Thinking About the Human Neuron Mouse
by Henry T. Greely, Mildred K. Cho, Linda F. Hogle, Debra M. Satz
2007. The American Journal of Bioethics 7(5):27

Dr. Irving Weissman, a professor in the departments of pathology and development biology at Stanford, approached one of us in February 2000 with ethical questions about interesting experiments he was considering. The most interesting experiment would begin with an inbred strain of mouse that attempted to form brains during very early fetal development, but, several days before birth, died as a result of the death of most or all of the developing neurons in their brains (the glial cells that make up about 90 percent of the brain are unharmed). Weissman proposed to transplant human brain stem cells into the fetal mice, just before their own neurons died. His hope was to produce a living mouse with a functioning brain made up of mouse glial cells and human-derived neurons. This mouse could then be used to study human neurons in vivo in a laboratory animal, similar to the way the SCID-hu mouse, which he had helped developed in the late 1980s, allowed the study of the human immune system inside laboratory mice.

Introduction

Tonight I ask you to pass legislation to prohibit the most egregious abuses of medical research  [which include] creating human-animal hybrids.

- George W. Bush, 2006 State of the Union Address (2006)

Dr. Irving Weissman, a professor in the departments of pathology and developmental biology at Stanford University (Stanford, CA), approached one of the authors of this article in February 2000 with ethical questions about interesting experiments he was considering. The most interesting experiment would begin with an inbred strain of mouse that begins to form brains during very early fetal development, but, several days before birth, died as a result of the death of most or all of the developing neurons in their brains (the glial cells that make up approximately 90% of the brain are unharmed). Weissman proposed to transplant human brain stem cells into the fetal mice, just before their own neurons died. His hope was to produce a living mouse with a functioning brain made up of mouse glial cells and human-derived neurons. This mouse could then be used to study human neurons in vivo in a laboratory animal, similar to the way the severe combined immunodeficiency (SCID)-hu mouse, which Weissman had helped developed in the late 1980s, allowed the study of the human immune system inside laboratory mice.

That conversation led to the creation of a five-person working group that, after meeting for more than one year, in early 2002 reported to Dr. Weissman that it believed his proposed experiments could be performed ethically, subject to some guidelines. The report has never been published and the experiments, for reasons not associated with the report, were never performed. Yet both the experiments and, to a lesser extent, the report have been subjects of discussion and debate (DeWitt 2002; Krieger 2005; Wade 2005; Weiss 2004), and the issue of human/non-human chimeras has only grown more controversial, leading even to proposed criminal legislation that has the "unambiguous" support of the President of the United States (S. 1373; Brownback 2005).

This article is a revised version of our report, updated to reflect nearly five years of debate about the ethical issues surrounding the creation and use of human/non-human chimeras. That debate has taken place in scholarly journals, important policy reports, and the halls of Congress. We believe our analysis has interest as one of the earliest efforts to come to grips with the implications of this scientific research and as an example of a "benchside consult," an effort to provide ethics-based advice on research in progress. More importantly, we also believe that it remains, with slight modifications, a useful approach to such experiments. Our report focuses on transplanting human neural progenitor cells into non-human brains and so falls well within whatever boundaries define "neuroethics," but it also has broader implications for the creation of other kinds of human/non-human chimeras, including some with non-biological components.

This article begins by describing the debate over human/non-human chimeras. It then focuses on our case study, Weissman's proposed experiments aimed at creating what we have called the "human neuron mouse." It provides some background on the experiments and discusses their potential benefits and their risks and costs before providing our recommendations to Dr. Weissman (and, now, others contemplating similar experiments). The article ends with some broader conclusions about the ethics of research with human/non-human chimeras.

Some readers will, no doubt, be disappointed that neither this article, nor the original report, attempts to answer the question whether conferring human-like mental characteristics on non-human animals is, or is not, ethically appropriate. We concluded that this fascinating question just was not plausibly raised by Weissman's proposed experiments. To emphasize that question in the context of these, or similar, experiments would give too much credence to a sensational misreading of this research; as we note in our last section, the question does need further work.

