Lifeline in an ethical quagmire: umbilical cord blood as an alternative to embryonic stem cells☆

Tissue-specific stem cell differentiation 

At first glance, it seems surprising that tissue-specific stem cells may also have the ability to be multipotent. Lineage restriction of cells may be an important mechanism that has developed over time as a way to ensure reproducible development of the embryo and that developmental restriction occurs due to cell response to external signals from neighboring cells rather than by inherent limitation of the potential of the tissue stem cells.

Can tissue-specific stem cells be altered to produce a self-renewing multipotent cell similar to an ES cell? In vivo studies using clonal cell populations or single cells strongly suggest that the answer is yes. However, more studies are needed to verify whether or not tissue-specific stem cells can replace damaged or diseased tissue with functional cells.

To use adult stem cells safely for tissue therapy, we must first determine the mechanism(s) of stem cell differentiation. Do stem cells respond to signals involved in tissue healing? When HSCs are transplanted into severe combined immunosuppressed (NOD/SCID) mice, the recipient mice are irradiated, resulting in tissue damage to the bone marrow, liver, and other tissues. The cell signaling in the damaged tissue (e.g., liver) that triggers cellular repair mechanisms may also trigger the differentiation of HSCs to the liver phenotype. There has recently been success in using hematopoietic stem cells to treat a mouse model of Parkinson's disease. Parkinson's disease occurs due to the loss of dopamine-expressing neurons in the substantia nigra due to cell death. It is possible that repair signals from the surrounding neurons may trigger differentiation of HSCs, which find their way to the substantia nigra after transplantation, into dopamine-producing neurons (11).

An alternative mechanism could be cell fusion rather than transdifferentiation. In other words, cells don't actually transform, but rather fuse with another cell and take on its characteristics. Recently, neural stem cells cultured together with a muscle cell line (C2C12 cells) resulted in the conversion of the neural stem cells to muscle, but this conversion occurred only if the neural stem cells were in contact with the C2C12 cells. Clusters of neural cells separated from C2C12 cells within the culture failed to generate muscle characteristics. This conversion was irreversible; once the neural stem cells differentiate toward mature muscle cells they become unable to differentiate back into neural cells (12). Regardless of whether the stem cell fuses or transdifferentiates, the ultimate result is still repair of the damaged tissue.

Clinical evidence for the multipotential of HSCS 

In an interesting clinical study, donor bone marrow cells became functional hepatocytes in the recipient's liver tissue after transplantation. This was a retrospective study in which the bone marrow transplant recipients and the donors were of different sex. Using a combination of sex chromosome-specific fluorescence in situ hybridization (FISH) combined with immunohistochemistry for liver-specific markers, it was possible to clearly show that bone marrow cells contributed to functional liver cells within the recipients (13). The recipients received high doses of radiation to deplete their bone marrow of cells; it is likely that the radiation caused some hepatic injury at the time of bone marrow grafting. This may have set the stage for appropriate signaling from the damaged tissue to result in transdifferentiation of the bone marrow stem cells to hepatocytes. Alternatively, the mechanism of hepatocyte formation may have been cell fusion. Careful studies using donor- and recipient-specific cell markers may help to determine which of these mechanisms is active.

Therapeutic cloning 

The birth of the cloned sheep Dolly in 1997 suggests that a terminally differentiated adult somatic cell nucleus, reprogrammed using somatic nuclear transfer (SNT), can cause a tissue-specific stem cell to transdifferentiate (14). Recently, Hochedlinger and Jaenisch (15) used B-cell nuclei in an SNT experiment to generate functional ES cells. These two studies prove that a terminally differentiated adult cell can be reprogrammed to produce multiple cell types. Therapeutic cloning has some attractiveness since it avoids the problem of tissue rejection due to HLA mismatches. However, clonally derived embryos may not be normal in different aspects, including epigenetic stability and telomerase function 16, 17, and will not reach the blastocyst stage to allow embryonic stem cells to be produced from an inner cell mass. An additional concern is that human oocytes are currently necessary for SNT. Both problems will likely make therapeutic cloning impractical if not unattainable.

Bone marrow-derived adult stem cells 

Bone marrow-derived stromal and hematopoietic cells are able to differentiate into multiple cell types, thereby demonstrating ES cell-like properties. Hematopoietic stem cells purified from bone marrow and transplanted into recipient mice can differentiate into hepatocytes (18) and rescue a liver defect, demonstrating functionality. In another study, HSCs were purified and clonal populations examined to determine if they could self-renew or differentiate. Long-term repopulation of irradiated hosts showed that these cells not only migrate to the bone marrow, but also can differentiate into epithelial cells of the liver, lung, gastrointestinal tract, and skin, albeit in small numbers (19). Mesenchymal progenitor cells from the stromal cells of the bone marrow also have the ability to produce many diverse cell types from a clonal population, both in vivo and in vitro (20).

Are umbilical cord blood hematopoietic stem cells multipotent? 

Umbilical cord blood was first used for a successful bone marrow transplant in a patient with Fanconi's anemia in 1988 (21). Hematopoietic stem cells are found in the fetal circulation, and in the 100 mL or so of blood in the placenta and umbilical cord, which are typically discarded after delivery. Within hours following delivery, the HSCs migrate to the bone marrow where they provide the progenitors of all the blood-forming elements, including erythrocytes, leukocytes, and platelets.

