News & Views | Published:


Blastomeres and stem cells

Naturevolume 444pages432435 (2006) | Download Citation


Generating human stem cells from a single cell recovered during preimplantation genetic diagnosis does not, in principle, harm the embryo. Can the approach be used in assisted reproductive technology programmes?

Elsewhere in this issue, Klimanskaya et al.1 report an exciting advance for the field of human embryonic stem (hES) cell biology. Readers may have a sense of déjà vu, however, for this paper first appeared as an online publication2 on 23 August. It was the subject of controversy, partly because of confusion over certain points. It now appears both in print (see page 481) and on the Nature website with an explanatory addendum.

The promise of regenerative medicine through stem-cell therapy is intoxicating. Stem-cell lines are pluripotent (that is, they can develop into various cell types); they are self-renewable; and they can differentiate into functional cells. Scientific hurdles remain, principally in inducing nascent stem cells into the desired differentiation pathway. But the potential for treating intractable human disease, by using stem cells to repair damaged tissue, is real. This, however, is a topic that is as much about ethics and politics as about science.

The approach of Klimanskaya et al.1 offers a possible route around the objection that the creation of hES cells involves destroying the embryo from which they arise. The authors propose that it could be applied as part of the established technique of preimplantation genetic diagnosis (PGD), which is used to identify genetic defects in embryos created through in vitro fertilization before they are implanted in the uterus.

Their procedure1 was as follows. During the early stages of embryo development, the fertilized egg starts to divide into cells called blastomeres. Klimanskaya et al. separated individual blastomeres from human embryos at the 8-cell stage, using a micromanipulation method widely used in PGD. Multiple blastomeres from a given embryo were singly placed in individual wells within a flask and allowed to divide. Approximately half did so, and were subject to further culture, as described in the paper. The original cohort of 91 blastomeres yielded 19 ES-cell-like outgrowths and two stable hES cell lines. This work with human blastomeres follows a demonstration by the same group that ES cells can be derived from single mouse blastomeres (Fig. 1a)3. In these earlier mouse experiments, an intact viable embryo developed that consisted of the seven remaining blastomeres; by contrast, in the work with human cells, multiple blastomeres were taken from the 8-cell stage and no embryos were allowed to remain in culture. This was a source of confusion in the earlier online publication2.

Figure 1: Production of embryonic stem-cell lines.
Figure 1

a, The experiment with mice reported earlier this year3. A blastomere removed at the 8-cell stage was used to produce stem cells; the remaining 7-cell-embryo was allowed to develop further, was implanted in the uterus and went on to form an intact embryo. b, The idea behind using human preimplantation genetic diagnosis (PGD) for hES cell production. The biopsy is the same as in a, but the single blastomere is allowed to divide. One cell is used for PGD and the other for hES cell production. This scheme shows the 7-cell-embryo being allowed to develop further, as it would in practice, but in the experiments of Klimanskaya et al.1 no embryos were produced because all blastomeres were removed. c, Production of hES cells from the inner cell mass (ICM) at the blastocyst stage. This approach is the best developed but involves destroying the blastocyst.

The main question raised by Klimanskaya et al.1 is whether a single human blastomere, already required for PGD, could be allowed to divide prior to genetic analysis: one daughter cell would then be used for clinical diagnosis, whereas the other would be used to derive an hES cell line. The remaining 7-cell ball of blastomeres would be implanted in the uterus to develop normally, as is PGD practice (Fig. 1b). The advantage of this procedure would be in creating additional hES cell lines. Parents could possibly bank hES cells from the child from which the blastomere came, so their child could benefit from patient-derived hES-cell therapies if they come to fruition. Furthermore, banking of large numbers of hES cells could be a way of providing immunologically compatible cell lines for a large proportion of the population.

Existing hES cell lines are derived from a later stage of embryo development — the preimplantation blastocyst, 5 to 6 days after conception (Fig. 1c). A blastocyst is needed because its inner cell mass (ICM) is necessary to generate stem cells; outer cells (trophoblasts) differentiate into the placenta. All hES cell lines 'approved' by the US National Institutes of Health are derived from an ICM, as, presumably, are the less publicized, unapproved hES cell lines. Yet there are deep divisions among both the public and politicians over whether it is ethical to use an ICM to create hES cells, even for therapeutic purposes. This is because of the need to destroy the embryo from which the ICM is obtained.

To obviate this (and perhaps help lift the embargo on federal research funding in the United States), various other alternatives have been proposed, as described in Box 1. The current contribution1 offers another route.

