The Human Genome Project: The Formation of Federal Policies in the United States, 1986-1990
Robert Mullan Cook-Deegan
The human genome project began to take shape in 1985 and 1986 at various meetings and in the rumor mills of science. By the beginning of the federal government's fiscal year 1988, there were formal line items for genome research in the budgets of both the National Institutes of Health (NIH) and the Department of Energy (DOE). Genome research budgets have grown considerably in 1989 and 1990, and organizational structures have been in flux, but the allocation of funds through line-item budgets was a pivotal event, in this case signaling the rapid adoption of a science policy initiative. This paper focuses on how those dedicated budgets were created.
This case is not about the genome project itself, because that is still a nascent enterprise, but rather about the process by which it was conceived, formulated, and ratified at several levels in various federal science agencies. Describing this process is an exercise in contemporary history, retaining the advantages of direct access to the principal decision makers but necessarily suffering from a lack of perspective that only decades can bring. There are three main sources of information.
Robert M. Cook-Deegan is a physician, formerly with the congressional Office of Technology Assessment and the Biomedical Ethics Advisory Committee. In 1991 he joined the Institute of Medicine as Director of the Division of Biobehavioral Medicine and Mental Disorders.
First, I conducted interviews with the people involved. The first formal round of interviews occurred in January and February 1987, with most of them conducted during a two-month travel marathon spent visiting many western cities in the United States to gather facts for the congressional Office of Technology Assessment (OTA). Several more interviews took place later in 1987, principally in Boston, New York, and Washington, D.C. A second major round of interviews took place in July and August 1988. Since July 1986, I have also attended dozens of scientific symposia, administrative meetings, hearings, and other public events related to the genome project. At those events, I have spoken with the individuals cited in this paper, as well as several hundred more, many on a regular basis (once per quarter or more frequently).
The second source of information consists of planning documents, memos, letters, and other information gathered first for OTA and later in preparation for a book funded by the Alfred P. Sloan Foundation. Many of the people I interviewed opened their files to me, and I have copied material from OTA, the National Research Council, the University of California at Santa Cruz (UCSC), Cold Spring Harbor Laboratories, the DOE genome offices in Germantown, Maryland, the Office of the Director, NIH, and the National Center for Human Genome Research at NIH. This study has been an extraordinary opportunity to sift through the history of a science program in its infancy. Staff in the agencies were extremely generous with their time and free in allowing me access to documents. I have by no means gone through all the material in all these places; rather, I copied those documents identified as critical by those who made the decisions, or I filtered out pertinent material from large file collections. Finally, I systematically surveyed the science and lay literature for articles referring to the human genome project through mid-1988. Press accounts document the story, but they are also a part of it, as the channels of communication are themselves mechanisms for producing action.
The first section of the paper deals briefly with the origins of mapping and sequencing technologies. This topic is discussed because these technical capacities are the reason the genome project exists at all. Technical advances are not the focus of the paper but rather a backdrop to understand the ensuing story about policy formation; consequently, the technical background section is brief but dense, and it may be rough going for nonscientists or scientists outside molecular biology and genetics. If this is the case, the NRC or OTA reports on mapping and sequencing explain the technical background at greater length in lay language (National Research Council, 1988; U.S. Congress, OTA, 1988a).
The new technical means led to bureaucratic adaptations in the science agencies, and the paper's second section describes the history of
the numerous individuals who originated different conceptions of systematic genome-scale research. It also tracks how the technical ideas were translated into science programs at DOE, NIH, and the Howard Hughes Medical Institute (HHMI). This history is mingled with the concomitant process by which these science programs were funded by Congress. Securing a budget for a new program is the first step that requires justification to a community beyond the science agencies, because the budget processes within the executive branch and the justification of budgets to the appropriations committees in Congress both entail considerable effort. Securing a budget requires convincing those with broad-ranging responsibilities well beyond a particular scientific community not only that something new is needed but that it is needed more than other items competing for funds in the federal budget. Programs must contend not only with other life science programs but also with broader national priorities within science and with any federal programs that entail annual appropriations.
