Few subjects in biomedical science have captured the imagination of both the scientific community and the public, as has the use of stem cells for the repair of damaged tissues. Because they may be able to replace cells that have atrophied or have been lost entirely, stem cells offer the hope of restoration of cellular function and relief from suffering associated with many disabling disorders. Beyond tissue repair, cultured stem cells might also find application in the analyses of disease mechanisms and normal development, as assays for screening new drugs, and as vehicles for gene therapy.
Each potential use of stem cells promises revolutionary advances. However, the word "promise" must be underscored – to date, no cures have been realized, no disease mechanisms have been uncovered, and no new drugs have been developed. Many in the international scientific community believe that the promise of stem cell-based studies or therapies will be realized only if we can derive new human embryonic stem cell (hESC) lines.
Now, the production of new cell lines involves destruction of preimplantation embryos at the 100-200 cell (blastocyst) stage. Debate currently centers on the moral status of these embryos, which are now stored at in vitro fertilization (IVF) clinics or created by somatic cell nuclear transfer (SCNT). What is the moral status of the blastocyst? Should blastocysts be protected under the same laws that govern research on human subjects? These and related questions are at the center of a debate that involves the lay public and the scientific community.
The outcome of this debate will have an impact on the way we conduct the science of hESCs, a field still very much in its infancy. Indeed, the integrity of the scientific process and its independence from politics and from fundamentalist dogma are at stake. It is important, therefore, to define the relevant terminology and discuss it objectively. Along the way, we must reduce the emotional valence of phrases such as "therapeutic cloning" and "destruction of embryos." To engage in this debate, it is important to have an overview of stem cell biology.

A stem cell is defined by two properties. First, a cell can divide indefinitely, producing a population of identical offspring. Second, stem cells can, on cue, undergo an asymmetric division to produce two dissimilar daughter cells. One is identical to the parent and continues to contribute to the original stem cell line. The other varies in some way. This cell contains a different set of genetic instructions (resulting in an alter-native pattern of gene expression) and is characterized by a reduced proliferative capacity and more restricted developmental potential than its parent. Eventually a stem cell becomes known as a "progenitor" or "precursor" cell, committed to producing one or a few terminally differentiated cells such as neurons or muscle cells.
The different types of stem cell populations can be illustrated by considering the earliest stages of embryogenesis. Soon after fertilization, the haploid nuclei of the egg and sperm merge to form a single nucleus with the diploid number of chromosomes. The zygote divides and its progeny divide several times thereafter to form a compact ball of cells called the morula (likened in appearance to a mulberry). Each of the 32-128 cells in the morula is totipotent in that each one can give rise to all cell types in the embryo plus all of the extraembryonic tissues necessary for implantation in the uterine wall. These cells are also at the center of preimplantation genetic testing.
As the morula is swept along the oviduct, the cells continue to proliferate and the morula enlarges to form a hollow sphere called a blastocyst (or blastula). During the final days in the oviduct and the first days in the uterus, a few cells delaminate from the surface layer of the blastula to form an inner cell mass (ICM) within the cavity. This cluster of cells is the source of embryonic stem cells. It is important to emphasize that the ICM forms prior to implantation. Blastocysts created in vitro contain an ICM even though the embryo was created and maintained in a test tube. It is possible to isolate cells from the ICM of human blastocysts and grow them in tissue culture (Figure 1), using techniques first developed 20 years ago for the manipulation of mouse embryos. Cells dissociated from the ICM are pluripotent in that they can become any of the hundreds of cell types in the adult body. They are not totipotent because they do not contribute to extraembryonic membranes or the formation of the placenta.

Figure 1. Pluripotent stem cells, isolated from the ICM in the blastocyst, have the ability to give rise to all types of cells in the human body, but not the placenta and other supporting tissues.

The time from fertilization to implantation in the uterine wall is approximately 14 days in humans. Soon after implantation, the blastocyst invaginates, much like a finger pressing into a round rubber balloon. A critical series of cell movements known as gastrulation results in the formation of the three germ layers of the developing embryo: the ectoderm, the endoderm, and the mesoderm. The basic plan of the human body is laid out during this remarkable process as the fate of many cells is determined: the endoderm gives rise to the vasculature and blood-forming organs; the mesoderm produces muscle; and the ectoderm gives rise to the skin and the nervous system.
Stem cells are present in each of the three germ layers. The spectrum of offspring from these stem cells is more restricted than that of cells derived from the ICM, so they are described as multipotent rather than pluripotent. Cells in one germ layer breed true; they do not ordinarily transdifferentiate to form derivatives of other germ layers. Indeed, there is strong evidence for a restriction of developmental potential with time throughout embryogenesis. However, the plasticity of adult stem cells is an issue of great interest, and it merits further investigation.

Stem cells have been identified in adult tissues including skin, intestine, liver, brain, and bone marrow. Bone marrow stem cells have been studied most extensively because a variety of cell surface and genetic markers have helped delineate various stages of their differentiation during hematopoiesis.
However, several drawbacks a priori, make adult stem cells less attractive than embryonic stem cells as sources for most of the uses described above. It has been difficult to isolate stem cells from adult tissues. The cells are few in number, and it is difficult to keep them proliferating in culture. To date, it appears that cultured adult stem cells give rise to only a limited number of cell types. Finally, they are adult cells, have been exposed to a lifetime of environmental toxins, and have accumulated a lifetime of genetic mutations.

