SEARCH

SEARCH BY CITATION

Keywords:

  • cancer treatment;
  • cell-based therapy;
  • human mesenchymal stem cells;
  • transformed mesenchymal stem cells;
  • spontaneous transformation

Abstract

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

The human population is increasingly facing various diseases, including types of cancer, that cannot be cured with conventional drugs. Advanced drug targeting of tumor cells is also often impossible when treating highly invasive and infiltrative tumors such as glioblastoma or pulmonary cancer, because of tumor cells' high migration and invasiveness. Pluripotent human mesenchymal stem cells (hMSCs) have been extensively studied, and strategies are being proposed for treating “incurable” cancers and injury/disease-affected organs. Because of their own intrinsic properties, involving homing and immunomodulatory potency, hMSCs could be used as an excellent cell/drug delivery vehicle in those cell-based therapies. Their unprecedented use has been shadowed, however, by their spontaneous transformation, which links them to cancer-initiating cells during tumor development. How malignant initiation proceeds in vivo, and what are the exact characteristics of the cancer-initiating cells, still remain to be investigated. In the present review, the authors summed up the most recent knowledge about hMSC characteristics, their malignant transformation, and outlined the possibilities of their safe use in novel cell-based therapies. Cancer 2010. © 2010 American Cancer Society.

Cells to be used in cell-based regenerative therapies should display certain properties and have a broad differentiation potential. Cells' differentiation potency is their ability to differentiate into cell precursors of all 3 germ layers and extraembryonic structures. Pluripotent cells, for that matter, can differentiate into derivatives of all 3 germ layers, but are risky to use for therapeutic administration; a donor-derived fetal stem cell tumor was observed in a patient with ataxia telangiectasia after transplantation.1 Pluripotent human embryonic stem cells, which are an in vitro equivalent of pluripotent cells residing in the inner cell mass of the blastocyst stage embryo, face tremendous ethical issues that are stalling their research. For this reason, isolation and characterization of pluripotent human mesenchymal stem cells (hMSCs) as putative adult analogues of human embryonic stem cells is of such importance. hMSCs are present in different tissues and have a potential to differentiate into numerous cell types.2-4 The ease of hMSC isolation from adult tissues and the fact they are not ethically restricted make them attractive candidates for use in cell/gene therapy for tissue regeneration and anticancer treatments. The characteristics that speak for and against hMCS use shall be addressed in the present review.

Isolation Efficiency and Origin-Dependent Growth of hMSC

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

Human mesenchymal stem cell niches reside in various tissues, from which different amounts of hMSCs can be isolated. With 100% efficiency they can be isolated from bone marrow (BM) aspirates, adipose tissues, and umbilical cords,5 whereas from fresh cord blood the isolation efficiency is only 30%5 to 63%.6, 7 The lowest 19.5% isolation efficiency was determined for cryopreserved cord blood samples.7 Recently, amniotic membrane became an attractive source of hMSCs, because this fetal tissue is discarded without any ethical issues, gives the greatest yields of hMSCs, and is devoid of vasculature that causes hematopoietic cell contamination.8 Improved hMSC isolation protocols and optimized culture/storage/transport conditions5, 9 allow the isolation of hMSCs from almost any identified source. Nevertheless, the growth rate of hMSCs seems to be origin dependent,10 and interclonal variability exists among the hMSC clones derived from the same origin in terms of their proliferation capacity and life span,11, 12 as also demonstrated for the 4 BM hMSC clones (Fig. 1). Novel devices containing a nonwoven fabric filter that selectively traps BM hMSCs by bone marrow affinity, not cell size, may give purer hMSC cultures with high expansion potential.13

thumbnail image

Figure 1. Human mesenchymal stem cells (hMSCs) display high interindividual variability in their growth rate. (a-e) Four different clones of hMSCs derived from the bone marrow of different individuals at Passage 5, 3 days after passaging are shown, and display a significant difference in growth rates. Regrading this difference, the hMSC clones were grouped into (b and d) fast and (a and c) slow proliferating clones.

Download figure to PowerPoint

Proliferation/expansion potential of hMSCs is affected by in vitro culture, which results in changed cell/culture morphology. That is, under established in vitro conditions, hMSCs grow as a monolayer, but when cultured in hypoxic atmosphere—a condition found in many tumors in vivo—they continue to proliferate during cell density increase and display a 30-fold higher expansion rate.14 This could be used when treating chronically hypoxic tumor tissues, where even low numbers of hMSCs present may display a beneficiary effect because of their proliferative activity. With scalable automated cell culture platforms, it is possible to optimize the expansion process and conduct a factorial screening (seeding density, serum concentrations, media quantity, and incubation time) of different hMSC responses (such as cell number, gene expression, STRO-1, CD105, and CD71 increase).15 Addressing differences accompanying hMSC growth under hypoxic conditions may reveal the true characteristics of hMSCs and their putative derivatives, tumor stem cells, which are causative for cancer development and persistence.16

