Abstract
The reconstitution of adult stem cells may be a promising source for the regeneration of damaged tissues and for the reconstitution of organ dysfunction. However, there are two major limitations to the use of such cells: they are rare, and only a few types exist that can easily be isolated without harming the patient. The best studied and most widely used stem cells are of the hematopoietic lineage. Pioneering work on hematopoietic stem cell (HSC) transplantation was done in the early 1970s by ED. Thomas and colleagues. Since then HSCs have been used in allogenic and autologous transplantation settings to reconstitute blood formation after high-dose chemotherapy for various blood disorders. The cells can be easily harvested from donors, but the cell number is limited, especially when the HSCs originate from umbilical cord blood (UCB). It would be desirable to set up an ex vivo strategy to expand HSCs in order to overcome the cell dose limit, whereby the expansion would favor cell proliferation over cell differentiation. This review provides an overview of the various existing HSC expansion strategies—focusing particularly on stem cells derived from UCB—of the parameters that might affect the outcome, and of the difficulties that may occur when trying to expand such cells.
Introduction
In the early 1970s ED. Thomas and colleagues started to utilize hematopoietic stem cells (HSC) for transplantation (65,66). Hematopoietic stem cells are needed to regenerate blood formation after medical procedures in the fields of hematology and oncology, which are most often performed on patients with diseases of the blood, bone marrow, or with certain types of cancer. Much has been accomplished in the last two decades with regard to the separation, culture, and expansion of HSCs originating from bone marrow (BM) and peripheral blood (PB) (37). In the transplantation setting, fresh HSCs from BM and PB are commonly used without prior expansion.
It would be desirable to first expand the stem cells gained from umbilical cord blood (UCB) ex vivo before storage and later use because of their limited number (46). This would ensure a sufficient quantity for backup in the case of a failed engraftment and for broader use in the adult setting (the amount of cells needed for a transplantation is commonly calculated on the basis of the recipient's body weight).
Historically, it was of secondary importance to report on animal and xeno-free cell culture protocols because mainly fresh and nonexpanded cells were used for transplantation.
This article provides a concise summary of current knowledge on cell culture and expansion strategies for HSCs used for transplantation (i.e., CD34+ and/or CD133+ progenitor cells) and focuses particularly on evolving xeno-free protocols for HSCs from UCB.
Knudtzon (25) was the first to demonstrate the presence of HSCs in UCB. Since then UCB has clearly gone from being a biological waste product to an important source for use in allogenic and autologous stem cell transplantations. In the pediatric setting, the data clearly establish UCB as a standard of care that has now surpassed the use of BM and PB cells. Opinion in the literature has recently changed, and there is now medical interest in UCB as a source for stem cell transplantation for larger patient groups—including adults. Without question, UCB is going to be more frequently used, particularly for the thousands of patients who might potentially benefit from such a therapy, but who cannot find an appropriate BM donor.
There is growing evidence that the therapeutic use of UCB in over 70 diseases, including cancer and immunodeficiency disorders, represents a potential medical breakthrough that will benefit millions of patients. UCB also holds the promise that it can be applied in regenerative medicine and be used to successfully treat a wider range of indications in the future (67).
Stem cells from UCB have several significant and clinically relevant advantages over bone marrow transplantations. The most important ones include:
Speed: UCB can be stored and made available to patients without delay, while many patients waiting for a matching BM-transplant die during the lengthy search process for a suitable donor (days versus months) (6,69).
Immaturity of cells: UCB, when transplanted in patients, bears less risk of graft-versus-host-disease (GVHD) because immune cells found in umbilical cords are still in a primitive state. The immaturity of cells is a considerable advantage in the transplantation setting because a “perfect match” between donor and recipient is not an absolute requirement, as is the case with BM or PB transplantations (60,67).
Very young chronological age of UCB donors: compared to BM donors with an average age of 40 years (58), UCB is derived from newborns. The younger the stem cell source, the better this is for recipients, especially in cases in which children require a stem cell transplant (42,63).
Very low risk of transmitting clinically important infections through latent viruses, such as the cytomegalovirus (CMV) and the Epstein Barr virus (61,62).
Distinctive proliferative advantages of UCB stem cells, including higher cell cycle rates, autocrine production of hematopoietic growth factors, and longer telomeres than are found in BM or PB HSCs (36,70).
