Skip Navigation


Carcinogenesis Advance Access originally published online on October 9, 2008
Carcinogenesis 2009 30(1):150-157; doi:10.1093/carcin/bgn234
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
30/1/150    most recent
bgn234v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Tasso, R.
Right arrow Articles by Pennesi, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tasso, R.
Right arrow Articles by Pennesi, G.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Development of sarcomas in mice implanted with mesenchymal stem cells seeded onto bioscaffolds

Roberta Tasso1,2, Andrea Augello1,2, Michela Carida’1,2, Fabio Postiglione1,2, Maria Grazia Tibiletti3, Barbara Bernasconi3, Simonetta Astigiano4, Franco Fais5, Mauro Truini6, Ranieri Cancedda1,2 and Giuseppina Pennesi1,2,7,*

1 Department of Oncology, Biology and Genetics, University of Genova
2 Laboratory of Regenerative Medicine, National Cancer Research Institute, Largo Rosanna Benzi 10, 16132 Genova, Italy
3 Department of Pathology, Ospedale di Circolo, Viale Borri 57, 20121 Varese, Italy
4 Embryogenesis and Tumorigenesis in Animal Models, National Cancer Research Institute, Largo Rosanna Benzi 10, 16132 Genova, Italy
5 Department of Experimental Medicine, University of Genova, Via Leon Battista Alberti 2, 16132 Genova, Italy
6 Department of Diagnostic Technologies, National Cancer Research Institute, Largo Rosanna Benzi 10, 16132 Genova, Italy
7 Present address: Stem Cells Laboratory, Advanced Biotechnology Center, Largo Rosanna Benzi 10, 16132 Genova, Italy

* To whom correspondence should be addressed. Tel: +39 010 5737511; Fax: +39 010 5737505; Email: pina.pennesi{at}istge.it


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 
Bone marrow-derived mesenchymal stem cells (MSCs) are precursors of bone, cartilage and fat tissue. MSC can also regulate the immune response. For these properties, they are tested in clinical trials for tissue repair in combination with bioscaffolds or injected as cell suspension for immunosuppressant therapy. Experimental data, however, indicate that MSC can undergo or induce a tumorigenic process in determined circumstances. We used a modified model of ectopic bone formation in mice by subcutaneously implanting porous ceramic seeded with murine MSC. In this new model, host-derived sarcomas developed when we implanted MSC/bioscaffold constructs into syngeneic and immunodeficient recipients, but not in allogeneic hosts or when MSCs were injected as cell suspensions. The bioscaffold provided a tridimensional support for MSC to aggregate, thus producing the stimulus for triggering the process eventually leading to the transformation of surrounding cells and creating a surrogate tumor stroma. The chemical and physical characteristics of the bioscaffold did not affect tumor formation; sarcomas developed either when a stiff porous ceramic was used or when the scaffold was a smooth collagen sponge. The immunoregulatory function of MSC contributed to tumor development. Implanted MSC expanded clones of CD4+CD25+ T regulatory lymphocytes that suppressed host’s antitumor immune response.

Abbreviations: GFP-Tg, green fluorescent protein transgenic; FISH, fluorescence in situ hybridization; HA, hydroxyapatite; MSC, mesenchymal stem cell; s.c., subcutaneously; Treg, regulatory T lymphocyte


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 
Tumor formation is a complex event involving the uncontrolled proliferation of cells and the failure of the antitumor immune response. Tumors are formed when embryonic stem cells are injected in recipient animals (1). It has been reported that also adult stem cells could trigger tumor formation. In particular, human bone marrow-derived mesenchymal stem cells (MSCs) have been shown to favor tumor growth by promoting neoangiogenesis (2) and preventing tumor recognition by the immune system (3). They also induce the metastasis of breast cancer cells constituting a stroma delivering a continuous tumorigenic stimulus for metastasis of breast cancer cells (4). On the contrary, other studies have observed that MSC inhibits tumor growth in murine (5) and rat models (6,7).

The tumorigenic potential of MSC has been thus far studied using experimental protocols where MSC are injected as cell suspensions into immunodeficient recipient animals, thus focusing specifically on the biological pathways of cell transformation. However, the use of MSC for tissue engineering approaches implies the presence of a bioscaffold to mold and facilitate the de novo development of a functional tissue. The plastic interactions between MSC and the bioscaffold could modify the oncogenic potential of MSC (8) or influence the host’s response to the cell graft. This is a critical issue since MSC are often being used in the clinical arena for tissue repair in combination with suitable bioscaffolds (9), and the range of potential therapeutic applications is widening [see (10) and (11)].


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Supplementary material
 Funding
 References
 
Mice
C57Bl/6 (MHC H2b haplotype), Balb/c (MHC H2d haplotype) and Swiss-Nude (nude) mice were purchased from Charles River Laboratories (Calco, Province of Lecco, Italy). C57Bl/6-Tg(ACTB-EGFP)1Osb/J mice [green fluorescent protein transgenic (GFP-Tg)] (MHC H2b haplotype) were purchased from The Jackson Laboratory (Bar Harbor, MA). Mice were used between 5 and 8 weeks of age. All animals were bred and maintained at the Institution’s animal facility in 12 h light–dark cycles and fed ad libitum. The care and use of the animals were in compliance with laws of the Italian Ministry of Health and the guidelines of the European Community and the Internal Ethic Committee had approved all animal experiments.

Cell isolation and culture
MSCs were obtained from C57Bl/6 or GFP-Tg mice. Murine bone marrow cells were collected and cultured as described previously.

Briefly, mice were killed and bone marrow cells were collected by flushing nucleate cells out of the femurs and tibiae with cold phosphate-buffered solution. Cells were cultured (10 x 106 nucleated cells per 10 cm Petri dish) in Coon’s modified F12 medium (Biochrom AG, Berlin, Germany) supplemented with 10% fetal calf serum (GIBCO, S.Giuliano Milanese, Milano, Italy), 2 mM L-glutamine, 50 µg/ml penicillin/streptomycin and 1 ng/ml basic fibroblast growth factor (standard medium). Cultures were incubated at 37°C in a 5% CO2 atmosphere. After 3 days, non-adherent cells were removed. When 80% confluent, adherent cells were trypsinized (0.05% trypsin/ethylenediaminetetraacetic acid at 37°C for 10 min) and expanded (P1 stage).