In addition, the proliferation of hMSCs could be increased by treated adhesion surface17 or with growth factor-supplemented hMSC culture media.18 Collagen I-coated scaffolds, in addition to improving cell attachment, enhance proliferation up to 3×.17 Moreover, hMSC proliferation increases in cocultures with glioma cells or their conditioned medium (Fig. 2).19 In contrast, the addition of recombinant protein wingless-type mouse mammary tumor virus (MMTV) integration site family member 3A (Wnt3a)20 and growth factors transforming growth factor (TGF)β1 and TGFβ221 decreases hMSC proliferation. hMSC proliferation ceases similarly during prolonged in vitro culture, because of increased endogenous expression of TGFβ, TGFβ-R1, and Smad3.21 Systemic quality engineering approaches are essential for the design of regulated cell therapy manufacturing processes because of their focus on identifying the sources and the control of in vitro culture variations.

thumbnail image

Figure 2. Proliferation response of human mesenchymal stem cells (hMSCs) and glioma cells in cocultures is shown. The proliferation index of hMSCs was defined as a ratio of (a) positive Ki-67 nuclei versus (b) the total number of hMSCs per vision field. (c) Glioblastoma cells U251 and their conditioned media (CM) significantly increased (P < .05) the proliferation index in 2 hMSC clones, whereas (d) hMSCs and their CM significantly decreased (P > .05) the proliferation index of U251 cells.

Download figure to PowerPoint

Immunophenotype and hMSC Differentiation

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

There is no single surface marker, but rather a panel of surface markers that define hMSCs, derived from fresh tissues or cryopreserved samples. Due to different hMSC tissue sources, differences exist among these cells. Therefore, the Mesenchymal and Tissue Stem Cell Committee of the International Society for Cellular Therapy proposed 3 minimal criteria to define hMSC: 1) the ability to adhere to plastic; 2) presence of >95% expression of hMSC-specific antigen markers and >95% absence of hematopoietic/endothelial marker expression; and 3) the differentiation of the hMSCs into osteoblasts, adipocytes, and chondroblasts in vitro.22 Thorough evaluation of those criteria still does not rule out the heterogeneity of hMSCs from different sources; however, detailed description of cell markers and behavior allows us to decide which cell source to use for a certain therapy in a certain individual.

In detail, hMSCs are negative for 3 hematopoietic markers, CD14, CD34, CD45, and antigens CD4, CD8, CD11a, CD14, CD15, CD16, CD25, CD31 (endothelial marker), CD33, CD49b, CD49d, CD49f, CD50, CD62E, CD62L, CD62P, CD80, CD86, CD106, CD117, cadherin V, glycophorin A, human leukocyte antigen (HLA)-DR, and HLA class II.3, 8, 23-25 Conversely, hMSC are positive for CD10, CD13, CD29 (β1-integrin), CD44, CD49e (α5-integrin), CD54 (ICAM-1), CD58, CD71, CD73 (SH3), CD90 (Thy-1), CD105 (SH2/endoglin), CD146, CD166, CD271, vimentin, cytokeratin 8, cytokeratin 18, nestin, and von Willebrand factor. They weakly express HLA class I and CD123, and have a variable expression of FLK1 (KDR), CD133/1, and CD133/2.3, 8, 23-25 With the refinement of isolation methods, a new “perivascular niche” of hMSCs was identified recently and defined as a common stem cell microenvironment for resident hMSC-like cell populations in different tissues.26 The hMSCs from the perivascular niche display also a STRO-1 marker,26 which classifies them for a more primitive hMSC subset of the bulk tissue hMSCs.27, 28 A new concept has emerged that a high-quality primitive subpopulation of hMSCs from different tissues may actually all originate from the perivascular niche.26-28 The origin-dependent level of surface antigen expression in hMSCs is still higher in early passages than in late passages,12 suggesting that the hMSC pool within an organism probably consists of a heterogeneous hMSC population.

Despite morphological and immunophenotypical similarities, hMSCs from different origins differ regarding their differentiation potential. BM hMSCs can differentiate along all known differentiation pathways.3 In contrast, umbilical cord blood (CB) hMSCs and umbilical cord (UC) hMSCs display a reduced sensitivity to undergoing adipogenic differentiation,6, 7 although they can differentiate into adipocytes.29 During adipogenesis, hMSCs' continuous secretion of interleukin (IL)-1β inhibits collagen II, aggrecan, and matrix metalloproteinase-13 expression30 and induces IL-6 secretion by p38MAPK activation.31 CB hMSCs' lower adipogenic commitment shows in the production of fewer and smaller lipid vacuoles, compared with BM hMSCs or adipose tissue (AT) hMSCs, whereas amniotic membrane hMSCs differentiate into advanced lipid-secreting aggregates.8 Independently of their origin, hMSCs' adipogenic potential is inversely related to the length of in vitro culture, and sharply declines when hMSCs become senescent.32 Contrary, prolonged culturing increases their osteogenic potential.12 hMSC in vitro expansion should therefore be performed with limited passaging, to avoid changes in cells' differentiation ability.

