Technical Caveats in Stem Cell Preparation and Administration 364
Standardization of Stem Cell Methodology and Quality Control 364
Heterogeneity of Stem Cell Preparations 364
Caveats of Stem Cell Markers and Tracing Methods 365
Contamination of Stem Cell Preparations 367
Storage of Mesenchymal Stem Cells 368
Strategies and Problems of Stem Cell Delivery 368
Potential Drug Effects on Stem Cells 369
Sick People May Have Sick Stem Cells: How to Choose the Right Donor 370
Safety Issues Regarding Culture of Embryonic Stem Cells 370
Potential Clinical Problems of Stem Cell Therapies: Tumor Formation, Ectopic Tissue and Other Unwanted Effects 371
Risks of Long-term Mesenchymal Stem Cell Culture 371
Potential for Malignant Transformation 371
Functional Changes Resulting from Aging and Senescence of Cultured Mesenchymal Stem Cells 371
Other Changes Associated with Long-term Culture 372
Stem Cells and Malignancies 372
Phenotypic Similarity and Local Proximity between Mesenchymal Stem Cells and Cancer Stem Cells 373
Mesenchymal Stem Cells Can Give Rise to Malignant Lesions 375
What Can be Done to Avoid Unintended Support of Tumors by Transplanted Mesenchymal Stem Cells? 375
Trophic Paracrine and Direct Cell-to-Cell Effects of Mesenchymal Stem Cells: Risks and Benefits 376
Non-malignant Maldifferentiation and Unwanted Actions of Stem Cells 376
Mesenchymal Stem Cells and Fibrosis 378
Unwanted Immunological Effects of Transplanted Stem Cells 378
A Different Concept: Stimulation of Intrarenal Resident Stem Cell Populations 379
The Human Risk Factor: Stem Cell Cures 379
Stem cell therapies are very likely to have a major impact on regenerative medicine, but for them to succeed we need to be aware of their specific challenges to avoid setbacks as experienced in the first gene therapy trials. In stem cell therapies, just as in gene therapy, malignancies are a major concern. Other concerns relate to contamination of cell preparations and unwanted formation of ectopic tissue (maldifferentiation), but also to the validity of experimental results, e.g. stem cell markers and heterogeneity of stem cell preparations. Consequently, this chapter will address both the medical safety aspects of stem cell therapies and the scientific safety of stem cell-related data. In reflection of the current literature on stem cell therapies in nephrology, the main focus here is on mesenchymal stem cells (MSCs) and other adult bone marrow preparations. However, with rapid developments in the field, safety aspects concerning stem cells other than MSCs may gain importance in the near future and therefore will be mentioned throughout this chapter.
No medical treatment can ever be considered completely safe. For example, much excitement greeted the prospects of gene therapy, but the field suffered a severe setback when several patients from the first clinical trials developed leukemia . Malignancies, mostly related to applications of embryonic stem cells (ESCs), have evolved to be a major concern in stem cell therapies as well. Other concerns relate to the validity of experimental results. For example, stem cell markers are often prone to technical misinterpretation or may be affected by heterogeneity of the cell preparations. Consequently, this chapter will address both the medical safety aspects of stem cell therapies and the scientific “safety” of stem cell-related data ( Fig. 23.1 ).
Currently, nephrological stem cell studies are almost exclusively performed in animal disease models. These mostly use rodents, sometimes crossing species borders by transplanting human stem cells into immunodeficient [e.g. severe combined immunodeficiency (SCID)] mice. Only a handful of clinical phase I studies on stem cells in renal diseases is on its way. The cells that usually come to mind when thinking about stem cell therapies are ESCs. But up to March 2010, no clinical studies using ESCs or induced pluripotent stem (iPS) cells in renal diseases were registered on the National Institutes of Health (NIH) site ( www.clinicaltrials.gov ). Rather, ongoing clinical trials in nephrology mostly investigate tolerance induction by hematopoietic stem cells (HSCs) or mesenchymal stem cells (MSCs) after kidney transplantation. In addition, very few studies involving MSCs target other renal diseases, such as a Chinese study on refractory systemic lupus erythematosus and a single-center study on acute kidney injury (AKI) . Consequently, this chapter will focus mainly on safety issues regarding MSCs. However, with the rapid developments in the field, including the first Food and Drug Administration (FDA) approval of an ESC study on spinal cord injury, safety aspects concerning stem cells other than MSCs may gain importance in the near future and will consequently be mentioned throughout this chapter.
Characteristics of Mesenchymal Stem Cells
MSC stands for mesenchymal stem cell or, more precisely, multipotent mesenchymal stromal cell, following the recommendations of a recent position paper by the International Society for Cellular Therapy (ISCT). MSCs have to fulfill certain consensus criteria regarding shape and surface markers, adherence, self-renewing capacity and in vitro differentiation potential, described in detail in an ISCT publication .
Although researchers still have to deal with the lack of indisputable criteria defining MSCs, making it necessary to constantly standardize the widely differing protocols for isolation, expansion and characterization, the emerging (and considerable) therapeutic potential of MSCs renders them strong candidates for human therapies. Recent studies have established animal serum-free culture conditions for human MSCs , a prerequisite for clinical use with respect to biological safety, and successfully established isolation and expansion of large numbers of MSCs derived by washings of bone marrow collection bags . If not mentioned otherwise, “MSCs” in this chapter will always refer to bone marrow-derived MSCs.