In the case of the genome project, the legislative and bureaucratic developments hinged on arguments, made principally by scientists themselves, about the merits of the enterprise. Some of the principal arguments and issues raised in the process of persuasion are teased apart in the final section of this paper. The justifications proffered for public funding of the genome project generated a set of obligations that the project will have to meet, and I briefly note these and discuss whether keeping such promises matters. Four brief appendices elaborate on certain specific issues mentioned only briefly in the text.
The future of the genome project remains in doubt. Public fears of how genetic information might be handled, discomfiture with the power of such intimate knowledge, and democratic distrust of powerful elites are all elements that could disrupt the working consensus that currently favors public support. Opposition to a new style of biology and research management has been intense since the beginning and shows little sign of abating. There is no disease-oriented constituency supporting the program, and so the genome program is largely a creature of the molecular biologists and human geneticists who conceived it and supported it in its early stages. The human genome project is thus, more than many other areas of biomedical research, under pressure to produce results, and it is likely to be held to greater standards of accountability than other projects for the initial promises made on its behalf. Whether the genome project is judged a social benefit is contingent on how its results are used, and there will be considerable uncertainty about this for several years at least.
The genesis of the human genome project highlights the complex interplay between people and the institutions in which they work, illumi-
nating how much difference a few individuals can make but also demonstrating how constrained those individuals are, how persistent they have to be, and how little power any one person has in the great march of science. A few individuals independently conceived a large-scale genome project, a much larger number of scientists reformulated the initial plans in a way that commanded greater support within the scientific community, and policymakers ratified the judgments made by scientists, thus providing resources to begin work. The genome project is now poised to begin in earnest, and its results over the next few years, and the uses to which those results are put, will determine the success or failure of the endeavor.
TECHNICAL AND SCIENTIFIC BACKGROUND
The genome project coalesced from a number of independent developments. Historical strands can be traced to evolutionary and population genetics, medical genetics, molecular biology, detection of mutations, advances in instrumentation, and computational biology. The principal factor was a meeting of the fields of human genetics and molecular biology. One field has long been largely clinical and descriptive; the other has been highly reductionist and focused on mechanics. As these two worlds came together in the 1970s and 1980s, each was fundamentally transformed, a process that continues today. The human genome project is a result of this collision.
Human gene mapping began in 1911, when researchers deduced that, because of its pattern of inheritance, color blindness lay on the X chromosome. For five decades thereafter, study of the odd inheritance patterns of X-linked disease was the only reliable mapping method. In the late 1960s, two technical developments occurred. First, somatic cell hybridization became a mapping strategy. This method mixed chromosomes by fusing together cells from humans and other organisms. The mixed chromosomes fragment and reorganize into metastable cell lines that retain various amounts of human deoxyribonucleic acid (DNA). It turned out that rodent-human cell lines, after a few generations, generally retained mainly rodent and only a small amount of human DNA and were relatively stable over time. By assembling large numbers of such cell lines, and devising clever ways to select only those cells that contained functional genes of interest, it became possible to map genes. During this period, it also became possible to differentiate the 24 distinct human chromosomes under the light microscope by staining them with DNA-binding dyes, producing a karyotype (normally 22 pairs of autosomes and either a pair of Xs, in females, or an X and Y, in males). In a photograph of the nucleus of a cell, the chromosomes
could be directly seen and large-scale deletions, rearrangements, and duplications detected. Somatic cell hybridization and karyotyping launched human geneticists on their quest for a complete gene map (McKusick, 1988).
In the mid-1970s, restriction enzymes, recombinant DNA techniques, and the enormous variety of molecular biological techniques for selectively cutting and copying DNA ushered in a new era in gene mapping. Recombinant DNA led to the isolation and cloning of hundreds of human genes, but there was another significant spinoff: mapping by linkage to DNA markers. The idea was to find landmarks along the human chromosomes that would allow geneticists to determine which parts of which chromosomes were inherited from which parent. Once located, the markers could be used to trace the inheritance of bits of chromosomes through families, so that the inheritance of markers could be compared with the inheritance of diseases or other traits.
