Nature Editor's overview

Somatic cell nuclear transfer (SCNT), also known as “therapeutic cloning”, is an intensive area of study for stem cell scientists. It potentially provides a way to make patient-tailored embryonic stem (ES) cell lines to avoid the problem of immune rejection of cells transplanted for therapy.

It has long been believed in the cloning field that successful SCNT requires the use of unfertilised eggs. Eggan and colleagues at Harvard University now publish a remarkable technique that proves that unfertilised eggs are not required, at least not in mice1. They show that very early embryos can be used as recipient cytoplasts for nuclear transplantation, allowing the production of cloned animals and somatic cell–derived ES cell lines similar to those needed for patient-tailored ES cell therapies.

The procurement of human oocytes is a thorny issue because of the invasiveness of the procedure to retrieve them from women's ovaries. The authors argue that it will be easier to obtain early human embryos from in vitro fertilization (IVF) clinics than unfertilised oocytes, thus overcoming one of the major logistic and ethical hurdles of SCNT. Early embryos are destroyed in the process, but they would be ones donated by couples who had excess embryos following IVF.

Read below to see a panel of experts' comments (in black) on this interesting article, and responses from the authors (in italics). Table and reference numbers refer to those in the research article.

Care to comment? Blog here on The Niche

The Experts' Corner: Experts in embryonic stem cells, nuclear reprogramming and mouse epigenetics weigh in

Embryonic stem cell expert

This is an interesting and inventive article. The authors convincingly demonstrate that injection of mitotic chromosomes from mouse zygotes, blastomeres, ES and somatic cells into mitotic mouse zygote cytoplasm consistently leads to the developing of morula, blastocysts and, in some cases, live pups.

Previously, the reprogramming necessary for these transitions had been thought to require exposure of the new genome to metaphase II (MII) oocyte cytoplasm. Although it seems clear that the extent of reprogramming in zygote cytoplasm is inferior to that seen in oocyte cytoplasm, it is sufficient nonetheless to allow the generation of reconstructed blastocysts from donor ES or somatic (skin cells) chromosomes. These can be then be used to make nuclear transfer (nt) ES cell lines. Where an ES donor cell was used, the resulting ntES cells were shown to make chimeras and populate the germ line: the use of somatic cell chromosomes (tail-tip fibroblasts) led to ntES cell lines able to populate chimeras, although no germ line data were presented. Although transplantation of the ES donor–derived cell chromosomes into mitotic zygotes led to six live births, all died soon after birth of complications probably related to incomplete reprogramming. No births were reported for the skin cell–derived chromosome transfers, but the authors do not comment on this.

Using these data, the authors speculate that human aneuploid zygotes generated in the course of IVF procedures and caused by polyspermy or failure to extrude the second polar body, could provide a substitute for ethically contentious MII oocytes in the quest to make customized ES lines. They also speculate that mitotic ES cells also might provide similarly useful reprogramming activities.

Author response

At present we have no indication that the reprogramming ability of the metaphase zygote is lower than that of the MII oocyte. Rideout et al.1 reported a rate of 13% development of fibroblast-derived clones to the blastocyst stage and a 3% rate of ES cell derivation, whereas Wakayama et al.2 reported a 38% rate of blastocyst development and a 6% rate of ES cell derivation using MII oocytes. We report here a 10% rate of blastocyst development after successful somatic cell chromosome transfer (26/231) and a 2% rate of ES cell derivation (5/231) using mitotic zygotes as recipients. This efficiency is slightly lower but well within a range that is similar to that previously observed. In our tables, as we indicate above, are included all data from the very first time we attempted this method to our most recent experiments. Therefore, we believe that the total efficiency we report is actually an underestimation of the real efficiency in an optimised setting such as that reported in the two articles that we cite above for oocyte nuclear transfer. As a result we were rather conservative with our statement that the efficiency of development after transfer of ES cell genomes into zygotes is “comparable to nuclear transfer experiments with ES cell donor nuclei and unfertilised oocytes.” In fact, we believe that nuclear transfer into arrested zygotes is at least as efficient or more efficient than nuclear transfer into oocytes. With an optimised chromosome transfer protocol, we now have an efficiency of development to the morula and blastocyst stage of 70–90% for successfully reconstructed clones derived from ES cell donor chromosomes (see last experiments in Table 1). This efficiency, achieved in our most recent experiments is much higher than the 34% reported by Wakayama et al.