At this passage, part of the cultures were seeded onto a bioscaffold and in vivo implanted, whereas cells of the same culture were kept for further experiments.

We checked the immunophenotype of MSC by flow cytometry using monoclonal antibodies to CD11b, CD14, CD34, CD44, CD45.2, CD106, CD140a, H2-Kb and H2-IAb (clone M1/70, rmC5-3, RAM34, IM7, 104, 429, APA5, AF6-88.5 and AF6-120.1, respectively) (BD Pharmingen, Milano, Italy) and using a polyclonal antibody to CD31 (Santa Cruz Biotechnology, Heidelberg, Germany).

Spleens from C57Bl/6 or Balb/c mice were collected, and splenocytes were obtained by mechanical shredding and collected in RPMI 1640 medium (Sigma, Milano, Italy) supplemented with 2-mercaptoethanol, glutamine, non-essential amino acids, sodium pyruvate, antibiotics (Sigma) and 1% normal mouse serum.

CD4+CD25+ regulatory T lymphocytes (Tregs) were separated from splenocytes derived from C57Bl/6 or Balb/c mice. We first performed a T lymphocyte enrichment using a Pan T Cell Isolation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany) that depleted the cell suspension of non-T lymphocytes. Then we isolated Tregs from this CD4+-enriched population using the CD4+CD25+ Regulatory T Cell Isolation Kit (Miltenyi Biotec). Cells were counted and examined by flow cytometry to evaluate the purity of the cell suspension using monoclonal antibodies to CD4 and CD25 (clone GK1.5 and PC61, respectively) (BD Pharmingen).

Cultures of sarcoma cells were established from biopsies obtained from primary tumors grown in immunocompromised or syngeneic implanted mice. Briefly, specimens were minced; tumor chips were plated in 10 cm Petri dishes and cultured in standard medium for further experiments.

MG-63 cells, a cell line derived from a human osteosarcoma (12), were cultured in standard medium and served as a positive control for the soft agar colony formation assay.

Surgical procedures
A highly porous ceramic support based on 100% hydroxyapatite (HA) with a Ca:P ratio of 1.66 ± 0.5 and 80 ± 5% porosity (Engi Pore) (FinCeramica, Faenza, Italy) and a cross-linked collagen sponge (Cymatese biomateriaux, Chaponost, France) were used as bioscaffolds.

After 3 or 4 weeks expansion, MSCs were detached from the dishes with 0.05% trypsin/ethylenediaminetetraacetic acid, washed in serum-free medium and resuspended at 2.5 x 106 cells per 20 µl fibrinogen (2.5 mg/ml) (Baxter, Milano, Italy). Each aliquot of MSC was then applied on the opposite faces of a heat-sterilized (180°C for 8 h) HA block or collagen sponge. Twenty microliters of thrombin (500 IU/ml) (Baxter) was applied to each scaffold.

After anesthesia, recipient Balb/c, C57Bl/6 and nude mice were subcutaneously (s.c.) implanted with MSC/bioscaffold constructs. More than five animals per group were used for each experiment, repeating experiments at least three times. Similar experiments were performed using MSC derived only from male C57Bl/6 mice and implanted into female C57Bl/6 recipients.

The tumorigenic potential of primary tumors grown in immunocompromised or syngeneic mice was assessed in 4-week-old nude mice. All animals were maintained in a sterile environment. Briefly, parts of primary tumors of ~2 mm of diameter were explanted in sterile conditions from immunocompromised and syngeneic recipients, s.c. implanted into the left flank of five nude mice and observed for 12 weeks. We performed three serial in vivo transplantation passages.

Animals were killed when the growth of a tumor mass was incompatible with the mouse well-being, usually 4 weeks after implantation of MSC/collagen sponge constructs and 8 weeks after implantation of MSC/HA combinations.

As control experiments, 1 x 106 MSCs isolated from GFP-Tg mice were s.c injected as a cell suspension on the back of three nude mice and observed for 12 weeks. Furthermore, 2.5 x 106 MSCs isolated from GFP-Tg mice were s.c injected as a cell suspension on the back of six nude mice and observed for 12 weeks.

In vitro assessment of cell tumorigenic potential
MSC and sarcoma cells were cultured in Coon’s modified F12 medium (Biochrom AG) supplemented with 2 mM L-glutamine, 50 µg/ml penicillin/streptomycin and 1 ng/ml basic fibroblast growth factor (serum-free medium). As a control, cells were cultured in standard medium. After 8 weeks, plates were washed, fixed with 1% formaldehyde and stained with 1% methylene blue.

Cellular potential for invasiveness of MSC and sarcoma cells was determined using six-well transwell system with a membrane coated with matrigel. Briefly, sarcoma cells and MSCs were seeded into the upper inserts of the chamber coated with 25 µg of matrigel at 2 x 105 per insert in serum-free F12 medium. Outer wells were filled with F12 containing 10% fetal calf serum (GIBCO, S.Giuliano Milanese), used as chemoattractant. Cells were incubated at 37°C, 5% CO2 for 24 h and then non-invading cells were removed by swabbing top layer. Membranes containing invading cells were stained with hematoxylin for 3 min, washed and mounted on slides. Invading cells contained in the membranes were counted under light microscope at x32 objective, using a grate and performing three different random counts per membrane.

Anchorage-independent growth ability of cells was evaluated by a soft agar colony formation assay. Briefly, 1 x 105 MG-63 cells, 1 x 105 sarcoma cells derived from biopsies obtained from primary tumors grown in immunocompromised or syngeneic recipients and 1 x 105 MSC were plated on a 0.6% agarose base in six-well plates in 1 ml of standard medium containing 0.3% agarose. After staining with 0.005% crystal violet (Sigma Chemical Co., St Louis, MO), colonies >50 µm were counted 21 days after plating.