In addition to differentiation pathways that define hMSCs, they can differentiate into other cell types. Their differentiation ability can be controlled by various means:

  • Media supplemented with growth factors, such as the addition of 1 ng/mL of fibroblast growth factor (FGF)-2 or 10 ng/mL of TGFβ3 improves osteogenesis33 and chondrogenesis, respectively.34 Similarly, a hepatogenic protocol to predifferentiate hMSCs includes the successive addition of FGF-2, FGF-4, epidermal growth factor (EGF), hepatocyte growth factor, nicotinamide, oncostatin M, dexamethasone, and insulin-transferrin-selenium.35

  • Microenvironment: hMSC coculture with other primary cells (ie, liver cells) enhances hepatic transdifferentiation,36 and constant oxygen levels or hypoxic preconditioning enhances osteogenesis.37

  • Also, biomaterial surface coatings affect hMSC differentiation in a lineage-dependent manner,38 and biodegradable scaffolds do not act only as a temporary matrix for cell migration, but have osteoinductive potential even without osteogenic media stimulation.17

  • Moreover, some noncoding RNAs together with transcripts of unknown function (TUFs) affect transcriptional kinetics during hMSC differentiation.39

  • Finally, mechanotransduction40 and electrostimulation41 have a stimulating effect on hMSC differentiation as well, by which the drawbacks of conditioned media could be avoided.

Genetic Fingerprint and Cytokine Profiling of hMSCs

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

With advanced gene expression analyses, it is possible to characterize morphology changes of hMSCs. Transcriptome analysis from early and late passage hMSCs shows significant alterations in the patterns of gene expression.42 Gene expression reflects a hierarchy that exists among hMSCs32 by confirming the highest level of transcription factor Oct-4 expression in amniotic membrane hMSCs.8 Oct-4 expression defines pluripotency of stem cells, and always decreases during differentiation. A broad differentiation potential and Oct4 expression place amniotic membrane hMSCs at the top of the differentiation potential hierarchy among all hMSCs.8 Second place in this hierarchy is occupied by CB hMSCs, which in turn have a higher differentiation potential than BM hMSCs.29 Thus, we believe that CB and amniotic membrane may in the future present the optimal source of high-quality hMSCs, where their heterogeneity may be made uniform by standardized protocols. Such high-quality hMSCs could be preserved in human tissue banks and used for either autologous or allogeneic cell therapy when needed.

In each organ, there is a location or niche enriched with tissue stem cells that resides within a subpopulation of stromal cells, which are closely associated and control the activity of stem cells. In niches, hMSCs secrete molecules to communicate and respond to other cell types. In vitro, hMSCs secrete several cytokines, ILs, and growth factors like epithelial cell-derived neutrophil-activating peptide-78, granulocyte-macrophage colony-stimulating factor, growth-related oncogene, IL-1β, IL-6, IL-8, monocyte chemoattractant protein-1, oncostatin m (OSM), vascular endothelial growth factor (VEGF), FGF-4, FGF-7, FGF-9, granulocyte chemotactic protein 2 (GCP-2), insulinlike growth factor-binding protein (IGFBP)-1, IGFBP-2, IGFBP-3, IGFBP-4, interferon-inducible protein-10, macrophage-inhibiting factor, macrophage inflammatory protein-3a, osteoprotegerin, pulmonary and activation regulated chemokine (PARC), phosphatidylinositol glycan class F (PIGF), TGF-β2, TGF-β3, tissue inhibitor of metalloproteinase (TIMP)-1, and TIMP-2.31 In addition to these, CB hMSCs also constitutively produce somatic cell-derived growth factor (SCF), leukemia inhibitory factor (LIF), macrophage colony-stimulating factor, IL-11, IL-12, IL-15, stromal cell-derived factor-1α, and hepatocyte growth factor,25 and after IL-1β stimulation, they secrete significantly more granulocyte colony-stimulating factor, SCF, and LIF.25 The basal cytokine profile of hMSCs from different origins is the same, differing only in concentration ranges. The cytokine secretion levels depend on hMSC morphology.43 Aggregated hMSCs or spheroids secrete up to 20× more VEGF, FGF-β, angiogenin, procathepsin B, IL-11, and bone morphogenetic protein type 2 compared with hMSC monolayers.43 The enhanced/inhibited production of cytokines in vivo might lead to interaction with cells at the implantation site. Therefore, affecting this secretory crosstalk would be recommended tool in regenerative medicine and oncology, and needs to be addressed in further studies.

