original research article cellular adipose tissue displays ... · contract grant sponsor:...

Adipose Tissue Displays TrophicProperties on Normal LungCellular Components WithoutPromoting Cancer Cells GrowthF. ANDRIANI,1 F. FACCHINETTI,1 S. FURIA,2 L. ROZ,1 S. BURSOMANNO,1 G. BERTOLINI,1

C. CARNITI,3 G. SOZZI,1* AND U. PASTORINO2

1Tumor Genomics Unit, Department of Experimental Oncology Molecular Medicine,

Fondazione IRCCS Istituto Nazionale dei Tumori, Milano, Italy2Unit of Thoracic Surgery, Department of Surgery, Fondazione IRCCS Istituto Nazionale dei Tumori, Milano, Italy3Hematology and Bone Marrow Transplant Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Milano, Italy

Surgical removal is the mainstay for early lung cancer treatment and persistent air leaks represent one of the most common clinicalcomplications after lung surgery. Adipose tissue transplantation has been proposed as a new strategy for regenerative therapy after breastcancer surgery; however its efficacy and safety of lung tissue healing after lung resections are unknown. The purpose of this study was totest the biological activity of adipose tissue to facilitate lung tissue healing and evaluate its effect on cancer cells growth, thus providinginsight for a possible clinical application. Different in vitro cellular models were used to prove the potential biologic effect of autologous fattissue (AFT) in repairing injured lung tissue, and in vivo xenograft models were used to evaluate tumor promoting potential of AFT onputative residual cancer cells. Treatment of both embryonic (WI-38) and adult lung fibroblasts and of normal bronchial epithelial cells(HBEC-KT)withAFT samples, harvested from subcutaneous tissue layer of 20 patients undergoing pulmonarymetastasectomy, improvedwound healing and cell proliferation indicating a trophic effect on bothmesenchymal and epithelial cell types. Conversely AFT-conditionedmediumwas unable to stimulate in vitro proliferation of a lung adenocarcinoma reporter cellular system (A549).Moreover, co-injection ofAFT and A549 cells in nude mice did not promote engraftment and progression of A549 cells. These preclinical findings providepreliminary evidence on the potential efficacy of AFT to accelerate lung tissue repair without undesired tumor promoting effects onputative residual cancer cells.J. Cell. Physiol. 228: 1166–1173, 2013. � 2012 Wiley Periodicals, Inc.

Adipose tissue as source of adipose-derived stemcells (ADSCs)has the ability to repair and maintain tissues (Schaffler andBuchler, 2007; Gimble andNuttall, 2011). ASCs release growthfactors such as transforming growth factor (TGF) (Rehmanet al., 2004), platelet-derived growth factor (PDGF) (Craft et al.,2009), fibroblast growth factor (FGF) (Bhang et al., 2009),hepatocyte growth factor (HGF) (Zhu et al., 2009), andmembers of epidermal growth factors family (EGF) that are ableto induce cell proliferation, differentiation, and migration of avariety of cell types such as fibroblasts, endothelial, andepithelial cells which together contribute to a regenerativeeffect in various organs. Human adipose tissue is considered areadily accessible source of mesenchymal stem cells and itsapplication has been frequently reported in the field of plasticand reconstructive surgery, especially in lipofilling after breastcancer resection (Billings and May, 1989; Kanchwala et al.,2009). However one of major question that remains to beanswered before this technique can be usedmore extensively inthe regenerative field, is the effect that adipose tissue couldhave on growth of residual cancer cells when transplanted incancer patients. In fact, together with adipocytes, fat tissue alsocontains a complex mixture of vascular and stromal cells(including ADSCs), named stromal–vascular fraction (SVF)(Zuk, 2010), which could contribute to create amicroenvironment with tumor promoting potential. However,two recent large clinical studies of lipofilling in breast cancerpatients showed no increased risk of local recurrence of canceror development of new cancer after fat injection procedures(Delay et al., 2009; Petit et al., 2011).

These studies prompted us to consider a possible applicationof this approach in the field of thoracic surgery to accelerate and

improve lung tissue healing and reduce air leaks after resectionsof pulmonary parenchyma performed for multiplemetastasectomies by laser or electrocautery techniques. In fact,in this particular surgery performed with curative intent, themost relevant complication is represented by persistent air-leaks requiring prolonged pleural drainage.

The authors declare that they have no conflict of interest.

Author’s contribution: All authors read and approved the finalmanuscript. F.F. carried out the in vitro experiments, includingmigration, proliferation, and scratch assays. F.A., G.B., and S.B.performed the in vivo animal experiments. F.A., L.R., and G.S.participated in designing the study. S.F. and U.P. providedlipoaspirate samples. F.A. drafted the manuscript in cooperationwith L.R., G.S., and U.P.

Co-last authors.

Contract grant sponsor: Associazione Italiana per la Ricerca sulCancro (AIRC).Contract grant sponsor: Fondazione Adele e Bruno Onlus(Tradate, Italy).

