Stem-cell competition

Article metrics

The debate continues over the relative merits of using embryonic and adult stem cells for research — and perhaps, one day, to treat patients. Two new papers look at the abilities of these remarkable cells.

Last August tiny cells in Petri dishes captivated primetime television audiences. In a carefully worded address, US President George Bush discussed the medical potential, risks and ethics of studying human embryonic stem (ES) cells. For proponents, these cells represent our greatest hope for treating devastating disorders such as Parkinson's disease, diabetes and spinal-cord injuries. But for those who are adamantly opposed to the use of cells derived from human embryos, stem cells from adults have been advocated as an ethically palatable and experimentally reasonable alternative.

Two papers in this issue rekindle the debate. On page 50, Kim and colleagues1 describe how they generated a specific class of neurons from cultured mouse ES cells and used the neurons to reverse symptoms of Parkinson's disease in rats. In other words, ES cells can generate specialized cell types that are therapeutically effective in animals. Meanwhile, Jiang and colleagues2 (page 41) have derived remarkably versatile cells from the bone marrow of adult mice, rats and humans. These two studies further the promise of stem cells even while they stoke the debate about whether such cells should be obtained from embryos or adults.

Stem cells, essential building blocks of multicellular organisms, have two defining properties — they can produce more stem cells and they can generate specialized cell types such as nerve, blood or liver cells. Stem cells come in different varieties, relating to when and where they are produced during development, and how versatile they are. Pluripotent stem cells give rise to all cell types. The best characterized are ES cells (Fig. 1), which are derived from very early mouse or human embryos. These cells proliferate indefinitely in culture, while retaining the potential to differentiate into virtually any cell type when coaxed. So, in principle, ES cells might be able to generate large quantities of any desired cell for transplantation into patients.

Figure 1


Embryonic stem cell — primetime exposure.

Stem cells collected from tissues of adults or older embryos are typically more restricted in their developmental potential and ability to proliferate. For example, blood-forming (haematopoietic) stem cells make all types of blood cells in vivo, but proliferate little in culture and have been thought not to make cells of other tissues. Recent studies have raised the possibility that some adult stem cells can give rise to cells outside their tissue of origin; however, these results are controversial3 and have often proved difficult to reproduce4. Nonetheless, those opposed to using human ES cells tout the possibility of pluripotent adult stem cells as a way of realizing medical gain without ethical pain. Although researchers generate ES cells from pre-implantation embryos in culture, and several countries have sanctioned deriving human ES-cell lines from 'surplus' embryos created through in vitro fertilization, some remain uncomfortable with the destruction of human embryos, even those destined never to be implanted in a uterus.

Beyond these ethical issues, there are technical obstacles to the use of ES cells. First, these cells can be obtained only from very early embryos and, although several human ES-cell lines have been made, they will not be immunologically compatible with most patients who require cell transplants. So researchers will need either to derive many more ES-cell lines or to customize ES cells on a patient-by-patient basis by 'therapeutic cloning'. Second, undifferentiated ES cells form teratomas — benign tumours containing a mixture of tissue types — after being transplanted. Thus ES cells must be reliably differentiated into the appropriate cell type in culture before transplantation.

Moreover, until now it had not been proved that specialized cells derived from cultured ES cells can actually function within tissues after transplantation5,6. For example, mouse ES cells produce insulin-secreting cells in culture, but such cells have not been shown to reverse high blood sugar levels in mice with symptoms of diabetes6. It is perhaps not surprising that cells generated in vitro might not be equivalent to those arising in vivo, given the extensive cellular interactions and 'education' that take place during development. But Kim et al.1 have now overcome this problem in their work on rats with symptoms of human Parkinson's disease.

Parkinson's disease is caused by the death of neurons, found in the striatal region of the brain, that produce the neurotransmitter dopamine and are involved in controlling movements. The symptoms of Parkinson's disease have been successfully treated in animals by transplanting dopamine-producing neurons into the striatum to replace lost neurons. In humans, however, results have been mixed7. A central problem has been the scarcity of suitable neurons. These neurons would be abundant if they could be produced from cultured ES cells, but this has proved inefficient and it has been uncertain whether the resulting neurons would be effective.

Enter Kim et al.1. To enhance the yield of dopamine-producing neurons, the authors expressed the Nurr1 protein — which is needed to generate these neurons in vivo — in cultured ES cells. The resulting neurons survived after being transplanted into rats with symptoms of Parkinson's disease, and showed appropriate electrophysiological properties. More impressively, the rats started to recover normal movements. Others have transplanted partly differentiated ES cells to ameliorate Parkinson's symptoms in rats, but observed tumours in some of the animals8. By demonstrating efficacy while avoiding tumour formation, Kim et al. have achieved a proof of principle, although ES cells that have been genetically modified in this way might not be desirable for use in people.