THE DEBATE OVER HUMAN/NON-HUMAN CHIMERAS

Although the definitions and meanings of chimeras are numerous and complex (Greely 2003), for the purposes of this article chimeras are creatures with cells, tissues or organs from individuals of two different species (interspecific chimeras). In spite of President Bush's language, hybrids are not chimeras but are, instead, the result of sexual reproduction involving individuals of different species, as a mule is a hybrid resulting from the mating of a male donkey with a female horse. Human/non-human chimeras can be created in two different directions, by putting human cells or tissues into non-human animals or by putting non-human cells or tissues into humans. This article discusses only the first; the second topic is more often referred to as xenotransplantation and is the subject of wide-ranging debates, mainly about its safety. (Interestingly, at least some experiments have transplanted non-human neural cells into human brains with long-term survival [Deacon 1997].) This section of the article reviews the scientific, ethical, and policy discussions that have taken place concerning the first method.

The Continuing Creation of Human/Non-Human Chimeras

The science and politics of human stem cells have combined to keep human/non-human chimeras a scientifically relevant issue. Weissman hoped to make human neuron mice largely so the mice could serve as model organisms for studying human cells. But as interest, scientific and popular, grows in human stem cell research, human/non-human chimeras are likely to take on broader uses. Before anyone makes new clinical use of human stem cells - or any clinical use of human embryonic stem cells - prudence (and the United States Food and Drug Administration [FDA]) are likely to require preclinical trials with the human cells in non-human animals. The result is likely to be a large number of human/non-human chimeras. When pluripotent embryonic stem cells are used instead of more differentiated stem cells, the concerns potentially become greater; a human embryonic stem cell, even if placed in the liver, might be able to become a neuron, a skin cell, or, ultimately, an egg or sperm.

Although Weissman has not performed the two experiments discussed in the report (he has continued some human/non-human chimera experiments), other researchers have continued to make human/non-human hybrids, in a wide variety of contexts, such as studying human tumor cells by transplanting them into mice. These chimeras receive little or no attention, but two researchers have received some publicity for work with chimeras, one involving neural cells, one with liver cells.

A group of Yale University (New Haven, CT) researchers led by Dr. Eugene Redmond have been experimenting with transplanting immature human neural cells into the dopamine-producing regions of the brains of green vervet monkeys. Those regions are associated with Parkinson's disease, and Edmond and his group hope that their research may ultimately be useful in understanding and treating humans with Parkinson's disease (Redmond 2002; Shreeve 2005).

Meanwhile, at the University of Nevada (Reno, NV), Dr. Esmail Zanjani has produced chimeras by transplanting human stem cells, mainly human blood-forming stem cells, into sheep. Zanjani has claimed that these human cells have been transformed into a variety of cell types in the sheep, in at least one case producing a sheep with a liver with 40% of the liver cells derived from human cells. According to Zanjani, these livers contained characteristically human structures and produced fully human proteins. Zanjani's work stirred up controversy with reports that the chimeric sheep had been given to a University-owned ranch that let the naïve research sheep out to graze as if they had been raised on the ranch, resulting in dead sheep and happy coyotes (Mullen 2005).

Bioethics

At the time of our report to Weissman, there was effectively no bioethics literature on human/non-human chimeras. That began to change in 2003 with the publication in the American Journal of Bioethics of a target article by Jason Scott Robert and Françoise Baylis (Robert and Baylis 2003). Robert and Baylis argued for caution in the creation of human/non-human chimeras, based on the possibility of creating confusion about the moral status of the resulting organism. Their article attracted many comments, of which those by Greely, Streiffer, Cohen, and Karpowicz were particularly interesting (Cohen 2003; Greely 2003; Karpowicz 2003; Streiffer 2003).

Phillip Karpowicz, Cynthia Cohen, and Derek van der Kooy published a useful article in 2005, following up in more detail on a 2004 article (Karpowicz et al. 2004; Karpowicz et al. 2005). They analyze four arguments against human/non-human chimeras: moral taboo, species integrity, unnaturalness, and human dignity. They find only the last argument convincing, but only if the human cells, when transplanted into the prenatal mouse or monkey, were to proliferate and develop into a whole human-like brain and if human-like capacities associated with human dignity were to emerge in such animals to some degree  (Karpowicz 2005, 123-124).