*Correspondence to: G. Sozzi, Tumor Genomics Unit,Department of Experimental Oncology and Molecular Medicine,Fondazione IRCCS Istituto Nazionale dei Tumori, via G. Venezian1, 20133 Milan, Italy.

Manuscript Received: 29 May 2012Manuscript Accepted: 18 October 2012

Accepted manuscript online in Wiley Online Library(wileyonlinelibrary.com): 5 November 2012.DOI: 10.1002/jcp.24270

ORIGINAL RESEARCH ARTICLE 1166J o u r n a l o fJ o u r n a l o f

CellularPhysiologyCellularPhysiology

� 2 0 1 2 W I L E Y P E R I O D I C A L S , I N C .

We started our research from limited pulmonary resectionsin patients affected by lung metastases, with the idea to extendthis approach to segmentectomy and lobectomy in primarylung cancer patients. Here we describe preclinical studiesperformed to assess the properties of adipose tissue, as awhole, in lung tissue repair using fat tissue collected from20 patients enrolled in a pilot clinical study consisting in theimplantation of autologous fat graft on lung resected areas forthe prevention of postoperative air leaks. The focus of thisworkwas to clarify the effect of lipoaspirates on wound healing usingdifferent cellular models including lung embryonic and adultfibroblasts as well as immortalized normal bronchial epithelialcells and their genetically modified variants. In addition to invitro wound healing assays, we also investigated the putativepro-tumorigenic potential of autologous fat tissue (AFT) byperforming in vitro and in vivo experiments using a reportercellular system of lung adenocarcinoma. Finally, to study themolecular mechanism underlying wound healing properties oflipoaspirates, we evaluated their effect on synthesis anddeposition of extracellular matrix by fibroblast cells, thatrepresents a critical feature in the healing events (Schultz andWysocki, 2009).

Materials and MethodsFat tissue collection

Fat tissue (lipoaspirate, LA) was collected from 20 patients (LT)enrolled in a pilot clinical study and undergoing pulmonarymetastasectomy between March and September 2010 atFondazione IRCCS Istituto Nazionale dei Tumori (Milan, Italy). Allpatients received chemotherapy after the first surgery interventionfor the primary tumor. Lipoaspirates were obtained from thesubcutaneous layer during thoracotomy. In the intra-operativephase, a specimen of fat tissue ranging from 40 to 120 cm3 washarvested in a syringe by low pressure aspiration. The sample wasthen centrifuged at 3,000 rpm for 3min to obtain a three layerseparation according to the density gradient: the top layer,containing damaged cells in the supernatant liquid; the intermediatelevel, consisting of the concentrated adipose tissue containing stemcells; and a bottom layer composed mainly of concentrated bloodcomponents.

Upon arrival in the laboratory the concentrated adipose tissuesamples were collagenase digested and the cellular yield wascalculated as the number of stromal–vascular cells obtained permilliliter of bioptical tissue. Within the SVF the concentration ofcells with stem cell properties defined as ADSCs was also

estimated using multi-parametric flow cytometry. ADSC werecharacterized by the expression of the following surfacemarkers CD45�, CD90þ, CD73þ, CD271þCD34 bright, andCD146þ/�.

For three patients no lipoaspirates were available and forpreclinical studies, lipoaspirates from two patients that did notparticipate to the clinical studies were collected (LT197 andLT198). For two patients the count of stem cell concentrationfailed. The patient’s characteristics are described in Table 1.

For the in vivo experiments lipoaspirates were diluted 1:4 withserum freemedium (RPMI), mechanically digested and immediatelyinoculated in mice. For the in vitro experiments samples werecentrifuged and stored at �208C.

Cell lines and cultures

A549 lung adenocarcinoma cells (American Type CultureCollection, ATCC) were propagated in RPMI medium with 10%fetal bovine serum and transducedwith a lentiviral vector carrying aGFP transgene to obtain A549-GFP.

Adherent A549 were also plated in serum-free mediumDMEM/F12 (Lonza, Milan, Italy), supplemented with commercial hormonemix, B27 (Gibco Invitrogen, Carlsbad, CA), EGF (20 ng/ml,PeproTech EC, London, UK), bFGF (10 ng/ml, PeproTech), andheparin 2mg/ml. Under these conditions the growth-factorresponsive cells proliferate and formfloating clusters of cells withina couple of weeks whereas most differentiating or differentiatedcells rapidly die. The floating cells, called A549s, give rise to sphereswith stem like-cells properties (Bertolini et al., 2009) and have hightumorigenic potential.

HBEC-KT (obtained from Prof. J. Minna, UTSW, Dallas, Texas)are normal bronchial epithelial cells immortalized with hTERT andCDK4 exogenous expression (Ramirez et al., 2004). HBEC-KT andtheir non-tumorigenic genetically modified variant HBEC-KT p53-/KRASG12V (Sato et al., 2006) (with down-regulation of p53 andexogenous expression of mutant KRAS) were cultured in KSFMmedium (Invitrogen) with EGF (0.2 ng/ml), and BPE (20mg/ml).