hMSC Proliferation Arrest and Senescence

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

Gradual shortening of the telomeres during a cell's life continues, until the presence of critically short telomeres triggers a p53/Rb senescence pathway,10 which results in proliferation arrest. Because of that, a normal human cell can only divide 50 to 100 times in in vitro conditions. hMSCs are no exception, with the telomere shortening rate of 50 bp per population doubling.10, 29 CB hMSC, however, have slightly longer telomeres than other hMSCs, and thus can be cultured for longer before they senesce.25 Proliferation arrest in hMSCs results in senescence, which is described by the appearance of large senescent cells with flat shape, circumscribed nuclei, increased lysosome compartment, and endogenous β-gal activity.12 These morphological changes are not restricted to the senescent stage only, but represent continuous alterations in the course of hMSCs' long-term culture.12

In senescent hMSCs, up-regulated genes are highly over-represented (P < .0001) in the categories integral to plasma membrane, receptor activity, cell adhesion, vacuoles, and lysosomes,12 which coincides with the enlargement of the plasma membrane during senescence.12, 23 Significantly up-regulated genes in senescent hMSCs are: human glycoprotein neuromedin B (NMB), regeneration-associated muscle protease homolog (RAMP), p16 (INK4A), p53, p53 apoptosis effector related to PMP-22 (PERP), p21 (Waf1), lymphocyte antigen 96 (LY96), STAT1, and prion protein (PRNP).12, 32 The expression levels of p16 (INK4A), Cdk inhibitor 2A, plasminogen activator inhibitor type I (PAI-1), and c-myc genes steadily increase with hMSC population doubling.12, 32 Presenescent and senescent hMSCs can therefore be distinguished from early passage hMSCs by differential gene expression profiles.

Proliferation arrest influences hMSCs' karyotype, and at least 30% of senescent hMSCs display trisomy of chromosome 8.23 hMSCs start to be polyploid (mainly tetraploid) at passage 20, and become aneuploid afterward.32 During senescence, deregulated genes in hMSCs were mainly found on the chromosome region 4q22-q23, which when inserted into immortal cells caused a loss of proliferation.12 Senescence is also associated with up-regulation of micro RNAs, namely hsa-mir-371, hsa-mir-369-5P, hsa-mir-29c, hsa-mir-499, and hsa-let-7f, which change the methylation pattern.12 With respect to therapy application, special attention should be given during culture to hMSC epigenetic changes and the appearance of senescence that could result in genomic abnormalities.

Spontanenous Transformation of hMSCs

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

Cell transformation (spontaneous or artificial) is the process, initiated by at least 2 genetic events (mutations), by which cells gain immortality. First studies of spontaneous transformation claimed that unlike hMSCs, only murine MSCs can spontaneously transform in culture.42 It was concluded that in vitro BM-hMSC expansion and their use in therapy are completely safe.11 But in parallel, Rubio demonstrated23 that after 4 to 5 months of in vitro culture, 50% of the postsenescent AT hMSC clones can escape the proliferation crisis, resume proliferation, lose contact inhibition, and become tumor-like transformed mesenchymal stem cells.23 It was argued nonetheless that the susceptibility to malignant transformation is dependent on hMSC origin, and that AT hMSCs derived from stem cell poor fat tissue are more prone to transformation than BM hMSCs derived from stem cell rich BM.11 This argument was banned when BM hMSC clones were shown to spontaneously transform as well.21 A 2-stage model of spontaneous transformation was proposed, according to which a senescence crisis with proliferation arrest always precedes the resumption of proliferation that occurs when hMSCs undergo spontaneous transformation.44 This model was widely accepted and challenged only by Wang et al, who argued that spontaneous transformation may already occur as early as at hMSC isolation.24

Spontaneously transformed hMSCs (transformed mesenchymal stem cells) are morphologically distinct from early passage hMSCs and senescent hMSCs,23 as we demonstrated for BM hMSCs (Fig. 3). Being actively proliferating and compact, transformed mesenchymal stem cells exhibit a higher nucleus to cytoplasm ratio, prominent nucleoli, and more desmosomes.45 Transformed mesenchymal stem cells derived from BM hMSCs may sometimes become intermixed with short spindle cells, which lack contact-independent growth and form foci with cells released into suspension.24 Transformed mesenchymal stem cells down-regulate expression of typical early passage membrane markers CD34, CD90, and CD105, and phenotypically become CD133+, CD34−, CD45−, CD105−, and VEGF receptor+.23, 24 Because of the antigen profile and obvious transformed mesenchymal stem cell's morphology, the progress of spontaneous transformation is easy to detect.

