Human amniotic fluid stem cells protect rat lungs exposed to moderate hyperoxia

Pediatric Pulmonology

Human Amniotic Fluid Stem Cells Protect Rat LungsExposed to Moderate Hyperoxia

Davide Grisafi, PhD, PharmD,1,2 Michela Pozzobon, PhD,3,4 Arben Dedja, PhD,1,4

Valentina Vanzo, MD,1 Rosella Tomanin, PhD,2 Andrea Porzionato, PhD, MD,5

Veronica Macchi, PhD, MD,5 Roberto Salmaso, MD,6 Maurizio Scarpa, MD,2 Emanuele Cozzi, MD,7

Ambrogio Fassina, MD,6 Filippo Navaglia, PhD,8 Claudio Maran, PhD,9 Maurizio Onisto, PhD,9

Luciana Caenazzo, PhD,10 Paolo De Coppi, MD,3,11 Raffaele De Caro, MD,5 Lino Chiandetti, MD,1

and Patrizia Zaramella, MD1*

Summary. Background: Treatment of bronchopulmonary dysplasia (BPD) remains as yet an

unmet clinical need and recently stem cells have been proposed as a therapeutic tool in animal

models. We investigated the role of amniotic fluid stem cells (AFS) in an adult rat model of

hyperoxia lung injury. Methods: Fifty Sprague–Dawley rats were, at birth, randomly exposed to

moderate hyperoxia or room air for 14 days and a single dose of human amniotic fluid stem

(hAFS) or human Fibroblasts (hF), cells was delivered intratracheally (P21). At P42 animals

were euthanized and lung tissue examined using histology, immunohistochemistry, PCR, and

ELISA. hAFS cells characterization and homing were studied by immunofluorescence. Results:

In rats treated with hAFS and hF cells 16S human rRNA fragment was detected. Despite a low

level of pulmonary hAFS cell retention (1.43 � 0.2% anti-human-mitochondria-positive cells),

the lungs of the treated animals revealed higher secondary crest numbers and lower mean

linear intercept and alveolar size, than those exposed to hyperoxia, those left untreated or

treated with hF cells. Except for those treated with hAFS cells, moderate hyperoxia induced an

increase in protein content of IL-6, IL-1b, as well as IF-g and TGF-1b in lung tissues. High

VEGF expression and arrangement of capillary architecture in hAFS cell group were also

detected. Conclusions: Treatment with hAFS cells has a reparative potential through active

involvement of cells in alveolarization and angiogenesis. A downstream paracrine action

was also taken into account, in order to understand the immunodulatory response. Pediatr

Pulmonol. � 2013 Wiley Periodicals, Inc.

Key words: hyperoxia; hAFS cells; transplantation; lung repair.

Funding source: Italian Ministry of Health grant, Number: CPDA081132/08, University of

Padova, Padova, Italy, Great Ormond Street Charity, Citta’ della Speranza Foundation.

1Neonatal Intensive Care Unit, Women’s and Children’s Health Depart-

ment, University Padova Hospital, Padova, Italy.

2Gene Therapy Laboratory, Women’s and Children’s Health Department,

University of Padova, Padova, Italy.

3Stem Cell Processing Laboratory, Women’s and Children’s Health

Department, University of Padova, Padova, Italy.

4Citta della Speranza Foundation, Padova, Italy.

5Department of Anatomy and Physiology, University of Padova, Padova,

Italy.

6Institute of Pathology, Department of Medical Diagnostic Sciences

and Special Therapies, University of Padova, Padova, Italy.

7C.O.R.I.T. (Consorzio per la Ricerca sul Trapianto di Organi), Padova,

Italy.

8Department of Laboratory Medicine, University Padua Hospital, Padova, Italy.

9Department of Experimental Biomedical Sciences, University of Padua,

Padova, Italy.

10Department of Environmental Medicine and Public Health, Legal Medi-

cine Section, University of Padova, Padova, Italy.

11Surgery Unit, Great Ormond Street Hospital and UCL, Institute of

Child Health, London, UK.

*Correspondence to: Patrizia Zaramella, MD, Women’s and Children’s

Health Department, Neonatal Intensive Care Unit, University Hospital of

Padova, via Giustiniani, 3, 35128 Padova, Italy.

E-mail: [email protected]

Received 18 July 2012; Accepted 17 February 2013.

DOI 10.1002/ppul.22791

Published online in Wiley Online Library

(wileyonlinelibrary.com).

� 2013 Wiley Periodicals, Inc.

INTRODUCTION

The pathogenesis of bronchopulmonary dysplasia(BPD) is multifactorial, suggesting that multipleapproaches might be considered to help reduce its inci-dence and severity. Hyperoxia is one of the primaryrisk factors leading to the onset of BPD, but the oxygentoxicity goes together with genetic risk factors1 and air-ways inflammatory cell responses.2 Clinically, BPD inpreterm infants is a challenging condition and disordersmay persist beyond childhood leading to lung diseasesin adults.3,4 Children and adolescents born pretermshow impaired lung function5 which in many casesleads to the development of chronic respiratoryillnesses.4,6

In rats, lung damage caused by 0.6 FiO2 exposure,namely moderate oxygen exposure,7,8 in contrast to0.95,9 during the first 14 days of life does not complete-ly arrest lung cell proliferation and is associated withparenchymal thickening, as observed in BPD. Isolateddevelopmental arrest in alveolarization may be causedby more subtle changes in signaling pathways withoutevidence of inflammation or necrosis, which are likelyto happen under moderate hyperoxia. Meanwhile, thereis increasing evidence of the value of stem cells as apotential treatment for neonatal lung diseases.10–12 Pres-ently adult mesenchymal stem cells (MSCs) and umbil-ical cord blood cells show particular promise forexperimental purposes in newborn murine models.9,13–

18 MSCs have cytoprotective effects mainly related to aparacrine mechanisms,17 and a MSC-conditioned mediahas recently been demonstrated to provide some benefi-cial results in a hyperoxia-challenged experimentalmodel of the BPD.19 While MSCs cannot always beisolated from cord blood, adult MSCs from BM arerare, difficult to expand and therefore banking of thesecells for therapy produces limited results.20 AFS cellswere demonstrated to resemble MSCs in manyrespects21 including surface marker expression and dif-ferentiation potential. AFS cells are likely to exhibitimmunomodulatory capabilities as MSCs. Direct con-tact with AFS cells inhibits lymphocyte activation andcell-free supernatants, derived from AFS cells co-cultured with total blood monocytes or IL-1b, a cyto-kine released by monocytes in the inflammatoryresponse, which also inhibited lymphocyte activation.