The WI-38 fibroblast cell line, derived from normal embryonic(3-month gestation) lung tissue, was obtained from ATCC andcultured in EMEM supplemented with 10% fetal bovine serum.WI-38 and HBEC-KT cell lines were both transduced with a lentiviralvector containing a GFP reporter gene to obtain the fluorescentderivatives (WI-38-GFP, HBEC-KT-GFP) used in scratch assaysand proliferation studies. Two cultures of normal fibroblasts wereisolated through enzymatic digestion with collagenase (5mg/ml)fromnormal lung tissue of patients undergoing surgical resection oflung metastasis and were grown in RPMI medium with 10% FBS.

TABLE 1. Clinical characteristcs of patients and cellular content of lipoaspirates

Patient (LT sample) Age Primary tumor Total cells (106/ml) %Adipose-derived stem cells (ADSCs)

— 59 Soft tissue sarcoma 5 0.350— 23 Testicle carcinoma 2.2 0.729— 43 Osteogenic sarcoma 0.9 0.125153 50 Sinovial sarcoma 0.47 0.209157 20 Ewing sarcoma 1.125 1.703158 69 Colorectal adenocarcinoma 0.93 2.338163 54 Colorectal adenocarcinoma 1.6 0.258168 57 Colorectal adenocarcinoma 0.05 0.729170 62 Larinx carcinoma 0.22 1.232175 63 Soft tissue sarcoma 0.92 3.571178 55 Uterus carcinoma 0.9 1.357180 65 Colorectal adenocarcinoma — 1.912181 75 Colorectal adenocarcinoma — 2.175182 29 Sinovial sarcoma — —183 43 Melanoma — 2.051184 48 Colorectal adenocarcinoma 0.8 0.710186 48 Colorectal adenocarcinoma — —188 54 Adenoid cystic carcinoma 0.35 1.842189 75 Larinx carcinoma — 1.074190 50 Kidney carcinoma 0.8 2.093

JOURNAL OF CELLULAR PHYSIOLOGY

P O T E N T I A L O F L I P O A S P I R A T E S I N L U N G R E P A I R 1167

To confirm their homogeneity, cultures of fibroblasts wereanalyzed by FACS for the expression of mesenchymal markers(CD73, CD105, CD90, and CD166). The cells strongly expressedmesenchymal markers and were negative for epithelial (CD326)and hematopoietic (CD45) markers to exclude potentialcontamination of different cell populations.

Cell migration and scratch assays

Lipoaspirates chemoattractive properties were assessed by a cellmigration assay using transwell plates (8-mm pore size). A total of1� 105 cells (WI-38 or HBEC-KT) were seeded in the upperchamber while the different lipoaspirates were placed in the lowercompartment. As control, cells were seeded in the top chamberwith medium 1% serum, while, in the lower chamber, medium 10%serum was added as chemoattractant. In this conditions cellsmigrated to the lower chamber that contain more serum. After48 h the filters were removed and stained with DAPI and themigrating cells were counted in three randomly chosen fields.

For the scratch assay, embryonic fibroblasts WI38-GFP ornormal bronchial epithelial cells (HBEC-KT-GFP), were seeded in12-well plates and allowed to grow until confluence. A ‘scratchwound’was then created in the cellmonolayerwith a sterile pipettetip and the wounded monolayer was rinsed with PBS to removenon-adherent cells. Different lipoaspirates were then added(dilution 1:20 in culturemedium)whilemediumcontaining 10%FBSwas used as control. Wound width was estimated by fluorescenceevaluation with a microplate reader (Infinite M1000, Tecan,Mannedorf, Switzerland) at 1, 3, and 6 days after wounding todetermine the rate of cell migration.

Invasion assay

To analyze the chemoattractant properties of lipoaspiratesthrough reconstituted basement membranes, a Matrigel transwellassay were performed (Albini and Benelli, 2007). Transwell plateswith a polycarbonate insert of 8mm pore size is coated withMatrigel (BD Biosciences, Franklin Lakes, NJ) diluted 1:10 in serumfree medium. After adding the diluted Matrigel on top of the filter,the plate were incubated at 378C in humidified atmosphere with5%CO2 for at least 3 h. After gelification, 1� 105 cells were seededon top the gelified Matrigel matrix. To attract the cells towardsthe bottom side of the porous membrane controls containingmedium 10% serum or lipoaspirates were applied in the lowercompartment. The plate was incubated at 378C in a humidifiedatmosphere for 3 days and the inserts were washed in tris-bufferedsaline (TBS) solution and fixed in ice-cold methanol for 10min.Nuclei are stained with DAPI and invading cells were counted usinga fluorescence microscope.