The preference for MSCs in nephrological research can be understood when comparing MSCs to ESCs. The apparently less powerful and versatile MSCs are favorable in many practical aspects: they can be cultured easily and without significant ethical concerns from adult bone marrow aspirates (and several other organs, e.g. adipose tissue and umbilical cord blood) and expanded under simple, inexpensive conditions in vitro. Their phenotypic stability has been considered superior to that of ESCs and, initially, no teratoma formation or other malignant transformation had been observed. However, recently, Tolar et al. described cytogenetic aberrations in mouse MSCs after several passages in vitro and (unwanted) sarcoma formation of transduced MSCs in recipient mice in vivo. Since then, several alarming observations on MSCs of different species have been published and will be discussed later (Potential Clinical Problems of Stem Cell Therapies, below).
Another appealing aspect of MSCs is that they can be obtained as syngeneic material from a patient before a calculated medical risk, e.g. major surgery, and later be readministered in case of organ failure (e.g. AKI). This idea led to the first phase I trial of allogeneic, intra-aortic MSC administration in AKI, mentioned above.
Allogeneic transplantation of MSCs seems feasible given a growing number of studies investigating and mostly confirming their tolerance-inducing effects by modulation of the immune response and their ability to escape T-cell recognition. These and other favorable properties of MSCs are currently being evaluated in several preclinical and clinical studies [reviewed in Ref. ].
Older studies on MSCs focused mainly on their ability to adopt non-mesenchymal phenotypes, e.g. neural precursors and cardiomyocytes. However, methods employed to verify differentiation of MSCs into other phenotypes in vivo are technically problematic and prone to misinterpretation (see Caveats of Stem Cell Markers and Tracing Methods and Fig. 23.3 , below). In more recent studies, beneficial effects of MSCs were largely related to paracrine effects, e.g. secretion of cytokines and growth factors, rather than to differentiation into new phenotypes . This is well in line with knowledge on the feeder function of MSCs for embryonic and hematopoietic stem cells.
The therapeutic potential of MSC-conditioned media opens up the prospect of cell-free stem cell therapy, which is potentially less risky than living cell injection.
Sometimes risks are easier to identify when recapitulating the natural processes. What is the physiological (and thus safe) contribution of stem cells to kidney regeneration? Bone marrow-derived stem cells in general may contribute to cell turnover and repair in the kidney . In an attempt to visualize this contribution, mice transgenic for enhanced green fluorescent protein (eGFP) were used as bone marrow donors and differentiation of bone marrow cells into glomerular cells expressing desmin was described in wild-type recipients . In rats with anti-Thy1.1 nephritis, differentiation of invading bone marrow cells into endothelial or mesangial cells was also reported . These data were interpreted to reflect a regular contribution of bone marrow to glomerular cell turnover, which is modified during disease and initiated the first studies which supplemented exogenous MSCs in renal disease models. Nevertheless, transplantation of MSCs is not a naturally occurring process. First, the cells are isolated from their niche, disrupting the regular microenvironment with its signals and interactions. Second, the cells are subjected to an artificial cell culture milieu. Third, the cells are forced mostly without preconditioning into a completely new environment, e.g. the myocardium or the glomerulus. It should not come as a surprise that the sum of these processes can sometimes lead to unexpected MSC behavior, as discussed later in this chapter (see Non-malignant Maldifferentiation and Unwanted Actions of Stem Cells and Fig. 23.6 A,B).
The potential of MSCs for renal repair has been shown in several rodent models of AKI, where the course of glycerol, cisplatin or ischemia–reperfusion-induced acute renal failure (ARF) was improved by MSC injection shortly after disease induction . Follow-up in these studies was usually short and no unwanted effects were noted. More recently, the first long-term studies in renal disease were published. Beneficial effects of bone marrow-derived cells or MSCs were noted in a model of chronic progressive renal fibrosis, i.e. in mice that are genetically deficient of the collagen α3(IV) chain (Alport mice). More studies investigating effects of MSCs in chronic fibrosis are coming up and will allow us to gain a more complete picture of possible long-term wanted and unwanted effects of these cells (see Potential Clinical Problems of Stem Cell Therapies, below).
Characteristics of Induced Pluripotent Stem Cells
A possible and powerful future alternative to MSCs is iPS cells. These cells were introduced in 2007 when they were created by two research teams by inserting genes encoding transcription factors into somatic human skin cells . Since then, various reprogramming methods for generation of iPS cells from a variety of tissue sources have been described. Nevertheless, iPS cells can lead to teratoma formation and considerable variations in the safety of iPS cell lines exist . In particular, reactivation of c-Myc was associated with tumor growth in iPS cell offspring and chimeric mice derived from c-Myc-negative iPS cells were much less prone to tumor formation .
There are no published studies using iPS cells in renal disorders yet, but initial data show that murine iPS cells can differentiate into cells expressing renal lineage markers in vitro. In particular, activin may enhance the differentiation of iPS cells to tubular cells .
Technical Caveats in Stem Cell Preparation and Administration
Standardization of Stem Cell Methodology and Quality Control
A recent publication described a “simplified culture and polymerase chain reaction identification assay for quality control performance testing of stem cell media products”. The authors wanted to offer a quick and sensitive assay/tool for validation and quality control for, in particular, MSC in research, and clinics and found it in a combination of culture assay, gene microarray and polymerase chain reaction (PCR). When this reportedly inexpensive low-technology assay indeed does live up to the expectations, it will certainly contribute to stem cell safety by facilitating to discriminate undifferentiated stem cells from early stage differentiating cells.
Apart from the ISCT position papers on MSCs mentioned above , a position paper on MSCs in solid organ transplantation with special focus on safety concerns has been published by MISOT, the Mesenchymal stem cell In Solid Organ Transplantation group and another paper on human stem cell research has been published by the World Federation of Neurology .
Not only in MSCs, but also in other progenitor cells small alterations in the culture conditions can induce a shift in cell phenotype. As an example, in cultured human fetal kidney epithelial progenitor cells, expression of E-cadherin was calcium dependent . Effects of needle diameter and flow rate on viability, phenotype and multilineage differentiation capacity of MSCs were tested in the past, and no significant interference was found .