With respect to the issue of producing cloned mice from somatic cells by this method, we have thus far not attempted that experiment and instead here focused more on the issues relevant to the production of human ES cells by SCNT, a central area of interest for our group.

Expert

The authors understandably have not referred to the South Korean SCNT work, but they also make no mention of citations from (amongst others) Schatten's group referring to particular difficulties of chromosome assembly and segregation in primate SCNT embryos.

Author response

It is true that we intentionally avoid referring to work that involved the South Korean team, because we are concerned about the veracity of both experiments that were retracted and those that have not been. In the work that the referee refers to, it was found that primate SCNT embryos did not undergo normal cleavage divisions and many of the resulting blastomeres were found to be aneuploid. It is true that this may be a problem in many or all SCNT experiments in primates. We have considered these results and feel that these findings might make our approach even more enticing. In normal SCNT a nucleus with a centriole is transplanted into the oocyte. In humans, the meiotic spindle in the oocyte is nucleated without a centriole, whereas the mitotic spindle, after introduction of the centriole with the sperm, is. It is speculative, but this could account for some of the spindle abnormalities observed in primate SCNT embryos. In our chromosome transfer approach a mitotic spindle is replaced with a full set of mitotic chromosomes and the associated centrioles. This might alleviate the problems with chromosome segregation that were previously observed.

Expert

I presume that there is information on the behaviour of donor and host chromosomes in the abnormal zygotes. Do all of the chromosomes assemble in one place? Is it easy to remove them? I am surprised that the authors neither comment on this nor provide any preliminary evidence that such abnormal zygotes in mouse (assuming they occur) can act as vehicles for reprogramming.

Author response

It is known that chromosomes in polyspermic zygotes often coalesce at a single spindle in both mouse and human and that such embryos can undergo normal cleavage development. We cite an article in our discussion that documents this in human preimplantation embryos. However, in response to the reviewer's comments, we ourselves investigated whether aneuploid mouse zygotes that were either fertilised by two sperm or failed to extrude the second polar body could be enucleated by our method and were suitable recipients for chromosome transfer. We have modified the text of the manuscript and added a new Figure 5 describing these experiments in detail, including how we obtained aneuploid mouse zygotes. These new data include evidence that all supernumerary chromosomes can easily and reliably be removed at the first embryonic mitosis and replaced with a euploid genome of an ES cell. We show that these chromosome transfer preimplantation embryos developed at an efficiency that is similar to clones obtained after chromosome transfer into euploid zygotes. These experiments prove in an animal model that the approach we propose for human somatic cell chromosome transfer using aneuploid embryos is feasible.

In addition, we have also revised our manuscript to comment on the use of aged failed fertilised human oocytes for SCNT. Since our submission, aged oocytes have been successfully used for SCNT in the mouse. However, attempts at nuclear transfer with failed fertilized human oocytes have thus far failed. One potential explanation is that 'failed-to–fertilise' human oocytes cannot be obtained earlier than 24 hours after ovulation, whereas the mouse oocytes were aged for just 6 hours.

Nuclear reprogramming expert

This article reports a remarkable technical advance, in that they can carry out nuclear transfers in mice using chromosomes on spindles, rather than complete nuclei. As far as I can see, they achieve results as successfully, by this means, as is done by transplanting somatic cell nuclei into enucleated unfertilised mouse eggs.

This article does not claim to have revealed a new insight into development. Rather, the special interest in this work appears to be that it could open up a means of doing nuclear transplantation from human somatic cells into human eggs or embryos. As the authors say, this could obviate the difficulties that currently exist in obtaining unfertilised eggs from humans together with consequential ethical concerns. However, ethical concerns about using human zygotes might be at least as severe as with unfertilised eggs. In fact, the problems could be more severe, because permission would be needed from a mother and father of such embryos.