Real-time PCR

At the end of the scratch, wound assay RNA was obtained bothfrom fibroblasts migrated in the wounded area and from theunaffected monolayer outside the scratch wound. All theexperiments were run on a GeneAmp1 7900 Sequence DetectionSystem using TaqMan1 technology (Applied Biosystems, Monza,Italy). Primers and probes for the different genes were available asready to use assays-on-demand. Real-time PCR was carried out intriplicate in a 20ml volume and assays were repeated on differentdilutions as needed. Two different endogenous controls (HPRTand B2M) were used as references to normalize the differentsamples and for relative quantification of gene expression. Datawere analyzed by the comparative Ct method (DDCt).

Proliferation assay

A549-GFP or HBEC-KT-GFP cells were placed on 96 blackpolystyrenewell plates at 1.0� 105 cells/well and then treatedwithdifferent lipoaspirates diluted 1:20 with appropriate medium. Theevaluation of growth kinetics was estimated by fluorescence

intensity (excitation at 480 nm, emission detection at 530 nm) witha microplate reader (Infinite M1000, Tecan).

The fluorescencewasmeasured on days 1, 3, and 6 of the cultureperiod for the different samples.

In vivo experiments

All experiments were carried out with female CD-1 nude mice orSCID mice, 7–10 weeks old (Charles River Laboratories,Wilmington, MA). Mice were maintained in laminar flow rooms,with temperature and humidity constant. Mice had free access tofood and water. Experiments were approved by the EthicsCommittee for Animal Experimentation of the Fondazione IRCCSIstitutoNazionale dei Tumori, according to institutional guidelines.

Subcutaneous injections of 104 A549S cells mixed with digestedlipoaspirates were performed into two nude mice in both flanks toobtain four measures per group. A549 cells suspended in 50ml ofRPMI serum freemediumweremixedwith 50ml of lipoaspirate and100ml of MatrigelTM, and injected subcutaneously. A549S cells andlipoaspirates alone resuspended in the same volume of RPMImedium and MatrigelTM were used as positive and negativecontrols respectively. Mice were monitored every week for3 months to assess tumor take and growth kinetics of establishedtumor nodules was recorded. Tumor volume based on calipermeasurements were calculated by the modified ellipsoidal formula(Tumor volume¼ (1/2)(length�width2).

Statistical analysis

Data are presented as the mean� SD. Means were compared byStudent’s t-test. Probability values of P<0.05 were interpreted todenote statistical significance.

ResultsAdipose tissue stimulates migration, wound healing andinvasion of normal lung fibroblasts

To test the chemotactic activity of different lipoaspirates, weinvestigated the migration ability of WI-38 embryonicfibroblasts, transduced with a lentiviral vector containing theGFP transgene, using a transwell assay. Cells were seeded in theupper chamber and lipoaspirate samples were used aschemoattractant in the lower chamber. Cells migrated in thelower chamberwere stainedwithDAPI and counted (Fig. 1A,a).Most lipoaspirates used (6/8, 75%) were able to significantlyincrease migration at 48 h compared with control medium(Fig. 1A,b).

In order to assess the migratory promoting effect of AFTdirectly on primary fibroblasts from the patients enrolled inthe clinical protocol, we performed migration assay on lungfibroblasts cell cultures established from two of the 20 patients.For the examination of migration levels, the number of cellsmigrated was evaluated. The basal migration rate of adult lungfibroblasts was lower compared to WI-38 and the presence oflipoaspirates did not significantly increase the migration ratecompared to control (Fig. 1B,a,b) suggesting that lipoaspiratestreatment might be more effective in inducing migration of cellsendowed with higher regenerative potential and likelyrepresenting tissue-specific progenitor cells (such as embryonicfibroblasts WI-38) rather than of mature adult fibroblasts.Indeed, direct comparison of expression levels of stemnessrelated genes (NANOG, OCT-4, FN1, VIMENTIN) confirmedthat WI-38 transcriptional profile is more similar tomesenchymal stem cells than to adult fibroblasts (data notshown).

Since fibroblasts involved in the wound healing aresusceptible to the effects of chemotherapy (Payne et al., 2008)to verify that treatment did not importantly affect themigrationactivity of adult fibroblasts derived from treated patients weperformedmigration experiments, in presence of lipoaspirates,


1168 A N D R I A N I E T A L .

in a normal fibroblasts cell line (LT237) obtained from a patientnot treated with chemotherapy. We did not observedifferences in migration rate between fibroblasts from treatedand non-treated patients, indicating that the migrationefficiency in our model is independent from on thechemotherapeutic treatment (Fig. 1B,c).

To investigate whether lipoaspirates behave aschemoattractant, invasion experiments usingMatrigel transwellplate, to detect the migratory activity associated with matrix

degradation, were performed. We found that lipoaspiratestriggered invasion of WI-38 cells confirming their chemo-attractant properties also through reconstituted basementmembranes (3/4; Fig. 2).