Heterogeneity of Stem Cell Preparations
Heterogeneity in a stem cell preparation can strongly influence both its therapeutic potential and its safety. Moreover, heterogeneity may be mistaken for true plasticity or transdifferentiation across tissue lineage boundaries. Bone marrow, apart from hematopoietic stem cells, contains subpopulations of tissue-committed stem cells and primitive pluripotent stem cells which are released from the marrow to the blood upon injury . Thus, bone marrow per se is a heterogeneous stem cell/progenitor cell preparation. Even if aiming at distinct bone marrow subgroups, current isolation protocols do not guarantee homogeneity. As an example, MSC preparations are notoriously heterogeneous ( Fig. 23.2 ). These cells can be isolated from a multitude of tissue types but cannot easily be defined by phenotypic or functional characteristics . Even in a seemingly homogeneous MSC preparation, cells may be rapid dividers, slow dividers or non-dividers. In human MSCs vascular cell adhesion molecule-1 (VCAM-1) positivity and fibromodulin (FMOD) positivity predicted low progenitor activity and limited differentiation capacity, whereas small, spindle-shaped cells divided very slowly and were therefore discussed to be very early, quiescent progenitors . The existence of distinct clonal subpopulations in cultures of bone marrow MSCs has been supported by several studies. Recently, proteomic profiling of clonal subpopulations has been performed and demonstrated differential expression of 11 out of 83 proteins within the clones. These proteins were related to cytoskeleton and cellular structure, calcium binding and intermediate filaments .
The cell source is another confounder to heterogeneity and, for example, MSCs from different human materials such as placenta, cord blood and bone marrow express different cytokine profiles .
Apart from the cell source, cell density, culture conditions (including the use or non-use of fetal calf serum) and the number of passages all affect the differential expansion of specific cell subsets and may affect the therapeutic potential of the final cell preparation. For example, a study attempting to standardize MSC protocols found it comparatively simple to repeatedly obtain homogeneous MSC preparations, but culture conditions then had a major impact on their proteome and transcriptome .
To complicate matters further, the nomenclature for MSCs differs widely. Attempts to clarify both nomenclature regarding MSCs and definition of minimal consensus criteria for their functional and phenotypic properties have been made by the ISCT, as mentioned above .
In addition to heterogeneity resulting from differences in technical aspects, there is intraspecies and interspecies variability. For example, in rodents cross-strain transplantation data suggest that MSCs from different strains possess different immunogenic potential and induce variable immunological responses . Another ISCT publication recently compared bone marrow-derived MSC lines from four frequently used rat strains (Fisher and Lewis, both inbred, and Wistar and SPRD, both outbred) . They analyzed cell surface markers, population doubling times throughout 10 passages, and their differentiation capacity and proliferation rates in cocultures with spleen cells. Findings in these four rat strains were much more uniform than in different mouse strains, where key factors such as differentiation potential and cell surface epitopes can differ widely . A publication from 2009 investigated human first passage MSCs from different people with regard to their proteomic profiles and found seemingly identical proteomic patterns and functionally similar properties of the cells .
Finally, it is often desirable to verify rodent data in large-animal models, e.g. pigs, before embarking on human studies. In this respect it is of note that a recent study could not confirm the beneficial effects of MSCs observed in rodents with renal ischemia–reperfusion injury in an analogous porcine model . In that study, porcine MSC preparations failed to exhibit the same immune-modulating abilities as human or rat MSCs and induced the secretion of proinflammatory interleukin-6 (IL-6) in the recipient. Similarly, immunosuppressive effects varied broadly when comparing human, monkey and mouse MSCs . In yet another large-animal study, autologous MSC injection in sheep with ischemic kidney injury caused no benefit, although the injected cells could be localized to the kidneys . All the above caveats are likely amplified in xenograft models, e.g. transplantation of human adipose-tissue derived MSCs in immunocompetent C57BL/6 mice with AKI .
In summary, there is an urgent need to develop reliable reagents, widely accepted guidelines and standards for MSC preparations together with a definition of precise molecular markers to define subtypes of MSCs.
Caveats of Stem Cell Markers and Tracing Methods
In their seminal work from 2002, Terada et al. shattered the common belief in almost unlimited plasticity, i.e. transdifferentiation of stem and progenitor cells, by demonstrating spontaneous cell fusion as a mechanism by which bone marrow cells (BMCs) can adopt the (unexpected) phenotype of differentiated cells from other lineages ( Fig. 23.3 A). Transdifferentiation so far had mostly been deduced from donor-specific genes such as Y-chromosomes in differentiated cells within a female cell recipient. The authors cocultured mouse BMCs expressing GFP and a puromycin resistance gene with mouse ESCs in the presence of IL-3. Later, ESCs were removed from the culture by addition of puromycin. The remaining GFP-positive clones resembled ESCs in many features and were first assumed to represent BMC-derived ESCs. Only genetic analyses revealed quite unexpectedly that coculture of BMCs with ESCs did not induce ESC formation from BMCs. Instead, the surprisingly tetraploid and hexaploid DNA content of the GFP-positive cell clones and consecutive PCR amplification of multiple microsatellites that were polymorphic between genomes of ESCs and BMCs suggested that the BMC-ESC clones were created simply by fusion of these two cell types. These findings were consecutively reproduced by the authors in four repetitions of the experiment and showed dependency on IL-3 . Which exact bone marrow cell type might have been responsible for the fusion events could not be clarified and the authors noted that fusion was a very rare event. In an equally important work, Ying et al. showed similar mechanisms in cocultures of GFP-positive, puromycin resistance gene-positive progenitor cells of the central nervous system with ESCs . Again, tetraploid cell clones were formed exhibiting the genome of both cell types.