However, HSCs from UCB still have some disadvantages. The major disadvantages are the limited number of HSCs and the amount that can be collected on a single occasion. HSCs form UCB, as compared to BM and PB, display a delayed engraftment due to the limited number of cells available for transplantation (especially in the adult setting, in which more stem cells have to be transplanted per kg body weight) and the higher risk of graft failure (3). Because the outcome of a transplantation is in direct proportion to the number of stem cells infused (16), one strategy to overcome the cell dose limit is to use more than one UCB unit in the same patient. Ex vivo expansion of the HSCs is being intensively investigated as an alternative. This would make UCB-HSC transplantation available to a larger patient collective, and decreased engraftment times would allow for faster posttransplantation immune reconstitution. Furthermore, the ability to expand or generate HSCs ex vivo would greatly facilitate the development of novel therapies (e.g., gene therapy) for hematopoietic disorders, for which no current therapy exists, or in situations in which the numbers of HSCs available are insufficient.
However, ex vivo expansion of HSCs is still not as efficient as one would desire and need for the transplantation setting (19,32).
EX VIVO Expansion
The major limitation to the application of HSCs in various clinical settings has been the absolute number of stem cells available in stem cell products (blood, aspirates, and selected cells). Many researchers have evaluated ex vivo manipulation/expansion of HSCs as a first step towards their clinical application (18). There are a few requirements that should be met for HSCs, at least by an ex vivo expansion protocol (37): 1) promotion of self-renewal/HSC expansion ex vivo, without (or minimal) differentiation, which increases the number of primitive cells in a stem cell product; 2) optimization of the composition of expanded cells in a way that helps shorten the time of full engraftment after transplantation; 3) a sufficient amount of cells should be obtained from a low starting population (e.g., avoid large-scale harvesting of aspirates; a single UCB product could be used to reconstitute an adult).
In our manuscript, we highlight the most important parameters and strategies to achieve these aims. The original articles should be consulted for more detailed information.
Parameters for Ex Vivo Expansion of HSCs
Collection of HSCs From UCB
There is still room for optimizing the sample collection of HSCs from UCB. Immediately after birth the umbilical cord is punctuated with a needle, and the blood is collected in a conventional blood collection bag. The UCB is taken either while the placenta is still in the uterus or after the delivery of the placenta. In the majority of cases, the time taken to collect the blood is too short. It has been demonstrated that the last batch of blood, collected before the cord is totally flushed, contains the highest amount of HSCs (CD34+ cells) without variation in nucleated cell content (33). Optimal collection of blood samples is the basis for retrieving high numbers of HSCs, which also benefits ex vivo cell expansion. Interestingly, it seems that there is a correlation between stress during delivery and the number of both nucleated cells and HSCs in UCB (31).
Storage of Blood and/or Cells
Another important topic is the overnight storage and cryopreservation of blood or selected cell fractions of blood. It is possible to achieve high survival rates of primitive progenitor cells after overnight storage. Such cells are of major interest as data suggest that these progenitors are not only important for long-term engraftment, but also contribute significantly to the early phase of hematopoietic engraftment after myeloablative therapy (50).
Freezing the samples is the only way of storing them for future use. During freezing, osmotic gradient and/or intracellular ice crystallization result in cell dehydration and possible injury to the cells. Different cryopreservation methods and different cryoprotectants for minimizing cell damage during the freeze–thaw process are used or under investigation. The typical cryoprotectant is dimethyl sulfoxide (DMSO). In higher concentrations DMSO is, however, toxic to cells, and it has been demonstrated that the infusion of DMSO-cryopreserved cells is associated with negative side effects in patients. For those reasons, alternatives to DMSO, such as trehalose and sucrose, are being tested (56,60).
Furthermore, every freezing and thawing cycle results in a loss of cells and leads to a reduction in their ability to expand. Despite this, the use of frozen products is standard, and freezing might be seen as a selection step, because only the fittest and most potent cells survive (12,13).
Is Preselection of HSCs Necessary?
Cell Preparations Used for Expansion
Magnetic activated cell sorting (MACS) strategies are mainly used for cell isolation and preselection. These systems are based on the use of an antibody covalent bound to microbeads made of an iron dextran complex. MACS-selected cells that are infused into patients immediately after selection and that are still coupled with the microbead-bound antibodies do not seem to be associated with negative side effects, such as allergic reactions (54).
It has been shown that when cells are preselected using MACS the use of 2% human plasma or human serum albumin (HSA) results in an enhanced enrichment, ex vivo expansion, and might, furthermore, improve the engraftment of selected cells (27,51).
There are two surface markers mainly used for the enrichment of HSCs: CD34 and CD133. HSCs are phenotypically characterized at least by the expression of the CD34 marker, a surface glycoprotein that identifies a nonhomogenous cellular population comprising more primitive hematopoietic progenitor cells.
It has been shown that preselection of CD34+ cells facilitates the ex vivo expansion of blood products (32,37,52). However, no correlation between the higher purity of CD34+ cells and the improved expansion of these cells could be shown (27).
