Select Committee on Science and Technology Written Evidence

Memorandum 15

Submission from Dr Robin Lovell-Badge, Head, Division of Developmental Genetics, MRC National Institute for Medical Research


  1.  Chimeras and hybrids are distinct types of entity (although they are often confused in the media). A chimera is composed of a mixture of cells derived from two individuals, which can be from the same or different species, whereas each cell of a hybrid carries genetic material derived from two individuals, comprising different strains, subspecies or species. Neither represent new ideas: genetic hybrids have been vital to plant and animal agriculture for millennia, while if grafts are classified as a type of chimera, they are widely used in plant agriculture and in modern medicine. For example, bone marrow transplant patients are chimeras. It should also be noted that human-human chimeras occur naturally on rare occasions when two early embryos mix together, and that most mothers are chimeras as they carry a few cells derived from their offspring when this was a fetus.

  2.  Both chimeras and hybrids have been and continue to be very important for many areas of basic, translational and applied research. However, the issues facing this Inquiry are largely restricted to chimeras involving human and animal cells or hybrids involving combinations of human and animal genetic material where the entity produced passes through stages of development typical of an early embryo and could be considered to be at least in part a human embryo. However, there is a wide range of possibilities where this could happen, some already used in research. Except for some obvious cases that are already illegal, it will be difficult to draw a fixed line between desirable and permissible research and that which should be forbidden, especially when public attitudes to this type of research may change dramatically.

  3.  In the following I have tried to put the issues in context, relating them to current research as well as what may or may not be feasible in the medium term. It is important to include some details as these reveal where the problems of regulation will be encountered.


4(i)   Grafts of human cells into postnatal animals

  These are used to explore the ability of human cells to differentiate and contribute to complex tissues. This includes using human ES (hES) cells to derive teratocarcinomas in mice, which is a standard assay for pluripotency, and the transfer of human neural stem cells into the postnatal brain of animals (mice to monkeys) as a test to see if they give rise to neurons and glia, and that these make functional connections. Grafts of this sort are also used to create "humanised" animal models, for example replacing the mouse immune system with the human immune system or mouse liver with human liver cells. The former is used to explore responses of the human immune system to pathogens, such as HIV, or to grafts, etc; the latter to test how human liver cells metabolize drugs or toxins, or infections with liver specific viruses. This type of animal model is restricted, however, to tissue types that can undergo significant replacement or regeneration after birth. Grafts of human cells to animals are also used as a test of tumourigenicity. It is likely that the latter will be a regulatory requirement for many stem cell treatments, whether using embryonic or adult stem cells or their derivatives.

5(ii)   Grafts of animal cells into postnatal humans

  These are usually referred to as xenotransplants. Concerns seem to focus more on transmission of animal viruses to humans than on the notion of making humans part animal.

6(iii)   Grafts of human cells into animal foetuses

  Again, as in (i) above, this can be used to explore the ability of human cells to differentiate and contribute to complex tissues or to create humanised animal models, but where it is critical to introduce the human cells during development. Such chimeras/grafts are also used to test developmental potential of human stem cells and questions of plasticity (one cell type changing into another). As methods of in vivo imaging, eg by ultrasound, improve, it is becoming possible to introduce cells or small pieces of tissue at earlier and earlier stages of embryo development. Perhaps, after giving the human cells some competitive advantage (see below), it would be possible to use these methods to increase the range of humanized animal models available for research, by substituting part of the resulting embryo with the equivalent human progenitor tissue.

7(iv)   Embryo aggregation or blastocyst injection chimaeras

  These are used frequently in research on mouse development and genetics, but it is also the ultimate test of pluripotency of an early embryo cell, such as an embryonic stem (ES) cell, as the latter should be able to contribute to all tissues of the resulting animal. One published attempt (in the USA) to introduce human ES (hES) cells into mouse blastocysts gave very poor contribution of human cells to early mouse embryos [James et al (2006) Dev Biol. 295, 90-102]. This result was fairly predictable as human embryonic cells divide much slower than those of the mouse, so they will be rapidly competed out by the latter. To have any significant contribution it would probably be necessary to have a strong selection for the human cells and against the mouse cells of a particular tissue. However, as embryonic development depends on coordinated growth of many tissue types, it will be a significant challenge to derive viable embryos with a significant human contribution. If such chimaeras are made between human and non-human primate embryos or ES cells, then the likelihood that these would give viable embryos is much greater. Experimentally produced human-human chimaeras would almost certainly develop normally. If any of these combinations are to be used merely to assay the developmental potential of hES cells, then some information could be obtained under the current 14 day rule, i.e. any such chimaera would be allowed to develop in culture for no more than 14 days before being destroyed. It would not be possible to say that the cells were competent to differentiate very far, even into the early embryonic cells typical of ectoderm, mesoderm and endoderm as these only form during gastrulation, let alone into any mature cell type. However, by taking individual cells or parts of these early embryos after their destruction, and culturing them or using them to make further chimeras of the type discussed above in (iv) might be informative.