Fusion phenomena have since widely been accepted as a major caveat in interpreting events that seemingly visualize transdifferentiation of a stem cell into a differentiated cell. It has become common knowledge that the presence of not only donor cell markers but also recipient cell markers in an apparently transdifferentiated cell needs to be evaluated. Since 2002, observations that had led to assumptions of plasticity and transdifferentiation have been revisited and analyzed more closely. Since then, paracrine effects of injected stem cells have widely superseded the belief in plasticity as the main mediator of their beneficial actions . It should be mentioned that spontaneous fusion between mammalian cells had already been described as early as 1961 .
The first report on fusion events between renal and bone marrow cells after unilateral ischemia–reperfusion injury using two transgenic, gender-mismatched mouse strains as cell donors and recipients was published in 2007 and confirmed the above findings. Notably, fusion was observed in only seven out of 10,000 tubular cells, whereas 180 out of 10,000 tubular cells were estimated to be donor derived. The authors demonstrated that injury was a prerequisite for fusion and no such event occurred in healthy recipients of bone marrow transplants.
Another concern regarding transdifferentiation events came in 2006 . Thymidine analogs such as bromodeoxyuridine (BrdU), which have been used for cell tracking by incorporating into DNA during S phase, can be transferred from labeled dead donor cells to (dividing) neighboring recipient cells, thus mimicking transdifferentiation ( Fig. 23.3 B). This observation was made when 3–12 weeks after transplantation of thymidine analog-labeled, GFP-positive live stem cells into the central nervous system, few local GFP-positive cells but many thymidine analog-labeled neurons and glia cells were found. A control group using thymidine analog-labeled dead progenitor cells caused the same strong thymidine analog staining in the surrounding tissue, as did thymidine analog-labeled fibroblasts, whereas supernatants of labeled cells did not. This was the first conclusive report on false labeling of neural cells and it certainly applies to stem cell transplantation in other organs as well. It is well documented that the majority of injected stem cells dies quickly, thus offering large quantities of thymidine analogs to be processed by endogenous DNAses and to be incorporated into dividing recipient cells.
Another method to identify putative stem cells is the efflux of certain dyes, most prominently Hoechst 33342 as described in the so-called side population cells and other hematopoietic progenitors . This method has been used to identify putative stem cells in the adult human kidney but, overall, cells sorted for dye efflux are quite heterogeneous with all the consequences arising from heterogeneity as described above.
Even old-fashioned markers such as β-galactosidase (β-gal) expression in transgenic stem cells came under scrutiny in 2005 when Duffield and Bonventre discussed whether prior detection of apparent transdifferentiation of β-gal-expressing stem cells into tubular cells might in reality mirror increased intrinsic β-gal activity in the injured tubules . Another traditional method of cell tracking, the technically demanding Y-chromosome detection by fluorescence in situ hybridization (FISH) in gender-mismatched stem cell transplantation in ARF was discussed by the same authors to be prone to misinterpretation through artifacts caused by superimposition of a tubular cell and an infiltrating bone marrow cell .
Another principle for stem cell detection is quiescence. One of the first descriptions of intrarenal stem cells was published by Oliver et al. in 2004 , who described slow-cycling, BrdU-retaining cells in the renal papilla of adult mice that re-entered the cell cycle in AKI, thus giving the impression that they were recruited for repair. Later, others have shown that even fully differentiated epithelial cells can retain BrdU, merely because they are quiescent, which renders BrdU retention alone an unreliable stem cell marker .
Another new marker predominantly used to characterize human glomerular progenitor cells is CD133 . Kidney-derived CD133 + cells were later found to be able to contribute to tumor angiogenesis . Nevertheless, is has to be kept in mind that CD133 is one of the most widely accepted markers for cancer stem cells (CSCs) and has been described previously in renal malignancies and Wilms’ tumor , demonstrating the need to isolate putative progenitor cells based not exclusively on expression of one marker, but rather on a combination of several markers (e.g. CD24 + ) .
Contamination of Stem Cell Preparations
Risk factors for microbial contamination of stem cells include the source/donor as well as the methods of cell collection, expansion and, maybe, manipulation ( Fig. 23.4 ). MSCs are susceptible to human herpes viruses and transmission from the donor to the severely immunocompromised recipient has been a concern. Nevertheless, Sundin et al. did not find human herpes simplex virus (HSV), cytomegalovirus (CMV), varicella zoster virus (VZV) or Epstein–Barr virus (EBV) in MSCs from herpes virus-seropositive donors . In 2008, a retrospective analysis of 20 MSC preparations from healthy human donors uncovered one preparation with persistence of parvovirus B19 DNA. The persistence of this virus is facilitated by the expression of B19 receptor and a coreceptor on human MSCs. These cells can transmit B19 to bone marrow (hematopoietic stem) cells in vitro, underscoring the importance of monitoring B19 transmission of MSC products . In another investigation of MSC susceptibility to specific viruses, all tested human MSCs were found to express the human herpes virus-6 virus receptor CD46, but none of the cell samples was found to harbor the virus .
In the limited and restricted ESC lines, much effort was invested in securing the absence of infection. The International Stem Cell Initiative 2007 published its comparative findings on 59 human ESC lines worldwide and found no indications of (significant) contamination with mycoplasma, cytopathic viruses or bacteria .
The recent development of protocols to replace fetal bovine serum (FBS) in clinical-scale expansion of MSCs is certainly a step in the right direction to prevent bovine prion, viral and zoonose contamination of the cell preparation and product. Human platelet lysates from buffy coats still hold certain risks, but are undoubtedly superior to FBS .