There are, very roughly, 3 million differences in DNA sequence “spelling” between any two people. Most of these differences have no detectable effect on the individuals, but they can be measured by direct analysis of DNA. If there are enough markers and enough people in a family to do the statistics, one can “link” the inheritance of a genetic character (a disease or trait) to the inheritance of a chromosomal marker. The closer a gene is physically to the marker being studied, the less often it will be separated in the process of producing sperm and egg cells, and the greater the statistical linkage to the marker. Because the marker's chromosomal location is known, at least approximately, this information locates the gene nearby.
Kan and Dozy first used linkage to a sequence difference to detect different variants of hemoglobin in 1978 (Kan and Dozy, 1978). The first published suggestion that a systematic collection of such markers be made occurred in 1979 (Solomon and Bodmer, 1979). A landmark paper published a few months later (Botstein, 1980) elaborated the idea in considerably more detail, initiating an explosion of genetic linkage mapping in the 1980s.
Methods of studying the inheritance of markers and genetic characters harken back to the mathematical genetics developed late in the last century and early in this one. The scientific approach is fundamentally classical genetics—the study of the inheritance of observable differences among individuals—supplemented by clinical observation to define the genetic characters under study and augmented by the modern tools of molecular marking. The process relies on the mathematics of probabilities to make correlations. The communities that studied evolutionary and population genetics immediately understood the significance of genetic linkage mapping. They were joined by a few
medical geneticists who were comfortable with the statistical techniques of linkage. When the method yielded success in locating the gene responsible for Huntington's disease in 1983 (Gusella et al., 1983) and the gene responsible for polycystic kidney disease in 1985 (Reeders et al., 1985), clinical genetic research quickly adopted it.
By the mid-1980s, genetic linkage mapping was part of the mainstream of human genetics. Newsweek magazine quipped in late 1987 that there was a disease a week being mapped by genetic linkage (Begley et al., 1987). Technical advances further extended the ability to work backwards from an approximate gene location, determined by linkage to a marker, to find the gene itself and identify its product (in most cases, a protein). The first successful search for a gene starting from its chromosomal location ended in 1987, with the cloning of a gene that causes the rare condition chronic granulomatous disease (Royer et al., 1987). This achievement was soon followed by location of the Duchenne's muscular dystrophy gene (Koenig et al., 1987) and retinoblastoma (Friend et al., 1986; Lee et al., 1987). In these cases, however, the location was known from patterns of inheritance (on the X chromosome for Duchenne's muscular dystrophy and on chromosome 13 for retinoblastoma), or from human-hamster hybrids, and the study of individual patients who had lost specific portions of the X chromosome.
The process of going from chromosomal location to isolated gene is slow, tedious, unreliable, and often frustrating. (Intensive work over seven years failed to produce the Huntington's disease gene, for example.) But many prevalent disease-causing genes have been isolated in this way, most notably the gene that causes cystic fibrosis (Kerem et al., 1989; Riordan et al., 1989; Rommens et al., 1989). Cystic fibrosis was the first case in which the gene was mapped initially by genetic linkage; then the regional DNA was studied until a gene was found and its product identified, a membrane protein thought to be involved in the regulatory flow of chloride ions into cells. The idea that mapping can be the critical first step in understanding genetic disease has thus been confirmed in principle and in practice, but there is a long way to go before the more than 4,000 known disorders have been correlated with genes and gene products. The late 1980s were the period that saw human genetics, and with it the study of genetic diseases, joined with molecular biology in happy matrimony.
Molecular biology is largely a post-World War II phenomenon. Its two seminal events are Avery, MacLeod, and McCarty's discovery in late 1943 of DNA as the “transforming principle,” conferring heritable traits (Avery et al., 1944), and Watson and Crick's revelation in 1953 of the double helical structure of DNA (Watson and Crick, 1953). These
are, indeed, two of the high points in twentieth-century science and culture.
The distinctive signature of molecular biology is its approach to understanding function through the study of molecular structure. The double helical structure of DNA is the touchstone of this approach, explaining at once how information can be transmitted from generation to generation or from cell to cell during development, and also how information can be decoded into cellular processes through DNA-directed synthesis of proteins and ribonucleic acid molecules. One tenet of molecular biology is to study simple systems of living things, and early work in molecular biology, Delbrück and Luria's “phage” group in particular, focused on the simplest—viruses that infect bacteria (Judson, 1979). Beginning in the 1960s, however, molecular biology invaded field after field, applying its increasingly powerful tools to questions of greater complexity. By the mid- to late 1970s, molecular genetics was applied with astonishing success to the study of cancer and resulted in the discovery of oncogenes.