As a consequence, in this work we have explored thetherapeutic effect of human amniotic fluid stem (hAFS)cells in a rat lung injury model, after hyperoxia expo-sure during the neonatal period. So far AFS cells havebeen shown to be capable to differentiating into severalcells type, such as respiratory epithelium, chondrocytes,adipocytes, neural cells, and endothelial cells.22–25 Totrigger lung injury and to inhibit alveolar growth,7,26 we

opted for moderate levels of hyperoxia rather than thehigher exposure levels used in previous studies.9,22

Our hypothesis was that hAFS cells might heal mod-erate hyperoxia-challenged rats in a late alveolarizationphase. Amelioration of lung histopathology and hAFScell retention were evaluated. A downstream paracrineaction was also taken into account, noting the immuno-modulatory response and angiogenesis.

MATERIALS AND METHODS

Cell Preparation

Human AFS cells were isolated and cultured fromhuman amniotic fluid, as previously reported.2325 Brief-ly, following the guidelines established by the HospitalEthical Committee (protocol number 451P/32887), con-senting volunteer donors gave fresh backup samples ofroutine amniocentesis (2–3 ml). The cells were platedin Petri dish (3.5 cm diameter) and once adherent cellsreached approximately 60–70% of confluence, theywere detached with Trypsin 0.05–EDTA 0.02 w/v (Bio-chrom AG, Berlin, Germany) and magnetically selectedfor human anti-mouse c-kit (Santa Cruz Biotechnology,Inc., Delaware, CA). Cells were characterized by cito-fluorimetric analysis using the following antibodies:CD73-PE, CD-90-FITC, CD105-FITC, HLA-ABC-FITC, and HLA-DR-PE all purchased from BD andSSEA-4 (Santa Cruz Biotechnology, Inc.) with second-ary antibody 488 conjugated (Invitrogen, Milan, Italy).Cells were differentiated in osteogenic and adipogenicmedium and Von kossa and oil-red-o staining25 wereperformed to verify the attained differentiations.

Human primary fibroblasts, hF, used as negative con-trol, were anonymously obtained from foreskin circum-cision samples and kindly provided by the Laboratoryof Gene Therapy, Women’s and Children’s HealthDepartment, University of Padua, Padua, Italy. Cellswere selected by adhesion and maintained in Petridishes (10 cm diameter) with complete DMEM and15% fetal bovine serum. When required cells weredetached with Trypsin 0.05–EDTA 0.02 w/v (LifeTechnologies, Milan, Italy) and resuspended in PBS inaliquots of 1.5 � 106 cells, ready for the injection.

Hyperoxia Exposure

Female wild-type Sprague–Dawley (SD) rats (HarlanItaly, Udine, Italy) and their offspring were housed andhandled in accordance with the recommendations of thePublic Health Office release of December 17, 2009,number 220/2009-B under the supervision of the localEthical Committee. The study was conducted on maleand female rat pups kept together with their nursingmother in conventional facilities. Adequate care for

2 Grisafi et al.


their health and well-being was provided in accordancewith Italian Law 116/92. Within the first 12 hr of birth(P0), pups and dams (or foster mothers) were placedin trasparent polished acrylic chambers providedwith software enabling the continuous monitoring ofO2 and CO2 (BioSpherix, OxyCycler model A84XOV,Redfield, NY). The normobaric oxygen concentrationwas set at FiO2 ¼ 0.6 until postnatal day (P) 14. Theanimals were fed ad libitum and exposed to alternating12-hour light/dark cycles, humidity level was main-tained at 50%, and ambient temperature at 248C.

Experimental Design

Fifty newborn rats were randomly distributed be-tween four experimental groups (Fig. 1A):

- Room air, n ¼ 10, pups were raised in normoxia atFiO2 ¼ 0.21, for 42 days (control group);

- Untreated hyperoxia, n ¼ 12, pups were exposed toFiO2 ¼ 0.6 from birth for 14 days, then to room airfor 28 days;

- Hyperoxia þ hAFS cells, n ¼ 18, pups wereexposed to hyperoxia for 14 days and room air for28 days, and intratracheally administered with hAFScells at P21;

- Hyperoxia þ hF, n ¼ 10, pups were exposed tohyperoxia for 14 days, then to room air for 28 days,and intratracheally administered with hF at P21.

At P42 (after 21 days in room air following hAFS orhF cell treatment) animals were euthanized and lungtissue was collected.

Transplant Procedure

Following an anterior neck incision under isofluoraneanesthesia, the trachea was punctured with a 30G nee-dle and each animal in the hAFS cell group receiveda single dose of 1.5 � 106 cells resuspended in 30 mlof phosphate buffered saline (PBS, pH 7.4); similarlyanimals transplanted with hF cells were administered1.5 � 106 cells in PBS, pH 7.4. At P42 all animalswere deeply anesthetized with a 1:1 combination ofzoletil and xylazine and euthanized. All efforts weremade to minimize suffering.

Histology and Immunofluorescence

Lung Histology, Morphometric Analysis andImmunohystochemistry

Following terminal anesthesia a catheter was placedinto the trachea, and the lungs were inflated and main-tained at 30 cmH2O pressure with 4% paraformalde-hyde in PBS for at least 45 min. A ligature wastightened around the trachea to maintain pressure afterremoval of the tracheal cannula. Lungs were immersed

in paraformaldehyde solution overnight, and a 0.5-cmthick section of the left lower lobe was embedded inparaffin. Lung sections 4 mm thick were stained withhematoxylin and eosin. For each subject, we examinedsix sections and three fields per section. Photomicro-graphs were obtained on a field of 864 � 648 pixelsat 20� magnification, with a Leica DM 4000B micro-scope (Leica, Solms, Germany) integrated with a cam-era (Leica DFC 280). Lung histopathology analyseswere performed by two independent researchers blindedto the treatment strategy, using ImageJ, a public domainJava image-processing program (http://rsb.info.nih.gov/ij),as previously reported.27 In particular in each section,

Fig. 1. A: Schematic outline of the experimental design.

From birth to postnatal day (P) 14: normoxia or hyperoxia

(FiO2 ¼ 0.6) exposure; from P15 to P42 room air recovery;

hAFS cell and hF cell injection were performed at P21. The

study groups, i.e., room air, n ¼ 10; untreated hyperoxia,

n ¼ 12; hAFS cells, n ¼ 18; and hF, n ¼ 10. The stage of lung

development corresponds to the saccular phase on P1 and

becomes mainly alveolar by P14. B: A box-and-whisker plot

displayed median (horizontal line), 25th and 75th percentiles

(box), and range (whiskers) of the averaged weights for each

studied groups, from P0, P14, P21, and P42 time-points. No

body weight differences were found.