  8.  A common concern with all these types of chimeras is whether they would lead to conferring human higher order brain function on an animal. It should be remembered that the size and complexity of our brain is probably an important determinant of our humanness, and this may not be achievable if human cells are developing in an animal, especially if this is much smaller and/or only a distantly related species. However, where should the line be drawn? Would it matter if parts of the peripheral nervous system or spinal cord of an animal were substituted with human tissue? Another potential concern is the production of human gametes in an animal, which could in theory be achieved by any of the methods above, including just by grafting pieces of human testis or ovary under the skin of an appropriate animal host. It would seem simpler and more appropriate to store gametes frozen than to use such methods, for example to allow an individual undergoing treatment for cancer to subsequently have children. But there may be valid experimental reasons for studying human germ cell development in animals if this would be the only way to better understand or control fertility.

  9.  In general it may be difficult to draw a clear line between what is acceptable and what is not, especially given shifting circumstances, or experimental details. Perhaps one clear distinction that could be made is if the experiments were to involve animal-human chimeras that could theoretically implant into a woman's uterus, ie if the outer layer of a chimaeric blastocyst contained human trophectdoderm cells, then it is reasonable to prohibit the transfer being done—indeed this is already illegal. However, I see no reason why such chimeras should not be studied in vitro, with the 14 day limit, before they are destroyed—as long as destruction would include disaggregation to allow further study of component cells, including deriving cell lines.


10(i)   Full genetic hybrids

  These involve the mating of two variants within a species or between species. The latter will usually be possible only between closely related species, and even then the hybrids are likely to be sterile. Such hybrids have a normal diploid set of chromosomes, one from each parent. Mitochondrial DNA is usually only inherited from the oocyte, as there is a special mechanism that eliminates mitochondria brought into the early embryo by sperm. However, in cross-species hybrids this mechanism does not always operate, so their mitochondrial type may not just represent that from the maternal species [Kaneda et al (1995) Proc Natl Acad Sci U S A. 92, 4542-6].

11(ii)   "Somatic cell hybrids"

  This is a term that usually reflects the fusion in vitro of two unrelated cell types (with respect to species and/or tissue of origin). However, "somatic" may be misleading as one (or both) of the cells could be an embryonic stem cell or a germ cell.

  12.  The resulting cells may be tetraploid, although with fusions between mouse and human cells, human chromosomes tend to be lost. It is also possible to use techniques such that one of the fusion partners contains only one or even part of a chromosome. Such hybrids have been used extensively to map genes onto specific chromosomes, indeed this work was essential preliminary work in the human genome project. They are also used to study questions of "dominance", and aspect of reprogramming concerned with asking which parental cellular phenotype does the hybrid possess and which "master regulatory" genes are responsible? They are also important to study the activity or function of genes or chromosomes from one species in another—for example to ask if the mechanism that leads to one of the two X chromosomes being inactivated in female cells is evolutionary conserved by transferring a human X chromosome into mouse ES cells. Somatic cell hybrids are also of very important practical use as they are the way most monoclonal antibodies are produced.

  13.  While such hybrids are mostly studied in tissue culture, mouse ES cells can be used as one of the fusion products. Chimaeric mice made using these ES cells can then be used to transmit the chromosome to offspring, to establish a strain of mice that carries it. This was the technique used to derive the "Downs Mouse" strain, which carries a copy of human chromosome 21 [O'Doherty et al (2005) Science 309, 2033-7]. Such animals are indeed mice, but show a number of characteristics typical of Down's syndrome and are proving to be a very valuable animal model of this syndrome.

  14.  In somatic cell nuclear transfer (SCNT) or "cloning" technology, the recipient enucleated unfertilized egg is known to contain factors that can reprogramme a somatic cell nucleus back to the zygotic (or one-cell) stage of embryonic development. Another type of animal-human hybrid that has already been used to give valuable information about this reprogramming process, involved transferring several human somatic cells into Xenopus frog eggs [Byrne et al (2003) Curr Biol 13, 1206-13]. These were not allowed to develop, and probably would not have done so normally as they would be polyploid (i.e. contain many more than the normal diploid set of chromosomes). However, the only way to ascertain if the reprogramming process had been complete would have been to allow embryo development to proceed. This would not need to involve human somatic cells if the purpose was to explore mechanisms in general, but it would if the goal was to ask if human somatic cells responded differently in any way compared to animal cells.