The first disease characterized at the molecular level was sickle cell anemia. In 1949, genetic studies by Neel showed that it was a recessive genetic disease, and biochemical studies by Pauling and colleagues (1949) indicated that it was caused by a chemical change in the structure of hemoglobin. In the mid-1950s, Ingram identified the difference between sickle and normal hemoglobin by breaking the protein into small fragments and looking for differences. He was able to establish that in the sickle cells a single glutamine amino acid had been replaced by valine in one of the two pairs of protein chains that make up hemoglobin (Ingram, 1957). This difference suggested a mutation in the DNA encoding of the beta chain of hemoglobin. Until the past few years, most of the tools of molecular biology were applied following this paradigm, that is, studying individual genes, one at a time, by biochemical analysis. Application of molecular techniques to chromosome mapping came from pushing molecular biological techniques at both ends—on the one hand, forcing chromosomal mapping to higher resolution, ultimately enabling direct decoding of the DNA base pair sequence, and on the other hand, developing techniques to separate and clone larger and larger fragments of DNA, culminating in reproduction of megabase stretches of DNA.
Before the development of these techniques, the handling of large fragments of DNA was difficult for several reasons. First, the manipulations involved in preparing it for analysis often sheared the long, fragile strands. Second, the widely used analytical techniques could only separate fragments of up to thousands of base pairs in length. Finally, because of the modified viruses and plasmid vectors used, the length of DNA that could be cloned was also limited to a range of from several
thousand to a few tens of thousands of base pairs. During the early to mid-1980s, it became possible to handle long strands of DNA without breaking them by manipulating them in gels rather than in solutions. It also became possible to separate DNA molecules of up to several million base pairs in length using pulsed-field gel electrophoresis, first developed by Schwartz and Cantor, who pioneered several innovations in electrophoretic separation methods (Schwartz and Cantor, 1984). Moreover, cloning vectors that could consistently contain 30,000 to 40,000 base pairs became standard fare through incremental improvements in dozens of laboratories. With these concomitant advances, it became possible to take DNA from chromosomes, clone it, and analyze it to reconstruct the order of cloned DNA fragments, so that eventually a complete map of the original DNA could be assembled. This kind of map had the enormous advantage that the chromosomal DNA would be not only mapped but also cloned and stored in the freezer for further analysis. If such a tool had been available for the tip of chromosome 4, for example, those searching for the Huntington's gene in that region could have studied the DNA directly as soon as they located the gene. Having the DNA cloned would make the search for closer markers and candidate genes much simpler and faster. The DNA sequence would be another leap forward. Work on both fronts is now proceeding.
Two groups began independently to apply the cloning and ordering strategy to make ordered maps of yeast, in Olson's laboratory at Washington University, and of nematodes, in Sulston and Coulson' s laboratory in Cambridge, U.K., and Waterston's at Washington University. Work began in the early 1980s and began to show promising results by 1986 (Coulson et al., 1986; Olson et al., 1986). The genome of the nematode is roughly the same size— 100 megabases—as a small human chromosome, and it thus became conceivable to map the human genome by extension of the nematode method. Such an extension followed a tradition in molecular biology to focus on a new problem an order of magnitude greater than one that has been solved before. (The exact size of the leap was open to discussion, but the principle was widely accepted.)
