The Prospects for Human Germline Engineering

How far are willing to go in reshaping the human form and psyche?

Science and medicine have moved from elucidating our genes to manipulating them. Human gene therapy - science fiction a mere decade ago - now boasts more than 500 approved human studies and a U.S. National Institute of Health budget of some $200 million a year. The ability to make genetic changes to our germinal cells will represent a major advance in such therapy, because changes to the first cell of the human embryo are copied into every cell of the body and can thus reach any tissue.

"Germline” therapy embodies the most profound possibilities and challenges of molecular genetics, because it promises (some would say threatens) eventually to transform our very beings as ever more significant genetic changes are introduced into our genomes. This technology will force us to re-examine even the very notion of what it means to be human, for as we become subject to the same process of conscious design that has so dramatically altered the world around us, we will be unable to avoid looking anew at what distinguishes us from other life, at how our genetics shapes us, at how much we are willing to intervene in life's flow from parent to child.

Not surprisingly, the media focuses on menacing distant possibilities of human germline engineering rather than mundane therapeutic ones that may develop in the immediate years ahead. This is hardly surprising; the technology reverberates with haunting echoes of eugenic cleansing and disturbing sci-fi images of superhumans. By raising the possibility of meaningful human design, germline engineering uniquely captures the challenge of our coming era; other potent developments may immerse us in a radically different external world, but they still will likely leave the essentials of our biology unchanged.

Though germline intervention may not be clinically feasible for several decades, there is little doubt its potential is immense. One day it may protect children from cancer, AIDS and other diseases, enhance their intelligence and even extend their life spans. But the technology also embodies a fundamental challenge: how far are willing to go in reshaping the human form and psyche? This question lies at the heart of the emerging international debate about the application of molecular genetics to humans. And as the evening news increasingly features breakthroughs from work on the Human Genome Project, in-vitro fertilization, animal cloning, and artificial chromosomes, human germline engineering will become an ever greater focus of discussion, soul-searching, and legislation. Through this technology, we will seize control of our own evolution, yet we have still barely begun to grapple with the consequences.

Until recently, germline engineering hadn't been discussed much since the early 90's, when it seemed distant, something theoretical and safe to talk about because it concerned our children or grandchildren but not us. But molecular biology has progressed so rapidly since then, that rudimentary germline manipulation in humans is already nearly possible, even if not yet with the safety and reliability we demand for human medical intervention. It is time for society as a whole to begin to look at the possibilities and implications of this technology. The ghosts associated with these concepts – the ardent pseudoscience of eugenic efforts to "improve” our species and Hitler's brutal attempt to create a master race – are so close that it is difficult to ignore them. But this powerful technology is so nearly upon us that we must begin to explore its possibilities and challenges.

Towards that end, in March of 1998, John Campbell, a colleague at the UCLA Medical School, and I organized a one-day symposium at UCLA entitled Engineering the Human Germline to probe these issues in as serious and public a way as possible. We felt that gathering together a critical mass of distinguished scientists and ethicists - French Anderson, James Watson, Lee Silver, Lee Hood, Daniel Koshland, Mario Cappechi, John Fletcher, and others - willing to speak frankly about this difficult topic – would stimulate broad, intelligent public discussion. The one-day event was free, open to the public, and broadly advertised, but given the controversial nature of the subject we moved forward with some trepidation. Our fears were unwarranted. A thousand people attended, and responded so thoughtfully to what was said that it became clear to us, that at least in the U.S. people are ready to discuss germline engineering's present and future possibilities in a serious way. Some of the ideas that emerged at the meeting were particularly interesting:

There was general agreement that the key question is not whether human germline manipulation will occur, but how and when it will. The fundamental discoveries that bring the technology into being will occur whether or not we actively pursue them, because they will come from research deeply embedded in the mainstream and not itself directed towards human germline engineering.