Author response

We agree that these ethical considerations surrounding oocyte and embryo donation are important to address. We feel that the main advantage to our approach is that women would no longer have to be recruited to donate their oocytes for research. We feel that our work will dramatically change the discussion on egg donation and stem cell research in the United States. Using this approach, women would no longer have to undergo the time-consuming, painful and potentially dangerous act of superovulation and egg retrieval to provide unfertilised oocytes solely for SCNT. Instead, couples that have completed their assisted reproduction treatment and that have frozen zygotes or cleavage-stage embryos in excess of their clinical need could donate them for stem cell/somatic chromosome transfer research instead of simply discarding them. These frozen preimplantation embryos have traditionally been the source used for deriving existing human ES cell lines. We believe that this expert is correct in suggesting that both members of the couple who were involved in the creation of the embryo (gamete providers) would have to consent for the use of the preimplantation embryo in this research. However, this is currently routinely done for donation of discarded embryos for human ES cell research and these proposed experiments should not be any different, nor any more difficult so long as the couple is interested and willing to participate in such research. In fact, we have already modified our human subjects protocols for derivation of human ES cells from discarded embryos and SCNT to allow for these experiments, and couples with excess embryos have already begun to donate to our program. Thus far, we have discovered that this is a much more ready source of recipient cytoplasts than dedicated oocyte donation for SCNT. It is true, that these experiments do not circumvent the question of the destruction of the human preimplantation embryo, which is problematic to some. However, that is an issue that is also raised by more traditional SCNT approaches.

Expert

I believe that the embryos resulting from IVF but not implanted back to a mother are selected at the blastocyst stage, rather than at the zygote or two-cell stage. It would therefore appear difficult to say that such embryos, selected at that stage, could not have developed normally.

Author response

During assisted reproduction treatment, preimplantation human embryos are frozen at several stages of development. These are, in order of most to least frequently, cleavage (four- to eight-cell) stages, the blastocyst stage and the zygote (one-cell) stage. When a couple completes their family after assisted reproduction treatment they often discard or donate their embryos. This occurs irrespectively of at what stage the embryos were frozen. In our program, in which couples donate discarded human embryos to us for human ES cell derivation, we have obtained significant numbers of human embryos at each one of these stages. We agree that blastocyst-stage embryos hold little promise for use in the chromosome transfer approach that we propose. However, as we suggest in the text, embryos frozen at the zygote or cleavage stages may be excellent candidates for chromosome transfer after thawing.

In addition, if as our experiments in Figure 5 suggest, aneuploid zygotes can be used as recipients for chromosome transfer, there may be tens of thousands of fresh human zygotes that would normally be discarded, and that could never give rise to a child , donated by couples interested in participating in stem cell research, available as recipients for chromosome transfer. For as long as clinicians continue to practice IVF, this source of human cytoplasts that would otherwise be discarded will remain available. We believe that all of these topics will be significant areas of discussion for the public after the publication of our work. In fact, some may find our approach of using aneuploid zygotes for chromosome transfer a particularly attractive one because these zygotes could not even from the moment of conception have given rise to a child.

Mouse development and epigenetics expert

This study represents a technical advance in animal cloning and the derivation of pluripotent ES cells. The authors found that it is possible to use zygotes for transplantation of genetic material from donor cells. The critical step in the procedure is to induce arrest of mitosis, to remove the recipient chromosomes, and then to introduce chromosomes from arrested donor cells into the zygote. The procedure is very clearly described, and the authors report derivation of cloned mice and ES cells from donor genetic material introduced into the enucleated zygotes.

This study shows why previous attempts to use zygotes failed, which was because zygotes were previously enucleated by the removal of pronuclei, and this would result in the depletion of 'reprogramming' factors from the recipient cell. The arrest of zygotes in mitosis as shown here will cause release of these factors back into the cytoplasm.

The frequency of development using this approach is comparable to the previous work using oocytes as recipients. The authors claim that this approach may be easily applicable to humans where a large number of zygotes (some karyotypically abnormal) are available and could be used to derive human ES cells from adult somatic donors. If so, this method will significantly advance efforts in this area. The new procedure used here could have an impact on the derivation of human ES cells from adult somatic nuclei. However, the suggestion that the procedure could work with similarly enucleated ES cells should have several caveats despite the results of fusion with somatic cells. With the zygote, the donor nucleus has the opportunity to undergo reprogramming over several cleavage divisions. The authors themselves suggest that early blastomeres may work just as zygotes, providing a prolonged period for reprogramming. The introduction of a somatic nucleus into an enucleated ES cell using a similar method to the one described here may not work if continuous presence of factors generated by the ES genome in hybrid cells over several divisions is required. Even under these circumstances, the frequency of reprogramming in hybrid cells is very low.

In conclusion, the study represents an important technical advance, which if it can be used successfully in humans, would increase the chances of deriving human ES cells from somatic cells. However, the mechanism of reprogramming itself still remains to be determined.

Author response

We agree with this interpretations of the current state of affairs and that further work is required to determine whether this new procedure can also be applied to mitotic blastomeres or mitotic ES cells. This exciting work is in progress and that we hope will be the subject of further publications.