DNA sequencing was developed by groups located in both Cambridges, more or less simultaneously, using entirely different approaches. Sanger's group in Cambridge, U.K., developed DNA sequencing after a dedicated and deliberate effort that started with protein sequencing, progressed to the sequencing of ribonucleic acid, and culminated in DNA sequencing. Sanger presented a partial sequence of a virus to an awestruck audience in May 1975 (Judson, 1987) and published a modified, simpler method in 1977 (Sanger et al., 1977). Maxam and Gilbert, working in Cambridge, Massachusetts, developed DNA sequencing from their at-
tempts to study directly the regulation of gene expression in bacteria. Gilbert's group had been early pioneers in the field. They isolated their first DNA segment and deduced their first DNA sequence during 1972-1974. This first sequence consisted of 24 base pairs and took two highly competent investigators two years to achieve (W. Gilbert, Harvard University, personal communication, July 1988). The next step was to use chemical modifications of DNA bases to study directly the DNA protein-bound segments that regulate gene expression. Maxam and Gilbert realized they had come upon an approach that, with some further work, would permit direct DNA sequencing (Kolata, 1980). By August 1976, they were ready to distribute the chemical recipes used in their sequencing reactions at a Gordon conference (one of many small, closed gatherings of scientists held in New Hampshire colleges each summer). They also published their method of DNA sequencing in 1977 (Maxam and Gilbert, 1977).
Molecular biology thus generated a cornucopia of technological tricks that allowed scientists to think seriously about constructing physical maps of chromosomes and determining their DNA sequence. Some human geneticists were quick to apply the developing techniques to study diseases, but with a few exceptions, molecular biology was a separate field from human genetics. Nonetheless, the two disciplines were rapidly converging, as molecular biology worked its way into yet another field ripe for the picking.
The early to mid-1980s also saw two other important technological developments: diffusion of the personal computer and automation of microchemical manipulation. The computer revolution was imported from other areas but quickly adapted to the needs of biologists. It was important because it put personal computers in thousands of laboratories that were unacquainted with them. It permitted more analysis of raw data, and there was a natural harmony with the digital analysis of linear DNA sequence information. As information processing became faster and cheaper by orders of magnitude every few years, biologists, including molecular biologists, began to use computers more and more.
Automation of microchemical processes made possible experiments that were too tedious to do by hand. Automation was successfully cultivated at only a few university centers and in companies that either were already selling instruments to biologists or had been newly formed to do so. Instruments were devised first to sequence and synthesize proteins for analysis of their amino acid building blocks. Analysis of DNA was the next step. Serious efforts to synthesize short segments of DNA, a capability essential to developing highly sensitive probes for analyzing genetic experiments, began in the late 1970s; they had by 1980-1981 proved successful.
Automation of DNA sequencing began around this time in both Japan and the United States. In the United States, the first efforts leading to the current generation of DNA sequenators began in 1980 at the California Institute of Technology, or Caltech, under a five-year grant from the Weingart Institute. The first government support, through the National Science Foundation (NSF), came only in 1984 after a successful prototype was developed. In Japan, the Science and Technology Agency in 1981 began to support a project to automate DNA sequencing that involved several corporate sponsors (Fuji Photo, Seiko, and Matsui Knowledge Industries). The automation effort at the European Molecular Biology Laboratory in Heidelberg began several years later, supported by several European governments.
All of these technological developments surged during 1980-1985. In their wake came ideas for a concerted genome project, and several farsighted people independently brought them forth.
ORIGINS OF DEDICATED GENOME RESEARCH PROGRAMS
The idea of a systematic gene map of human chromosomes was not new. Human geneticists had talked of it for decades. The notion of large-scale sequencing was also discussed soon after sequencing techniques became widespread in 1977. Several groups talked of sequencing the HLA region involved in immune regulation and the regions encoding antibody protein genes. The European Molecular Biology Laboratory seriously discussed a dedicated project to sequence Escherichia coli in 1980-1981. Solomon and Bodmer (1979) mentioned the benefits of finding DNA markers throughout the chromosomes, and Botstein and colleagues analyzed in detail the significance of a systematic effort in this area (Botstein, 1980) in a landmark 1980 paper. Yet none of these ideas for a collective effort took hold within the federal government. DNA sequencing was widely used, but it remained the province of thousands of small laboratories focused on small regions. The Cambridge, U.K., group, under Sanger and then Barrell, were almost alone in sequencing the entire genomes of progressively larger organisms.