Four specific realms of activity stand out, each driven by its own powerful dynamics:

  1. Medicine: Somatic gene therapy, which is directed at non-sexual cells and thus not passed to future generations, was pioneered by French Anderson and others. While it has not yet led to significant new treatments of disease, there is considerable enthusiasm and financial support to pursue the many possibilities it has opened up. Gene therapy is a fresh approach for treating hitherto untreatable diseases, and much of the technology being developed in this arena will be applicable to germinal cells .
  2. Fertility research: When Louissa Brown was born by in-vitro fertilization in 1979, she was labeled a "test tube baby,” and there was considerable hand-wringing about this "dangerous” technology. But IVF is now the obvious choice for tens of thousands of couples who would not otherwise be able to have children, and the success rate in some clinics is over 40% per cycle, higher than for natural conception. An enormous amount of energy is going into the refinement and extension of IVF, because society as a whole does not oppose expanding the reproductive options of infertile couples. Ultimately, human germline engineering will simply be an adjunct of this technology.
  3. The human genome project: Whether this effort is completed in the six years remaining in the original timetable or in the three years recently mentioned by The Institute for Genomic Research (TIGR), it is clear that the results will revolutionize biology and medicine. Uncovering the details of human genetics will create the foundation for all future gene therapy and present a myriad of enticing possibilities for germline and somatic intervention.
  4. Animal research: Recent progress in applied molecular genetics and reproductive biology has been driven by diverse laboratory work in academic institutions throughout the world, as well as by corporate efforts to produce cheaper pharmaceuticals, enhance crops, and breed new strains of livestock. Many results that emerge from these efforts - which have already produced breakthroughs such as knock-out mice, cloning, and artificial chromosomes - will be applicable to humans.

To argue that "what can be done will be done” is unconvincing because a lot that can be done is not done. But it is hard to believe that something that can be done easily and cheaply by people all over the world, and that furthermore is desired by many people with significant resources will not be done. In the not-too-distant future, germline engineering is likely to be in just this position: the technology will be feasible in hundreds of laboratories throughout the world and there will be genetic interventions that many people find alluring.

But this idea of inevitability should be balanced by the realization that widespread use of human germline engineering and other advanced reproductive technologies will come slowly. It's a long way from early research to viable clinical procedures that can be applied broadly. Twenty years after the advent of human IVF, only some 10,000 births a year in the U.S. rely on this procedure, about one in 400 babies. Procedures of this sort remain expensive and difficult for many years; there will be at least a generation between the time the procedure is safe and reliable and the time it is broadly enough available to affect more that a relative handful of people. Mario Capecchi was on the mark when he said that we always overestimate what is possible in five years and underestimate what is possible in twenty-five (Final Report, Engineering the Human Germline). Our blindness to unforeseen complications with new technologies is matched by our inability to anticipate the dramatic breakthroughs that will overcome major obstacles.

The bottom line is that we will have several decades to decide how best to handle these things. We can't know what the most important challenges of germline engineering will be, because we don't know how the technology will unfold. But once the challenges become clear, there will be time to figure out how to deal with them. Without question, this germline engineering embodies enormous possibilities and challenges, but until they become more concrete we cannot regulate the technology wisely. It is time to discuss and explore incipient possibilities, not to constrain the uncertain future. To imagine that we can see, much less judge, the consequences of this technology at this point would be a dangerous conceit.

A familiar debate about human genetic engineering concerns the distinction between therapy and enhancement, but the distinction will never be a clean one. Aging is a good example: if we succeed in unraveling the genetics of aging sufficiently to retard the process with various gene cassettes housed in an artificial chromosome placed in the germline, that would obviously be a germline enhancement, and yet it would also be germline therapy – a preventative for various age-dependent diseases. The same logic applies to cancer. If we could inoculate ourselves against cancer, it would be an enhancement, and yet at the same time it would be a therapy, a preventative against one of our biggest killers. Both examples are essentially "therapeutic enhancements,” and most germline interventions plausible in the immediate future will also be therapeutic in one way or another.

Two technical approaches

Two distinctly different technical approaches to germline engineering can be delineated. The first is homologous replacement, which is the modification of an existing gene in an existing chromosome. Such intervention, though nearly feasible now, seems unlikely ever to be used much in humans. Homologous replacement simply cannot compete with simpler technologies, such as preimplantation testing and embryo selection. Why use germline engineering to correct an aberrant gene, when an embryo selected for a correct version of this same gene can be implanted. Indeed such an approach has already been done for parents at risk of having a child with cystic fibrosis.

The second method, labeled "double addition” by John Campbell, consists of the insertion of additional genes into an additional chromosome that is added to the cell. This approach – especially with multiple genes - will likely be cheaper, safer, and more versatile than homologous replacement. Because there would be no direct interference with a cell's existing genes, double addition would be less likely to result in unwitting perturbations of the linkages among them. Ideally, a stable artificial chromosome with a series of docking sites that could each be loaded with its own independent cassette of genes would be injected into the target egg. Rudimentary chromosomes of this sort already have been developed and several companies are competing to refine them.