The article continues to make specific recommendations for limiting chimera research to avoid the risks of "developing whole human-like brains" or "human-like capacities."

Also in 2005, a working group organized by Ruth Faden and Guy McKhann at Johns Hopkins (Baltimore, MD) reported its conclusions concerning transplanting human neural stem cells into the brains of non-human primates (Greene et al. 2005). This working group concluded that such research should try to minimize the risk that the resulting animal would have more human-like cognitive capacities. It argued that such experiments be subjected to special review and that the reviewing bodies should consider six factors:

(i) the proportion of engrafted human cells, (ii) neural development, (iii) NHP [non-human primate] species, (iv) brain size, (v) site of integration, and (vi) brain pathology (Greene 2005, 386).

The Johns Hopkins working group recognized that there was "no simple relationship between these factors and, thus, no formula for making evaluative judgments." It did use the six factors to argue that grafting a large number of human cells into developing great apes would be more troubling than transplanting small numbers of human cells into the brains of healthy, adult monkeys most distant from humans.

Still later in 2005 Robert Streiffer took the position that the most important ethical considerations in such research should arise out of concern for the chimeras created by such research. He argued that such research might significantly enhance the moral status of those chimeras, which he sees as, on its face, good, but that the chimera will not be treated as its higher status demands. Recognizing the very substantial uncertainties about what qualities such research would create - and what ethical significance those qualities would have - he argues that, at this time, policies that require the early termination of such chimeras or that forbid the introduction of pluripotent stem cells into non-human primate blastocysts (the position taken by the National Academy of Sciences [NAS] Guidelines) are appropriate (Streiffer 2005).

Finally, in August 2006 a private organization called the Scottish Council for Human Bioethics published a report on "animal-human mixtures" (Scottish Council 2006). The report covered a wide range of ways in which human genes, cells, or reproductive processes might be mixed with those of non-humans. Two of this group's recommendations are particularly relevant:

1. The incorporation of human stem cells into post-natal animals should proceed with extreme caution. Moreover, such a procedure should only take place if it can be demonstrated that the cells cannot contribute to the germline or give rise to specifically human brain functions in the animals.

2. The incorporation of human stem cells into post-blastocyst stages of non-human embryos should only take place if it can be demonstrated that they cannot contribute to the germline or brain cells of the animal (Scottish Council 2006, 1).

The recommendation concerning embryos is more restrictive (cannot contribute to "brain cells" versus cannot contribute to "specifically human brain functions") because of greater concern about introducing cells at an earlier stage of development. The report makes these recommendations after reviewing many of the discussions of chimeras, which it says demonstrate the special sensitivity of possibly affecting a non-human's brain (or germ) cells.

Policy

The issue of human/non-human chimeras has led to at least two efforts to create guidelines or rules for this activity. The first comes from the National Research Council (NRC) and Institute of Medicine (IOM) in the United States. The second is legislation proposed in the United States. There may appear to be at least one more effort, as Canada prohibited the creation of "chimeras" in 2004 when it adopted its Assisted Human Reproduction Act. That act, however, defines chimeras only as human embryos into which cells from other human or non-human entities have been placed, and thus would not cover mice with human cells. (Assisted Human Reproduction Act 2004). (We have not tried to survey all legislation around the world in search of bans on chimeras; other such bans may exist.)

In April 2005 a committee created by the NRC and the IOM produced a report with guidelines for how to conduct human embryonic stem cell research (NRC 2005). The October 2004 meeting of this committee had included testimony from several scholars about the creation of chimeras, including Irving Weissman, David Garbers and Bridgid Hogan on the scientific issues and Henry Greely, Cynthia Cohen and William Hurlbut on the ethical issues (NRC 2005, Appendix C). These guidelines included the recommendation that an embryonic stem cell research oversight committee review and approve all research "involving the introduction of hES [human embryonic stem] cells into nonhuman animals at any stage of embryonic, fetal, or post-natal development." The guidelines further urged that "particular attention should be paid to the probable pattern and effects of differentiation and integration of the human cells into the nonhuman animal tissues." The guidelines also stated that no animals in which human embryonic stem cells had been introduced should be allowed to breed and no such cells be introduced into the blastocysts of non-human primates (NRC 2005). The text of the report addressed specifically the issue of putting human embryonic stem cells into the brains of non-human animals:

Perhaps no organ that could be exposed to hES cells raises more sensitive questions than the animal brain, whose biochemistry or architecture might be affected by the presence of human cells. Human diseases, such as Parkinson's disease, might be amenable to stem cell therapy and it is conceivable, although unlikely, that an animal's cognitive abilities could also be affected by such therapy . Protocols should be reviewed to ensure that they take into account those sorts of possibilities and that they include ethically sensitive plans to manage them if they arise (NRC 2005, 50).