Histological analysis and immunohistochemistry
For histological examination, tumors were surgically removed and fixed in 10% neutral-buffered formalin, dehydrated and embedded in paraffin. Formalin-fixed HA blocks were decalcified. Four micrometer serial sections were stained with hematoxylin and eosin or processed for immunohistochemistry. After hydration, antigen retrieval was heat induced in citrate buffer pH 6.0 and endogenous peroxidases were blocked with 3% H2O2 in water. Non-specific binding was inhibited, by incubating the slides in 10% normal goat serum (Sigma–Aldrich, Milano, Italy). The following primary antibodies were used: rabbit polyclonal anti-GFP antibody (Molecular Probes Europe BV, Leiden, The Netherlands), mouse monoclonal anti-vimentin (clone V9), mouse monoclonal anti-smooth muscle actin (clone 1A4), rabbit anti-desmin (Dako Cytomation, Milano, Italy), rabbit anti-CD14 (Abcam, Cambridge, UK), rat monoclonal anti-mouse CD31 (clone MEC 13.3) (kindly provided by Dr P. Castellani) and rat monoclonal anti-mouse F4/80 (clone A3-1) (AbD Serotec, Milano, Italy).

Negative controls with preimmune serum and positive controls were run in parallel. As a control for the detection of GFP-positive MSC, we performed additional reactions on sections of GFP-Tg (positive controls) or C57Bl/6 (negative controls) mouse tissues.

After extensive washing in Tris-buffered saline, slides were incubated with either biotin-conjugated goat anti-rabbit or anti-mouse (BioSpa, Milano, Italy). After washing, horseradish peroxidase-conjugated streptavidin (BioSpa) was added and incubated for 10 min. Slides were then stained with diaminobenzidine chromogen (Lab Vision, Fremont, CA). Counterstaining was performed with hematoxylin. Images were captured by transmitted light microscopy with an Olympus C3030 digital camera and Camedia Master Olympus software.

Fluorescence in situ hybridization analysis
Interphasic fluorescence in situ hybridization (FISH) analysis was performed on 4 µm thick slides cut from paraffin blocks. The presence of tumor cells was evaluated in each tissue block on hematoxylin and eosin-stained slides cut before and after the section used for FISH analysis. FISH experiments were carried out as described previously (13) using painting DNA probe CY3 labeled specific for mouse Y and chromosome painting DNA probe fluorescein isothiocyanate labeled specific for mouse X chromosome (Cambio, Cambridge, UK). Interphasic FISH analysis was also performed on isolated nuclei from 15 µm sections (14). More than 50 nuclei from at least five to eight different areas selected for well-preserved cellular and nuclear morphology were examined in each case to ensure representative samples and to avoid a nuclear truncation in paraffin-embedded sections. Only experiments with 90% hybridization efficiency were considered. Histological sections of male mouse tissues were used as technical controls. FISH analysis was performed as recommended by the International System for Human Cytogenetic Nomenclature (15).

In vitro expansion of CD4+CD25+ T regulatory cells
CD4+CD25+ and CD4+CD25– T lymphocytes were isolated from naive C57Bl/6 mice and were plated at a density of 1 x 106 cells in the upper insert of a six-well transwell system (Corning B.V., Schiphol-Rijk, The Netherlands). The lower part of the system was filled with 1 x 106 MSC and cultured in presence or absence of 0.4 µM phytohemagglutinin or 1 x 106 sarcoma cells. The cultures were maintained for 10 days at 37°C, 5% CO2. After 10 days, at the end of the experiment, cells cultured in the upper inserts were harvested, counted, examined by flow cytometry and used in proliferation assays.

Proliferation assay
We tested proliferation of splenocytes derived from sarcoma-bearing mice to stimulation with 0.2 µM of a mitogenic stimulus, phytohemagglutinin (Sigma), with 5 x 105 {gamma}-irradiated syngeneic or allogeneic splenocytes and {gamma}-irradiated syngeneic tumor cells, in the presence or absence of {gamma}-irradiated syngeneic Tregs.

We plated 5 x 105 responder cells per well in round-bottomed 96-well plates (Corning BW Lifescience, Pero, Milano, Italy) in proliferation assay. Cultures were incubated for 72 h and pulsed with [3H] thymidine (1.0 µCi per well) (Amersham Biosciences, Milano, Italy) for at least 16 h. Cells were harvested and cell proliferation was evaluated by counting thymidine uptake and the averaged proliferation rate was measured as counts per minute.

In another experimental setting, we tested proliferation of splenocytes derived from mice bearing sarcomas to stimulation with 5 x 105 {gamma}-irradiated syngeneic sarcoma cells, in presence or absence of 5 x 105 {gamma}-irradiated Tregs in vitro expanded as described previously.

Statistical analysis
Statistical significance of observed differences of tumor formation after in vivo implants among different experimental groups of mice were calculated by Fisher’s exact test. A two-tailed t-test was used to evaluate statistical significance of observed results of the invasion assay, in vitro expansion of CD4+CD25+ Tregs and cell proliferation assays. P-values ≤0.05 were considered to be statistically significant.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Supplementary material
 Funding
 References
 
Characterization of MSC
MSC expressed major histocompatibility complex (MHC) class I molecules (average 96% ± 3 SD), but not MHC class II antigens (Figure 1). A large majority of MSC constitutively expressed CD44 (average 97% ± 3 SD), CD106 (average 81% ± 4 SD) and CD140a (platelet-derived growth factor receptor {alpha}) (average 96% ± 2 SD) (Figure 1), considered to be markers of progenitor cells of the mesenchymal lineages. The expression of CD45 was always <1%, indicating the absence of contaminating myeloid cells. CD34 was expressed in an average of 21% ± 9 SD. Less than 1% of MSC expressed CD11b, whereas CD14 and CD31 were widely expressed (average 47% ± 17 SD and average 22% ± 5 SD, respectively) (Figure 1).


Figure 1
View larger version (22K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Phenotype of MSC population. Flow cytometric analysis of MSC derived from C57Bl/6 mice and harvested after one in vitro passage. Gray areas represent expression of a specific marker on tested cell populations. The uncolored areas under the black line serve as negative control (cells stained with isotype antibody). Shown are the results of a representative experiment.