The main focus of our laboratory is to determine whether cells from umbilical cord blood have similar properties to human ES cells. In preliminary experiments, human umbilical cord blood CD34+ cells were grown in vitro in serum, showed mesenchymal cell morphology, and were positive by PCR for bone (TRAP), muscle (desmin), neural (nestin), and astrocyte (Gfap) markers. The cells also had positive staining for the vimentin antibody, confirming the mesenchymal morphology. These results demonstrate the multiple cell potential of umbilical cord blood cells and confirm previously reported findings.

Cord blood stem cell expansion 

There are enough hematopoietic stem cells in an umbilical cord to replace the bone marrow of a child; only 25% of stored samples in our bank contain sufficient numbers of cells to transplant an adult. Stem cell expansion in vitro, therefore, is an important goal, not only for increasing the use of cord blood HSCs for replacing bone marrow transplants, but also for tissue therapeutics. Having enough cells to make clinical therapeutics possible will be a major hurdle to overcome. Hematopoietic stem cells will divide, producing one daughter stem cell and a progenitor cell, which will then produce mature blood cells. Reproducing this division process in vitro will only be sufficient to maintain levels of the starting population. However, it is difficult to cause hematopoietic stem cells to proliferate because the process also causes differentiation, as growth factors have both mitogenic and differentiation properties. The key barrier to in vitro cell expansion, therefore, is the loss of self-renewing stem cells that occurs during induced cell proliferation. For one stem cell to give rise to two new stem cells, the differentiation pathway must be blocked by triggering proliferation before the onset of the cells' internal differentiation program (22).

Efficiency of differentiation of various types of stem cells to mature cells 

Although the percentage of cord blood HSCs demonstrating nonblood markers was low in our experiments, other types of stem cells have a surprisingly consistent low rate of differentiation to mature cells. Murine embryonic stem cells have the highest efficiency of production of specific mature cells among all stem cells. Dang et al. (23) were able to increase the production of embryoid bodies to 42% with a continued differentiation to hematopoietic cells (as measured by hematopoietic colony forming unit [CFU]) to about 5% of the input ES cells. The conversion of murine ES cells to cells with neuronal markers was only 0.2% of the input ES cells (24), while others (25) demonstrated that 1% of the cells originally isolated from day 9.5 murine embryos produced neural spheres, with 20% of the primary spheres being capable of producing secondary spheres. Of these secondary spheres, 80% of the spheres are positive for oligodentricytes, neurons, and astrocytes; therefore resulting in a differentiation rate of less than 16% of the primary sphere population.

About 5% of skin cells isolated from a skin biopsy will proliferate in culture and demonstrate the ability to form neural cells (26). With mesenchymal cells from bone marrow, about 1/10,000 of adherent mononuclear cells (CD45−) will produce a proliferating colony with multiple potential. Of these isolates, when placed into differentiation cultures, 90% become endothelial or neural and 60% become albumin positive in cultures that favor hepatocyte development (20).

Our studies demonstrate that about 10% of our enriched population of umbilical cord blood stem cells will proliferate in culture and demonstrate an expanded range of differentiation. Of these proliferating cells, 80% will be positive for neural, endothelial, muscle, or mesenchymal cell markers in vitro.

Stem cells from different sources, therefore, demonstrate similar rates of differentiation. What distinguishes them from each other, as the prime candidate for cell therapeutics, is the ability to proliferate sufficiently to overcome the low rates of differentiation. Currently, ES cells have the most efficient proliferation in vitro but in the future, conditions for high levels of proliferation of tissue-specific stem cells are likely to be developed. Our research has already yielded a 15- to 20-fold increase in cord blood HSCs in culture; this increase is likely to improve in the future.

Ethical issues 

The use of human ES cells for tissue and cell therapeutics will ultimately be limited by ethical concerns since these cells are derived human embryos. The need for HLA matching of ES cell-derived tissues for allogeneic transplantation will require the production and banking of several thousand ES cell lines to make tissue therapeutics practical. This requirement would, therefore, require tens of thousands of starting embryos since ES cell derivation from human embryos is an inefficient process at present. We believe that a nonembryonic source of stem cells will help to overcome the ethical and practical issues associated with the use of human embryonic stem cells. Of adult stem cell studies so far, umbilical cord blood hematopoietic stem cells are the most attractive because they are obtained from material that is generally discarded following delivery. The development of several thousand umbilical cord blood samples to address the HLA matching problem is a simple matter, limited only by the number of deliveries in any particular region and the willingness of parents to donate cord blood.

Conclusion 

Studies by our laboratory demonstrate that human umbilical cord blood hematopoietic stem cells appear to have the capacity to proliferate in vitro without differentiation, to acquire markers consistent with an ES cell phenotype, and to transdifferentiate into nonblood cell types such as hepatocytes, neurons, and muscle. This observation raises the exciting possibility of replacing human ES cells for tissue and cell therapeutics with umbilical cord blood hematopoietic stem cells that are normally discarded with the placenta after delivery. In addition, the ease of creating umbilical cord blood banks containing tens of thousands of samples realistically ensures the ability to achieve appropriate HLA matching for recipients of such therapy, in contrast to the practical and ethical concerns of creating sufficient ES cell lines for proper HLA matching. We anticipate that future studies by ourselves and others will confirm the clinical usefulness of umbilical cord blood HSCs in tissue therapeutics and gene therapy.