Stem Cells Repair Hyperoxic Lung 3


from specific plug-in which leads to the evaluation ofthe skeletonized air spaces into each high-power field(hpf) the averaged alveolar size was evaluated by con-sidering the alveolar minimum and maximum diameterand excluding the areas of large airways or vesselsfrom analysis. The intra-alveolar distance was measuredas the mean linear intercept by standard method, utiliz-ing the same plug-in, by dividing the total length oflines drawn across the lung section (grid), by the num-ber of intercepts encountered. A cell counter wasapplied for assessing the secondary crests number/hpf.Lung histopathology analyses were performed by twoindependent researchers blinded to the treatment strate-gy, using ImageJ.

Anti-CD31 (1:50; DAKO Italia s.p.a., Milan, Italy)immunohistochemistry was performed. After incubationwith anti-mouse serum overnight (Envision, DAKO),3,30-diaminobenzidine (DAB, Sigma–Aldrich, Milan,Italy) containing H2O2 was used and counterstainedwith haematoxylin. Sections incubated without the pri-mary antibody showed no immunoreactivity, confirmingthe specificity of the immunostaining. Analysis of themicrovessel density (MVD) was blindly performed in10 hpfs at 20� of one representative slide. Vascularstructures were considered those with positive reactionand a visible lumen or well-defined linear vessel shape.They were mainly small capillaries, with narrow lumenand very thin walls, located in the alveolar walls. Themean MVDs, expressed as number of positive structuresfor hpf, were calculated for each case and for the entiresamples.

Immunofluorescence Staining and hAFSCells LungHoming

Anti-human mitochondria antibody Ab-2 (anti-hMIT)clone MTC02 (1:100; Thermo Fisher Scientific, Fre-mont, CA), rabbit anti-human C-protein surfactant anti-body (hSFTPC) (1:200; ProteinTech Group, Chicago,IL), rabbit anti-rat fibronectin polyclonal antibody(1:200; Abcam, Cambridge, UK) were detected byimmunofluorescence in 4 mm sections, incubated for45 min at 258C, then rinsed in PBS.

Alexa Fluor 488-labeled goat anti-mouse isotype-spe-cific antibodies diluted 1:3000 (Molecular Probes,Eugene, OR) and Alexa Fluor 594-labeled goat anti-rabbit IgG antibodies diluted 1:2000 (Molecular Probes)were used as secondary antibodies. The percentage ofhuman anti-mitochondria-positive cells was determinedamong lung cells (number of DAPI stained nuclei) perfield. Ten random fields were chosen for each lung (onsix different sections). Immunofluorescence was evalu-ated under the Leica DM 4000B microscope, integratedwith LAS (Leica Application Suite) software using3 dapi/green/orange filters. Specific stains for hSFTPC

and anti-hMIT were taken using a laser scanner confocalmicroscope (Model TCS-SL; Leica, Wetzlar, Germany).

Duplex PCR

We applied a method based on a duplex PCR productamplified from mitochondrial DNA corresponding tothe common human and rat cytochrome b, cyt b,(359 bp) and a 16S rRNA (157 bp) fragment, human-specific. DNA was extracted from the lung tissue usingthe Chelex Method28 or CST FORENSIC DNA PURIFKIT (Invitrogen) following the manufacturer’s instruc-tions. The sequences of primers used in a single PCRreaction are: cyt b forward 50-CCA TCC AAC ATCTCA GCA TGA TGA AA-30; cyt b reverse 50-GCCCCT CAG AAT GAT ATT TGT CCT CA-30; 16SrRNA forward 50-CAA TTG GAC CAA TCT ATCACC-30; 16S rRNA reverse 50-GTG AGG GTA ATAATG ACT TGT-30. After an initial denaturation step of1 min at 948C, samples were amplified in a PerkinElmer 9700 thermocycler, for 40 cycles of 5 sec at948C, 30 sec at 508C, and 40 sec at 728C, followed bya final elongation step of 3 min at 728C. The PCR reac-tion mix was as follows: 0.4 mM each of cyt b primers,0.6 mM each of 16S rRNA primers, 1.25 U Taq poly-merase (Applied Biosystems, Foster City, CA), 0.2 mMeach dNTPs, 2.5 ml 10� Taq Buffer (Applied Biosys-tems), in a final volume of 25 ml.

RT-PCR

For angiogenesis analisys the total RNA in the sam-ple was extracted using RNA Trizol according to themanufacturer’s protocol (Invitrogen) and the concentra-tion was measured (Nanodrop, Wilmington, DE) at260 nm. One microgram of total RNA from each lungsample was reverse transcribed into cDNA and real-time PCR-amplified with VEGF and GAPDH primersby the SYBR Green I method (Applied Biosystems).Reactions (12.5 ml 2� SYBR Master Mix, 300 nM pri-mers and template in a total volume of 25 ml) under-went denaturation at 958C for 10 min, followed by 40cycles of 958C for 30 sec and 608C for 1 min. Plasmidscontaining VEGF and GAPDH fragments were used astemplates for standard curves. VEGF mRNA levelsfor each sample were normalized using GAPDH as aninternal control. The sequence of primers used was asfollows: VEGF (for- 50-ATGACGAGGGCCTGGAGT-GTG-30; rev- 50-CCTATGTGCTGGCCTTGGTGAG-30)and GAPDH (for- 50-ACACCCACTCCTCCACCTTT-30; rev- 50-TCCACCACCCTGTTGCTGTA-30).

Cytokines ELISA Immunoassay

At the experimental endpoint, lungs were harvestedand homogenized in 300 ml of phosphate buffer pH 7.4

4 Grisafi et al.


and centrifuged at 15,000g for 10 min at 48C with pro-tease inhibitor, and were finally centrifuged to removedebris and supernatant before storage at �808C. Totalprotein content was measured in the supernatant byBradford’s method,29 (Bio-Rad Laboratories, Hercules,CA). The supernatant was then diluted to a final volumeof 500 ml with 0.9% NaCl saline containing 4% BSA(Sigma–Aldrich, St Louis, MO).

Concentration of rat IL-6, IL-1b, IFN-g, and TGF-b1 in lung homogenates was measured using ELISAassay kits according to the manufacturer’s instructions(Bender Medsystems, Vienna, Austria). Assay standardconcentration ranges were 31–2,000 pg/m.