To sidestep the disadvantages of ES cells, it would be ideal to identify a pluripotent adult stem cell that proliferates indefinitely in culture. This is just what Jiang et al.2 seem to have done. Prior work9 indicated that there is a population of stem cells in bone marrow, known as mesenchymal stem cells, that can form muscle, cartilage, bone and fat. Taking a similar approach, Jiang et al. started with non-haematopoietic bone marrow cells, cultured them, and isolated a population that they called multipotent adult progenitor cells (MAPCs).

In culture, single mouse MAPCs proliferated indefinitely and differentiated into many specialized cell types. Upon injection into mouse blastocyst-stage embryos, individual MAPCs contributed widely to developing tissues. And when injected intravenously into adult mice they gave rise to several blood and epithelial cell types. These stunning findings cannot readily be explained by incorrect identification of the progeny of the transplanted stem cells. But it remains possible that the MAPCs might have fused with blastocyst cells10,11, explaining their versatility in that experiment.

Are MAPCs the adult equivalent of ES cells? More data are needed, yet so far it seems that there are both similarities and important differences between these cells (Fig. 2). What, then, are MAPCs? They might be very rare pluripotent stem cells that persist from the embryo into adult life. To prove this it would be necessary to identify these cells prospectively in vivo (rather than retrospectively in vitro) by the marker proteins they express, and to purify them without an intervening culture step.

Figure 2: Comparison of embryonic stem (ES) cells, and the multipotent adult progenitor cells (MAPCs) described by Jiang et al.2.

ES cells (and mouse MAPCs) need LIF for growth; other cultured cells do not, so this seems to be a unique feature of pluripotent stem cells. The expression of Oct-4 in ES cells correlates with their versatility; if MAPCs are similar to ES cells one might expect comparable expression of Oct-4. Germ cells are eggs or sperm. Dosage compensation inactivates one X chromosome in females, so that males (which have one X and one Y chromosome) and females (XX) express X-chromosome genes to the same degree. Autotransplantation refers to the possibility of taking stem cells from patients, deriving the required specialized cells, and transplanting them back in the patients.

Alternatively, MAPCs might not actually exist in vivo. The extended period of culture might have triggered certain bone marrow cells to regress to a more primitive state, just as primordial germ cells — from which eggs and sperm are produced — can be reprogrammed in culture to acquire properties like those of ES cells12,13. (In fact, ES cells do not exist as such in embryos; they arise after being cultured.) If so, we have much to learn about how cells can be reprogrammed in culture.

MAPCs might prove useful in treating diseases irrespective of their origin. But, until they have been identified in vivo, it's premature to speculate that they might have a natural role in repairing injured tissues. Moreover, although MAPCs can generate many specialized cell types, it remains to be seen whether those cells function normally and could be used to treat animals, as Kim et al.1 used ES-cell-derived neurons. Such experiments will now be a priority, given the possibility of isolating MAPCs from the bone marrow of any patient, and transplanting the progeny of these cells back to the patient without risk of rejection.

The potential benefits of treating diseases by using specialized cells generated in vitro are enormous. But much further work is needed. Possible risks include the tendency of ES cells to form teratomas, and the unknown hazards of using cells — whether ES cells or MAPCs — that have been cultured for long periods. And, although MAPCs seem to have normal chromosomes, it is important to establish that the pathways governing cell proliferation are unperturbed. Otherwise, short-term gains might fall prey to long-term complications.

The work of Kim et al.1 and Jiang et al.2 will not resolve the debate over embryonic versus adult stem cells. Rather, it underscores the need for research in this area to continue unfettered by political concerns. Only then will the public have a chance to get what it deserves: novel, validated and safe treatments for intractable diseases.


  1. 1

    Kim, J.-H et al. Nature 418, 50–56 (2002); advance online publication, 20 June 2002 (doi:10.1038/nature00870).

  2. 2

    Jiang, Y. et al. Nature 418, 41–49 (2002); advance online publication, 20 June 2002 (doi:10.1038/nature00900).

  3. 3

    Wells, W. J. Cell Biol. 157, 15–18 (2002).

  4. 4

    Morshead, C. M., Benveniste, P., Iscove, N. N. & van der Kooy, D. Nature Med. 8, 268–273 (2002).

  5. 5

    Kyba, M., Perlingeiro, R. C. R. & Daley, G. Q. Cell 109, 29–37 (2002).

  6. 6

    Lumelsky, N. et al. Science 292, 1389–1394 (2001).

  7. 7

    Freed, C. R. et al. New Engl. J. Med. 344, 710–719 (2001).

  8. 8

    Bjorklund, L. M. et al. Proc. Natl Acad. Sci. USA 99, 2344–2349 (2002).

  9. 9

    Pittenger, M. F. et al. Science 284, 143–147 (1999).

  10. 10

    Ying, Q.-Y, Nichols, J., Evans, E. P. & Smith, A. G. Nature 416, 545–548 (2002).

  11. 11

    Terada, N. et al. Nature 416, 542–545 (2002).

  12. 12

    Matsui, Y., Zsebo, K. & Hogan, B. L. M. Cell 70, 841–847 (1992).

  13. 13

    Donovan, P. J. Curr. Top. Dev. Biol. 29, 189–225 (1994).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.