 
Tumor development in mice implanted with murine MSC seeded onto bioscaffolds
MSCs were isolated from the bone marrow of C57Bl/6 mice. After a brief period of expansion, one in vitro passage (P1), MSCs were seeded onto porous ceramic (HA) scaffolds. According with the mouse model of ectopic bone formation under unloaded conditions (16), we grafted combinations of MSC/bioscaffold constructs s.c. into syngeneic, immunocompromised (nude) or allogeneic recipients. Syngeneic and nude recipients were not expected to mount an allospecific immune response to the grafted cells, whereas implants into allogeneic, genetically mismatched and not immunosuppressed recipients could elicit a strong immune response. As a control experiment, we injected MSC as cell suspensions s.c. on the back of immunodeficient recipient mice.

We observed the formation of a tumor in 10/12 (83%) syngeneic recipients and 8/10 (80%) nude mice, whereas none (0/20) of the allogeneic recipients developed a tumor (P < 0.0001 comparing the number of tumors formed in syngeneic or nude animals with the number of tumors formed in allogeneic recipients) (Figure 2A). We killed the mice when the tumors were ~0.8–1 cm of diameter (Figure 2B). At the gross examination, the tumors appeared as highly vascularized masses, with necrotic areas (Figure 2B and C). Angiogenesis was evident as early as 2 weeks after implantation of MSC/collagen combinations in syngeneic recipients, but not when MSC/collagen constructs were implanted in allogeneic mice (Figure 2C). Tumors were defined as undifferentiated sarcomas positive for vimentin and desmin (Figure 2C), but negative for {alpha}-actin and smooth muscle actin (Figure 2C). Newly formed vases were evident by a positive staining with anti-CD31 monoclonal antibody (Figure 2E). Negative and positive controls of immunohistochemistry can be found online in the supplementary Figure 1 (available at Carcinogenesis Online).


Figure 2
View larger version (59K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. Incidence and features of tumors induced by MSC loaded onto bioscaffolds. (A) Histograms show the incidence of tumors induced by MSC/HA implants (left) or by MSC/collagen sponge implants (right) grafted into allogeneic, syngeneic and nude recipients. (B) Gross examination of sarcomas. The left panel shows a representative image of the tumor external appearance, and the right panel shows the internal part of the tumor including massive areas of necrosis (arrowhead) near the HA bioscaffold (asterisk), that is integrated inside the tumor mass. (C) Figure shows a massive angiogenetic process in MSC/collagen combinations implanted in syngeneic recipients at 2 weeks post-implantation (left panel) and the absence of a substantial vascularization in the MSC/collagen constructs implanted into allogeneic recipients at 2 weeks post-implantation (right panel). (D) (left upper panel) Hematoxylin and eosin (H&E) staining of a sarcoma developed in syngeneic recipients (magnification x500). The insert shows a higher enlargement of the same slide displaying mitotic figures (magnification x1000); (right upper panel) representative images of sarcomas derived from syngeneic implanted mice immunohistochemically stained with anti-vimentin antibody (magnification x500); (left bottom panel) with anti-desmin (magnification x500) and (right bottom panel) anti-{alpha} smooth muscle actin (magnification x1000). In this latter staining, tumor is negative but the entrapped vessel offers an internal positive control. (E) Anti-CD31 immunohistochemical staining of a sarcoma grown in a syngeneic recipient showing the presence of newly formed vases (magnification x20).

 
Since the stiff walls of the porous ceramic scaffold could have triggered a chronic inflammatory stimulus by brushing against the subcutaneous tissues, thus favoring the formation of tumors originating from the host’s tissues surrounding the implants, we performed further experiments by implanting combinations of MSC and collagen sponges. The smoothness of this scaffold, however, did not precluded tumor formation. In fact, we observed that 8/8 (100%) of syngeneic recipients and 6/6 (100%) of immunocompromised hosts developed tumors, whereas none (0/4) of the cell/scaffold combinations implanted in allogeneic mice degenerated into tumors (P < 0.0001 comparing the number of tumors formed in syngeneic or nude animals with the number of tumors formed in allogeneic recipients) (Figure 2A).

Tumors originated from the implants of MSC/collagen sponges constructs were undifferentiated sarcomas histologically similar to tumors formed within the porous ceramic. The different incidence of tumors formed using the two different scaffolds was not statistically significant. The latency time of tumor appearance was shorter using collagen sponges than using HA. When collagen sponges were used, tumor masses began to be evident as early as 4 weeks post-implant.

Tumor cells maintained their oncogenic potential when serially reimplanted into syngeneic or nude hosts where they formed new sarcomas histologically similar to the primary tumor (supplementary Figure 2 is available at Carcinogenesis Online). However, the latency time decreased after each in vivo passage; tumor masses were evident ~2 weeks after the first reimplantation passage and 1 week after the second and the third passage.

Sarcoma formation was not observed in any of the allogeneic not immunosuppressed recipient (Figure 2A), and the exogenous MSC seeded onto the bioscaffold appeared acutely rejected by the second week after implantation (R.Tasso, A.Augello, S.Boccardo, S.Salvi, M.Carida', F.Postiglione, F.Fais, M.Truini, R.Cancedda and G.Pennesi, submitted for publication) (Figure 3).


Figure 3
View larger version (76K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3. Acute rejection of MSC/porous ceramic combinations grafted in allogeneic immunocompetent recipient. Shown is the histological appearance of grafts of MSC/HA combinations in syngeneic and allogeneic recipients examined at 2 weeks post-implantation. Fibroblastoid cells and new vases invaded the pores of the scaffold in syngeneic implants (left panel), whereas only necrotic cells are evident within the scaffold of the allogeneic combination (right panel) (hematoxylin and eosin staining; magnification x20).

 
Tumor development was not observed in mice receiving MSC as cell suspensions.