The NAS Guidelines and discussion dealt only with human embryonic stem cells, but the issues they raise apply to all human stem cells that can give rise to central nervous system cells.

It is not clear how widely the NAS Guidelines are being followed by United States institutions performing human embryonic stem cell research. Those following stem cell research generally believe that the NAS Guidelines are widely used, although firm evidence is lacking. In California, the NAS recommendations on embryonic stem cell research oversight committee review of chimeras have been largely adopted both as regulations by the California Institute for Regenerative Medicine, which manages the stem cell research funding provided by California's Proposition 71, and by the California Department of Health Services Advisory Committee on Human Stem Cell Research, which is charged with recommending guidelines for human embryonic stem cell research in California not funded by California Institute for Regenerative Medicine.

In early 2005, The New Atlantis, a conservative journal that calls itself "an effort to clarify the nation's moral and political understanding of all areas of technology," published an editorial entitled "The Bioethics Agenda and the Bush Second Term," in which it called for an aggressive legislative campaign both to ban the creation and destruction of human embryos for research purposes and to "defend and advance the dignity of human procreation and the human family" (Cohen 2005). In March 2005 Senator Sam Brownback (R-KS) proposed legislation that would largely implement the recommendations made by the editorial in pursuant of its second goal - and that would ban the creation of some forms of human/non-human chimeras.

The Human Chimera Prohibition Act of 2005, originally introduced in March 2005 and reintroduced that July as S. 1373, would make it a felony, punishable by 10 years in prison and a civil fine of at least $1 million, to create, attempt to create, transport, or receive for any purpose a "human chimera" (S. 1373, § 302(2)). The Act defines "human chimera" in eight different ways, most of which appear to deal with hybrids of various sorts rather than chimeras (S. 1373, § 301(1)). The eighth subsection appears to have been aimed directly at Dr. Weissman's proposed experiment:

(H) a non-human life form engineered such that it contains a human brain or a brain derived wholly or predominantly from human neural tissues (S. 1373, § 301(1)(H)).

(The Senator may have missed his target, as Weissman's most ambitious experiment, even if successful, would not affect the 90% of the mouse brain that was made up of glial cells instead of neurons.) It is this legislation that President Bush was endorsing in his 2006 State of the Union address, legislation that has not been voted on in the Senate or the House of Representatives, or received committee hearings.

BACKGROUND TO THE WORKING GROUP REPORT

When the original report was written for Dr. Weissman, researchers had recently reported finding human brain stem cells - cells that can become all or most of the cell types found in the human brain, including neurons and glial cells. These cells, thought to be very infrequent in adults, were isolated from the brains of human fetuses. This discovery opened the possibility of studying human neurons in vivo in laboratory animals by creating mice whose brains were made up, in part or in whole, of human neurons.

Although such a "human neuron mouse" would not stand and talk like a cartoon character, its possible creation raises important and interesting ethical questions about research in human neuroscience. The next section lays the groundwork for evaluating these issues by discussing human brain stem cells, examining completed and planned experiments involving transplantation of these cells into mice, and finally by describing our working group and its general approach to the questions before it.

