncogene2012;32:2747–55.

24. Zhou W, Fong MY, Min Y, Somlo G, Liu L, Palomares MR, et al. Cancer-secreted miR-105 destroys vascular endothelial barriers to promote metas-tasis. Cancer Cell 2014;25:501–15.

25. Greening DW, Gopal SK, Xu R, Simpson RJ, Chen W. Exosomes and theirroles in immune regulation and cancer. Semin Cell Develop Biol 2015;40:72–81.

26. Gu J, Qian H, Shen L, Zhang X, Zhu W, Huang L, et al. Gastric cancerexosomes trigger differentiation of umbilical cord derived mesenchymalstem cells to carcinoma-associated fibroblasts through TGF-b/Smadpathway. PLoS One 2012;7:e52465.

27. Webber JP, Spary LK, Sanders AJ, Chowdhury R, Jiang WG, Steadman R,et al. Differentiation of tumour-promoting stromal myofibroblasts bycancer exosomes. Nature 2014;34:319–31.

28. Ringuette-Goulet C, Bernard G, Chabaud S, Couture A, Langlois A, NeveuB, et al. Tissue-engineered human 3Dmodel of bladder cancer for invasionstudy and drug discovery. Biomaterials 2017;145:233–41.

29. Webber J, Clayton A. How pure are your vesicles? J Extracell Vesicles 2013;2:19861–6.

30. Schageman J, Zeringer E, Li M, Barta T, Lea K, Gu J, et al. The completeexosome workflow solution: from isolation to characterization of RNAcargo. BioMed Res Int 2013;2013:253957.

31. Casey TM, Eneman J, Crocker A,White J, Tessitore J, StanleyM, et al. Cancerassociated fibroblasts stimulated by transforming growth factor beta1(TGF-b1) increase invasion rate of tumor cells: a population study. BreastCancer Res Treat 2008;110:39–49.

32. Shangguan L, Ti X, KrauseU,Hai B, Zhao Y, Yang Z, et al. Inhibition of TGF-b/Smad Signaling by BAMBI blocks differentiation of human mesenchy-mal stem cells to carcinoma-associated fibroblasts and abolishes theirprotumor effects. Stem Cells. 2012;30:2810–9.

33. Derynck R, Zhang YE. Smad-dependent and Smad-independent pathwaysin TGF-b family signalling. Nature 2003;425:577–84.

34. Webber J, Steadman R, Mason MD, Tabi Z, Clayton A. Cancer exosomestrigger fibroblast to myofibroblast differentiation. Cancer Res 2010;70:9621–30.

35. Chowdhury R, Webber JP, Gurney M, Mason MD, Tabi Z, Clayton A.Cancer exosomes trigger mesenchymal stem cell differentiation intopro-angiogenic and pro-invasive myofibroblasts. Oncotarget 2015;6:715–31.

36. Li Q, Zhang D, Wang Y, Sun P, Hou X, Larner J, et al. MiR-21/Smad 7signaling determines TGF-b1-induced CAF formation. Sci Rep 2013;3:2038.

37. Beach JA, Aspuria PJ, Cheon DJ, Lawrenson K, Agadjanian H, Walsh CS,et al. Sphingosine kinase 1 is required for TGF-b mediated broblast- to-myofibroblast differentiation in ovarian cancer. Oncotarget 2016;7:4167–82.

38. Kalra H, Simpson RJ, Ji H, Aikawa E, Altevoqt P, Askenase P, et al.Vesiclepedia: a compendium for extracellular vesicles with continuouscommunity annotation. PLoS Biol 2012;10:e1001450.






2018;16:1196-1204. Published OnlineFirst April 10, 2018.Mol Cancer Res Cassandra Ringuette Goulet, Geneviève Bernard, Sarah Tremblay, et al.

SignalingβFibroblasts through TGFExosomes Induce Fibroblast Differentiation into Cancer-Associated

Updated version

10.1158/1541-7786.MCR-17-0784doi:

Access the most recent version of this article at:

Material

Supplementary

http://mcr.aacrjournals.org/content/suppl/2018/04/10/1541-7786.MCR-17-0784.DC1

Access the most recent supplemental material at:

Cited articles

http://mcr.aacrjournals.org/content/16/7/1196.full#ref-list-1

This article cites 33 articles, 4 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/16/7/1196.full#related-urls

This article has been cited by 1 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://mcr.aacrjournals.org/content/16/7/1196To request permission to re-use all or part of this article, use this link



http://mcr.aacrjournals.org/lookup/doi/10.1158/1541-7786.MCR-17-0784

http://mcr.aacrjournals.org/content/suppl/2018/04/10/1541-7786.MCR-17-0784.DC1

http://mcr.aacrjournals.org/content/16/7/1196.full#ref-list-1

http://mcr.aacrjournals.org/content/16/7/1196.full#related-urls

http://mcr.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://mcr.aacrjournals.org/content/16/7/1196


Exosomes Induce Fibroblast Differentiation into Cancer … · Tumor Microenvironment Exosomes...

Documents

Transcript of Exosomes Induce Fibroblast Differentiation into Cancer … · Tumor Microenvironment Exosomes...