We also performed scratch assays to investigate the ability oflipoaspirates to stimulate wound closure by WI-38-GFPfibroblasts. In particular, this assay allows analysis not only forcell motility studies but also for cell–matrix and cell–cellinteraction levels such as tissue reorganization, or matrixdeposition. Five out of seven (71%) lipoaspirates samples wereable to induce wound healing more efficiently than controls,indicating a positive effect on wound repair capacity. Theacceleration of wound closure was already observed at 1 daypost-treatment and persisted at 3 and 6 days (Fig. 3, parts A andB).

Finally, a slight, but not significant, increase in proliferationrate of WI-38-GFP cells was detected in most of the samples,further supporting a potential role of adipose tissue to stimulatethe proliferation of progenitor cells at the site of damaged tissue(Fig. 4).

Interestingly overall no correlation between lipoaspirateseffect and content of mesenchymal stem cells was found (datanot shown).

Adipose tissue stimulate migration of lung fibroblastswith increased expression of extracellular matrixcomponents

To better clarify thewound healing potential of lipoaspiratesweevaluated lipoaspirate-induced modulation of expression ofgenes related to the extracellular matrix on fibroblastsmigrated in the wounded area compared with fibroblastsoutside this area.

To this purpose we performed a scratch assay on WI-38fibroblast cells and treated them with control medium orlipoaspirates. Cellsmigrated in thewound area and cells outsidethe wound area were collected separately and the level ofexpression of extracellular matrix components, such ascollagen 1, fibronectin or elastin fiber components (elastin,fibrillin, fibulin) were evaluated by Real-time PCR.We analyzedexpression levels of these transcripts comparing fibroblastsmigrated inside the wounded area with fibroblasts outside this

Fig. 1. Presence of lipoaspirates results in increased migration rateof WI-38 embryonic fibroblasts in vitro. Part A: (a) DAPI staining ofmigrated cells. (b) Migration assay experiments with eight differentlipoaspirate samples. A significantly increased number (MP<0.05) ofcells migrated across the transwell membrane at 48 h in six out ofeight lipoaspirates tested compared with the control group.Migration rate is expressed as percentage relative to control (100%).Each experiment was performed at least three times. Part B: (a)Primary adult normal fibroblasts are not responsive to lipoaspirates.Migration assay on normal fibroblast cell line LT158NF with fourdifferent lipoaspirates. No statistical difference between samples andcontrolwasobserved. (b)Normal fibroblast cell line isolated fromonepatient (LT190NF) treated with the autologous lipoaspirate (LT190-LA). No significant difference between control and treated cells wasfound. (c) Migration assay on normal fibroblast cell line LT237NFderived from one patient that did not received chemotherapy usingfour lipoaspirates. No statistical difference between samples andcontrol was observed.

Fig. 2. Lipoaspirates have affect invasion ofWI-38 throughMatrigel.Invasion assay experiments with four different lipoaspirate samples.A significantly increased number (MP<0.05) of cells migrated acrossthe transwell membrane at 72 h in three out of four lipoaspiratestested comparedwith the control group. Invasion rate is expressed aspercentage relative to control (100%). Each experiment wasperformed at least three times.



area and observed that 7/10 (70%) extracellular matrixassociated genes were differentially expressed (Fig. 5). Inparticular migrated fibroblasts showed higher levels ofexpression of FN1 (P¼ 0.006), FBN1 (P¼ 0.0004), FBNL1(P¼ 0.04), FBNL5 (P¼ 0.002), ITG-4 (P¼ 0.01), ITG-6

(P¼ 0.01), and ELN (P¼ 0.003) (Fig. 5). This activatedphenotypewas also observed inmigrating cells from the controlmedium treated assay (data not shown),whichwere however inremarkable lower number compared with lipoaspirate treatedsamples (Fig. 3). These results indicate therefore that

Fig. 3. Lipoaspirates increase wound healing rate in WI38-GFP fibroblasts. Part A. Phase contrast image of the scratched monolayer in cellstreatedwith a lipoaspirate sample and in control atDay1andDay6after scratching. PartB.Woundhealing assay evaluation after treatmentwithlipoaspirates or medium alone. Most lipoaspirates (5/7) induced a significantly increased rate of wound closure compared with medium alone.Representative examples of increased (left and middle part) or steady (right part) rate of closure are shown.

Fig. 4. Lipoaspirates induce a slight increase of proliferation rate ofWI-38 cell line. Microplate reader results of proliferation test in WI-38 GFP cell line treated with different lipoaspirates. No significantdifference between cells treated with lipoaspirates and controls wasfound except for three samples (LT175-LA, LT183-LA, and LT188-LA) evaluated after 2, 4, and 8 days. Growth rate is expressed asfluorescence intensity measurement (FI).

Fig. 5. Expression of extracellular matrix related genes in migratedfibroblasts. Expression levels of a subset of extracellular matrixrelated genes in fibroblasts migrated in the wounded area comparedwith fibroblasts outside the wound. MP<0.05 for differences betweenthe two groups.



lipoaspirates can actively increase the migration of fibroblastsendowed with extracellular matrix deposition potential.