Human Brain Stem Cells

In 2000, researchers claimed to have isolated human brain stem cells from the brains of fetuses aborted after 12 weeks of development (Uchida 2000). Research with these cells in vitro showed that they could form many different kinds of human brain cells - not just neurons in their various types but also other cells that play essential roles in the brain, such as glial cells. They seem, therefore, to be multi-potent cells. The isolation of these cells opened the possibility of growing and transplanting mature brain cells, particularly neurons, into patients with such debilitating neural degenerative disorders as Parkinson's disease. In 2006, the FDA granted an investigative new drug exemption to one firm to perform such transplants for a rare childhood disease (Batten's disease); an institutional research board at the Oregon State Health University (Portland, OR) recently approved a trial of this approach (StemCells, Inc., Palo Alto, CA). Whether this kind of neural regenerative medicine will prove safe or effective remains, of course, unknown. Stem cell therapy with hematopoietic stem cells is regularly used, with frequent success, to build or rebuild a patient's blood and immune systems; it remains of speculative value in other contexts.

Dr. Irving Weissman at Stanford Medical School (Stanford, CA) was one of the researchers who helped isolate these brain stem cells. Weissman had long worked with stem cells and had been instrumental in the isolation of human hematopoietic stem cells. Working with those cells and other human tissues, in 1988 he and Dr. Joseph M. McCune created the so-called "SCID-hu" mouse. This work started with an inbred strain of mouse born with severe combined immune deficiency (SCID). These mice, as a result, had severely impaired immune systems. Weissman and McCune transplanted human hematopoietic cells (in later experiments, human hematopoietic stem cells) as well as the tissues that support for the formation of blood and cells of the immune system (human fetal bone, thymus and liver) into these SCID mice. The weak immune systems of the mice did not attack the human cells as alien and those cells were able to colonize the human fetal bone and liver, and later thymus, to create in them a human blood-forming and immune system. The result was a laboratory animal model of the human blood-forming and immune systems, on which experiments could be done that could not ethically be done with the only other creatures with an in vivo human immune system, living humans.

Using these mice the human hematopoietic stem cell was first isolated and gained FDA approval for trials that showed these cancer-free stem cells could regenerate the blood-forming and immune systems that had been depleted by cancer therapies. These animals were also used to infect a human immune system with patient isolates of HIV, the first time one could show definitively that HIV caused the changes that characterize AIDS in humans.

The Mouse Transplant Experiments

As part of the research leading to the isolation of human brain stem cells, Weissman, Uchida and other colleagues at the firm StemCells Inc. began transplanting human brain stem cells into the brains of SCID mice with normal murine brains. (SCID mice were again used to avoid an immune system attack on the human cells.) The human brain stem cells were placed in a brain structure called the lateral ventricle, which, in mice, connects to their brains' quite large olfactory bulbs. Weissman's group was able to show that the human neuronal stem cells engrafted in a brain stem cell niche called the subventricular zone, near the injections. Those cells also migrated to a second niche, the dentate gyrus of the hippocampus. In these niches the human cells divided and many of them migrated toward the olfactory bulb (Tamaki et al. 2002; Uchida et al. 2000). Samples of the brains of these mice showed that the human neurons had survived and had connected to the mouse brain. Mouse brains have a (relatively) much larger olfactory bulb than human brains and new mouse neurons are regularly migrating to their olfactory bulbs; the human-derived cells did the same. Examination also showed that the human neurons had moved into other areas of the murine brains and made up more than 1% of the neurons in some regions. This research could not, however, determine whether the human neurons were actually functioning as part of the mouse brain, let alone whether they were functioning normally.

Weissman then wanted to do two other experiments transplanting human brain stem cells into mice. These proposed experiments were the subjects of our report.

The first proposed experiment would work with an inbred strain of mouse that lost all the neurons in its cerebellum several weeks after birth. In both mice and men, the cerebellum is involved in fine motor skills, coordination and balance. Other roles of the cerebellum, including any role in consciousness, are unknown. A mouse, or a person, without a functioning cerebellum can survive but with substantial motor deficits. Friedrich's ataxia is a human disease caused by the death of cerebellar cells; this strain of mice displays symptoms similar to those of humans with Friedrich's ataxia. Weissman wanted to transplant human brain stem cells into such a mouse just before its cerebellar neurons began to die. He hoped that the human cells would differentiate into neurons, would migrate into the mouse's cerebellum, and would begin to function. Unlike his earlier experiment with putting brain stem cells into the lateral ventricles of mice, this experiment would be able to determine whether the human neurons were not only surviving but were functioning, in part through seeing whether and to what extent the treated mouse showed signs of normal cerebellum activity. Based on Weissman's previous experiment, he also expected that the human cells would appear at low concentrations in other parts of the mouse's brain.