The ability of hESCs to proliferate indefinitely in tissue culture and the wide range of cell types to which they give rise make these cells unique. They become even more valuable as new molecules that trigger their differentiation in vivo are discovered. It has proven easier to mimic the normal sequence of development than to reverse this process in an attempt to have cells dedifferentiate.
In 1998, capitalizing on nearly twenty years of experience with mouse embryonic stem cells, scientists isolated stem cells from the ICM of human blastocysts and grew them in tissue culture for prolonged periods of time. Under the right conditions, several types of mature cells appeared in the cultures, including nerve cells, muscle cells, bone cells, and pancreatic islet cells (Figure 1). This paper has led to an explosion of research on hESCs.
Results obtained from studies with mouse ESCs raise the possibility that clinical trials with hESCs are not far off. Mouse ESCs have been steered to become spinal cord motor neurons, dopaminergic neurons, and many other types of cells. One example must suffice here to emphasize their therapeutic promise. In one of the most thorough and elegant studies published to date, mouse ESCs were steered to differentiate into spinal cord motor neurons by successive exposure to retinoic acid and sonic hedgehog, a protein known to trigger the differentiation of motor neurons in developing embryos. When treated cells were injected into the spinal cord of a chick embryo, they migrated to their proper location in the ventral horn.
Some cells sent axons out of the spinal cord to invade the developing limb (Figure 2) and form synapses on target muscle fibers. This type of research lends hope to individuals suffering from amyotrophic lateral sclerosis, spinal muscular atrophy, spinal cord injury, and related disorders.

Figure 2. Integration of transplanted mouse embryonic cell-derived motor neurons into the spinal cord in vivo. Transverse section through the lumbar region of the spinal cord reveals that enhanced GFP+ axons exit the spinal cord via the ventral root and project along nerve branches that supply dorsal and ventral limb muscles. The pathway of axons is detected by neurofilament (NF) expression.

Another argument for support of stem cell research follows from the success of transplanting intact human tissues. Pancreatic islets have been implanted into patients with type 1 diabetes to restore them to insulin independence. Islet transplantation works. Likewise, implantation of fetal mesencephalic brain tissue into the brains of patients with Parkinson disease resulted in measurable improvement in some indices of motor performance. Both implantation studies, however, were limited by tissue availability, and, in the Parkinson disease study, there were serious side effects (e.g., dyskinesias). Both studies call for further work with hESCs.

In Somatic cell nuclear transfer (SCNT), the nucleus from a mature cell is injected into the cytoplasm of an oocyte from which the original (haploid) nucleus has been removed. As in the union of haploid sperm and egg nuclei, one ends up with a diploid number of chromosomes, but in the case of SCNT, all of the chromosomes originate from the donor nucleus. The great advantage of SCNT is that the ESCs derived from blastocysts so created will be genetically similar to the cells of the individual who donated the nucleus. It is less likely, therefore, that the expressed proteins will be recognized as foreign and evoke an immune response in the host.
It is difficult to reproduce the course of early embryonic development by reprogramming an adult nucleus. Factors in the egg cytoplasm that regulate gene expression in the hours after a zygote is formed must act in the adult nucleus to produce the same patterns of gene activation that are critical for early embryonic development. The striking success in deriving new hESC lines from embryos created by SCNT has been widely noted.
These ESCs may not be entirely "normal," in the sense that subsequent coordinated development is impaired, but they can serve as starting points for the production of specific cell types.
The use of SCNT for the purposes of creating stem cell lines seems to be an innocuous process, but intrauterine implantation of blastocysts created by SCNT might lead to a live birth, a process known as reproductive cloning. Dolly the sheep was generated using this technology. However, successful reproductive cloning is an extremely improbable event. Most embryos created by SCNT are malformed and die in utero. It required 277 attempts to create Dolly, and there is strong evidence that Dolly exhibited many pathologies (e.g., arthritis, obesity) throughout her life. No one knows how many attempts it would take to create a live human being or what genetic abnormalities such an individual would bear.
Mice have been cloned from adult nuclei, even from postmitotic, terminally differentiated olfactory neurons. Although the full diversity of olfactory neurons was present in the offspring, more work is needed to define the limits of normalcy in such animals.
Years of experience with animals makes it clear that attempts at reproductive cloning of humans is scientifically unjustifiable at this time. Moreover, there are no compelling medical reasons to pursue this research. "Cloning" means to copy, and the word evokes an image of an identical replica. Given all of the epigenetic events that must occur during differentiation, it is inconceivable that an exact replica of an individual animal let alone a human being can be made. Parents of identical twins easily recognize the enormous number of differences between them.
The terms "research cloning" and "therapeutic cloning" have been applied to the creation of blastocysts by SCNT. These terms are unfortunate, as they have become confused in the public's mind with reproductive cloning. They share a common word, and they (wrongly) evoke the worst connotations of the oversimplified image of cloning.
The need for hESC research is extraordinary. We are on the doorstep of a new type of restorative therapy that goes beyond treating disease symptoms. Disorders in which the lesions are focal will be the first to undergo stem cell therapy. Replacing cells in the pancreas, motor neurons in the spinal cord, and dopaminergic cells in the basal ganglia are the most obvious examples. We must weigh the obligations of the moral imperative to help suffering individuals against the inherent value of preimplantation blastocysts.
We have many examples in history where attempts to outlaw fields of study have led to terrible and terrifying consequences (from Galileo to Lysenko). Conversely, many technological break-throughs now highly valued by both the scientific and lay communities, such as IVF or heart transplants, were once thought to be too dangerous or were seen as "playing God".
Finally, this effort should go forward because we simply will not know the answers unless we do the research. The desire to know is intrinsic to humans and has a survival value as well as a moral one.