Adipose tissue increases proliferation of normalbronchial epithelial cells

We reasoned that a coordinated and regulated interactionbetween epithelial cells and mesenchymal cells should berequired for the activation of the repair process of the damagedlung. To determine whether treatment with lipoaspirates alsoincreases epithelial cells migration to the injured area weperformed a scratch assay using as substrate HBEC-KT-GFP(HBEC-1) and HBEC-KT-p53-/KRASG12V-GFP (HBEC-6)cells. After treatment with lipoaspirates cell monolayersrecovered and healed the wound during a time course of24–72 h. However, most lipoaspirates (4/5, 80%) induced only

a slight increased migration rate of both HBEC-KT-GFP andHBEC-KT-p53-/KRASG12V-GFP cells compared to controlsamples (Fig. 6A). Conversely, the proliferation rate of HBEC-KT-GFP cells and their variant was significantly increased afterexposure to all the lipoaspirate samples (7/7) (Fig. 6B).

Adipose tissue does not affect cell proliferation and invivo tumor growth of lung cancer cell line A549

To evaluate whether treatment with lipoaspirates in vitro wasassociated with an increased proliferation potential of lungcancer cells, A549-GFP cells were treated with the differentlipoaspirate samples for 1 week and growth kinetics evaluatedat 1, 3, and 6 days. We found that most of the lipoaspiratestested (7/9, 77%) did not increase the proliferation rate of

Fig. 6. Lipoaspirates increase proliferation on normal bronchial epithelial cells without significantly affecting wound healing closure. Part A.Lipoaspirates exposure induces a slight but not significant increased migration rate on normal bronchial epithelial cells compared with control.PartB.Significantly increasedproliferation rateafterexposurewith lipoaspirates comparedwithcontrol representedby treatmentwithmediumalone. The number of proliferating cells was evaluatedwith a fluorescencemicroplate reader and is expressed as fluorescence intensity (FI). Theexperiments were performed at least three times.



A549-GFP cell line (Fig. 7A) as compared to the growth rate ofcells treated with medium alone.

To assess whether lipoaspirates promote tumor growthin vivo in nude mice, we injected subcutaneously (sc) in nudemice 104 A549S cells alone or in combination with lipoaspiratesat 1:1 volume ratio. Lipoaspirate samples alone were used asnegative controls. We tested 10 lipoaspirate samples availablefrom patients. Interestingly none of the mice co-injected withA549 tumor cells and lipoaspirates (0/10) showed an increasedtumor growth rate and size compared to A549s injected alone(Fig. 7B) suggesting that fat treatment does not promoteengraftment and/or progression of cancer cells in vivo.

Discussion

Adipose tissue is considered an abundant and easily accessiblesource of autologous mesenchymal stem cells that have beenshown to have regenerative capability in different fields ofmedicine (Satija et al., 2009). Although the clinical applicationson airway defects are limited so far, the use of adipose tissue assource of mesenchymal stem cells could open new prospectsto avoid post-operative complication also in thoracic surgeryafter pulmonary resection.

The present study was designed to support, through in vivoand in vitro studies, the potential clinical use of adipose tissueto improve pulmonary surface healing after metastasectomyand also to rule out a putative pro-tumorigenic effect ofadipose tissue on residual cancer cells in the resected area.

In order to closely mimic the clinical study and to evaluatethe contribution of both cellular and fat components, we usedin all experiments lipoaspirate samples as a whole, withoutpurification of adipose cell fractions.

To study repair mechanism after woundingwe used differentcellular models such as lung embryonic fibroblasts, primaryadult normal fibroblasts cultures from patients, and bronchialepithelial cell lines that allowed us to show the involvement ofdifferent cell types and lineages in the repair process.

We found that lipoaspirate treatment has an effect in thepromotion of lung fibroblasts migration. The results ofmigration assays in transwell plates showed that the number offibroblasts migrating through the membranes increased in thelipoaspirate-treated cell lines thus indicating a chemoattractantactivity. Interestingly we have not observed any correlationbetween lipoaspirates biological effect and content of ADSCssuggesting that the repair mechanism is not linked to theadipose stem cellular components of lipoaspirate but rather isassociated to an interaction between lipoaspirate and residentfibroblasts or, more importantly, to the recruitment of bonemarrow derived progenitors to the site of injury (Opalenik andDavidson, 2005). Some of these signals may be facilitated bychemokines and growth factors present in adipose tissue whichare associated not only to adipose stem cells but also to othercell types including endothelial cells and macrophages. Theefficacy of lipoaspirates could be thus more related to solublemediators concentration that is variable among patients than tothe adipose stem cells content. Moreover, wound healing assay