thumbnail image

Figure 3. Human mesenchymal stem cells (hMSCs) changed morphology and spontaneously transformed during prolonged in vitro culture. (a-h) Transformation of the hMSC clone derived from bone marrow and grown in 20% fetal bovine serum Dulbecco modified Eagle medium is given as an example. (a and b) Early passage hMSCs at Passage 5 display spindle shape morphology with parallel orientation at high confluence. (c) Above Passage 18, the cells enter a presenescent stage and begin to acquire a flat shape. (d) After intensive cell death during Passages 19-23, the remaining senescent cells stop to proliferate. (e) Small proliferating cells, spontaneously transformed hMSCs (transformed mesenchymal stem cells), begin to emerge at the ends of senescent hMSC cell clusters after 3 weeks and form a (f) colony beside large senescent cells. (g) Through the following passages (Passages 25-35), the senescent cells are lost from the culture, and (h) a morphologically homogenous population of transformed mesenchymal stem cells is obtained at around Passage 35.

Download figure to PowerPoint

Spontaneous transformation is accompanied by distinct transcriptomic changes. Cells bypass the senescence crisis and transform by up-regulation of c-myc expression, repression of p16 levels, acquisition of telomerase activity, Ink4a/Arf locus deletion, and Rb hyperphosphorilation.44 Transformed mesenchymal stem cells display increased transcription and translation of Cdk1, Cdk2 Cdk4, Cdk6, cyclins B1 and D2, RAD51α, DNA-PKcs, ERCC4, DNA ligase IV, DNA polymerase,44, 45 and VEGF.23, 24 In contrast, they display down-regulation of cyclin D1, cadherin 6, Snai1, ACTA2, matrix metalloproteinase (MMP)-2, collagens Iα2 and Va2, syndecan2, fibulin2, connective tissue growth factor, and FGF-2.44, 45 A low level of TGFβ expression in transformed mesenchymal stem cells compared with hMSCs indicates that TGFβ signaling is not active in transformed mesenchymal stem cells because of overexpression of Smad7, telomerase, and c-myc.45 Although transformed mesenchymal stem cells posses a high level of telomerase activity compared with hMSCs, they soon become aneuploid, with various intrachromosomal translocations.24 Karyotype changes are consistent among different transformed mesenchymal stem cell isolates, and they include a translocation between chromosome 3 and 11, intrachromosomal rearrangements of chromosome 5, and the occasional presence of an isochromosome 8.23 Whether the genetic predisposition of an individual is responsible for these genomic changes, or they result from in vitro culture conditions, remains to be investigated.

Homing Potential and Migration of hMSC Tropism to Tumor Cells

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

Homing potential is the ability of hMSCs residing in a tissue niche to actively migrate in response to chemokines. hMSCs engraft and remain detectable only at injured sites, and persistently colocalize with the sites of tumor development.46 Because of that, they could be used as drug carriers for treatment of tumors that are difficult to reach and remove. In vivo BM hMSCs efficiently home to brain tumor and human glioma xenografts (U87, U251, and LN229).47, 48 Unfortunately, culture-expanded hMSCs exhibit much lower tropism to bone marrow and damaged tissue,43 because in culture they lose CXCR4, which is a key receptor responsible for hematopoietic stem cell homing.43 However, in vitro 3-dimensional culturing of hMSCs as spheroids increases stromal cell-derived factor-1 (SDF-1) (CXCL12) signaling, which restores functional expression of its receptor CXCR4 and homing potential,43 that is crucial for hMSCs' therapeutic application.

Homing potential of hMSCs to desired cells/tissues was proved in vitro by hMSC coculture with glioma cells or glioma-conditioned medium (Fig. 4) that increased hMSC invasiveness, similar to the addition of VEGF-A.49 The in vivo homing potential of hMSCs circulating in the bloodstream to the sites of injury/inflammation can be regulated by adhesion of hMSCs to endothelium, achieved by pretreatment of endothelial cells with proapoptotic agents.50 Other angiogenic and inflammatory cytokines and growth factors, such as IL-8, neurotrophin-3, TGFβ, IL-1β, TNFα, platelet-derived growth factor, EGF, and SDF-1 enhance hMSC migration as well.48, 51 Many of those are secreted by tumor cells,48 so this way tumors could attract hMSCs in vivo, where hMSC migration/invasion toward tumor cells is accompanied by their secretion of matrix-degrading enzymes MMP-2, membrane type 1 MMP, and MMP-9, and their inhibitors TIMP-1 and TIMP2.51 With controlled/maintained concentration gradients of cytokines that recruit hMSCs from BM via peripheral blood to injured tissue, hMSC homing could be optimized. In vivo myocardial injury was shown to establish the SDF-1 gradient toward the heart, which recruited hMSCs from BM.52 Moreover, in regenerating skeletal tissues, the hMSC homing may be improved with growth factor delivery, as combined hMSC and erythropoietin infusion gave better results in limb ischemia treatment.53