Origin of the tumors
To test whether tumor formation was due to the presence of tumorigenic cells in the initial population of MSC and under the assumption that only transformed and immortalized cells can survive and proliferate when cultured in absence of serum, we cultured them in media not supplemented with serum (starvation condition). As a result, we observed that the initial MSC populations did not survive when cultured under starvation conditions for 8 weeks (Figure 4A). Moreover, cancer cells can infiltrate the surrounding tissues and this property can be in vitro tested by checking the ability of the cells to migrate through a porous membrane coated with a gel of matrix proteins (matrigel). The number of cells passing through the matrigel is considered proportional to the number of invasive cancer cells present within the tested population (17).


Figure 4
View larger version (82K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4. MSC cultures under starvation conditions and invasion assay. (A) Absence of cell growth in MSC cultured with serum-free medium for 8 weeks (left panel). Growth of MSC populations in standard medium (right panel). Shown are images of representative cultures. (B) Absence of MSC invading the pores of a matrigel-coated transwell membrane (left panel). Colored cells from a sarcoma passing through the matrigel coat and invading the membrane pores (right panel). (C) Anchorage-independent growth ability of sarcoma cells and primary culture of MSC. The histogram shows the number of colonies formed in soft agar by tumor cells derived from primary sarcomas grown in an immunocompromised recipient (tumor cells 1) or in a syngeneic host (tumor cells 2), cells from primary cultures of MSC before implantation (MSC) and a control human osteosarcoma cell line (MG-63). An asterisk indicates P < 0.0001 comparing MG-63, Tumor cells 1 and Tumor cells 2 with MSC. #Indicates P < 0.0001 comparing MG-63 with Tumor cells 1, Tumor cells 2 and MSC.

 
When comparing the invasive ability of the initial MSC population with that of cells isolated from the tumors derived from the implants, we found that no cells from the original MSC culture could pass through the matrigel coat, whereas an average of 69% ± 0.04 SD of the pores of the matrigel-coated culture dishes were occupied by tumor-derived cells (P < 0.0001) (Figure 4B). Furthermore, we examined the anchorage-independent colony-forming capacity in soft agar of MSC, sarcoma cells derived from biopsies obtained from primary tumors grown in immunocompromised or syngeneic recipients, and MG-63 cells, a cell line derived from a human osteosarcoma, used as a positive control. After 21 days of culture, both sarcoma cells obtained from primary tumors grown in immunocompromised or syngeneic recipients were able to form colonies of a diameter >50 µm (averaged colony number 24 ± 5 SD and 31 ± 5 SD, respectively) (Figure 4C). However, they grew with a reduced efficiency if compared with the positive control MG-63 cell line (averaged colony number 58 ± 6 SD) (tumor cells versus MG-63 P < 0.0001) (Figure 4C). As expected, no MSC culture was able to form colonies in soft agar (Figure 4C). Taken together, these results tended to exclude the presence of tumorigenic cells within the MSC population prior seeding and implantation.

Using MSC isolated from GFP-Tg mice, we could determine that sarcomas contained a few clusters of GFP-positive cells of donor origin scattered throughout a mass of GFP-negative cells originated from the recipient’s tissues (Figure 5A). To further confirm that tumors stemmed from the recipient’s tissues, we implanted MSC from male donors seeded onto a bioscaffold into female recipients, sorting the origin of the tumor cells using dual-color FISH with specific probes for the murine Y (labeled in red) and X chromosomes (labeled in green). We analyzed all the sarcoma cells on histological sections and in addition 400 extracted nuclei obtained from 15 µm sections. In all the analyzed nuclei no Y sequences were found. Two chromosome X was observed in 57% of cells, whereas aneuploidies of X chromosome were identified in the remaining 43% of tumor cells. These FISH results demonstrated the female origin (then the recipient origin) of tumor cells (Figure 5B) (18). Interestingly, the sarcoma cells showed chromosome instability that is a peculiar feature of undifferentiated solid tumors. A healthy tissue surrounded the sarcoma; composed by small cells inside of an abundant stroma were Y negative and euploid (Figure 5C) (a representative image of FISH-positive control can be found online in supplementary Figure 3, available at Carcinogenesis Online).


Figure 5
View larger version (71K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5. Origin of tumor cells. (A) Anti-GFP immunohistochemical staining of a sarcoma induced by implants of GFP-positive MSC/HA combinations grafted into GFP-negative syngeneic recipients. Only a few scattered GFP-positive cells are present through the tumor mass (magnification x20). (B) Dual-color FISH using Y chromosome (red labeled) and X chromosome (green labeled) probes. Shown is a representative image of nuclei extracted from sarcoma cells indicating the presence of multiple (two to three) green spots corresponding to X chromosomes. No red-labeled Y chromosome was observed (magnification x100). (C) FISH analysis of nuclei isolated from non-cancerous cells surrounding the tumor mass showing euploid female karyotypes with two green-labeled X chromosomes (magnification x100). (D) (left panel) Anti-CD14 immunohistochemical staining of a sarcoma developed in a syngeneic recipient explanted 4 weeks after MSC/HA construct implantation (early stage) (magnification x20); the insert shows a higher enlargement of the same slide displaying the distribution of CD14+ cells in areas at the periphery of the tumor mass (magnification x40). (right panel) Anti-F4/80 immunohistochemical staining of a sarcoma developed in a syngeneic recipient explanted 4 weeks after MSC/HA construct implantation (early stage) (magnification x20).

 
It has been shown that macrophage can induce chromosome instability (19). For this reason, we checked whether macrophages infiltrated tumor mass during early and late stages of tumor formation by immuhistochemical staining. At the early stages of tumor formation, i.e. within the first 4 weeks after implantation, a few CD14+ cells surrounded the tumor mass and were present near the necrotic areas suggesting their likely role as scavenger cells (Figure 5D). The same distribution of CD14+ cells was observed also at later stages of tumor growth, 8–9 weeks after implantation immediately before the killing of the implanted animals (data not shown). A subpopulation of macrophages expressing the marker F4/80 could promote tumor growth by suppressing host immune reaction (20). To test this possibility, we checked the presence of F4/80+ cells within the early and late stages of tumors. Immunohistochemistry revealed the absence of mononuclear myeloid-derived suppressor cells at the early stage of cancer formation (Figure 5D) and at later stages (data not shown). Positive controls of immunohistochemistry can be found online in the supplementary Figure 1 (available at Carcinogenesis Online).