Storage of Mesenchymal Stem Cells
Therapeutic use of MSCs is hard to imagine without cryopreservation. One study found a viability of approximately 90% in cryopreserved human MSCs from several donors and no negative influence of cryoconservation for up to 3 years on osteogenic potential . Chin et al. tested the feasibility and safety of off-site MSC culture for therapeutic use in ischemic cardiomyopathy. MSCs isolated from bone marrow aspirates were submitted to freeze-controlled cryopreservation with dimethyl sulfoxide (DMSO) and rapidly thawed immediately before intramyocardial injection in three patients. Again, cell viability was around 90% and patients were followed up for 1 year, remaining well . Others confirmed the resistance of MSC towards freezing and storage during ex vivo culturing , suggesting that cells derived from the same bone marrow aspiration might be used in the same patient at different time-points . Thus, cryopreservation of MSCs can probably be considered a safe and convenient storage method for both therapeutic and research applications, at least with respect to viability and osteogenic potential. However, with respect to immunomodulation, a recent study suggests that human MSCs should be stored for no longer than 6 months (and pooled from different donors) for optimal results . Nevertheless, even freezing for over 30 months allowed MSCs to suppress lymphocyte proliferation . In contrast to these data, liquid storage of human MSCs in saline or medium at +4°C or room temperature rapidly reduced their viability .
Strategies and Problems of Stem Cell Delivery
Different routes of stem cell delivery have been tested in various animal and disease models; not all have been efficient. In broad categories, investigators have delivered stem cells either systemically (usually intravenously) or locally.
For intravenous stem cell therapy, the capillary network of the lung imposes a major obstacle to site-directed delivery of stem cells (pulmonary first pass effect). For example, in rats the majority of labeled, intravenously injected cells, be they MSCs, multipotent adult progenitor cells (MAPCs), bone marrow-derived mononuclear cells (BMMCs) or neural stem cells, are trapped in the lung and only a few cells reach the arterial system, rendering any calculation of therapeutic dose questionable . MSCs performed worst in this respect and cell size is likely to be the key determinant . This is a major confounder in xenotransplantation experiments, when human stem cells are injected into immunodeficient mice. The issue of cell size also stresses the importance of using cells from early passages, as aging (stem) cells increase in size (see above), and the importance of optimized trypsinization and separation of adherent stem cells so that single-cell suspensions are obtained before injection. MSCs also express adhesion molecules such as intercellular adhesion molecule-1–3 (ICAM-1–3) and VCAM-1 and blocking them may decrease the first pass effect. Vasodilators may also be used to decrease the pulmonary first pass effect .
Following the primary entrapment in the lung, there may be an unpredictable time-curve of MSC rebound , resulting in a gradual redistribution of MSCs within 48 h to other organs, e.g. liver, spleen, kidney and bone marrow . There are also safety concerns that stem cells injected intravenously can damage the lung tissue itself, resulting in cystic structures with adjoining fibrosis . Finally, an (unwanted) influence of lung tissue on the MSC phenotype cannot be ruled out.
In other experimental and clinical situations, stem cells were applied locally. In general, strategies for local stem cell delivery are naturally more prone to side-effects such as bleeding, embolization, infection and even overdosing.
In the present authors’ hands, intravenous injection of MSCs in a rat model of acute glomerulonephritis did not exert any benefit, whereas intra-arterial delivery accelerated glomerular healing . In particular, in kidney transplantation local delivery of MSCs is feasible, for example via preflushing of the organ with an MSC solution before engraftment. However, in AKI models benefits from stem cells did not necessitate local delivery.
Another factor in stem cell delivery that may decide over success or failure is timing, especially in studies where immunomodulation by MSCs is needed. MSCs need to be licensed, i.e. they must spend time in an appropriate microenvironment with certain cytokines to allow them to acquire their immunosuppressive phenotype. Supernatants of cultured MSCs alone did not exert immunosuppressive effects, whereas supernatants from MSCs cocultured with activated T cells did . These findings are supported by an in vivo study in mouse autoimmune encephalitis, where administration of MSCs before, but not after disease induction could abrogate the disease .
Measures aimed at enhancing homing of exogenous or endogenous stem cells could also be useful. Stem cell migration and engraftment have been studied in various publications, especially in myocardial infarction . The basic requirements, e.g. adhesion to endothelial cells, extravasation of MSCs and the molecules involved, have been described by Ruster et al. .
A systematic analysis of the route of administration in regenerative cell therapy in the kidney is still pending, but such an effort was accomplished in 2010 for intravenous cell delivery in neurological disorders. That meta-analysis of 60 preclinical studies encouragingly revealed large treatment effects. Use of established stem cell lines, genetically modified cells and overall cell dose proved to be especially important .
Potential Drug Effects on Stem Cells
When asking whether stem cell preparations are safe for the recipient, one may also ask whether recipients are safe for the donated stem cells. It comes as no surprise that high-dose chemotherapy for treatment of hematological malignancies induces HSC damage, and it has been shown recently that the damage extends to MSCs as well . Autologous MSC cotransplantation in hematopoietic stem cell rescue could consequently be of lower value than expected. Chemotherapy targets rapidly dividing cells, and thus slow-cycling stem cells may remain largely unaffected in the short run. Nevertheless, even seemingly innocent drugs such as vitamin D 3 or triiodothyronine can dose-dependently induce stem cells to adopt a certain phenotype that may not have been wanted in the first place .
The steroid dexamethasone was shown to act antiproliferatively on MSCs in vitro . Other drugs with unwanted MSC side-effects are the thiazolidinediones, i.e. insulin sensitizers. Unexpectedly, they can cause anemia, possibly by inducing adipogenesis in MSCs and thus disturbing the normal bone marrow environment . Another example is the mammalian target of rapamycin (mTOR) inhibitor rapamycin, which modulates the MSC differentiation potential and inhibits the development of an osteogenic phenotype , whereas it was reported to promote the osteoblastic differentiation of human ESCs . Recombinant vascular endothelial growth factor (VEGF) protein inhibits the MSC expression of bone morphogenetic protein-2 (BMP-2), resulting in an inhibition of osteogenic differentiation .