The library of tested gene cassettes suitable for insertion in this way eventually could be enormous. An insert proposed by John Campbell provides a good example of what might be possible. His proposed genetic preventative for prostate cancer requires two genes. The first, which codes for an ecdysone-dependent transcription factor that controls the expression of the second gene, is itself preceded by transcription controls that turn it off except in prostate ectodermal cells (those prostate cells at risk of becoming cancerous). Thus, the ecdysone-dependent factor created by the first gene is present only in the prostate, and this factor, which allows the expression of the cell suicide gene, is active only in the presence of ecdysone. When this suicide gene is turned on, it creates a toxin and kills the cell in which it was expressed, but a person would only come into contact with ecdysone (an insect pheremone) if he got an injection of it. Thus, if someone with this gene cassette were diagnosed with prostate cancer, he could take a dose of ecdysone to stimulate a precise cellular surgery in his prostate to extirpate all the cancerous ectodermal cells there.

This same approach might be used to treat breast and other cancers as well, and it is not as distant as one might imagine, each individual element of the approach has already been tested in animals. Of course, there would be unforeseen complications, perhaps an exogenous signal other than ecdysone or a more sophisticated way of selectively activating the cell-suicide gene. The important point is that entirely new ways of approaching disease are possible with germline engineering.

Germline and somatic engineering

A key idea that came up at the UCLA symposium was that germline engineering would eventually be much easier than somatic engineering. One reason for this is that germline engineering takes place outside of the body, which allows much greater control much easier access to the nucleus. A further advantage is that germline modifications need not be particularly efficient because they can be verified in cell culture before a human embryo is created. If a desired modification occurs correctly in only 1 in 1,000,000 cells, that's enough, because that one cell can be selected and multiplied in culture. With an artificial chromosome, any proposed modification could be tested in mice or chimps before it was introduced into human cells. Such testing would be essential to achieve the kind of safety and reliability we demand in human medicine.

Germline intervention will eventually be profoundly different from the somatic therapy done today. The goal of somatic therapy is to physically insert a modified gene into some target tissue without putting it anywhere else. This is reasonable for accessible tissues like the lining of the lung mucosa (cystic fibrosis) or white blood cells (ADA deficiency), but most tissue is not readily accessible. How is one to reach the heart muscle selectively or cells inside the prostate? Using the circulatory system leads to very tricky targeting problems if expression is to be limited to a particular tissue, so the challenge of transferring genes efficiently is inherent to somatic therapy.

With germline therapy, gene modifications are automatically copied into every cell in the body, and transcription controls must ensure that the genes are expressed only in the proper place and at the proper time. This is the most "natural” way of doing gene therapy, because it is how our own bodies work - by regulating and controlling our genes. The differences between our skin and muscle cells arise not from differences in the genes they contain but from differences in how those genes are expressed. If inserted genes are to be effectively managed, they will have to be accompanied by DNA control sequences, which is why artificial chromosomes with their huge capacity seem such promising delivery vehicles.

Etical challenges

Many ethical concerns about germline engineering have been raised as though they are independent of how the technology is implemented. But they are not. Ethical challenges to germline engineering often hinge upon unjustified and unstated assumptions about how germline engineering would be done. For example, it has been implicitly assumed that no "prior consent” from subjects of germline intervention could possibly exist. But an intervention such as the prostate-cancer preventative described above, effectively allows such consent, because the key genes that are inserted in the therapy remain inactive until an exogenous signal turns them on. So, though the embryo may have received its genes at the direction of its parents, the ensuing adult will be able to choose whether to activate those genes. A man with prostate cancer could decide not to activate the genetic surgery described earlier and to rely on other treatments instead. Thus "delayed consent” is possible with the right implementation of germline engineering, at least for interventions that are active only in the adult.

An even more important ethical issue has been heritability. Once again it has been assumed that germline intervention must be heritable almost by definition. But germline interventions could actually be made non-heritable. Mario Capecchi described a cre-lox recombinase system that would allow an artificial chromosome in sexual cells to lose its centromere in response to an exogenous signal, and no longer be passed to the next generation. This construct would transform germline engineering into "whole body somatic engineering” - essentially, somatic therapy on the first embryonic cell. John Campbell and I have labeled this kind of non-heritable germline therapy, "genomic extension,” because it essentially extends the genome of a single individual.