Tumor immune escape
Splenocytes of tumor-bearing mice actively proliferated when in vitro challenged with cells from the autologous tumor in a mixed lymphocyte reaction, suggesting that the immune system of the mice developing the tumors was potentially able to recognize tumor-specific antigens carried by the sarcoma cells (Figure 6A). However, we found that CD4+CD25+ Tregs were present with an averaged incidence of 4.17% ± 0.35 SD in the spleen of tumor-bearing mice, whereas these cells showed an averaged incidence of 1.61% ± 0.45 SD in non-implanted naive syngeneic mice (P < 0.001) (Figure 6B), indicating that the presence of CD4+CD25+ Tregs could have blocked the in vivo antitumor response. Indeed, Tregs isolated from tumor-bearing mice significantly blocked the proliferation of host’s splenocytes challenged with syngeneic tumor cells, but not the proliferation of the same responder cells to a different cellular stimulus (Figure 6A).


Figure 6
View larger version (26K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 6. Generation of tumor-specific CD4+CD25+ Treg by exogenous MSC. (A) The histogram shows the compiled results of proliferation assays where responder splenocytes from tumor-bearing mice were in vitro challenged with a mitogenic stimulus (phytohemagglutinin, PHA), syngeneic or allogeneic splenocytes and syngeneic tumor cells in presence or absence of Treg generated from the same tumor-bearing mouse. An asterisk indicates P < 0.05 comparing the proliferation of each condition with the baseline, ‘medium’ bar. (B) Incidence of CD4+CD25+ T lymphocytes in the spleen of a representative naive mouse (left panel) and in the spleen of a representative syngeneic tumor-bearing mouse (right panel). (C) Incidence of CD4+CD25+ T lymphocytes generated in transwell cultures of CD4+CD25– T cells by MSC alone (left panel) and by MSC + syngeneic tumor cells (right panel). Shown are the results of a representative experiment. (D) Compiled results of proliferation assays where splenocytes from tumor-bearing mice were challenged with syngeneic tumor cells in presence of CD4+CD25+ T lymphocytes generated in transwell cultures of CD4+CD25– T cells and MSC alone, MSC + phytohemagglutinin and MSC + syngeneic tumor cells, respectively. An asterisk indicates P < 0.01 comparing the proliferation of each condition with the baseline, medium bar.

 
MSC expanded clones of tumor-specific Tregs by the action of soluble factors. When we cultured CD4+CD25– T lymphocytes on the top of a transwell plate and we stimulated them with MSC and tumor cells put on the bottom part of the plate, we could retrieve CD4+CD25+ lymphocytes (Figure 6C) able to suppress the proliferation of splenocytes derived from tumor-bearing mice in vitro challenged with syngeneic tumor cells (Figure 6D). We could not retrieve CD4+CD25+ Tregs when we stimulated CD4+CD25– T lymphocytes (on the top) with MSC alone or MSC and the mitogen agent phytohemagglutinin (on the bottom).


   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Supplementary material
 Funding
 References
 
Repairing tissues by a tissue engineering approach using exogenous cells seeded onto biomaterials involves actively transforming interactions between three players: (i) the progenitor cells that constitute the base elements for tissue regeneration and are both the producers and the targets of different growth factors; (ii) the biomaterial that provides the tridimensional scaffold where the engineered tissue develops, and that is gradually reabsorbed, and (iii) the recipient organism supplying the microenvironment allowing a correct tissue development. An engineered tissue should develop and integrate with adjacent host tissues and provide some minimal level of function immediately after implantation that should improve progressively until normal function has been restored (21). When the functional interactions between these three players are not balanced, a functional tissue would not develop and adverse reactions could happen.

A tumorigenic potential of human and mouse MSC was reported in in vitro and in vivo experimental settings using cell populations expanded for five to six passages in plates (22–24). Aneuploid karyotypes were observed in human and mouse MSC cultures after several (P9–P15) in vitro passages (25,26). The spontaneous transformation of MSC could be possibly due to the relative resistance of these cells to the telomerase-dependent mechanisms controlling the cell proliferation (27,28). For these reasons, we decided to use MSC after one passage of in vitro expansion (P1), assuming that a brief cycle of expansion did not allow any MSC to undergo spontaneous transformation into tumor cells. The risk of using this first-round cultured MSC was to have a highly heterogenous cell population including cells with angiogenic and inflammatory properties, such as macrophages and endothelial progenitor cells, able to induce precancerous changes in the tissues. The phenotype of the implanted MSC showed the presence of contaminating macrophages, but not endothelial precursor cells. The CD14+ fraction of macrophages contains cells with different functional activity, including osteoblast activity (29), and we cannot discern which functions were likely to play the implanted CD14+ cells. Moreover, the immunohistochemical analysis of the cancer tissue at the time when the tumorigenic injury was likely to occur, within the 4 weeks after implantation, showed the presence of CD14+ cells at the periphery of the tumor mass and near the areas of necrosis, indicating a more likely function of macrophages as scavenger cells.

The sarcomas observed in the mice implanted with MSC/bioscaffold constructs originated from the recipient’s tissues, thus pointing the attention to the paracrine effect of MSC on other cell populations.

In our experiments, the capacity of MSC to provide a local proliferation stimulus to the adjacent cells was enhanced by the presence of the scaffold creating the tridimensional architecture where the cells could develop into a structure similar to a tumor stroma. It has been established in vitro that growing cells in tridimensional scaffolds creates a microenvironment permissive to tumor development (8,30). Both the composition and the stiffness of the scaffold that mimics the extracellular matrix of the tumor stroma have major effects on cell signaling and behavior. Cells can sense the stiffness of the stroma matrix through bidirectional interactions mediated by cell surface integrin receptors and the contractile cytoskeleton. As a consequence, they can alter the functional profile of activated intracellular and extracellular molecular pathways inducing cell proliferation or cytokine production (31,32). Recent reports indicate that MSCs are recruited in large numbers to the stroma of developing tumors (3,33), and there they act enhancing the motility, invasion and metastasis ability of adjacent cancer cells (4).