The second proposed experiment would use a different inbred strain of mice, developed by Fred Alt's laboratory. These mice form the beginnings of the nervous system, creating the structures that support the movement of early stage neural progenitors, but all these developing neurons die, leading to the death of the early stage fetuses. This mouse strain also has a mutation that makes them SCID mice, so that the resulting mice would not reject human cells. For this experiment Weissman planned to transplant the human brain stem cells in utero shortly before the murine neurons were expected to begin dying. He hoped that the human cells would differentiate, migrate to the places where the murine neurons are dying, and take their places. The result would be a mouse brain, the neurons of which were mainly human in origin. This experiment could have at least two different end points. In one version, the mice could be aborted as fetuses shortly before birth and have their brains examined on autopsy to see whether the human neurons had populated their brains and, if so, what kinds of brain structures - mouse, human or mixed - they formed. Alternatively, the mice could be allowed to go to term and, in addition to examination of their brains, by neuroimaging while alive or by autopsy, their functioning and behavior could be observed for variations from the mouse norm. If the mice were viable, they might be the neuronal equivalent to the SCID-hu mouse in terms of being a laboratory animal that could be used for experiments on living, in vivo, human neurons.

Although, as subsequently described, the working group's report concluded that these experiments could be undertaken ethically, at this writing neither experiment has been performed. Weissman discovered that the Friedrich's ataxia model mouse did not, in fact, have the characteristics he needed for his experiment. One study, however, did follow up some of the questions involved in the cerebellar experiment. Dr. Fred Gage's laboratory reported in late 2005 that they had transplanted stem cells into the brains of rats and had been able, using patch clamping, to determine that the cells derived from the transplanted human cells were actually firing. The fact that these human-derived neurons were firing does not necessarily mean that those cells were functioning properly, either individually or as part of a larger unit, in the rat. But if it had been the case that the human-derived neurons were never firing, they clearly could not be functioning normally (Muotri et al. 2005).

As to the second experiment, there were problems with breeding the mouse strain with complete neuronal death. Weissman has also been occupied with other work, not only with other aspects of his own research but with administrative and advocacy work around human stem cell research. He also needed to find a graduate student or post-doctoral fellow interested in doing the work; the fellow who was interested at the time had gone on to other work. Weissman continues to say that he might try the second experiment, but he also from time to time refers to it as a "thought experiment." It is not clear to us, and perhaps not to him, whether or not he will return to this experiment.

The Working Group and Its Approach

Weissman was aware of the sensitivity of these planned experiments, both ethically and in terms of public reaction. He may well have had visions of a headline reading "Stanford Scientist Creates Mouse with Human Brain." As a result he asked one of the authors of this article (Greely) to consider putting together a group to examine the ethical issues in these proposed experiments. Greely pulled together this ad hoc group, with representation from several disciplines. We met several times during 2000 and 2001, interviewed Weissman, studied the scientific literature, and discussed the questions - and how we could approach those questions - at length. We concluded that the experiments did raise interesting and important, but manageable, ethical issues

In general, we approached the questions by asking about the potential benefits and the potential costs or risks of the proposed experiments. We first examined the costs to see if any of them might categorically rule out the experiments. We next considered ways in which the experiments might be undertaken to limit the costs of risks involved. We weighed the potential benefits of the research, with or without modifying conditions, against the potential costs or risks. We concluded that the experiments could proceed ethically, subject to careful staging and monitoring.

POTENTIAL BENEFITS

United States government regulations and international agreements on ethical research agree that research on human beings is only permissible if there are potential benefits, to applied or to basic science, from the research that outweigh the potential harms and risks. A similar, though weaker, standard applies in federal law to the use of many laboratory animals, including mice. Researchers obviously can do things to laboratory mice that they may not do to humans, including routinely maiming or killing them. They may not, however, do such things without a good reason. Both because living animals were to be used and because of the nature of the human cells being used, Weissman's proposed experiments could be justified only if the experiments were likely to offer some benefits.