15(iii)   Transgenic animals possessing one or more human genes

  These are very common in research and are used to study gene activity, structure and function, to make animal models of human diseases, or to produce valuable human proteins. An example of the latter is the recently publicised work leading to human proteins being made in chicken eggs. None of these are really hybrids, but they do contain a mix of human and animal DNA.

16(iv)   Combinations involving nuclear DNA of one species with cytoplasm, including mitochondrial DNA, of another or of both species

  Many such experiments have been conducted with tissue culture cells in vitro. These have traditionally been referred to as "cybrids" and were used to ask if reprogramming factors or cell type determinants are located within the cytoplasm or the nucleus, or to explore aspects of the biology of mitochondria, (see below).

17(v)   Use of enucleated animal eggs in nuclear transfer experiments with human somatic cells

  These are a form of cybrid as above and have similar uses, except they are also a potential way to derive patient-specific human ES cell lines. Such cell lines offer a way to study genetic diseases in the lab, especially those that are difficult to study using material derived directly from the patient, eg tissues that can't be accessed, or where the affected cell type has already undergone pathological changes, or is lost altogether. They also offer a way to explore the effect of genetic background and environment on disease progression and a way to screen potential therapeutic molecules. If they are derived from a mixture of animal and human material then it is unlikely that they would be considered suitable as a source of cells for cell-based therapies (after correction of the genetic defect). However, they could be used to explore such therapies using animal models of the disease as hosts.

  18.  Such hybrids can also be used to explore methods to obtain efficient reprogramming of the somatic cell nucleus and to explore the mechanisms by which this happens.

  19.  These aims could be explored using human rather than animal oocytes, and this is the ultimate aim and perhaps a necessity for any cell-based therapies to be put into practice. Indeed, the HFEA has already issued licences permitting this. However, from all the work carried out to date, there has been only minimal success and it is clear that the methods are very inefficient. This may in part arise from the poor availability of good quality human ooctyes for research. It could also reflect an aspect of human oocyte biology or early human development that we do not understand. (It is also possible for the human origins of the somatic cell to be at fault in these SCNT experiments.) Certainly while it remains an inefficient process, in my view it would be unethical not to explore the use of animal eggs for these studies, rather than wasting valuable human oocytes that are also in demand for fertility treatments and research.

  20.  Several different animal species have been proposed as sources of oocytes, in part because of the ease of obtaining them, but also because they share some aspect of early development in common with humans. For example, rabbit embryos show a similar time course of development in cleavage stages to human embryos, including the time that the embryonic (zygotic) genome is activated—when genes first become expressed and development is not just running on maternal products (RNA and protein) laid down in the oocyte. However, it is not known which animal species will be the best, or whether this matters at all. Xenopus eggs and early development is clearly very different from that of mammals, yet it is clear that at least aspects of reprogramming are conserved. Some studies will need to focus on which species is best to use in these animal-human hybrids.

  21.  Such animal-human hybrids made by SCNT will result in entities that have human nuclear DNA, but a mixture of animal and human mitochondrial DNA—the latter because a whole somatic cell, including its cytoplasm, is fused into the enucleated animal egg. Whether they should be classified as early human embryos, and therefore regulated by the HFEA has been one concern. No human gametes are involved. They will also begin with mostly animal proteins and RNA molecules, which will only be replaced as development proceeds and the human DNA begins to express genes appropriate for this. Indeed, some animal proteins are likely to persist throughout the stages of preimplantation development up at least to the blastocyst (several such proteins are known from studies of mouse development, eg Avilion et al (2003) Genes Dev 17, 126-40). So it can be argued that these entities only gradually become human embryos. They do possess a human genome, which may have the potential to direct human development, but it is not clear that this potential could ever be realised for biological reasons, and in my opinion it would be very foolish to try to find out (see below). Given that these hybrids are a purely experimental tool, designed to study very important questions in vitro, as preimplantation embryos and then as cell lines, and where it would already be illegal to implant them, it was my view on scientific grounds that they do not need to fall under the HFEA's remit. However, it a valid argument to say that public fears are more likely to be allayed if they fall under a regulatory body for control, and as the entities are mostly human, the HFEA is appropriate. But this does raise the possibility that if they had never been called human, and were just considered to fall amongst all the other types of experimental animal-human hybrids that are used for research, there would be no major issue to debate as to their creation.