Data Analysis and Statistics

Data are given as means � SE. Box plot was used toshow the median, the interquartile ranges and mini-mum/maximum values on average of the body weights.Scatter-plot graphics were made for morphometric datadistribution. The ANOVA test and the Neuman–Keulsmultiple comparison test were used for assessment ofdifferences among groups. A P-value <0.05 was con-sidered statistically significant (GraphPad Prism 5.04Software San Diego, CA).

RESULTS

All the 50 newborn rats randomly distributed be-tween the four experimental groups (Fig. 1A) survivedboth during oxygen exposure and after cell administra-tion procedures. They all gained weight throughout theP42 after birth with no differences among the groups(Fig. 1B).

hAFS Cells Colonize Damaged Lungs

Human AFS cells used in this study were character-ized by the stable presence of stromal cell markers asCD90 (Thy-1), CD105 (endoglin, TGFbeta receptor),and CD73, highly expressed (90%) along the entire pe-riod of culture. More than 30% of the AFS cells at thetime of injection were also Stem cell Embryonic Anti-gen-4 (SSEA-4) positive and they also expressed HLA-ABC while were negative for HLA-DR confirming theirlow immunogenicity profile (Fig. 2A). hAFS cells usedin the study showed, after appropriate culture condi-tions, to be able to undergo adipogenic and osteogenicdifferentiation (Fig. 2A).

Immunofluorescence revealed retention of humancells positive for anti-hMIT in all lung sections of ani-mals treated with hAFS cells at P42. Remarkably, cellsco-expressing hMIT and hSFTPC were found only inthe lungs of animals treated with hAFS cells. At theconfocality they appeared in close relationship with sec-ondary crests inside alveoli (Fig. 2B). The percentage

Fig. 2. A: hAFS cells pluripotence marker expression and

examples of flow cytometry profile of the injected stem cells;

hAFS cells used in the study showed, after appropriate culture

conditions, to be able to undergo adipogenic and osteogenic

differentiations. B: (a) Retained hAFS cells appeared in rat

lung sections stained with anti-human mitochondria (green).

Scale bar ¼ 15.9 mm, magnification 40�. B: (b) Lung sections

co-labeled (yellow) with anti-human surfactant C-protein (red),

and anti-human mitochondria (green), is shown in two repre-

sentative confocal images identifying the localization of hAFS

cells in treated animals. Scale bar ¼ 15.9 mm, magnification

40� (top); scale bar ¼ 23.8 mm, magnification 60� (bottom).

B: (c) Polyacrylamide-gel electrophoresis of a duplex PCR

product amplified from mitochondrial DNA. The hyperoxia

group not treated with human cells, lacks 16S rRNA.



of hAFS cells retained in the lungs accounted for anaverage of 1.43 � 0.2% (range: 1.07–3.21%) of thelung alveolar cells (Fig. 2B, b). In hAFS cells and hF-treated rats the presence of two bands (157 and 359 bp)indicates the presence of human cells, whereas a singleband (359 bp) in hyperoxia rats indicates a non-humanorigin of the samples (Fig. 2B, c).

hAFS Cells Ameliorate Lung Histopathology

hAFS cells treated animals highlighted a histologicalpattern with a marked improvement of homogenousalveoli, with the consequent recovery of alveolargrowth, in comparison with the untreated hyperoxiagroup. There was no evidence of inflammatory cells,but the septa were thicker in the untreated and speciallyin hF-treated hyperoxia groups, with respect to theroom air and hAFS cell groups. In contrast hF-treatedanimals showed no improved alveolar growth and per-sistent areas of fibrosis (Fig. 3A).

In the morphometric analyses, on average, the alveo-lar size (mm2) was lower in the room air group(1,049 � 60.89), when compared to the other groups(1,398 � 101.8 in hyperoxia, and 2,199 � 237.6 in hFgroups); hAFS cells (1,264 � 160.4), showed interme-diate values between room air and hyperoxia, thoughlower than hF, P < 0.0001; the value of mean linearintercept (mm) was lower in the room air group(71.64 � 1.74), when compared to hyperoxia only aswell as hF groups (79.69 � 1.84 and 94.25 � 2.46, re-spectively), hAFS-cells treated rats showed intermediatevalues between room air and hyperoxia (78.97 � 1.51),though lower than hF group, P < 0.0001.

Similarly, the secondary crests count (n/hpf) washigher in room air and hAFS cell-treated groups(40.36 � 2.56 and 31 � 2.36, respectively), rather thanhyperoxia and hF-treated groups (27.75 � 1.84 and5.6 � 2.01, respectively), P < 0.0001 (Fig. 3B).

hAFS Cells Sustain Rearrangement of CapillaryArchitecture and Exert an ImmunomodulatoryResponse

Anti-CD31 immunohistochemistry showed a morehomogeneous distribution of the capillary network inthe normoxia-exposed and hAFS cells treated animalswith respect to the other hyperoxic groups. In particu-lar, in normoxia there were numerous small capillaries,with narrow lumen and very thin walls, located in thealveolar walls. In untreated hyperoxia and hF-treatedgroups the alveolar walls showed some thickenings,where clear capillary structures were more rare and dif-ficult to identify. In the hAFS group, the capillariesof the alveolar walls were more clearly appreciable(Fig. 4B). Morphometric analysis showed a decreasedMVD in the untreated hyperoxia (83.72 � 7.57/hpf)

and hF-treated group (67.10 � 4.55/hpf) in comparisonto the room air group (146.7 � 6.95/hpf) and hAFSgroup (104.1 � 6.44/hpf) (P < 0.0001) (Fig. 4A).

Interestingly, VEGF gene expression in the lungs ofroom air rats (121.5 � 31.43) and exposed to hyper-oxia, but treated with hAFS cells (90.77 � 18.44) washigher than in untreated hyperoxia (16.77 � 4.24) andhF-treated (31.49 � 4.18) rats (P < 0.0004; Fig. 4A).The concentration of tissue IL-6, IL-1b, IF-g, andTGF-1b decreased at a value close to the room airgroup level in the hAFS cell group, when compared tothe untreated hyperoxia and hF groups, P-values arefrom 0.01 to 0.038 (Table 1).