This invention is likely to become extremely important, because heritability would be an undesirable property for the germline modifications envisioned today. It hardly needs to be mentioned that no one would want to pass an error to the next generation, but interestingly, few people will want to pass on a success either. By the time recipients of even the best engineered chromosome are ready to have children, it will be twenty or thirty years after they themselves were conceived. Their once state-of-the-art artificial chromosome will be hopelessly out-of-date, and they'll want to give their child the latest gene cassettes and artificial chromosomes. It's not so different from upgraded software; they'd want the new release.

At the UCLA symposium this March, I posed a critical question. "If for a reason entirely unrelated to germline engineering, you were going to have a child by IVF, and your obstetrician offered to add an artificial chromosome with a cluster of genes that would extend your child-to-be's life expectancy by a decade or two, would you add that chromosome?” Given a safe and reliable procedure, about 2/3 of the audience said yes, a handful said no, and the rest did not respond. If the technology develops to where safe, compelling gene constructs are available, many people will clearly want them.

Achieving reliability and safety

For more than isolated human germline engineering, two things will be required: compelling gene constructs and a delivery vehicle that can safely and reliably get them into the embryo. Until both these exist, sporadic rogue activity may take place, but responsible physicians will not use the technology. It is not obvious how safe such a procedure needs to be before it is used, but absolute safety is too much to demand, because the alternative - natural conception - is a hazardous and difficult process. Natural conception is the obvious baseline to use when considering safety, and some 75% of naturally fertilized embryos never reach term.

Human artificial chromosomes appear to be the best way of achieving reliability and safety. Considering the large amount of work going on to refine the various rudimentary chromosomes already developed, it seems likely that a safe, reliable delivery vehicle for human gene cassettes - one that has been thoroughly tested in the laboratory and successfully applied in mice and primates - will exist within a decade.

The second requirement for human germline engineering is the existence of genetic constructs that people find compelling. The time of their arrival is difficult to predict, but in light of the enormous energy going into somatic gene therapy and the impending completion of the human genome project, it seems likely that such constructs will appear within the next decade or so. After all, almost any genetic intervention developed through somatic work on AIDS, cancer, and other diseases would be transferable to the germline, and key alleles linked to specific traits in the population will be uncovered as gene-chip technology matures. The first germline interventions with widespread appeal, however, may well be those targeting aging. Germline engineering, by its very nature, must be directed towards conditions in the adult that can be reasonably anticipated prior to conception. Cancer is one such condition, but the threat of cancer pales in comparison to the assaults to our health that attend aging.

However comfortable people may feel about their allotted lifespans, it is hard to imagine that the demand for a germline genetic intervention that could meaningfully retard "aging,” would not be immense. We do not yet know whether aging will be subject to easy genetic manipulation, but there are indications from work on fruit flies and nematodes that it might be. Within a decade or two we should know the answer and it clearly will have enormous implications for germline gene therapy.

Separate fantasy from reality

So where do we go from here? More research would help us better separate fantasy from reality in germline engineering. Discussion has been driven too much by fantastical imaginings rather than the therapeutic possibilities within our immediate grasp. It would be a grave mistake for us now to try to shape the distant future of this technology 50 or 100 years hence. The future possibilities are too unclear at present, so we must trust our children and grandchildren to regulate that era by making decisions then that they feel will best serve their needs. Our focus now should be on the benefits and dangers lying immediately ahead. It makes no more sense for us to try to usurp our children's choices about germline engineering than it would have been for those living at the time of the Wright Brothers to have tried to regulate today's aviation industry.

But it is crucial to begin now a serious, broad, and well-grounded discussion of the possibilities and challenges that germline gene therapy will bring. This technology may arrive before we imagine, and to make wise decisions about it we must not let it catch us by surprise as cloning did.

Various policies would contribute significantly to increased reflection on germline technology. To get a realistic feel for the possibilities and challenges germline engineering embodies, it would be useful to encourage the development of concrete research proposals for review. In the U.S., the Recombinant Advisor Commission (RAC) could begin to entertain germline as well as somatic genetic engineering proposals. In Germany, a more open exploration of the possibilities of human genetic engineering could begin. In the world of today, unilateral action by a country to block this key realm of research merely hands management of it to the rest of the world.

A broad exploration of the critical questions raised by human germline engineering is needed. We have our heads in the sand. The most difficult, the most challenging, and the most provocative consequence of the human genome project is the potential it brings for tinkering with our biological blueprint to alter our genetic design. We need to look at the ways this could be done and see which will best serve us. We need to consider how to ensure that genetic engineering procedures are safe and reliable, how to ensure that they are available to those who seek them, how to recognize and minimize their misuse. The important issues to grapple with are not whether germline engineering should occur, but how, when, where, for whom, and to what extent it should be used.