  22.  Clearly a practical, but also a scientific, concern is what happens to both the animal and human mitochondria and their DNA in such animal-human hybrids. This is a complex issue where predictions are difficult to make and more experimentation is needed. Some of this work can be done using animal-animal combinations, or by fusing animal cytoplasm with hES cells. However, the specifics of using enucleated animal eggs and human somatic cells will need to be addressed, not just for any long-term use, but to know whether the approach is worth pursuing at all for practical reasons. Published data, notably that using rabbit eggs and human somatic cells, which reported the generation of several ES cell lines, already clearly indicates that the approach is worth following. This is in itself a justification for initiating the research as soon as possible in the UK. And this needs to be done on several fronts, for example, the outcome may depend on the animal species used, whether methods are employed to deplete or increase the number of either animal or human mitochondria prior to fusion, etc.

  23.  Research on such hybrids will also give valuable information on the biology of mitochondria. For example: (i) The interaction between products encoded by nuclear and mitochondrial genomes and the consequences of having components from two species. (ii) The importance of mitochondria to specific cell types including their contribution to the generation of "energy" (ie ATP). (iii) Mitochondria also have other functions within the cell, notably for the synthesis of steroids and in apoptosis (programmed cell death). The components for these processes are encoded by genes within the nucleus, so the relevant animal proteins should eventually be replaced by human ones, however, we do not know over what timescale, nor the consequences of any mismatch, etc. (iv) How mitochondrial genome division is controlled by genes within the nucleus, and whether the proteins involved will work across species. (v) How mitochondrial numbers per cell type are regulated, how they segregate during cell division, how mutations arise, DNA repair mechanisms, etc, etc. These studies will also be important for understanding mitochondrial diseases in humans and to exploring therapies.

  24.  As mentioned already, not all of these studies would necessarily require the generation hybrids in this way, but this knowledge is pertinent to how such hybrids will develop and to the properties and usefulness of any cell type they give rise to. In many ways the use of animal eggs and human somatic cell nuclei is special, so it is also possible that these experiments will necessitate using such hybrids to begin with. At the moment, the most important questions to be answered seem to be: what happens in cases where two or more mitochondrial types (in this case from two species) are present within the same cell (a condition referred to as "heteroplasmy") of an early embryo and does this have different consequences for different cell types? For example, in the preimplantation embryo and probably in ES cells derived from them, there is little requirement for energy (ATP) generated by mitochondria. Such a requirement probably only comes with the differentiation of the ES cells into specific cell types. One prediction is, therefore, that the early embryo and ES cells will contain both animal and human mitochondria as there is no selective pressure for one of them to be lost. Indeed, this was the result reported for the hybrids made from enucleated rabbit eggs and human skin cells [Chen et al (2003) Cell Res. 13, 251-63]. On differentiation to a cell type requiring high levels of energy, it is then probable that only cells that have lost the animal mitochondrial genome will survive, giving rise to cells that are entirely human. If true, this would have another practical consequence, in that the original hybrid embryos could probably never develop postimplantation, as there would be too much cell loss to maintain integrity of the embryo (assuming anyone was foolish enough to try to implant them).

  25.  It is this consideration of loss of the animal mitochondria, due to bottle necks, incompatibility between nuclear and mitochondrial encoded products, and a specific requirement for efficient mitochondria in some cell types, that suggests that there may be a gradual or step-wise predominance of human mitochondria as development of such hybrids proceeds. Such cells should be physiologically normal, and entirely appropriate for pursuing research on genetic diseases and their treatment.


  26.  There are many reasons to pursue research using animal-human hybrids and chimeras. Much of this research has been underway in the UK for many years. It is accepted to be beneficial, not least under regulations concerned with experiments on animals. So far it has not fallen under the HFEA remit, even if it involves similar types of mix between animal and human that are now being considered. While accepting that some of the proposals before the HFEA do verge into their territory, great care must be taken in deciding how the research should be regulated and where to draw the line on experiments that should not be permitted. A great deal of harm could be done to UK research, and critically to the aims of that research, which is motivated by the desire to alleviate suffering, if mistakes are made.

  27.  Where the research proposal is scientifically justifiable, then I would argue that there is an ethical imperative to allow it to proceed. Of course, it is always necessary to consider alternatives, especially when the experiments involve animals or human embryos. But good alternatives do not exist for the types of research on genetic disease that form the focus of the current proposals before the HFEA. Moreover, I hope I have illustrated the wide range of experimental approaches involving hybrids and chimeras that could generate improved animal and in vitro cell culture models of human diseases. These could be very beneficial in the search for therapies for a wide range of conditions. I do not wish to raise false hopes by saying this as research can never guarantee cures, but without embarking on the research we will never know.

  28.  My view is that it is far better to control such research activities under a good regulatory system through careful consideration of proposed experiments by scientific and ethical review panels, than it is by prohibitive laws that are likely to be both too restrictive and leave dangerous loopholes, especially in this rapidly advancing field of science.

January 2007

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2007
Prepared 5 April 2007