DISCUSSION

The role of stem cell therapy in animal models ofBPD has been studied mainly during the neonatal peri-od.9,22,30 Unfortunately limited data are available on themedium and long term follow-up in a context of moder-ate hyperoxia and stem cell administration.22 Therefore,our model has been developed to focus on the chroniclung disease rather than the acute injury. Cell adminis-tration was performed at P21, after 2 weeks of moder-ate hyperoxia and monitored up to P42. This designmimics the late phase of the BPD in a more advancedphase of the disease, which beginning in infancyextends to childhood and adulthood. For the in vivoprocedure the optimal number of stem cells proposedand the proper time for instillation were set up as previ-ously investigated at the beginning of the weaningphase (P21; data not shown). Like in postnatal surfac-tant instillation in neonates, hAFS cells were injectedintratracheally, this being shown to be the most efficientway for cell delivery and recruitment.18

Rats exposed to moderate hyperoxia for 14 days,showed a variable response to injury in several lungregions. At P42 the lung histopathology was rather wellcharacterized by an arrested alveolarization and dys-plastic vascular architecture, similar to the patternsreported in human BPD.31 Impaired alveolar growthand heterogeneous emphysematous areas were alsoconfirmed in our histopathology panels.

In this study we have described the effects of stemcell administration on lung mean linear intercept, alveo-lar size, number of secondary crests, pulmonary contentof VEGF, and the cytokines profile. In terms of remod-eling, hAFS cells triggered an interesting pulmonarymicrovascular development that is essential for an effi-cient gas exchange. This was also supported by thealveolar number and size more similar to the lungs ofthe room air group, with a marked improvement of ho-mogenous alveoli and a consequent recovery of alveolargrowth. We detected hAFS cells around the terminalbronchiolar epithelium, in the interstitial spaces and in

6 Grisafi et al.


the alveolar walls, indicating a selective lung injury tro-pism. In our study, alveolar type II cells might be spec-ulated to be differentiated, in the co-stained surfactantC-protein and human mitochondria cells. The anti-sur-factant antibody stained the inner alveolar and bronchi-olar terminal walls; cubic or cylindrically-shaped cellsco-stained with anti-human mitochondria and anti-

surfactant antibodies, showed a thick cytoplasm edgetowards the alveolar space. As we found no evidence ofmultinucleate hAFS cells, we support the hypothesis ofa cell epithelial differentiation rather than fusion. Dif-ferently sections from hF-treated rats did not showenough fluorescence signal to be detect, therefore wewere unable to assess their percentage of retention,

Fig. 3. A: Representative lungs from the studied groups: Room air, hyperoxia, hAFS and hF

cell treated groups. H&E, scale bar ¼ 100 mm, magnification 20�. Lung histopathology analy-

sis at P42. Representative H&E-stained lung sections show that untreated and hF-treated

hyperoxia groups were associated with arrested alveolarization, inducing changes in lung

morphology with patchy areas of parenchymal thickening and enlarged air spaces (black

arrows). Lung sections of hAFS cell treated rats contained smaller, more numerous alveoli

(black arrows) and were comparable with those of the room air group. B: Scatterplots of aver-

age alveolar size, mean linear intercept, and secondary crests. At P42 the mean alveolar size

(mm2) was significantly smaller in the room air and hAFS cell groups, when compared to the

hF-treated hyperoxia group; P < 0.0001; �P < 0.05 room air versus hyperoxia and hF-treated

groups, §P < 0.001 hyperoxia versus hF group. The mean linear intercept (mm) was lower in

the room air group than in the untreated and hF-treated hyperoxia groups, though the latter

was higher than that of the hAFS cell group; P < 0.0001; �P < 0.05 room air versus the three

other groups, §P < 0.001 hyperoxia versus the hF-treated group. The secondary crests (n/hpf)

were higher in the room air group than in the untreated and hF-treated hyperoxia groups,

hAFS cell-treated showed an increase in secondary crests regarding the hF-treated group,

P < 0.0001; �P < 0.001 room air versus the hyperoxia and hF-treated groups, §P < 0.001

hyperoxia versus the hF-treated group.



which was assumed to be very low. The molecular anal-ysis performed with PCR on mitochondrial DNA inlung parenchyma of hAFS and hF-treated rats, corre-sponding to the common human and rat cyt b, and a

16S rRNA fragment, suggest that the human cells arelikely to move from the site of injection, from the tra-chea to the broncholi and hence to the interstitium andalveoli. Although both human cell lines were detectable

Fig. 4. A: Left: VEGF gene expression rose in hyperoxic animals due to the hAFS cell trans-

plantation, reaching the room air level, P ¼ 0.0004; �P ¼ 0.004 room air versus hyperoxia and

hF-treated groups, §P ¼ 0.04 hyperoxia versus hAFS cells-treated groups. Right: representa-

tive data columns of the microvessel density from the studied groups confirmed that the

hAFS cell-treated group rose in CD31 positive cells to an intermediate value when compared

to the room air and untreated or hF-treated hyperoxia groups, P < 0.0001; �P < 0.001 room air

versus the three other groups, §P < 0.01 hyperoxia versus hAFS cell-treated groups. B: Immu-

nohistochemistry staining for CD31 demonstrates an higher number of capillaries in the alve-

olar walls of the room air and hAFS cells treated groups in comparison to the hyperoxia

untreated and hF-treated groups. Note the presence in the latter two hyperoxic groups

of thickenings of the alveolar walls with very few capillary structures (Scale bar ¼ 100 mm;

magnification 20�).

TABLE 1— IL-6, IL-1b, IF-g, and TGF-1b Concentrations Were Estimated in the Studied Groups (n ¼ 10 Biopsies inRoom Air Group, n ¼ 8 in Untreated, and hF-Treated Hyperoxia, n ¼ 7 in hAFS Cell-Treated Group)

Cytokines

(pg/mg of lung tissue) Room air Hyperoxia hAFS hF

IL-6 261.9 � 35.75� 465.3 � 53.62§ 296.1 � 62.62 481.3 � 56.92

IL-1b 464.7 � 58.77� 803.1 � 62.52§ 548.4 � 57.32 819.2 � 164.0

IF-g 210.3 � 20.60� 313.5 � 31.02 240.4 � 35.76 362.7 � 70.42

TGF-1b 420.5 � 59.74� 644.8 � 57.72§ 455.3 � 59.54 673.9 � 65.65

hAFS cell-administered animals showed intermediate values between untreated or hF-treated hyperoxia groups and room air group.�P-values from <0.01 to 0.038 room air versus hF-treated groups.§The same P-values hyperoxia versus room air groups.

8 Grisafi et al.


with PCR, a weaker intensity of the amplified band wasfound in hF-treated samples than hAFS cells. It seemsencouraging that the percentage we found of thedetected hAFS cells was consistent with that observedin previous works. In fact, the recruitment rate of exog-enous stem cells into the lungs remains controversial32

and is estimated to be from 0.01% to 0.1%,33 thoughnot more than 5%12 with the intratracheal route whichseems to enhance the retention rate to 5–10%, whencompared to the intravenous route.12,34 Unfortunatelyfluorescence techniques, although considered the goldstandard, may be a limit and are likely to explain manyof the conflicting results.35 The lack of cell division andproliferation tests could be a limitation of our study andhence this field shall have to be elucidated in furtherwork.