Besides, preliminary data suggest that CD133 may identify cells that are less mature than those that express CD34. The surface marker CD133 is often associated with a very primitive state of HSCs, rapid self-renewal, and low apoptosis (19).
There are a few issues to keep in mind concerning the preselection of cells. Selection is expensive, time-consuming, and may also lead to a loss of the most primitive HSCs, which are presumed to be negative for CD34 and/or for CD133 (15,43,72). The expression of surface markers, including CD34, can change depending on the activation status of cells and does not provide information on the functionality of the cells in vivo. Granted, there is the possibility that in a few years other markers will be assigned to stand for pluripotent HSCs, in which case the wrong fraction of cells will have been frozen.
One strong argument for the preselection of cells is the reduction in storage space, which is also reflected in lower storage costs.
Szabolcs et al. (64) demonstrate an additional strategy of preselection. Alloreactive T cells from UCB grafts were depleted by the combined use of trimetrexate and interleukin-2 immunotoxin under preclinical serum-free conditions. UCB grafts devoid of alloreactive potential might provide an ideal starting population for ex vivo expansion and for adoptive transfer as donor leukocyte infusions to reduce opportunistic posttransplantation infections that arise in the T-cell lymphopenic period.
Choice of Optimal Medium and Additives
Basic Media and Additives Used for UCB HSC Cultivation
BSA, bovine serum albumin; FBS, fetal bovine serum; FCS, fetal calf serum; HAS, human serum albumin.
The relatively cheap medium IMDM seems to be a good choice as a basis medium. It is an upgraded version of the basic medium DMEM and additionally contains amino acids, vitamins, and selenium. Media that are available commercially are well adapted to special cell types, but are extremely expensive.
Animal or human sera are added to support optimal cell growth. Sera are composed of up to 1,000 different substances in various and often unknown concentrations. They are undefined natural products of alternating composition whose substance content is unsteady. Besides the problem of the unknown composition of individual batches of serum (undefined culture conditions, regulatory problems, uncontrolled selection of cell types, etc.), sera also bear the risk of being contaminated with viruses or bacteria and might cause xenotransfection of nonhuman proteins [e.g., fetal calf serum (FCS) could be contaminated with prions, the infectious agents causing bovine spongiform encephalopathy, which is believed to also cause the variant Creutzfeldt-Jakob disease in humans]. Furthermore, it has been shown that transplanted cells exposed to animal material during their cultivation increase the risk of graft rejection (34).
In consideration of all these disadvantages, especially the input of nonhuman proteins, serum-free culture medium for the clinical use of HSCs has been proposed. Special media that are commercially available are often designed to be used without serum, but they are very expensive and not necessarily free of animal components (i.e., xeno-free).
Many recombinantly produced cytokines and many growth factors originate from cell cultures that are fed with serum containing media, and there is the risk of transfection of nonhuman proteins to those products. It is important to keep in mind that it is indispensable to supplement serum-free media with a few basic substances, that is, at least with insulin, transferring, and albumin, which are also recombinant or of allogenic origin.
Interestingly, a comparison of FCS with cultures containing autologous plasma has shown that in the case of FCS a concentration of 1% autologous plasma was as effective in growing cells as a concentration of 10% FCS (5).
Growth of Cells
The inoculum density is very important for effective expansion. It seems that paracrine cytokines strongly influence cell proliferation. For that reason, it has been strongly suggested that the density of cell numbers is important (26). Working densities depend on the composition of the media, but they should be in the range of 1 × 104–1 × 106 cells/ml. Generally speaking, if serum is used, only a low density of cells is required for expansion; if the culture medium is serum free, more cells are needed to start the culture.
After a lag phase of several days, HSCs usually start to proliferate exponentially and reach plateau levels on day 10–14. Most protocols work with cultivation times of 7–10 days. A short expansion time could be of advantage when planning chemotherapy and transplantations.
Scientists working in this area may be divided into two groups: those who work with floating cells in suspension culture, and those who work with the adherent cells of mononuclear cells or the CD34+ fraction, or who use the adherent fraction to support the growth of the floating population.
Some scientists use Teflon-covered bags to avoid cell adherence (4,12,40,41,48,49,52,53,59), whereas others try to establish the best conditions for adherence (10,24,47).
Under normal physiological conditions, the O2 concentration in UCB is approximately 5%, and efforts are under way to take this into account when working ex vivo (also under such conditions) (e.g., hypoxy-systems). The effects of different levels of oxygen in culture are still under investigation: one report (9) suggests that hypoxia improves the expansion of HSCs, others do not see this positive effect (11,35,73). Ex vivo expansion experiments with UCB cells under 2–3% O2 show that these conditions affect the cell cycle and result in a slower growth of cells without compromising their stemness potential (21).