Even common drugs such as acetyl salicylic acid can induce apoptosis via the Wnt/β-catenin pathway in MSCs . Non-steroidal anti-inflammatory drugs (NSAIDs; e.g. diclofenac or parecoxib) did not inhibit MSC proliferation or their direct osteogenic differentiation potential, but blocked chondrogenesis. They may thereby ultimately contribute to bone loss . These observations may be of particular relevance for orthopedic MSC-based studies, since orthopedic patients frequently are placed on NSAIDs in parallel.
Exogenous growth factors and hormones may modify stem cell functions as well. A prominent example is erythropoetin (EPO). MSCs express EPO receptors and respond to EPO with increased proliferation in vitro . As will be shown later in this chapter, MSCs can promote tumor growth via enhanced angiogenesis, which applies to EPO as well.
Apart from drugs, stem cells can be affected by physical challenges such as irradiation. Ionizing radiation in mice induced marked (and long-term) phenotypic and functional changes in hematopoietic progenitor cells . MSCs exhibit reduced osteogenic potential after irradiation and this may explain why irradiation of soft tissue (which is known to contain MSCs) can suppress ectopic calcification in rats .
Sick People May Have Sick Stem Cells: How to Choose the Right Donor
Stem cells can be affected by systemic diseases or show signs of a disease on their own. For example, MSCs from osteopenic rats showed less osteogenic differentiation in vitro and MSCs from patients with thrombocytopenic purpura showed an impaired proliferative capacity and a lower inhibitory potential on activated T-cell proliferation compared to healthy controls . However, others did not find significant differences in bone marrow progenitors from patients with or without hematological diseases, and the clonogenic potential of MSCs and HSCs was not altered .
Of particular relevance for nephrology are observations that renal failure has a pronounced effect on various stem cell functions. In patients with uremia, the frequency of functionally active peripheral blood endothelial progenitor cells is markedly reduced . Renal transplantation could normalize the phenotype of these cells . Similarly, c-kit + CD34 + bone marrow cells from patients with advanced age, anemia and renal failure exhibited impaired functions, in particular a reduced proangiogenic potency, compared with controls. After xenotransplantation into mice with ischemic limbs, the animals that received c-kit + CD34 + bone marrow cells from such patients showed worse blood reflow recovery than those from controls or patients with diabetes or hypertension . Other variables in addition to advanced age, anemia and renal function that led to reduced proangiogenic activity in this model included high serum triglycerides, C-reactive protein and IL-6 levels .
Similar to uremia, in patients with systemic sclerosis autologous MSCs were found to have a significantly reduced angiogenic potential . However, altered MSC behavior in immunological diseases is not universal and MSCs of patients with active Crohn’s disease maintained their immunomodulatory capacity, thus rendering them potentially useful for syngeneic cell therapy .
Gender may also affect stem cell functions. In a study on bone healing, female rats yielded significantly fewer MSCs in bone marrow isolations and the reduced number was associated with delayed bone formation, although functional characteristics of male and female MSCs were similar . Furthermore, in patients with leukemia who had received a stem cell transplant, 10 years later, secondary tumors developed significantly more often when the bone marrow donor had been female compared to male (4.6% vs 1.8%). These results indicate that stem cells from women may be functionally different, but the data are too preliminary to draw firm conclusions.
Another issue governing stem cell frequencies and functions is critical illness and/or exposure of blood to artificial surfaces. One study recovered mesenchymal, hematopoietic and epithelial progenitor cells from patients on extracorporeal membrane oxygenation . MSCs with trilineage differentiation capacity were present in 18 out of 58 blood samples, but not in normal controls. Thus, potentially a stress reaction of the bone marrow to severe shock, inflammation and/or the foreign surface led to altered stem cell frequency in the peripheral blood. The latter is supported by data showing that mechanical stimulation or shear stress can alter the phenotype of MSCs towards an endothelial differentiation .
In summary, testing of stem cells for functional abnormalities in patients with non-hematological, non-malignant diseases has just begun. Currently, healthy young male donors without medication seem the most reliable choice.
Safety Issues Regarding Culture of Embryonic Stem Cells
Human ESCs are typically grown on mouse feeder layers in the presence of animal-derived serum replacements. This can lead to the incorporation and expression of immunogenic non-human sialic acid on the stem cells . Sera from most humans contain antibodies against this sialic acid (Neu5Gc), which were demonstrated to damage ESCs in vitro and would probably lyse the cells in vivo as well. For human use, stem cell culture conditions eliminating all animal material therefore seem mandatory. The first stem cell lines maintained exclusively under cell- and serum-free conditions have been published .
Concerning the comparability among various ESC lines, several differences in the expression of lineage markers were detected and reported by the International Stem Cell Initiative 2007 . Several imprinted genes showed generally similar allele-specific expression patterns, but some gene-dependent variation was observed. Notably, some female lines expressed readily detectable levels of X-inactive specific transcripts (XIST), important for the inactivation of the X-chromosome, whereas others did not.
Potential Clinical Problems of Stem Cell Therapies: Tumor Formation, Ectopic Tissue and Other Unwanted Effects
Risks of Long-term Mesenchymal Stem Cell Culture
The lack of specific single MSC markers and their low frequency in the bone marrow usually necessitate their expansion in vitro before use in vivo. The longer MSCs are cultured, the higher the risk of unwanted changes in phenotype or loss of differentiation capacity, e.g. replicative senescence.