Since a growing number of studies7,36 have focusedon the alveolar vasculature and on factors that may reg-ulate vascular development in BPD, we determined thelung gene expression of VEGF and the content ofCD31. In the mature lung, alveolar capillaries are abun-dant and lie close to the alveolar epithelium, creating athin air–blood barrier. The mechanisms behind the re-pair processes may involve a paracrine effect that nor-malizes VEGF content and improves the alveolarcapillary bed architecture. The lower lung VEGF geneexpression found after oxygen exposure in untreatedanimals, confirms that VEGF signaling is disrupted inour model. Since normal VEGF gene expression playsa crucial role in the proper metabolism of the endotheli-al and alveolar barrier being a pro-angiogenic andendothelial survival factor,30 it is worth noting thatin hAFS cell-treated group the profile we found mayreflect a ‘‘healthier’’ state of the lung, more similar tothat of the room air controls. Nevertheless some cautionneeds to be exerted with respect to this interpretation ofVEGF content because there is no evidence presentedthat increased VEGF is coming from the retained hAFSor alveolar type II cells. Finally we are aware that afurther limitation of our work is that only a single timepoint was studied and expression could have been quitedifferent at earlier time points.

It is acknowledged37,38 that cell therapy contributeswith stem cells as producers of growth factors or modi-fiers of tissues with generalized effect, a fact todayknown behind the ‘‘stemness concept’’. We confirmedthat c-kit selected hAFS cells differentiate, when condi-tioned with adipogenic and osteogenic media, as previ-ously reported23 and express the characteristic surfacemarkers. It is worth noting that thanks to the low ex-pression of HLA-DR these cells may trigger a lowerimmune-response,39 which was also confirmed by theabsence of inflammatory areas in the recipients. For thisreason we then decided to avoid the immunosuppres-sant drugs, to preclude interference in stem cell

activity.40,41 In addition the low immune-response andthe reported absence of tumorigenesis justify in ourexperimental design the preclusion of a specific controlgroup in healthy animals administrated with hAFScells.23,25,42

We have also looked at the role of apoptosis in lungremodeling since epithelial cell apoptosis may be a keyfactor involved in alveolar epithelial regeneration andrepair.8 In our study, we found that the apoptotic indexchanged significantly in hyperoxia exposed rats versusroom air group (data not shown). This is quite an inter-esting point, as already demonstrated,43,44 because epi-thelial cell death is followed by remodeling processes,which consist of epithelial and fibroblast activation, cy-tokine production, activation of coagulation pathway,neoangiogenesis, and re-epithelialization. We thereforeinvestigated four cytokines normally involved in the im-mune and inflammatory responses43 and in the experi-mental lung injury condition caused by moderatehyperoxia. Interestingly, while IL-6, IL-1b, IFN-g, andTGF-1b, are significantly increased in hyperoxia andhF-treated rats at P42, the administration of hAFS cellsseems to decrease them. Angelini et al.45 has proventhe ability of stem cell therapy to counterbalance thenegative effect of the pro-inflammatory response 7 daysafter the experimental lung injury. Nevertheless a de-crease in all the cytokines we studied may still not nec-essarily indicate proper long term anti-inflammatoryeffects, instead it may indicate a ‘‘less injured’’ state ofthe lung, which was achieved by the AFS cell treat-ment. Here we cannot demonstrate if this late effect issecondary to a citoprotective action or a paracrine se-cretion mechanism. In light of this we consider that thelack of the assessment of anti-inflammatory cytokines’wide panel, that is, IL-10, IL-17A,46 or conditioned me-dia,21 etc. may represent presently an important limita-tion of our work and therefore deserves furtherinvestigation.

Considering the in vivo models of lung injury,9,17

the paracrine immunomodulation of stem cells and theprotective action in the parenchyma and vascular lunginjury result crucial for pulmonary disease. Chronicpulmonary disease occurence, mainly corresponding tothe moderate/severe forms of new BPD, was associatedwith increased pro-inflammatory and pro-fibrotic/angio-genic cytokines, while mild forms of new BPD werecharacterized only by the increase of pro-fibrotic/angio-genic cytokines, suggesting a different balance of twopathogenic mechanisms in different phases of the dis-ease.47 Taken together both cytokine pattern and cellretention are encouraging results, although the low per-centage of human cell uptake among alveolar cellsdetected and the specific mechanism of the anti-inflammatory effect needs to be clarified. The authorsconsider that the effects they found in the lung, are



predominantly related to a paracrine action of the hAFScells: further investigation will be needed to ascertainthe potential interaction between administration ofhAFS and their conditioned media on lung.21 Taken to-gether these encouraging results candidate the presentwork to represent a vital first step in demonstrating theimpact that this unique cell population may have oncellular based therapies for lung diseases.

In summary, postnatal treatment with hAFS cells in amoderate hyperoxia exposure, significantly reduceslung pathology disease in young rats by improvingalveolarization. Injection of hAFS cells might have areparative potential, improving angiogenesis with con-sequent organization of capillary networks. Intratrachealdelivery of hAFS cells succeeded to achieve someimmunomodulatory effects even for a relatively longtime. Taken together, this evidence is encouraging fromthe perspective of new approaches for the treatment oflate BPD disease in humans.

ACKNOWLEDGMENTS

The authors also thank the personnel of the Anatomyand Physiology Department, Dr. G. Sarasin, A. Ram-baldo, D. Guidolin, and the Section of Pathology, of theDepartment of Oncological and Surgical Sciences, inparticular Dr. V. Guzzardo, the NICU personnel and theAssociazione Pulcino. M.P. is supported by Citta’ dellaSperanza. P.D.C. is supported by the Great OrmondStreet Hospital Charity. D.G., R.T., A.D., A.P., V.M.,R.S., M.S., A.F., R.D.C., L.C., and P.Z. are supportedby the University of Padua (CPDA081132/08).

REFERENCES

1. Kwinta P, Bik-Multanowski M, Mitkowska Z, Tomasik T,

Legutko M, Pietrzyk JJ. Genetic risk factors of bronchopulmo-

nary dysplasia. Pediatr Res 2008;64:682–688.

2. Cheah FC, Pillow JJ, Kramer BW, Polglase GR, Nitsos I, Newn-

ham JP, Jobe AH, Kallapur SG. Airway inflammatory cell

responses to intra-amniotic lipopolysaccharide in a sheep model

of chorioamnionitis. Am J Physiol Lung Cell Mol Physiol 2009;

296:L384–L393.