thumbnail image

Figure 4. Invasion potential of (a) human mesenchymal stem cell (hMSC) and (b) glioblastoma U251 cell changes in coculture conditions (Boyden chambers) is shown. (a) U251 cells or their conditioned media (CM) significantly increase the invasion potential of hMSCs (P < .05). (b) On the contrary, hMSCs or their CM glioblastoma cells significantly decrease the invasion potential of glioblastoma U251 cells (P < .05). MTT indicates 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) test.

Download figure to PowerPoint

The hMSC engraftment at the site of injury is not always optimal, and there is room for improvement. In vitro hMSC chondrogenic predifferentiation apparently increases engraftment during in vivo cartilage repair,34 and genetic modification enhances hMSC engraftment as well. In an animal model, transgenic hMSC-MCN-2 more efficiently engrafted and activated a pacemaker function of the damaged dog heart.54 In addition, hMSC engraftment is improved by sustained delivery of growth factors to injected hMSCs,34 their coinjection,55 or cotransplantation of hMSCs with healthy tissue parts.56

Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

Cell-based therapies for anticancer treatment rely on hMSCs' ability to home actively to distant tissues and their intrinsic ability to affect the cells at the target site. hMSCs can inhibit growth of the tumor in vivo57 and decrease the proliferation/invasion of hepatoma cells (H7402, HepG2)57 and of glioma cells (Fig. 4) in vitro. hMSCs inhibit the malignant phenotype of the cancer cells by lowering their oncogene expression (c-myc, proliferating cell nuclear antigen, survivin, and β-catenin) and increasing latent time of tumor formation.57

Unfortunately, hMSCs are also considered candidates for cancer initiation, because of their self-renewal ability, which is similar to cancer cells, and their ability to spontaneously transform in vitro,23, 24, 45 with higher frequency than other cell types.58 Such spontaneous genetic changes, if occurring in vivo, could lead to the establishment of a malignant (stem) cell population.59 Indeed, in vivo cancer metastatic cells acquire an invasive phenotype via a similar epithelial to mesenchymal transition, which also occurs during the 2-step process of hMSC spontaneous transformation in vitro.45 The hypothesis was proposed by Matushansky et al60 that tumors may develop from highly invasive transformed mesenchymal stem cells. They showed that stem progenitors of malignant fibrous histiocytoma (MFH) cells can be derived from transformed mesenchymal stem cells, and oppositely, that hMSC-like cells can be derived from MFH cells and differentiate in the hMSC's manner.60 However, hMSC and cancer cells still differ significantly regarding the expression of proliferation genes,61 which argues against the hypothesis of hMSCs being direct predecessors of cancer (stem) cells.

Furthermore, Rubio et al45 hypothesized that during spontaneous transformation, hMSCs undergo a dedifferentiation process, resulting in transformed mesenchymal stem cells, which overexpress podocalyxin antigen and stage-specific embryonic antigen-4—a pluripotent marker of embryonic stem cells. Those markers give transformed mesenchymal stem cells a more primitive, embryonic stem cell (ESC)-like characteristic compared with hMSCs.45 Similar to ESCs, transformed mesenchymal stem cells are able to integrate into developing blastocysts and contribute to normal placental tissue during embryonic development, without tumor formation.45 Their repair role was also demonstrated in age-related tumors, where by fusion of transformed mesenchymal stem cells with host cancer cells, the latter converted into nonmalignant cells.58 Nevertheless it remains that unlike hMSCs, injected transformed mesenchymal stem cells form poorly differentiated, solid tumors with epithelial phenotype in nearly all organs of mice,23 where they grow aggressively and display large areas of necrosis.23, 45 In tumors, transformed mesenchymal stem cells contribute to the tumor vasculature and AT, and recruit host BM-derived cells to the site of tumor formation.24, 58 The unpredictable transforming ability of hMSCs casts doubt on their use in in vivo therapies.

hMSC Genetic Modification for Anticancer Treatment

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

Because of hMSCs' ability to home to distant tumors, strategies of using them in anticancer therapy are being developed. BM is still the preferred tissue source for hMSC isolation, but the low frequency of hMSCs in BM necessitates in vitro expansion before clinical use. At the same time, senescence-associated growth arrest occurring during in vitro culture limits the hMSC numbers that could be generated. Together with the transformation phenomenon, this questions in vitro expansion of hMSCs, which is necessary to prepare hMSCs for anticancer treatment via genetic modification.