The key idea in genome projects was a dedicated effort to map and sequence whole organisms or significant parts of their genomes (e.g., an entire chromosome or chromosomal region). Government support for such technically focused efforts was slow to develop. Several attempts to entice NIH to construct a genetic linkage map were rebuffed, in part because the logical mechanism was a service contract or other nongrant mechanism. These were considered highly suspect because, in part, they had been used to support cancer research of only marginal usefulness. One corporation, Collaborative Research, Inc., and a private philanthropy,
the Howard Hughes Medical Institute, stepped in to fund laboratories dedicated to generating DNA markers. These two laboratories, under the direction of Helen Donis-Keller (Collaborative Research) and Ray White (HHMI, Utah), contributed more than half the DNA markers that existed on the human genetic map in 1987 (Donis-Keller et al., 1987). The idea of focused mapping received some support when it fit into the format of a small scientific project, as in the case of physically mapping yeast, but several proposals to apply these methods to human chromosomes were rejected by scientific review groups in the mid-1980s.
In 1985 and 1986, however, several groups began to buck the tide. The first discussion of a large dedicated genome project came at a workshop convened at the University of California at Santa Cruz in June 1985. In the fall of the same year, Norman and Leigh Anderson proposed that sequencing the genome and cataloging all known genes should be a concerted national effort, but the idea was recorded in a relatively obscure journal and never caught fire (Anderson and Anderson, 1985).
By a curious twist, the history of the genome project is connected to the Keck telescope that now graces Mauna Kea, joining the cluster of other large telescopes on a Hawaiian mountaintop. The story behind this connection merits a digression because it is a classic example of how the quest for funds breeds scientific entrepreneurship and how thinking about Big Science infiltrated the field of biology.
Robert Sinsheimer was chancellor of UCSC in the fall of 1984. He was a biologist who wanted to leave a mark on his institution. In his own words, he “wanted to put Santa Cruz on the map in biology. ” He was also faced with a problem: he knew about a pot of money but had no way to spend it.
The events leading to this development were initially tied not to biology but to astronomy. The UCSC astronomy department had become extremely enthusiastic about building the largest optical telescope in the world. (UCSC had an excellent international reputation in astronomy, which was a great source of pride for the university.) One problem, however, was the prohibitive cost of producing the mirror for such a telescope. This problem was solved in principle by packing together 36 small hexagonal mirrors, rather than producing a single large mirror, which lowered the estimated costs from $500 million to $70 million. With this development, UCSC decided to seek funding for a telescope from private donors. A story was run in the San Jose Mercury, and the university received a call from a person familiar with the newly formed Hoffman Foundation, created after the death of Max Hoffman, the U.S. importer of Volkswagen and BMW automobiles. After further inquiry, the foundation indicated that Hoffman's wife might be interested in
contributing $36 million to help finance the world's largest telescope. David Gardner, president of the University of California system, was contacted, and the money was accepted. It was the largest single contribution to the University of California in its history. Mrs. Hoffman died the next day.
The university continued to search for funds, but it began to have difficulty securing the additional donations, in part because it was a state university largely supported with taxpayer dollars and in part because the agreement with the Hoffman Foundation included naming the telescope for Max Hoffman. Caltech was approached to explore the possibility of a joint effort, assuming it could help raise the requisite funds. Caltech secured an additional $15 million from among its trustees, and then it contacted the W. M. Keck Foundation, established with monies from Superior Oil. The Keck Foundation was willing to help but wanted to fund the entire effort and have the telescope named after Keck. According to this plan, the $36 million from the Hoffman Foundation and other prior donations could be used as operating capital. When the Hoffman trustees were approached with the idea as well as another proposal to build twin telescopes, however, both overtures were rejected. The University of California returned the check for $36 million. The Keck telescope saw first light in December 1990 managed by the California Association for Research in Astronomy, the University of California-Caltech group established for the project.