3. Trachsel D, Nichols DE, Kidd S, Hadorn M, Baumberger F.

4-aryl-substituted 2,5-dimethoxyphenethylamines: synthesis and

serotonin 5-HT(2A) receptor affinities. Chem Biodivers 2009;

6:692–704.

4. Wong PM, Lees AN, Louw J, Lee FY, French N, Gain K,

Murray CP, Wilson A, Chambers DC. Emphysema in young

adult survivors of moderate-to-severe bronchopulmonary dys-

plasia. European Respir J 2008;32:321–328.

5. Filippone M, Sartor M, Zacchello F, Baraldi E. Flow limitation

in infants with bronchopulmonary dysplasia and respiratory

function at school age. Lancet 2003;361:753–754.

6. Cutz E, Chiasson D. Chronic lung disease after premature birth.

N Engl J Med 2008;358:743–745, author reply 745–746.

7. Balasubramaniam V, Mervis CF, Maxey AM, Markham NE,

Abman SH. Hyperoxia reduces bone marrow, circulating,

and lung endothelial progenitor cells in the developing

lung: implications for the pathogenesis of bronchopulmonary

dysplasia. Am J Physiol Lung Cell Mol Physiol 2007;292:

L1073–L1084.

8. Yi M, Masood A, Ziino A, Johnson BH, Belcastro R, Li J, Shek

S, Kantores C, Jankov RP, Tanswell AK. Inhibition of apoptosis

by 60% oxygen: a novel pathway contributing to lung injury in

neonatal rats. Am J Physiol Lung Cell Mol Physiol 2011;

300:L319–L329.

9. van Haaften T, Byrne R, Bonnet S, Rochefort GY, Akabutu J,

Bouchentouf M, Rey-Parra GJ, Galipeau J, Haromy A, Eaton F,

Chen M, Hashimoto K, Abley D, Korbutt G, Archer SL, The-

baud B. Airway delivery of mesenchymal stem cells prevents

arrested alveolar growth in neonatal lung injury in rats. Am J

Respir Crit Care Med 2009;180:1131–1142.

10. Alphonse RS, Thebaud B. Growth factors, stem cells and bron-

chopulmonary dysplasia. Neonatology 2011;99:326–337.

11. Alphonse RS, Rajabali S, Thebaud B. Lung injury in preterm

neonates: the role and therapeutic potential of stem cells. Anti-

oxid Redox Signal 2012;17:1013–1040.

12. Thebaud B. Update in pediatric lung disease 2010. Am J Respir

Crit Care Med 2011;183:1477–1481.

13. Balasubramaniam V, Ryan SL, Seedorf GJ, Roth EV, Heumann

TR, Yoder MC, Ingram DA, Hogan CJ, Markham NE, Abman

SH. Bone marrow-derived angiogenic cells restore lung alveolar

and vascular structure after neonatal hyperoxia in infant mice.

Am J Physiol Lung Cell Mol Physiol 2010;298:L315–L323.

14. Tropea KA, Leder E, Aslam M, Lau AN, Raiser DM, Lee JH,

Balasubramaniam V, Fredenburgh LE, Alex Mitsialis S, Kour-

embanas S, Kim CF. Bronchioalveolar stem cells increase after

mesenchymal stromal cell treatment in a mouse model of bron-

chopulmonary dysplasia. Am J Physiol Lung Cell Mol Physiol

2012;302:L829–L837.

15. Pierro M, Thebaud B. Mesenchymal stem cells in chronic lung

disease: culprit or savior? Am J Physiol Lung Cell Mol Physiol

2010;298:L732–L734.

16. Bernardo ME, Cometa AM, Locatelli F. Mesenchymal stromal

cells: a novel and effective strategy for facilitating engraftment

and accelerating hematopoietic recovery after transplantation?

Bone Marrow Transplant 2012;47:323–329.

17. Aslam M, Baveja R, Liang OD, Fernandez-Gonzalez A, Lee C,

Mitsialis SA, Kourembanas S. Bone marrow stromal cells atten-

uate lung injury in a murine model of neonatal chronic lung

disease. Am J Respir Crit Care Med 2009;180:1122–1130.

18. Chang YS, Oh W, Choi SJ, Sung DK, Kim SY, Choi EY, Kang

S, Jin HJ, Yang YS, Park WS. Human umbilical cord blood-

derived mesenchymal stem cells attenuate hyperoxia-induced

lung injury in neonatal rats. Cell Transplant 2009;18:869–886.19. Hansmann G, Fernandez-Gonzalez A, Aslam M, Vitali SH,

Martin T, Mitsialis SA, Kourembanas S. Mesenchymal stem

cell-mediated reversal of bronchopulmonary dysplasia and asso-

ciated pulmonary hypertension. Pulm Circ 2012;2:170–181.20. Pozzobon M, Ghionzoli M, De Coppi P. ES, iPS, MSC, and

AFS cells. Stem cells exploitation for pediatric surgery: current

research and perspective. Pediatr Surg Int 2010;26:3–10.

21. Moorefield EC, McKee EE, Solchaga L, Orlando G, Yoo JJ,

Walker S, Furth ME, Bishop CE. Cloned, CD117 selected

human amniotic fluid stem cells are capable of modulating the

immune response. PLoS ONE 2011;6:e26535.22. Carraro G, Perin L, Sedrakyan S, Giuliani S, Tiozzo C, Lee J,

Turcatel G, De Langhe SP, Driscoll B, Bellusci S, Minoo P,

Atala A, De Filippo RE, Warburton D. Human amniotic fluid

stem cells can integrate and differentiate into epithelial lung

lineages. Stem Cells 2008;26:2902–2911.23. De Coppi P, Bartsch G, Jr., Siddiqui MM, Xu T, Santos CC,

Perin L, Mostoslavsky G, Serre AC, Snyder EY, Yoo JJ, Furth

ME, Soker S, Atala A. Isolation of amniotic stem cell lines with

potential for therapy. Nat Biotechnol 2007;25:100–106.

10 Grisafi et al.


24. Perin L, Sedrakyan S, Da Sacco S, De Filippo R. Characteriza-

tion of human amniotic fluid stem cells and their pluripotential

capability. Methods Cell Biol 2008;86:85–99.

25. Rota C, Imberti B, Pozzobon M, Piccoli M, De Coppi P, Atala

A, Gagliardini E, Xinaris C, Benedetti V, Fabricio AS, Squar-

cina E, Abbate M, Benigni A, Remuzzi G, Morigi M. Human

amniotic fluid stem cell preconditioning improves their regener-

ative potential. Stem Cells Dev 2012;21:1911–1923.

