Which Approach to Take?
Human embryos in excess of clinical need and donated with informed consent are cultured to the blastocyst stage. The surrounding trophoblast layer is selectively removed by exposing cells successively to anti-human antibodies and complement, a component of the immune system.
Illustration kindly provided courtesy of a collaboration between Monash University, the National University of Singapore, and the Hadassah Medical Center.
At the most basic level, we still need to characterize the stem cells of all human tissues. To begin with, we need molecular markers that can separately identify the small pool of stem cells from the far larger number of their descendant cells. We also need more information about the interactions between stem cells and the niches in which they live, and how those niches respond to the body's needs. Such information, which we currently have only for the hematopoietic, or blood-producing, stem cells of the bone marrow, may suggest therapies to increase the number of residual stem cells in a damaged tissue. Already, such knowledge allows us to recreate a person's blood system from a few harvested hematopoietic stem cells.
What about a worst-case scenario, in which a chronically ill patient has lost most of the stem cells in a tissue and needs replacements to survive? Today, the most feasible option would be to supply stem cells from the same kind of tissue, but obtained from an unrelated donor. This approach involves the same serious risks of rejection associated with any organ transplant from an unrelated donor.
A better approach would be to supply so-called autologous stem cells, those that are genetically identical to the patient. This is not currently feasible, but we have ideas about how to accomplish the feat. One way would be isolate and grow stem cells from a different tissue of the same patient, such as the bone marrow or skin, and reprogram them in vitro. To learn how to reprogram stem cells efficiently we need to study a whole range of experimental systems in which previously silent genes are reactivated and active ones switched off. Clues may come, for example, from studies on how the cells of an early embryo become restricted to different lineages. If we can understand the genetic circuits that control normal development, it may become easier to flip the switches in the laboratory.
A second approach is to use pluripotential stem cells lines derived from embryos at the blastocyst stage, reached soon after fertilization of the egg and before implantation into the uterus. Blastocysts, which consist of about 100 cells, contain a few unspecialized stem cells that can be coaxed to multiply indefinitely in culture (see Figure at left). Under appropriate conditions these cells will give rise to many different cell types. The first human pluripotential stem cells were derived from blastocysts obtained from an in vitro fertilization clinic because they were in excess of clinical need. This milestone occurred in 1998 in James Thomson's lab at the University of Wisconsin, Madison. A group at Monash University in Australia has recently achieved similar results (see Figure on next page). Both groups are currently characterizing the cells and their differentiated descendants.
These studies will provide invaluable data about gene function during early human development. Woefully little is known about this topic at present, due in part to restrictions on federal funding for embryo research. Although developmental mechanisms have been highly conserved in evolution, enough differences in detail have been seen among vertebrate species to suggest that not all genes will function identically in mouse and man. Thus, research with animals only cannot reveal everything we need to know about how to manipulate human stem cells.