To return to the genome thread of the story, Sinsheimer's problem late in 1984 was what he might do to recoup the Hoffman funds. He decided to develop a proposal for a big, attractive project. He began by considering what opportunities might be lost in biology because of an exclusive focus on projects that could be done by small groups without special facilities. After rejecting a number of possibilities, he hit upon the idea of sequencing the human genome. He called in UCSC biologists Robert Edgar, Harry Noller, and Robert Ludwig to discuss setting up an institute at UCSC for this purpose. At first the three were stunned by the idea, thinking it ludicrous in its audacity, but after some discussion they felt it was worth further consideration. Edgar and Noller then prepared a position paper, dated Halloween 1984, that described the genome sequencing institute as
a noble and inspiring enterprise. In some respects, like the journeys to the moon, it is simply a “tour de force;” it is not at all clear that knowledge of the nucleotide sequence of the human genome will, initially, provide deep insights into the physical nature of man. Nevertheless, we are confident that this project will provide an integrating focus for all efforts to use DNA cloning techniques in the study of human genetics. The ordered library of cloned DNA that must be produced to allow the genome to be sequenced will itself be of great value to all
human genetics researchers. The project will also provide an impetus for improvements in techniques that have already revolutionized the nature of biological research
As the next step in the project, the UCSC group decided to call a meeting of experts from around the world. Noller wrote to Frederick Sanger, the two-time Nobel laureate whose DNA sequencing methods were described above, and with whom Noller had worked early in his career. Sanger wrote back: “It seems to me to be the ultimate in sequencing and will probably need to be done eventually, so why not start on it now? It's difficult to be certain, but I think the time is ripe.”
The meeting was held on May 24 and 25, 1985. The group assembled included those pushing the limits of DNA sequencing (Bart Barrell, Leroy Hood, and George Church), some originators and practitioners of genetic linkage mapping (David Botstein, Ronald Davis, and Helen DonisKeller), large-scale physical mappers (John Sulston and Robert Waterston), mavens of large DNA fragment analysis (Leonard Lerner and David Schwartz), and a mathematician concerned with analysis of DNA sequence (Michael Waterman). An important addition, however, was made at the last minute.
While the meeting was being organized, Walter Gilbert, co-inventor of the chemical modification DNA sequencing method and one of the most highly respected minds in molecular biology, was off in the Pacific after having resigned as chief executive officer of Biogen, Inc. The Santa Cruz group strongly wanted his blessing, and after some effort Edgar finally reached him in late March. Gilbert was in transition back to his faculty position at Harvard, and he agreed to come. His presence became central to the unfolding genome story.
After the meeting, Sinsheimer summarized its conclusions, which Steven Hall later reported, capturing the modesty of the meeting in “Genesis, the Sequel” (Hall, 1988). The group agreed that it made sense to pursue systematic development of a genetic linkage map, a physical map of ordered clones, and the capacity for large-scale DNA sequencing (Sinsheimer, 1989). The sequencing effort early on should focus on automation and development of faster, cheaper techniques. This summary was sent to several potential funding sources, including HHMI and the Arnold and Mabel Beckman Foundation, but there was no response. Donald Fredrickson, then president of HHMI, was at that time also hearing ideas about a different variety of genome project from Charles Scriver, then a member of the HHMI medical advisory board. HHMI decided to investigate further but did not agree to fund the Santa Cruz proposal.
Gilbert was an extraordinarily articulate science visionary. In a
memo to Sinsheimer two days after the workshop, he translated the Santa Cruz group's ideas into specific operating plans and became the torchbearer for the effort to generate enthusiasm, taking the ideas generated at the workshop into the power centers of molecular biology. He gave informal presentations on sequencing the genome at a Gordon conference and at the first international conference on genes and computers in August 1985. Gilbert was extremely well connected and infected several of his colleagues with his enthusiasm. (Two of them in particular—Paul Berg and James Watson—figure later in the story. Both Nobel laureates like Gilbert himself, they were counted among the most respected and powerful figures in molecular biology.) Gilbert also gave the genome project much greater notice than it would otherwise have achieved, earning feature stories on his role in it from U.S. News and World Report (McAuliffe, 1987), Newsweek (Begley et al., 1987), Boston magazine (del Guercio, 1987), Business Week (Beam and Hamilton, 1987), Insight (Holzman, 1987), and the New York Times Magazine (Kanigel, 1987). In addition, he and Leroy Hood wrote supporting articles for a special section in Issues in Science and Technology published by the National Academy of Sciences (Gilbert, 1987; Hood and Smith, 1987), and he and Walter Bodmer wrote editorials for The Scientist (Gilbert, 1986; Bodmer, 1986b). Gilbert thus kept the steam up in the genome project engine, even as Sinsheimer's attempts locally at UCSC were meeting bureaucratic resistance from the University of California system.