Potential for Malignant Transformation
While the potential of ESCs for malignant transformation is well established, it has been noted only since 2005 that adult stem cells can do the same after about 8 months in culture . This may in part relate to the observation that (at least mouse) MSCs with an increasing number of population doublings develop chromosomal instabilities ( Table 23.1 ). However, in cultured rat MSCs, a markedly aneuploid karyotype and progressive chromosomal instability has been detected in all passages analyzed .
|Murine MSCs||Mouse||Immunosuppressive effects of MSCs favor tumor growth when cotransplanting MSCs and a melanoma cell line.|
|Human ESCs||Rat||Teratocarcinoma formation by xenogenic ESCs in experimental stroke|
|Mouse BMDCs||Mouse||Helicobacter infection of the stomach recruits BMDCs that later form epithelial cancers|
|Mouse vascular progenitors from BM||Nude mice||Recruitment by vascularized tumors (U-87 glioma cell line derived)|
|Adipose tissue-derived MSCsfrom young children||–||Spontaneous malignant transformation in long-term culture|
|hTERT immortalized human MSCs||–||Give rise to CSCs on long-term culture|
|Murine MSCs||Mouse||Accumulated chromosomal instability leading to malignant transformation/sarcoma formation|
|Human MSCs||Mouse||MSCs within tumor stroma of a weakly metastatic human breast cancer cell line promote metastasis|
|Murine MSCs, non-virally labeled||Mouse||MSCs cause sarcoma formation|
|Murine MSCs expressing PAX-FKHR||Mouse||In combination with secondary mutations giving rise to alveolar rhabdomyosarcoma|
|MSCs||Mice with autoimmune diabetes||MSC transplantation causes malignant lesions|
|Human neural stem cells||Patient with ataxia teleangiectasia||Intracerebral tumor formation by donor cells|
|MSCs||Malignant transformation regulated by p53|
|Rat MSCs||–||Malignant transformation long-term culturing|
|Human MSCs||Immunoincompetent mice||Long-term culture causes frequent malignant transformation, MSCs form lung tumors upon transplantation into mice|
|Human CD133 + renal progenitor cells||SCID mice||CD133 + progenitors cotransplanted with renal carcinoma cells contributed to tumor vascularization in SCID mice|
In MSCs of children in vitro analyses for properties leading to oncogenic transformation (i.e. alterations in p53, p16, RB, H-RAS, hTERT, kariotype) remained negative and no tumor formation was shown after transplantation into SCID mice . In contrast, human MSCs cultured for 5–106 weeks underwent spontaneous malignant transformation in 11 out of 24 cultures and formed lung tumors when transferred into immunodeficient mice . Apparently, MSC transformation depends on several factors, among these culture conditions, the origin of the cells and the time of culture. A central role for p53 could be shown in controlling proliferation, differentiation and spontaneous transformation of MSCs. Knockouts for p53 exhibited signs of genomic instability and changes in expression of c-myc .
Further links between stem cells and cancer will be discussed in a later section (Stem Cells and Malignancies).
Functional Changes Resulting from Aging and Senescence of Cultured Mesenchymal Stem Cells
Donor age reduces the proliferative capacity of MSCs. Mostly independent of donor age, another process called replicative senescence can lead to changes in cultured cells that may resemble aging. Nevertheless, senescence and aging must be distinguished and MSCs from an old donor may not equal senescent MSCs from a young donor. Aging is “the sum of primary restrictions in regenerative mechanisms of multicellular organisms” , limiting cell replenishment and lifespan, whereas senescence relates more to functional changes inducing an irreversible arrest of cell division but with the (functionally deranged) cell staying alive, as opposed to apoptosis . Thus, senescene represses genes that support cell cycle progression and upregulate its inhibitors, like p53/p21 and p16/RB . This fits the hypothesis that senescence may represent a fixed program where genes involved in the proliferation machinery are downregulated, and this has indeed been shown in specimens from several donors, suggesting a common molecular program . Ultimately this might have been caused by accumulation of cellular defects, e.g. oxidative stress, telomere loss and DNA damage. Whatever the cause, senescence-associated changes in phenotype, global gene expression and miRNA expression profiles are continuously acquired and increase from the onset of in vitro culture, even before reaching senescent passage numbers (shown to start within passage 7–12 in human MSCs) . This emphasizes the importance of short-term culture.
Others reported human MSCs to undergo replicative senescence after in vitro culturing for several months as evidenced by prolonged cell population doubling times, cell enlargement, increase in podia, accumulation of actin, loss of differentiation capacity and DNA methylation pattern changes . Senescent cells also release metalloproteinases, inflammatory cytokines and growth factors that can alter the tissue environment and may contribute to malignant transformation . In a study investigating age-related changes in MSCs of Wistar rats aged 3, 7, 12 or 56 weeks, MSCs from 56-week-old donors not only showed reduced proliferation and fibroblast colony-forming units, but also accumulated oxidized proteins and lipids and had a decreased activity of antioxidative enzymes . Oxidative damage thus may affect the functionality of MSC preparations from old donors just as much as decreased proliferation. Other groups reported a decrease in colony-forming unit potential of MSCs from aged donors [summarized in Ref. ] and those cells too exhibited a decreased growth rate . Most groups also report age-related decreases in osteogenic differentiation and it has been assumed that this is accompanied by an increase in adipogenic potential , culminating in senile osteoporosis. For example, secretion of osteoblast-stimulating transforming growth factor-β (TGF-β) was found to be reduced in MSCs from aged mice and, together with reduced BMP-2/4, may contribute to their adipogenic switch . However, there is also a report on an early loss of adipogenic differentiation potential during MSC aging .