Others have reported that MSC induced only a transient tumor transformation of adjacent cells when intravenously injected in immunodeficient mice (3,4). On the contrary, our results showed that MSC, when loaded onto bioscaffolds, induced a permanent tumor transformation of adjacent cells, thus creating a tumor niche. Moreover, the composition and stiffness of the scaffold determined the latency of tumor occurrence. Sarcoma developed faster when MSC were loaded onto collagen sponges than onto a stiffer and less resorbable HA block. We can speculate that collagen constitutes a natural component of tumor stroma (34), thus creating a more physiological microenvironment permissive of cancer growth. In alternative, we can think that the superior thickness and slower resorbability characteristic of porous ceramic delays the direct contact between implanted MSC and neighboring cells that constitutes the primary stimulus for tumor formation.

Control of tumor growth by the immune system is a complex issue (35). It has been demonstrated that host immunity can kill tumor cells by the activation of different subsets of natural killer cells and T lymphocytes (36). However, tumor itself and cells in the tumor microenvironment could provide the factors to block this immune reaction. There are evidences that different molecular and cellular pathways induce the differentiation of tumor-specific CD4+CD25+ Tregs able to block the antitumor-specific immune response (37,38). In our experimental model, tumor immune escape could be due to the expansion of specific Tregs that impaired the activation of T lymphocytes against sarcoma cells. The presence of tumor-specific Tregs, spontaneously occurring in different types of human tumors (37,38), could have been enhanced in our system by the implanted MSC (2). The presence of the bioscaffold did not modify the ability of MSC to generate T lymphocytes with suppressing activity (39,40).

In this paper, we reported the tumorigenic effect of the implantation of murine bone marrow-derived MSC in syngeneic animals. It is important to stress that, up to date, none of the patients treated with implants of autologous MSC seeded onto a ceramic bioscaffold for bone repair have been reported to have developed any type of tumors (41). Moreover, there are no published reports of tumor development in any patient implanted with autologous MSC both seeded and not seeded on a scaffold.

The murine model of ectopic bone formation under unloaded conditions is one of the most widely used approaches to study the development of an engineered bone tissue. The classical model involves the use of MSC of xenogeneic origin, i.e. human or sheep, implanted in combination with a bioscaffold into immunodeficient recipient mice. No tumor formation has been thus far reported under these experimental conditions. This discrepancy could be attributed to species-specific susceptibility of cells to tumor transformation determining an impairment of the cross talk between factors released by the donor and recipient cells (42).

Although the observed tumorigenic effect of implanted MSC could be species restricted, our results induce a more critical approach to (i) the use of MSC/bioscaffold constructs, in terms of biological interactions with the recipient; (ii) the source of MSC, autologous or allogeneic and (iii) the selection of patients, using an extreme caution in treating patients with a familial or personal history of malignancies.


   Supplementary material
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
Funding
 References
 
Supplementary Figures 1–3 can be found at http://carcin.oxfordjournals.org/


   Funding
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
References
 
Italian Ministry of Education, University and Scientific Research (FIRB RBNE01YANS); Italian Ministry of Health.


   Acknowledgments
 
Conflict of Interest Statement: None declared.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Supplementary material
 Funding
 References
 