The sudden recognition that "human cloning” may be near at hand has stimulated a variety of resolutions to block this occurrence. But human cloning is most significant as a symbol: it has served notice that humanity is going to change more than the landscape we inhabit; the powerful technologies we are developing are reflecting back upon ourselves and will intrude into the most private and intimate aspects of our lives. Incredibly, we are becoming the objects of conscious design. Whether or not human cloning is banned will have little impact on that critical transformation because biotechnology is racing ahead on a broad front. Most countries have strong controls in place to discourage if not prevent irresponsible human experimentation, and the threat of litigation makes any premature commercial human cloning extremely unlikely in the U.S. No ban on cloning is needed to save the world from widespread human duplication, and if any such legislation is proposed, it should at the very least be directed at the "act of cloning a human” and explicitly avoid prohibiting any basic research.

Humanity has moved out of its childhood and into its adolescence

A number of fears about germline genetic engineering have been articulated. One of the biggest is that it will dramatically change the world by sweeping aside important anchors to our lives and creating ethical and emotional challenges for which we are unprepared. This fear is neither unreasonable nor unique to germline engineering. Technology is sweeping us forward and changing our world so profoundly that many aspects of human life will be entirely different a century hence. Advances in genetics and medicine may greatly alter human reproduction, but developments in telecommunications and computer intelligence will likely affect us as strongly in other ways.

We are intervening in realms hitherto beyond our influence, and we can seek only limited guidance from the past. Humanity has moved out of its childhood and into its adolescence, and it must recognize its growing powers and take responsibility for them. We have no choice. We are beginning to play god in many realms and cannot turn back. Some have suggested we pause until we have the wisdom to proceed. But even were that possible, it would be a flawed approach. We will gain the wisdom to make wise decisions about our newfound capabilities not by fearfully trying to avoid them, but by feeling our way forward, by probing and gathering more information, by making mistakes and responding to them, and by fully engaging in a collective decision-making process about how to go forward.

A visceral fear is that germline engineering will be abused by some totalitarian regime. But the same risk applies to any technology, because the danger is political not technological. Germline engineering would be a subtle and difficult tool to wield at a social level; there are far more direct and efficient ways for despots to try to destroy an ethnic or religious group. Hitler's eugenics programs were a minor part of the evils he brought upon the world. To avoid potent technologies ripe with possibilities simply because we imagine they might be abused by tyrants is a perilous course.

Another concern about germline engineering is that it will be available only to the wealthy and will therefore further stack the deck against the poor and disadvantaged. This fear is partially based on a misunderstanding of the way technology evolves. Yes, these technologies will initially be difficult and expensive, but as they mature they will become increasingly accessible to the broad population. Such has always been the case with new technologies. Televisions provide a trivial but highly illustrative example. The best TV the richest person in the world could buy a generation ago cannot hold a candle to today's cheap mass-marketed sets. Genetic engineering procedures will progress in the same way, and the gulf between the rich and poor of one generation will be dwarfed by the chasm between that generation and the next.. The early users of this technology may well be those wealthy enough, eager enough, and brash enough to apply it to themselves. But who could be better for such R&D than well-informed, affluent volunteers so obviously free from economic or governmental coercion?

The intergenerational gulfs brought be germline engineering will be an enormous challenge, but they should not be confused with the stereotypic divisions between social classes. Indeed. the biggest gulf of all may open up between people who embrace the idea of human evolution through genetic engineering and those who for philosophical or religious reasons do not. Strangely, germline engineering may ultimately make it possible for cultures, peoples, and families to physically manifest their philosophies about human evolution, because humans may eventually have to choose between biological stasis and progression.

The fundamental challenge of germline engineering was nicely captured by James Watson in the panel discussion at UCLA, when he said: "And another thing, because no one really has the guts to say it, if we could make better human beings by knowing how to add genes, why shouldn't we do it?”

Gregory Stock
Director, Program on Medicine, Technology, and Society
School of Medicine
University of California, Los Angeles

Note: This article has been adapted from an article by the same title that will appear in volume 7 of Jahrbuch für Recht und Ethik, "Der analysierte Mensch," Verlag, Duncker & Humboldt, Berlin, 1999. (Gregory Stock)