Fig. 7. Lipoaspirates do not promote proliferation in vitro and tumour growth in vivo of A549 lung cancer cells. Part A. Analysis of proliferationrateofA549-GFPadherentlungcancercell lineafterexposureof lipoaspiratesfor6days.Thereisnostatisticaldifferencebetweencellstreatedwithmediumaloneordifferent lipoaspiratesexceptfortwosamples(LT182-LAandLT197-LA,MP<0.05).PartB.Tumorgrowthatdifferenttimepointsafter co-injection of 104 tumour cellswith the samevolumeofwhole lipoaspirate. Positive control and negative control are represented byA549scell line or lipoaspirate injected alone, respectively. No significant difference between positive control group and co-injection experimentswas found.



showed a more efficient wound closure activity in embryonicfibroblasts rather than HBEC-KT cells. Conversely, bronchialepithelial cells were more responsive to proliferation afterlipoaspirate treatment than fibroblasts. These findings suggestthat fibroblasts activation may occur in the early phases afterlung injury, whereas proliferation of epithelial cells populationcould be more relevant in a later phase of tissue repopulationand remodeling. Moreover we found that genes associated toextracellular matrix such as fibronectin, collagen, fibulin,fibrillin, and elastin were highly expressed in migratedfibroblasts compared to non-migrated fibroblasts suggestingthat lipoaspirates stimulate extracellular matrix remodeling.

As previously proven in other biological settings (Kannan andWu, 2006; Delgado et al., 2011), cells endowed withprogenitor-like phenotype (WI-38) were the cell type moreinvolved in the repair mechanism. In fact, lipoaspirate-inducedmigration of normal adult fibroblasts was not as efficient as theone observed in embryonic fibroblasts. In addition, the highresponsiveness to proliferation of HBEC-KT, could possibly berelated to the expression of p63, a p53 protein homologueessential for the regeneration process in epithelial development(Ramirez et al., 2004).

Although there is a growing body of literature demonstratingthat adipose stem cells could create amicroenvironment proneto enhance the growth of cancer cells (Park et al., 2000; Yu et al.,2008; Martin-Padura et al., 2012), no reports exist on thebiological activity of adipose tissue as a whole. Testing whetheradipose tissue is conductive to the expression of neoplasticpotential from putative residual cancer cells is particularlyrelevant in the context of pulmonary resections for lung cancer.We established an original model of lung adenocarcinoma cellsthat are responsive to growth promoting effects of primaryfibroblast cell lines isolated from lung cancer patients when co-injected at low cellular doses in mice (Andriani et al., 2011;AACR). Conversely, in this study the whole lipoaspiratesamples obtained from 10 patients, co-injected in mice withthese lung adenocarcinoma cells did not show any promotion oftumor growth.

Taken together these findings provide preclinical evidencefor biological activity of adipose tissue in inducing lung tissuerepair using different lung cellular models without promotinglung cancer cells proliferation in vitro and tumor growth in vivoand support a potential clinical use in thoracic surgery. Futureclinical studies are needed to confirm safety and efficacy ofadipose tissue treatments in lung resections.

Acknowledgments

This work was supported by grants from Associazione Italianaper la Ricerca sul Cancro (AIRC) to G. Sozzi and L. Roz andfrom Fondazione Adele e Bruno Onlus (www.adelebruno.org,Tradate, Italy) to U. Pastorino.

Literature Cited

Albini A, Benelli R. 2007. The chemoinvasion assay: Amethod to assess tumor and endothelialcell invasion and its modulation. Nat Protoc 2:504–511.

Andriani F, Bertolini G, Caserini R, Bursomanno S, Gasparini P, Pastorino U, Sozzi G,Roz L. 2011. Contribution of fibroblasts isolated from primary lung cancers andnormal lung tissue to tumor growth. Proceedings of 102nd AACR Annual Meeting.Abstract nr.530.

Bertolini G, Roz L, Perego P, TortoretoM, Fontanella E, Gatti L, Pratesi G, Fabbri A, AndrianiF, Tinelli S, Roz E, Caserini R, LoVS, Camerini T,Mariani L, Delia D,Calabro E, PastorinoU,Sozzi G. 2009. Highly tumorigenic lung cancerCD133þ cells display stem-like features andare spared by cisplatin treatment. Proc Natl Acad Sci U S A 106:16281–16286.

Bhang SH,Cho SW, Lim JM,Kang JM, LeeTJ, YangHS, SongYS, ParkMH,KimHS, YooKJ, JangY, Langer R, Anderson DG, Kim BS. 2009. Locally delivered growth factor enhances theangiogenic efficacy of adipose-derived stromal cells transplanted to ischemic limbs. StemCells 27:1976–1986.

Billings E, Jr., May JW, Jr. 1989. Historical review and present status of free fat graftautotransplantation in plastic and reconstructive surgery. Plast Reconstr Surg 83:368–381.

Craft RO,Rophael J,MorrisonWA,VashiAV,MitchellGM, PeningtonAJ. 2009. Effect of local,long-termdelivery of platelet-derived growth factor (PDGF) on injected fat graft survival insevere combined immunodeficient (SCID)mice. J Plast ReconstrAesthet Surg 62:235–243.