Use of Cytokines for Expansion
Expansion strategies for HSCs have been investigated for more than two decades. Many unresolved issues still remain, such as, for example, whether cytokines should be included in the expansion medium and, if so, the type of mix (39). For many years scientists have believed that there are certain growth factors that can stimulate the proliferation of HSCs without differentiation and that can therefore control the self-renewal of HSCs. If such factors do exist, it might be easily possible to expand HSCs ex vivo and produce large numbers of HSCs for clinical application.
List of Different Cytokine Combinations
All data in ng/ml unless otherwise noted. Angptl5, angiopoietin-like 5; CXCL12, chemokine ligand 12; EPO, erythropoietin; FLT3L, fetal liver tyrosine kinase 3 ligand; G-CSF, granulocyte colony-stimulating factor; GM-CFS, granulocyte macrophage colony stimulating factor; IL, interleukin; LIF, leukemia inhibitory factor; MCP-1, monocyte chemoattractant protein-1; MGDF, megakaryocyte growth and development factor; MIP-1a, macrophage inflammatory protein-1a; SCF, stem cell factor; TPO, thrombopoietin; VEGF, vascular endothelial cell growth factor.
Such multifactor studies have al]so revealed that combinations and concentrations of growth factors are critical and suggest that careful growth factor dosing is important. The different effects of the cytokines may provide evidence of links between hematopoietic cytokines and key transcriptional regulators of HSC function (57).
The question also remains whether the expanded cells might go into apoptosis after culture in the presence of growth factors if the appropriate growth factors are not presented in vivo. Cells undergoing apoptotic death may induce inflammation through the release of inflammatory cytokines or chemokines.
It is also important to keep in mind that multiple cytokines and growth factors may interfere with the engraftment process itself and lead to unwanted side effects once infused into patients, such as the factors contained in stem cell suspension. Besides, the toxicity profile for many of these cytokines has either been shown not to be acceptable for transplantation or has not been determined (2,37).
The critical interplay between intrinsic and extrinsic factors is still unclear (7). It will be difficult to carry out ex vivo expansion of these cells until we understand the extracellular and intracellular signals that govern their fates.
Use of Feeder Cells
It has also been shown that direct cell–cell contact with adequate feeder cells favors cell expansion. At the very least, feeder cells produce supporting cytokines. Stromal feeder cells are known to produce low levels of several cytokines (GM-CSF, IL-3, IL-6, SCF, FLT3L, TPO) (8). These cytokines are also used to expand cells. For that reason, the positive effect of feeder cells is easy to explain.
In the beginning, feeder cells were mesenchymal stem (stromal) cells (MSCs) from the BM of mice (8) or humans (11,32,55); later, MSCs from human UCB (20,22,71) or stromal cells from the placenta (74) were also used. Only human feeder cells are acceptable for further clinical application (for a detailed discussion, see Choice of Optimal Medium and Additives above).
It is not currently known whether noncontact conditions are sufficient for ex vivo expansion at all or whether stromal binding is required.
Clinical Use of Expanded Cells
The major aim in the field of HSCs is to use the cells for different kinds of cell therapies. However, the relative inability to expand HSCs ex vivo imposes major limitations on the current use of HSCs in transplantation. Despite scientists having worked on this topic for decades, it is still a fact that most culture conditions favor differentiation over expansion. Indeed, most ex vivo culture conditions lead to symmetrical HSC division and thus to differentiation, which promptly results in their depletion.
The plenitude of ex vivo studies has led to in vivo experiments and to the first clinical trials (12,28,44,48,49,53). However, standardized regulation of cell processing will have to be established and all supplements will have to be available at clinical grade if expanded cells are to be used in human therapy (38).
It has been shown that expanded CD34+ cells engraft more rapidly, but that they lack secondary and tertiary engrafting potential. While ex vivo culture stimulates the differentiation of cells, it does result in the loss of long-term engrafting potential. Clinical protocols may require the transplantation of two fractions of cells: an expanded fraction to provide rapid short-term engraftment and an unmanipulated fraction to provide stem cells for long-term engraftment (41).
Astori et al. (2) suggest expanding HSCs in two different culture conditions: the first culture condition promotes cell expansion, which is associated with lineage commitment and loss of stemness. The second culture condition induces cells to grow more slowly, which is associated with the survival and replication of primitive cells.
The research on HSC expansion is gradually leading to an effective method of expanding HSCs for clinical use. It is expected that new techniques and findings in the future will focus on the optimal setting for triggering HSC self-renewal ex vivo.
Footnotes
Acknowledgments
The authors acknowledge support by Medical University of Graz, Dr. H. Zech GmbH, and European Union commission grant LSHB-CT-2006-037261 ‘CRYSTALL’ (to A.H.Z.).