The most clear potential benefit is the creation of a non-human animal in which human neurons can be studied in a living brain. Many experiments on human neurons, and on the diseases of those neurons, cannot ethically be performed in humans. These experiments involve risks too high to be permissible for a human subject to bear or, in many cases, the killing of the human subject and the subsequent examination of his or her brain. Such research with human subjects is, of course, not morally acceptable.

This benefit, in effect, would come from the creation of a brain equivalent to the SCID-hu mouse. Thousands of SCID-hu mice have been used in research on the human immune system, particularly but not solely with respect to HIV infection. More than 100 grants from the National Institutes of Health (NIH) have involved the use of SCID-hu mice and, over the years, the NIH has contracted for the production of more than 1,200 SCID-hu mice.

Having a laboratory animal for studying human neurons might have substantial benefits, both for basic science and for clinical applications. For example, the methods by which various pathogens or exposures damage human neurons could be directly studied in a living brain without risking harm to a human subject. New drugs or other treatments could be first tested for their effects on human neurons in mice rather than in human subjects. Steps in the in vivo functioning of human neurons could be analyzed without risking harm to living people.

None of these benefits is assured. These experiments may fail, or, whether they fail or succeed, a human neuron mouse may prove impossible to create. Given the vast and thus far poorly understood number and type of interactions between cells that take place in the brain, we would be surprised if human neurons could function properly in all the roles necessary to create a properly working mouse brain. Even if a human neuron mouse proved possible, research with it might not be substantially better than existing alternatives. Studies of human neurons outside the brain through in vitro research or in vivo studies of mouse neurons in mouse brains might prove just as illuminating of human brain function as the study of human neurons in a mouse brain. Nonetheless, the potential for substantial scientific and even medical benefits seemed significant to us. Because of these anticipated benefits, the experiments seemed reasonable and, in the case of the experiment that could create a murine brain composed entirely of human cells, necessary steps to assess that potential.

RISKS AND COSTS

We identified five areas of concern that need to be examined and, if found significant, weighed against the potential benefits. These concerns include: 1) the sources of the human brain stem cells; 2) the potential for pain and suffering to the mice; 3) the propriety of this use of human tissues (particularly brain tissues); 4) the risks of possibly conferring some degree of humanity on another species; and 5) the risks to public support of science.

It is interesting, in retrospect, to compare those concerns with those subsequently expressed in the literature on human/non-human chimeras. Most of the issues that concerned us have been largely or entirely ignored in subsequent discussions. In one form or another, the question of "conferring humanity" has been the focus of the discussion, although generally expressed in terms of either human dignity (Karpowicz 2003) or avoiding moral confusion (Robert 2003). Streiffer's position is more complicated; he argues that the successful conferring of a higher moral status on a human-mouse chimera would not be wrong in itself, but would likely be wrong because we would not treat the chimera in a way consistent with that higher status (Streiffer 2005). A little has been said on the other issues. The Johns Hopkins group on transplanting human neural tissue to non-human primates did discuss briefly the issue of harm to the subject animals (Greene 2005); Karpowicz did discuss and reject at least one form of the public relations argument (Karpowicz 2005).

We did not discuss in our report some of the moral taboo arguments rejected by Karpowicz, or integrity of species borders and unnaturalness arguments, rejected by both Karpowicz and the Johns Hopkins group. Our internal discussions had already considered and rejected all of those arguments and our report described only arguments we found potentially plausible.

Aborted Fetuses as the Source of the Human Brain Stem Cells

The human brain stem cells that Weissman uses were derived from the brains of human fetuses that had been intentionally aborted. Use of such tissue has been controversial in the United States because of its link to voluntary abortion. The issue of using human fetal tissue in research and medicine was discussed widely in the late 1980s, spurred in part by evidence that transplants of fetal brain tissue into the brains of people with Parkinson's disease could lead to improvement in their condition. (As it happens, this therapeutic application of human fetal tissue has since been shown, at least so far, to be neither safe nor effective.) For research and medical purposes, tissues from intentionally aborted fetuses were greatly preferred to tissues from spontaneous abortions or stillbirths because of the much greater risk that the cells and tissues from the latter had suffered from fatal genetic conditions, had been contaminated by pathogens, or had died in the long period between the in utero death of the fetus and the collection of the tissues ...