Delay E, Garson S, Tousson G, Sinna R. 2009. Fat injection to the breast: Technique, results,and indications based on 880 procedures over 10 years. Aesthet Surg J 29:360–376.

DelgadoO, Kaisani AA, Spinola M, Xie XJ, Batten KG, Minna JD,WrightWE, Shay JW. 2011.Multipotent capacity of immortalized human bronchial epithelial cells. PLoS ONE6:e22023.

Gimble JM, Nuttall ME. 2011. Adipose-derived stromal/stem cells (ASC) in regenerativemedicine: Pharmaceutical applications. Curr Pharm Des 17:332–339.

Kanchwala SK, Glatt BS, Conant EF, Bucky LP. 2009. Autologous fat grafting to thereconstructed breast: The management of acquired contour deformities. Plast ReconstrSurg 124:409–418.

Kannan S, Wu M. 2006. Respiratory stem cells and progenitors: Overview, derivation,differentiation, carcinogenesis, regeneration and therapeutic application. Curr Stem CellRes Ther 1:37–46.

Martin-Padura I, Gregato G, Marighetti P, Mancuso P, Calleri A, Corsini C, Pruneri G,Manzotti M, Lohsiriwat V, Rietjens M, Petit JY, Bertolini F. 2012. The white adipose tissueused in lipotransfer procedures is a rich reservoir of CD34þ progenitors able to promotecancer progression. Cancer Res 72:325–334.

Opalenik SR, Davidson JM. 2005. Fibroblast differentiation of bone marrow-derived cellsduring wound repair. FASEB J 19:1561–1563.

Park CC, Bissell MJ, Barcellos-Hoff MH. 2000. The influence of the microenvironment on themalignant phenotype. Mol Med Today 6:324–329.

PayneWG,NaiduDK,WheelerCK, BarkoeD,MentisM, Salas RE, SmithDJ, Jr., RobsonMC.2008. Wound healing in patients with cancer. Eplasty 8:e9.

Petit JY, Lohsiriwat V, CloughKB, Sarfati I, Ihrai T, RietjensM, Veronesi P, Rossetto F, ScevolaA, Delay E. 2011. The oncologic outcome and immediate surgical complications oflipofilling in breast cancer patients: Amulticenter study-milan-paris-lyon experience of 646lipofilling procedures. Plast Reconstr Surg 128:341–346.

RamirezRD, Sheridan S,Girard L, SatoM,KimY, Pollack J, PeytonM,ZouY,Kurie JM,DimaioJM, Milchgrub S, Smith AL, Souza RF, Gilbey L, Zhang X, Gandia K, Vaughan MB, WrightWE, Gazdar AF, Shay JW, Minna JD. 2004. Immortalization of human bronchial epithelialcells in the absence of viral oncoproteins. Cancer Res 64:9027–9034.

Rehman J, Traktuev D, Li J, Merfeld-Clauss S, Temm-Grove CJ, Bovenkerk JE, Pell CL,Johnstone BH, Considine RV, March KL. 2004. Secretion of angiogenic and antiapoptoticfactors by human adipose stromal cells. Circulation 109:1292–1298.

Satija NK, Singh VK, Verma YK, Gupta P, Sharma S, Afrin F, SharmaM, Sharma P, Tripathi RP,Gurudutta GU. 2009. Mesenchymal stem cell-based therapy: A new paradigm inregenerative medicine. J Cell Mol Med 13:4385–4402.

SatoM, VaughanMB, Girard L, PeytonM, LeeW, ShamesDS, Ramirez RD, SunagaN, GazdarAF, Shay JW, Minna JD. 2006. Multiple oncogenic changes (K-RAS(V12), p53 knockdown,mutant EGFRs, p16 bypass, telomerase) are not sufficient to confer a full malignantphenotype on human bronchial epithelial cells. Cancer Res 66:2116–2128.

Schaffler A, Buchler C. 2007. Concise review: Adipose tissue-derived stromal cells—Basicand clinical implications for novel cell-based therapies. Stem Cells 25:818–827.

Schultz GS, Wysocki A. 2009. Interactions between extracellular matrix and growth factorsin wound healing. Wound Repair Regen 17:153–162.

Yu JM, Jun ES, Bae YC, Jung JS. 2008. Mesenchymal stem cells derived from human adiposetissues favor tumor cell growth in vivo. Stem Cells Dev 17:463–473.

Zhu XY, Zhang XZ, Xu L, Zhong XY, Ding Q, Chen YX. 2009. Transplantation of adipose-derived stem cells overexpressing hHGF into cardiac tissue. Biochem Biophys ResCommun 379:1084–1090.

Zuk PA. 2010. The adipose-derived stem cell: Looking back and looking ahead. Mol Biol Cell21:1783–1787.



original research article cellular adipose tissue displays ... · contract grant sponsor:...

Documents