hMSC's limited lifespan can be prolonged by viral transduction of human telomerase reverse transcriptase (hTERT).62 Lentivirally transduced hMSCs never senesce or malignantly transform, and apart from ectopic hTERT expression, show no endogenous telomerase activity.62 The tumor suppressors Rb, p21, and p53 are not down-regulated in transduced hMSCs, suggesting that growth of those cells is still normal and under the control of those known gatekeepers.62 Lentiviral hTERT transduction thus seems to be a safe way to generate immortal nonmalignant cells for further gene modification and therapy applications. For invasive types of cancer, a novel therapeutic approach using a combination of hMSCs as intermediate carriers and conditionally replicating adenoviruses has emerged.63 Due to hMSC intracellular support to conditionally replicating adenovirus amplification, more efficient tumor cell elimination was observed in distant tumors compared with direct viral injection.64

The main aim of tumor growth inhibition strategy is to genetically modify the hMSCs directly to act as a drug vehicle that exerts tumor cell's (tumor stem cell) toxicity and causes complete irreversible elimination of tumor. Treatment of human colon cancer xenografts (HT-29Inv2, CCS) in mice with hMSCs expressing a suicidal tyrosine kinase gene (HSV1-Tk) resulted in tumor elimination and postfestum healthy stroma formation.65 Therefore, genetically modified hMSCs with active homing potential and a preserved helping role in tissue regeneration could potentially be used for treating infiltrative tumors that cannot be cured with conventional drugs.

Delivery of hMSCs In Vivo and Monitoring of Their Behavior

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

Delivery routes and cell numbers required should be optimized for in vivo use of hMSCs for regenerative or anticancer therapy. In this respect, peripheral and direct injection of hMSCs into the injured organ are still preferred routes to be used in regeneration therapies.54 Bone fractures in rats can thus be cured successfully with paraspinal injection of 106 transgenic hMSC-BMP9,66 whereas cell doses of about 7 × 105 cells are sufficient for restoring a pacemaking activity in dogs.54 Excessively small cell numbers, however, may prove insufficient and fail to show the positive effect.67 hMSC numbers should always be optimized precisely to reach a positive outcome, and they depend heavily on the delivery route to be taken during therapy.

Direct hMSC injection is useful at the sites of soft tissue injury, where by proper cell differentiation and function acquisition the organ function can be successfully restored. But when regenerating bone or cartilage, extracellular matrix components and their tensile strength need to be restored for organ full function. In an attempt to solve that, initial studies focused on hMSC in vitro differentiation before transplantation34 and hMSC embedding into scaffold structures, which can enhance chondrogenic or osteogenic hMSC differentiation.38 After a variety of biodegradable gel scaffolds were tested,34, 68 natural silk and wool-like microfibers combined with nanoparticles evolved as the latest development in nanofibrous scaffolds for tissue engineering therapy.69 Because those scaffolds require mild aqueous conditions during their construction, they represent an option for labile cytokine or recombinant protein (bone morphogenetic protein type 1) incorporation.70 In addition to serving as a release medium for cytokines and growth factors, they act as an immune barriers70 and could therefore be efficiently used during allogeneic hMSC transplantation.

For designing new anticancer gene/cell therapies, however, the possibility of in vivo, noninvasive, whole-body tracking during tumor targeting or tissue regeneration with hMSCs needs to be addressed. Past assessments of hMSC behavior in recipients have relied on visual detection in host tissue after sacrifice, thus failing to monitor in vivo hMSC dispersion. Therefore cytocompatible, optimized labeling systems for hMSC detection evolved, which allow tracking of the injected cells' behavior (migration) in vivo.71-74 Recent ones are based on quantum dots—small, light-emitting semiconductor nanocrystals, typically in the size range of 2 to 10 nm, metal nanoparticles, or nanonatural fibers.71-73 Their ability to either bind to selected integrins on the membrane of hMSCs or reside in the cell organelles73, 75 makes them useful for prolonged in vivo cell tracking. In case of metal nanoparticles like super paramagnetic iron oxide73 or ferocarbuten,75 magnetic resonance imaging tracking can be used, whereas hMSCs modified to express firefly luciferase can be monitored by bioluminescent imaging during tissue engraftment.46