  1. Park IH, et al. Reprogramming of human somatic cells to pluripotency with defined factors. Nature (2008) 451:141–146.[CrossRef][Medline]
  2. Djouad F, et al. Immunosuppressive effect of mesenchymal stem cells favors tumor growth in allogeneic animals. Blood (2003) 102:3837–3844.[Abstract/Free Full Text]
  3. Ramasamy R, et al. Mesenchymal stem cells inhibit proliferation and apoptosis of tumor cells: impact on in vivo tumor growth. Leukemia (2007) 21:304–310.[CrossRef][Web of Science][Medline]
  4. Karnoub AE, et al. Mesenchymal stem cells within tumour stroma promote breast cancer metastasis. Nature (2007) 449:557–563.[CrossRef][Web of Science][Medline]
  5. Maestroni GJ, et al. Factor(s) from nonmacrophage bone marrow stromal cells inhibit Lewis lung carcinoma and B16 melanoma growth in mice. Cell. Mol. Life Sci. (1999) 55:663–667.[CrossRef][Web of Science][Medline]
  6. Nakamura K, et al. Antitumor effect of genetically engineered mesenchymal stem cells in a rat glioma model. Gene Ther. (2004) 11:1155–1164.[CrossRef][Web of Science][Medline]
  7. Ohlsson LB, et al. Mesenchymal progenitor cell-mediated inhibition of tumor growth in vivo and in vitro in gelatin matrix. Exp. Mol. Pathol. (2003) 75:248–255.[CrossRef][Web of Science][Medline]
  8. Yamada KM, et al. Modeling tissue morphogenesis and cancer in 3D. Cell (2007) 130:601–610.[CrossRef][Web of Science][Medline]
  9. Quarto R, et al. Repair of large bone defects with the use of autologous bone marrow stromal cells. N. Engl. J. Med. (2001) 344:385–386.[Free Full Text]
  10. http://www.clinicatrials.gov.
  11. http://www.controlled-trials.com.
  12. Heremans H, et al. In vitro cultivation of human tumor tissues. II. Morphological and virological characterization of three cell lines. Oncology (1978) 35:246–252.[Web of Science][Medline]
  13. Tibiletti MG. Specificity of interphase fluorescence in situ hybridization for detection of chromosome aberrations in tumor pathology. Cancer Genet. Cytogenet. (2004) 155:143–148.[CrossRef][Web of Science][Medline]
  14. Paternoster SF, et al. A new method to extract nuclei from paraffin-embedded tissue to study lymphomas using interphase fluorescence in situ hybridization. Am. J. Pathol. (2002) 160:1967–1972.[Abstract/Free Full Text]
  15. Shaffer LG, et al. ISCN 2005: International System for Human Cytogenetic Nomenclature: Recommendations of the International Standing Committee on Human Cytogenetics Nomenclature (2005) Basel, Switzerland: Karger, S.
  16. Muraglia A, et al. A nude mouse model for human bone formation in unloaded conditions. Bone (1998) 22:131S–134S.[CrossRef][Web of Science][Medline]
  17. Ho MM, et al. Side population in human lung cancer cell lines and tumors is enriched with stem-like cancer cells. Cancer Res. (2007) 67:4827–4833.[Abstract/Free Full Text]
  18. Lengauer C, et al. Genetic instabilities in human cancers. Nature (1998) 396:643–649.[CrossRef][Medline]
  19. Wang X, et al. Extracellular superoxide production by Enterococcus faecalis promotes chromosomal instability in mammalian cells. Gastroenterology (2007) 132:551–561.[CrossRef][Web of Science][Medline]
  20. Nausch N, et al. Mononuclear Myeloid-Derived "Suppressor" Cells express RAE-1 and activate NK cells. Blood (2008) doi:10.1182/blood-2008-03-143776.
  21. Vunjak-Novakovic G. Tissue engineering: basic considerations. In: Culture of Cells for Tissue Engineering—Vunjak-Novakovic G, Freshney RI, eds. (2006) Hoboken, NJ: John Wiley & Sons, Inc. 131–155.
  22. Zhou YF, et al. Spontaneous transformation of cultured mouse bone marrow-derived stromal cells. Cancer Res. (2006) 66:10849–10854.[Abstract/Free Full Text]
  23. Aguilar S, et al. Murine but not human mesenchymal stem cells generate osteosarcoma-like lesions in the lung. Stem Cells (2007) 25:1586–1594.[CrossRef][Web of Science][Medline]
  24. Tolar J, et al. Sarcoma derived from cultured mesenchymal stem cells. Stem Cells (2007) 25:371–379.[CrossRef][Web of Science][Medline]
  25. Rubio D, et al. Spontaneous human adult stem cell transformation. Cancer Res. (2005) 65:3035–3039.[Abstract/Free Full Text]
  26. Miura M, et al. Accumulated chromosomal instability in murine bone marrow mesenchymal stem cells leads to malignant transformation. Stem Cells (2006) 24:1095–1103.[CrossRef][Medline]
  27. Burns JS, et al. Tumorigenic heterogeneity in cancer stem cells evolved from long-term cultures of telomerase-immortalized human mesenchymal stem cells. Cancer Res. (2005) 65:3126–3135.[Abstract/Free Full Text]
  28. Zimmermann S, et al. Effects of telomerase modulation in human hematopoietic progenitor cells. Stem Cells (2004) 22:741–749.[CrossRef][Web of Science][Medline]
  29. Kuwana M, et al. Human circulating CD14+ monocytes as a source of progenitors that exhibit mesenchymal cell differentiation. J. Leukoc. Biol. (2003) 74:833–845.[Abstract/Free Full Text]
  30. Fischbach C, et al. Engineering tumors with 3D scaffolds. Nat. Methods (2007) 4:855–860.[CrossRef][Web of Science][Medline]
  31. Pelham RJ Jr, et al. Cell locomotion and focal adhesions are regulated by substrate flexibility. Proc. Natl Acad. Sci. USA (1997) 94:13661–13665.[Abstract/Free Full Text]
  32. Paszek MJ, et al. Tensional homeostasis and the malignant phenotype. Cancer Cell (2005) 8:241–254.[CrossRef][Web of Science][Medline]
  33. Hall B, et al. The participation of mesenchymal stem cells in tumor stroma formation and their application as targeted-gene delivery vehicles. Handb. Exp. Pharmacol. (2007) 180:263–283.[CrossRef][Medline]
  34. Tomkowicz B, et al. Interaction of endosialin/TEM1 with extracellular matrix proteins mediates cell adhesion and migration. Proc. Natl Acad. Sci. USA (2007) 104:17965–17970.[Abstract/Free Full Text]
  35. June CH. Adoptive T cell therapy for cancer in the clinic. J. Clin. Invest. (2007) 117:1466–1476.[CrossRef][Medline]
  36. Hammond-McKibben D, et al. Immunologic pathways in a quantitative model of immunosuppression based on rejection of an allogeneic or xenogeneic tumor graft. Transplantation (2005) 79:889–896.[CrossRef][Web of Science][Medline]
  37. Li J, et al. Functional inactivation of EBV-specific T-lymphocytes in nasopharyngeal carcinoma: implications for tumor immunotherapy. PLoS ONE (2007) 2:e1122.[CrossRef][Medline]
  38. Vence L, et al. Circulating tumor antigen-specific regulatory T cells in patients with metastatic melanoma. Proc. Natl Acad. Sci. USA (2007) 104:20884–20889.[Abstract/Free Full Text]
  39. Augello A, et al. Cell therapy using allogeneic bone marrow mesenchymal stem cells prevents tissue damage in collagen-induced arthritis. Arthritis Rheum. (2007) 56:1175–1186.[CrossRef][Web of Science][Medline]
  40. Prevosto C, et al. Generation of CD4+ or CD8+ regulatory T cells upon mesenchymal stem cell-lymphocyte interaction. Haematologica (2007) 92:881–888.[Abstract/Free Full Text]
  41. Marcacci M, et al. Stem cells associated with macroporous bioceramics for long bone repair: 6- to 7-year outcome of a pilot clinical study. Tissue Eng. (2007) 13:947–955.[CrossRef][Web of Science][Medline]
  42. Costea DE, et al. Species-specific fibroblasts required for triggering invasiveness of partially transformed oral keratinocytes. Am. J. Pathol. (2006) 168:1889–1897.[Abstract/Free Full Text]
Received May 12, 2008; revised September 11, 2008; accepted October 3, 2008.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
30/1/150    most recent
bgn234v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Tasso, R.
Right arrow Articles by Pennesi, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tasso, R.
Right arrow Articles by Pennesi, G.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?