Senescence-associated changes do not exclusively affect MSCs but also affect human hematopoietic progenitor cells (HPCs). However, the overlap of differentially regulated genes between MSCs and HPCs was only limited , suggesting that senescence-induced changes in expression profile vary among the affected cell types.
Other Changes Associated with Long-term Culture
It has been shown that MSCs from low passages secrete more VEGF than those from later passages. The latter were also less protective in a mouse myocardial infarction model . Another functional impairment of MSCs after repeated passaging in culture is a loss of homing capacity . Finally, ex vivo expanded MSCs lost part of their therapeutic potential in ischemic stroke when cultured for extended periods . This may relate to their paracrine activity, since tissue levels of trophic factors in a rat stroke model were higher after injection of human MSCs from passage 2 than after injection of those from passage .
Stem Cells and Malignancies
Nephrology has been spared reports about stem cell-induced malignancies or fatalities in humans so far. In rodents, it has been known for a while that injections of pluripotent embryonic cells or embryonic stem cell-derived precursor cells into the central nervous system can lead to malignant transformation of these cells , mostly forming teratomas or teratocarcinomas . Nevertheless, several stem cell types have been used in preclinical studies for neurological diseases. A randomized phase II clinical trial investigated the stereotactic transplantation of cultured human neuronal cells (5–10 million), derived from the human precursor cell line NT2/D1, into the brains of patients with motor deficits after strokes . The maximum follow-up was 2 years with some benefit in daily living and only rare major adverse effects such as strokes. Nevertheless, hopes for stem-cell based therapies encountered a major setback in February 2009, when Amariglio et al. described a donor-derived brain tumor after neural stem cell transplantation in an Israeli boy with ataxia teleangiectasia. The multifocal glioneuronal neoplasm was first diagnosed 4 years after human fetal stem cells had been injected intracerebellarly and intrathecally in a Russian clinic. Microsatellite and human leukocyte antigen (HLA) analysis could track the tumor to at least two different donors , and indeed the boy had received repeated injections.
Another example of a possible causative role of stem cell transplantation in the subsequent development of cancer is the presence of second solid cancers after allogeneic stem cell transplantation . Patients who had received graft donations from women were at a higher risk for second solid cancers when compared not only to the general population but also to recipients of male grafts.
Apart from these clinical reports, most concerns about a putative role of stem cells in direct or indirect formation of malignancies stem from experimental data. In addition, the striking similarities between stem cells and CSCs have led to much discussion.
Phenotypic Similarity and Local Proximity between Mesenchymal Stem Cells and Cancer Stem Cells
The existence of CSCs was first hypothesized more than a century ago by Rudolf Virchow and S. Paget . It took until 1964 for the first cells with CSC characteristics to be isolated from germ cell-derived teratomacarcinomas . The prospective identification and purification of CSCs were successfully accomplished in patients with leukemia in 1997 and extended to solid tumors such as breast cancer .
CSCs have for a long time been considered a very rare population within a tumor and at the same time the only tumor cells with the capacity for unlimited self-renewal . Their existence could explain why most cancers return although they had been dramatically reduced in size by conventional anticancer drugs until no longer detectable. Explanations for the persistence of CSCs after therapy include their dormant state or low cycling frequency and expression of both antiapoptotic and multidrug resistance proteins .
Very recently it was found that not only very rare cells, but actually 27% of human melanoma cells could seed a tumor in severely immunocompromised mice , and the respective cells did not all express the standard stem cell markers such as CD133. These findings have challenged the hitherto existent understanding of CSCs as rare intratumor cells with unique tumor-forming potential and it remains unclear whether CSCs are a real entity, whether they represent “stem cells gone bad” or whether mature but abnormal cells somehow acquired stem cell-like characteristics. Indeed, stem cells and CSCs share many features, such as self-renewal (asymmetric cell division ensuring that the stem cell numbers remain constant). One of the possible mechanisms leading to formation of a malignant phenotype from a “regular” stem cell is loss of polarity control. Asymmetric division occurs perpendicular to the cell’s assumed polarity, resulting in both a stem cell and a more committed daughter cell . In theory, if non-asymmetric stem cell division generates only committed daughter cells, depletion of the original stem cell pool ultimately occurs. If non-asymmetric stem cell division generates only (dysfunctional) replicating stem cells, rapid tumor growth may occur .
Another theoretical model for interactions between stem cells and CSCs is the “seed and soil” model, as first proclaimed by S. Paget in 1889 . According to that theory, stem/progenitor cells are recruited to the circulation upon tumor signaling and home to organs, where they prepare a premetastatic niche (the “soil”), giving circulating tumor cells the successful extravasation, survival and growth in these locations (the “seed”). In 2007, MSCs were shown to act in accordance with this theory in a mouse model of breast cancer ( Fig. 23.5 ). Not surprisingly, their potent proangiogenic potential, mediated by secretion of angiogenic factors such as VEGF , can promote growth of vascularized tumors . Another mode by which MSCs accelerate tumor growth is in the allogeneic setting through their immunosuppressive effects . MSCs are likely to have chemotactic properties and their tropism for the tumor microenvironment has been acknowledged for a while, possibly because damaged tissues, wounds and tumor stroma secrete many of the same inflammatory molecules . Irradiation of (tumor) tissue can significantly enhance the homing of MSCs by amplifying the local inflammatory signaling . This is eligible when modified MSCs carrying therapeutic genes are used but otherwise destructive because it would facilitate survival of remaining CSCs within the irradiated area when supported by homing MSCs. Examples of successful studies using tumor tropism of genetically modified MSCs exist, e.g. Loebinger et al. employed tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)-expressing MSCs injections to eliminate metastases in a lung cancer model . Oncologists are currently evaluating the safety of such approaches for modulation of carcinogenesis in MSC-mediated gene therapies , aiming at delivering anticancer agents within the tumor.