hMSC-Based Cell Therapy and Reality

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

Understanding the aging effect on hMSCs is crucial for autologous therapy of degenerative diseases and cancer, because elderly people are primary candidates for hMSC-based therapies. By using a wide range of donor ages, measuring various hMSC parameters, a significant reduction of hMSCs' fitness in adulthood relative to childhood was confirmed.76 hMSCs undergo a decline in cell numbers, differentiation, and expansion capacity during aging.76, 77 Moreover, hMSC pools of disease-afflicted people may be deficient, as MSCs from a mouse model of systemic lupus erythematosus display defects in gap junctions and impaired differentiation,78 and the elderly may simply have inadequate BM hMSC pools because of its depletion caused by osteoporosis or metastatic bone disease.66 This is problematic for the use of autologous hMSC in therapy. In these cases, allogeneic cell sources should be considered. Obtaining sufficient numbers of allogeneic hMSCs for all treatments needed by patients represents a limit to modern hMSC-based therapies. In this respect, one should look for alternative cell sources. Examples may be articular cartilage-harboring stem cells79 and fetal bone cells that are positive for Stro-1 (an hMSC marker). Because the latter are more differentiated than stem cells, fetal bone cells in particular represent an interesting allogeneic cell source in tissue engineering applications.80

But could allogeneic cell therapy be as safe as autologous cell therapy? General concern over immune response and hMSC rejection diminished when experimentally transplanted hMSCs were detected in most of animals' organs, indicating a lack of hMSC immune recognition, clearance,81 and graft-versus-host disease.81 hMSCs indeed alter the cytokine secretion profile (tumor necrosis factor-α, interferon-γ, IL-4, and IL-10) of immune cells (dendritic cells, naive T helper [TH] type 1 cells, effector TH2 cells, and natural killer cells) and promote anti-inflammatory phenotype.81 On their own, hMSCs produce prostaglandin E2, which is causative for immune response modulation81 and B-cell proliferation.82 Also, a Stro-1–enriched population of hMSCs showed a 10× larger inhibitory effect on lymphocytes than nonenriched hMSCs.27 Altogether, this may be why infused hMSCs are only weakly immunogenic in humans, which speaks strongly in favor of a valid clinical use of hMSCs from HLA-mismatched donors.83 Nevertheless, in future therapies allogeneic hMSCs could be either artificially encapsulated84 or modified by engulfment of microcapsules slowly releasing immunomodulatory cytokines34 to more efficiently escape patients' immune systems.

Conclusions: Expectations Too High?

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES

A great deal of hope has been put into hMSC research and therapy development, but in reality the beneficiary effects of hMSCs may not live up to our expectations. hMSC transplantations may for instance not result in adequate hMSC engraftment, as proven in animal models,10 or hMSCs could migrate to other organs and act there in a harmful manner. Moreover, even if recruited correctly to the damaged organ/tumor, hMSCs could fail to repair the organ function or cure the tumor. Examples are predifferentiated cardiomyocytes, which failed to repopulate murine heart85 and porcine hMSC infusion, which failed to repair renal lesions.86 However, most ongoing clinical trials for the bone disease osteogenesis imperfecta,87-89 heart stroke,90 neurodegenerative disorder caused by amyotrophic lateral sclerosis,91 and graft-versus-host disease88 are giving promising results. But so far, we are still not able to fully predict hMSCs' behavior, homing potential, and their ability to spontaneously transform. Last is a problem at developing hMSCs for tumor treatment when genetic modification and prolonged culture are needed. By optimized/standardized culture conditions and the use of high-quality hMSCs (from amniotic membrane or CB), spontaneous transformation, unless genetically predisposed, should be avoided in vitro. However, this does not guarantee that the injected cells will not transform on in vivo administration. But we believe that with the structural and comparative genomic characterization of cells, we should in the future be able to answer how vital cellular functions, stem cell differentiation, homing, and spontaneous transformation are related and why proposed tissue endogenous hMSCs transform in vivo into tumor-associated fibroblasts or carcinoma-associated fibroblasts that play an important role in the growth of epithelial solid tumors.16 In this way, cell-based therapies for targeting tumor stem cells and treating currently incurable cancers may meet success and advance as accepted regenerative therapies where hMSCs are already being used.

REFERENCES

  1. Top of page
  2. Abstract
  3. Isolation Efficiency and Origin-Dependent Growth of hMSC
  4. Immunophenotype and hMSC Differentiation
  5. Genetic Fingerprint and Cytokine Profiling of hMSCs
  6. hMSC Proliferation Arrest and Senescence
  7. Spontanenous Transformation of hMSCs
  8. Homing Potential and Migration of hMSC Tropism to Tumor Cells
  9. Tumor Formation and Putative Involvement of hMSCs and Transformed Mesenchymal Stem Cells
  10. hMSC Genetic Modification for Anticancer Treatment
  11. Delivery of hMSCs In Vivo and Monitoring of Their Behavior
  12. hMSC-Based Cell Therapy and Reality
  13. Conclusions: Expectations Too High?
  14. CONFLICT OF INTEREST DISCLOSURES
  15. REFERENCES