In PGD, the most common technique involves removing a single blastomere from the 8-cell (3-day) embryo. The blastomere is analysed to detect genetic abnormalities; the 7-cell embryos that have no abnormalities are then transferred to the mother's uterus for implantation and pregnancy. A variant involves analysis of the first and second polar bodies, products of early oocyte division, whose genotype can predict that of the oocyte that will contribute the female lineage4. In both approaches, embryo transfer takes place 24–48 hours after removal of a blastomere or polar body, although transfer can be delayed until the blastocyst stage (day 5 or day 6). The main point, however, is that there is a limited time window for blastomere removal, genetic analysis and transfer.

Is the procedure proposed by Klimanskaya et al.1 feasible in this context? Certainly there are ample cases of PGD, which is an increasingly attractive complement to traditional prenatal genetic diagnosis and is often preferred for identifying chromosomal abnormalities. Use of PGD decreases miscarriages, increases implantation rates and almost certainly increases livebirth rates in properly selected cases5,6.

For blastomere-derived hES cells to become a preferred option, all concerned would need assurance that deferring genetic analysis by allowing the blastomere to divide would decrease neither the already low pregnancy rate of in vitro fertilization (only 25–30%) nor diagnostic accuracy. Is the likelihood of providing accurate PGD diagnosis compromised as a result of waiting for cell division, potentially risking cell degeneration? In fact, only half the blastomeres divided in the current report1.

Safety is another concern. PGD apparently does not increase the frequency of birth defects in babies, but more than the usual manipulation might be necessary in generating blastomere-derived ES cells. Sequential PGD biopsies (first polar body, second polar body, 8-cell embryo) surprisingly seem not to diminish pregnancy rates4. If diagnosis is compromised by cell division, however, the temptation might arise to remove two blastomeres, one for PGD and one for hES cells. This surely would adversely affect embryo viability.

Potentially more problematic are the consequences of certain technical nuances used in Klimanskaya and colleagues' method1. Multiple isolated blastomeres from a single embryo were cultured together, presumably to derive benefits from cell–cell interactions. Because only one blastomere would be removed in clinical PGD, the authors propose that the blastomere should be co-cultured with its embryo; the embryo would later be transferred into the uterus to produce a pregnancy. The safety of this approach may depend on whether deleterious effects arise as a result of embryo culture, for allowing the blastomere to divide prior to genetic analysis means that the embryo has to remain longer in vitro. Perturbations of genomic imprinting due to the possibly longer time needed in culture are the underlying scientific concern. In imprinting, the expression of certain genes depends on whether they come from the mother or father; culture media may or may not mimic the milieu of the human reproductive tract in directing correct parental expression.

Despite these reservations, the derivation of hES cell lines from a single human blastomere using clinical PGD is realistic, and offers an attractive way out of an ethical conundrum. As Irving Weissman remarked in a previous News & Views article7, some politicians are already advocating that generating hES cells from the ICM should cease. Research would be redirected towards blastomere-derived or 'inhibited' embryonic lines (Box 1) — if indeed stem cells of embryonic origin are not abandoned altogether in favour of cell lines solely of adult origin, should that more demanding approach prove feasible. Yet efficacy of the alternatives to ICM-derived lines remains unclear. The reality is that ICM-derived hES cell lines exist, and in increasing number. The same cannot be said about the alternatives: research with ICM-derived hES cells should not be held in abeyance.

In the long term, will the method of deriving hES cells matter? Perhaps not. If we achieve a successful example of human stem-cell therapy with ICM-derived cells, the controversy over the provenance of the cells is likely to dissipate in large part. After all, opprobrium was initially heaped on advocates of prenatal genetic diagnosis in the 1960s, of in vitro fertilization in the 1970s, and of PGD in the 1990s. Each success was followed by widespread acceptance of these procedures8. Time and public opinion move on.


  1. 1

    Klimanskaya, I., Chung, Y., Becker, S., Lu, S.-J. & Lanza, R. Nature 444, 481–485 (2006).

  2. 2

    Klimanskaya, I., Chung, Y., Becker, S., Lu, S.-J. & Lanza, R. Nature doi: 10.1038/nature05142 (2006).

  3. 3

    Chung, Y. et al. Nature 439, 216–219 (2006).

  4. 4

    Verlinsky, Y. & Kuliev, A. Practical Preimplantation Genetic Diagnosis (Springer, New York, 2005).

  5. 5

    Munné, S . et al. Reprod. BioMed. Online 7, 91–97 (2003).

  6. 6

    Gianaroli, L. et al. Reprod. BioMed. Online 10, 633–640 (2005).

  7. 7

    Weissman, I. L. Nature 439, 145–146 (2006).

  8. 8

    Simpson, J. L. Reprod. Biomed. Online (2006).

  9. 9

    Meissner, A. & Jaenisch, R. Nature 439, 212–215 (2006).

Download references

Author information


  1. the Departments of Obstetrics and Gynecology, and Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, Houston, 77030, Texas, USA

    • Joe Leigh Simpson


  1. Search for Joe Leigh Simpson in:

About this article

Publication history

Issue Date


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.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing