The Flawed Scientific Basis of the Altered Nuclear Transfer-Oocyte Assisted Reprogramming (ANT-OAR) Proposal

Abstract

First put forth in June 2005, the altered nuclear transfer-oocyte assisted reprogramming (ANT-OAR) proposal has been promoted as an ethically-acceptable alternative to the embryo-destructive methods now used to obtain embryonic stem cells. According to its proponents, the goal of ANT-OAR is to use the cloning process to create a pluripotent stem cell. This would be achieved through overexpression of the transcription factor Nanog (or a hypothetical substitute) both in the enucleated egg cell and in the somatic cell prior to transfer of its nucleus. Although the ethical acceptability of ANT-OAR has been publicly debated, its scientific feasibility has not. This paper aims to help rectify this situation. It argues that ANT-OAR, as currently conceived, cannot realistically work. It presents evidence from the scientific literature showing that Nanog cannot single-handedly establish pluripotency in cells, but rather works together with a network of other transcription factors to maintain pluripotency. It argues that ANT-OAR is based on a flawed understanding of stem cell biology, and emphasizes that, in this debate about embryonic stem cells, scientists must strive to accurately and realistically assess the feasibility of the embryo research strategies they propose.

Appendix

Box 1 Controversy over the Timing of Cdx2 Expression

Recently, there has been some controversy regarding whether or not Cdx2 might be present much earlier than the morula stage. A report by Deb et al. [23] seemed to show that Cdx2 mRNA is present in the zygote and, indeed, even in the oocyte, where its asymmetrical distribution determines, after fertilization, which cell of the two-celled embryo becomes trophectoderm (the late-dividing cell does), and which becomes the inner cell mass (the early-dividing cell) of the blastocyst. These results contradicted the two accepted models for pre-implantation development derived from lineage tracing experiments [24]. The first model predicts that one blastomere of the zygote has a bias toward becoming what is known as the embryonic part of the blastocyst, which contains the inner cell mass (ICM) and polar trophectoderm, and the other blastomere has a bias toward forming the abembryonic part, which contains the mural trophectoderm and more superficial ICM. Both blastomeres contribute to both parts; thus, blastomere fates are not strictly determined. Which blastomere contributes more to which part depends on subsequent cleavage patterns [25]. The second model predicts that the contributions of the two-cell blastomeres to the parts of the future blastocyst are completely random [26]. The Deb et al. paper contradicts both of these models because it predicts that the lineage fates of the blastomeres are already determined-not biased or random-at the two-cell stage. Notably, however, it has since come to light that the results of the Deb et al. paper may not be reliable. The editor of Science, in which the work was published, recently issued an “editorial expression of concern” [27]. The most recent news is that the paper likely will be retracted [28]. Nicanor Austriaco, in his recent paper on ANT-OAR ([9]; see main text), cites the Deb et al. paper in order to show that Cdx2 is present from the one-celled stage onward and therefore is not first expressed at the morula stage. Thus, in this scenario, a lack of Cdx2 in the ANT-Cdx2 entity would be “felt” from the very beginning. This would resolve the ethical problem that some had with ANT-Cdx2 and that led to the proposal of ANT-OAR, namely, the delay in the expression of Cdx2 (see main text). Nevertheless, the results of the Deb et al. paper also would have presented a conundrum, for how could the absence of a factor that is important so early on not have a profound effect on development? In such a situation, how could the embryo develop to the blastocyst stage at all? How could viable stem cells, which are extracted from the inner cell mass, have been obtained by Meissner and Jaenisch [3] in their study? The likelihood that the Deb et al. results are not reliable clears up this apparent conundrum; Cdx2 is, after all, expressed first in the morula, not the zygote.

Box 2 Ethical Problems Associated with ANT-Cdx2 and ANT-OAR

An ethical issue related to the experimental testing of both ANT-OAR and ANT-Cdx2 is this: even if they are found to work in mice–an extremely unlikely proposition for ANT-OAR, although ANT-Cdx2 apparently does work in mice [3]—this does not mean that they will work in humans. Recently, it has become evident that the transcriptional network that regulates pluripotency in human embryonic stem cells is not exactly the same as the network that regulates pluripotency in mouse embryonic stem cells [29]. Moreover, there are differences between species with regard to some of the signaling pathways that regulate pluripotency and self-renewal. This suggests that what works in mice might not work in humans. To find out if the procedures work in humans, they will have to be tested using human embryos. (Note that any conceivable form of ANT has this difficulty). This is not a trivial issue, as Davor Solter has pointed out [30]. The procedures will have to be experimentally verified in the laboratory, which involves a trial-and-error approach. How could this be done ethically on human embryos? During the verification process, and even after successful verification, there inevitably will be some individual “failures” in that some normal human embryos will be produced. What will be done with these embryos? Though apparently normal, they nevertheless might harbor defects that will go undetected. What woman would knowingly allow such embryos to be implanted and gestated in her uterus? (Moreover, would this not be an example of reproductive cloning, which most agree is ethically unwise?) How will these embryos be saved? Clearly, the ANT solution is not an ethically acceptable solution at all.