26. Dauger S, Ferkdadji L, Saumon G, Vardon G, Peuchmaur M,

Gaultier C, Gallego J. Neonatal exposure to 65% oxygen dura-

bly impairs lung architecture and breathing pattern in adult

mice. Chest 2003;123:530–538.

27. Grisafi D, Tassone E, Dedja A, Oselladore B, Masola V, Guz-

zardo V, Porzionato A, Salmaso R, Albertin G, Artusi C, Zani-

notto M, Onisto M, Milan A, Macchi V, De Caro R, Fassina A,

Bordigato MA, Chiandetti L, Filippone M, Zaramella P.

L-Citrulline prevents alveolar and vascular derangement in a rat

model of moderate hyperoxia-induced lung injury. Lung 2012;

190:419–430.

28. Bataille M, Crainic K, Leterreux M, Durigon M, de Mazancourt

P. Multiplex amplification of mitochondrial DNA for human

and species identification in forensic evaluation. Forensic Sci

Int 1999;99:165–170.

29. Bradford MM. A rapid and sensitive method for the quantitation

of microgram quantities of protein utilizing the principle of pro-

tein–dye binding. Anal Biochem 1976;72:248–254.

30. Balasubramaniam V, Ingram DA. Endothelial progenitors in the

risk of developing bronchopulmonary dysplasia: can we include

endothelial progenitor cells in BPD risk assessment? Am J

Respir Crit Care Med 2009;180:488–490.

31. Baraldi E, Filippone M. Chronic lung disease after premature

birth. N Engl J Med 2007;357:1946–1955.

32. Loebinger MR, Aguilar S, Janes SM. Therapeutic potential of

stem cells in lung disease: progress and pitfalls. Clin Sci 2008;

114:99–108.

33. Krause DS. Engraftment of bone marrow-derived epithelial

cells. Ann N Y Acad Sci 2005;1044:117–124.

34. Wong AP, Keating A, Lu WY, Duchesneau P, Wang X, Sacher

A, Hu J, Waddell TK. Identification of a bone marrow-derived

epithelial-like population capable of repopulating injured mouse

airway epithelium. J Clin Invest 2009;119:336–348.

35. Leblond AL, Naud P, Forest V, Gourden C, Sagan C, Romefort

B, Mathieu E, Delorme B, Collin C, Pages JC, Sensebe L, Pit-

ard B, Lemarchand P. Developing cell therapy techniques for

respiratory disease: intratracheal delivery of genetically engi-

neered stem cells in a murine model of airway injury. Hum

Gene Ther 2009;20:1329–1343.

36. Sueblinvong V, Weiss DJ. Stem cells and cell therapy

approaches in lung biology and diseases. Transl Res 2010;156:

188–205.

37. Bernardo ME, Ball LM, Cometa AM, Roelofs H, Zecca M,

Avanzini MA, Bertaina A, Vinti L, Lankester A, Maccario R,

Ringden O, Le Blanc K, Egeler RM, Fibbe WE, Locatelli F.

Co-infusion of ex vivo-expanded, parental MSCs prevents life-

threatening acute GVHD, but does not reduce the risk of graft

failure in pediatric patients undergoing allogeneic umbilical

cord blood transplantation. Bone Marrow Transplant 2011;46:

200–207.

38. Heijnen CJ, Witt O, Wulffraat N, Kulozik AE. Stem cells in

pediatrics: state of the art and future perspectives. Pediatr Res

2012;71:407–409.

39. Bollini S, Pozzobon M, Nobles M, Riegler J, Dong X,

Piccoli M, Chiavegato A, Price AN, Ghionzoli M, Cheung KK,

Cabrelle A, O’Mahoney PR, Cozzi E, Sartore S, Tinker A,

Lythgoe MF, De Coppi P. In vitro and in vivo cardiomyogenic

differentiation of amniotic fluid stem cells. Stem Cell Rev 2011;

7:364–380.

40. Broekema M, Harmsen MC, Koerts JA, van Kooten TG, Uges

DR, Petersen AH, van Luyn MJ, Navis G, Popa ER. Ciclosporin

does not influence bone marrow-derived cell differentiation to

myofibroblasts early after renal ischemia/reperfusion. Am J

Nephrol 2009;30:73–83.

41. Nifontova I, Svinareva D, Petrova T, Drize N. Sensitivity of

mesenchymal stem cells and their progeny to medicines used

for the treatment of hematoproliferative diseases. Acta Haema-

tol 2008;119:98–103.

42. Gosemann JH, Kuebler JF, Pozzobon M, Neunaber C, Hensel

JH, Ghionzoli M, de Coppi P, Ure BM, Holze G. Activation of

regulatory T cells during inflammatory response is not an exclu-

sive property of stem cells. PLoS ONE 2012;7:e35512.

43. Chang YS, Choi SJ, Sung DK, Kim SY, Oh W, Yang YS, Park

WS. Intratracheal transplantation of human umbilical cord blood

derived mesenchymal stem cells dose-dependently attenuates

hyperoxia-induced lung injury in neonatal rats. Cell Transplant

2011.

44. Harijith A, Choo-Wing R, Cataltepe S, Yasumatsu R, Aghai

ZH, Janer J, Andersson S, Homer RJ, Bhandari V. A role for

matrix metalloproteinase 9 in IFNgamma-mediated injury in de-

veloping lungs: relevance to bronchopulmonary dysplasia. Am J

Respir Cell Mol Biol 2011;44:621–630.

45. Angelini A, Castellani C, Ravara B, Franzin C, Pozzobon M,

Tavano R, Libera LD, Papini E, Vettor R, De Coppi P, Thiene

G, Vescovo G. Stem-cell therapy in an experimental model of

pulmonary hypertension and right heart failure: role of paracrine

and neurohormonal milieu in the remodeling process. J Heart

Lung Transplant 2011;30:1281–1293.

46. Andreev K, Graser A, Maier A, Mousset S, Finotto S. Thera-

peutical measures to control airway tolerance in asthma and

lung cancer. Front Immunol 2012;3:216.

47. Vento G, Capoluongo E, Matassa PG, Concolino P, Vendettuoli

V, Vaccarella C, Frezza S, Zuppi C, Romagnoli C, Ameglio F.

Serum levels of seven cytokines in premature ventilated new-

borns: correlations with old and new forms of bronchopulmo-

nary dysplasia. Intensive Care Med 2006;32:723–730.



Human amniotic fluid stem cells protect rat lungs exposed to moderate hyperoxia

Documents