Sinsheimer, however, was also sounding out his colleagues about the idea of a genome sequencing institute. He spoke to James Wyngaarden, director of NIH, at a meeting in Washington, D.C., sometime in late February or early March 1985. His personal note about this conversation stated that Wyngaarden was quite supportive and urged Sinsheimer to put together a proposal to the National Institute of General Medical Sciences (NIGMS) after the May workshop. Wyngaarden judged that “it would not be too difficult to get congressional funding for the project, through NIGMS,” according to Sinsheimer. Two years later, Wyngaarden recalled this conversation only vaguely but agreed that he would probably have said something like what Sinsheimer reported.
It would have been logical to seek funding from NIH, but such a course presented several problems. The cost estimates from the Santa Cruz meeting—$25 to $40 million to build an institute, with an annual budget of roughly $10 million—were far too high for a grant or standard research program. The project would require a special appropriation, which raised the difficulties of approaching Congress. This step also required the approval of the president of the University of California system. Sinsheimer judged that the university system' s support was
contingent on getting a large private donation to start things off and proposed that he approach the Hoffman Foundation with his new idea. The president's office, however, stalled for several months on the proposal, perhaps because it did not accord the genome institute a high priority or because it was concerned about conflict among the various campuses in the system, many of which could argue that they were better positioned to house a genome institute. Whatever the reason, the approach was never made, and the initial impetus for the sequencing idea—the potential availability of funding from Hoffman—proved a dead end. Moreover, no other private donor ever materialized. Sinsheimer later lamented, “I thought the extraordinary significance of the project would be more self-evident to some of the prospective donors than proved to be the case” (R. Sinsheimer, personal communication, September 1988). In the end, without the needed private support, the idea of a genome institute at UCSC died a slow, quiet death.
THE DEPARTMENT OF ENERGY PLAN
A more successful seed was planted in December 1984, when DOE sponsored a meeting at Alta, Utah, to discuss how to measure heritable mutations in humans (Cook-Deegan, 1989). Ray White of the University of Utah organized the meeting at the behest of Mortimer Mendelsohn of Lawrence Livermore National Laboratory and David Smith of DOE. The specific question to be addressed was whether new DNA-based methods were sensitive enough to detect any increase in mutations among survivors of the Hiroshima and Nagasaki atomic bomb blasts. A group of scientists engaged in developing new DNA analytical techniques were invited to participate. The conclusion of the meeting was that the methods could not yield an answer with the scale of effort that was currently feasible, but the workshop had a more lasting effect as a result of the coincidence of several events. The workshop was in progress just as Schwartz and Cantor were producing the first data using pulsed-field gel electrophoresis for mapping, as George Church was beginning to think of new approaches to DNA sequencing directly from DNA in the native genome of an organism, and as Maynard Olson's physical mapping efforts in yeast were beginning to bear fruit. Its timing was propitious.
OTA staff were preparing a report on technologies to measure heritable mutations in humans because the issues of exposure to Agent Orange, environmental toxins, and radiation were beginning to come before congressional committees (U.S. Congress, OTA, 1986). Mike Gough, the OTA project director, was present at the meeting and discussed the various technologies in a draft report that was sent to DOE
for review. Charles DeLisi, as newly appointed head of DOE's Office of Health and Environmental Research, reviewed the draft and recalled looking up from its pages with the idea for a dedicated project focused on DNA sequencing and computation (DeLisi, 1988).
DeLisi and David Smith of DOE moved quickly on many fronts in the December lull of 1985. They asked the biology group at Los Alamos National Laboratory for its comments on DeLisi's idea, and just before Christmas the group responded with a dense, somewhat scattered, but extremely enthusiastic five-page memo. The principal author was physician Mark Bitensky. The memo concerned sequencing the entire human genome and barely mentioned physical or genetic mapping. It provided estimated costs, noted that such a project could become a “DNA-centered mechanism for international cooperation and reduction in tension,” and extolled the potential technical and human health benefits. The Los Alamos group even persuaded Frank Ruddle to agree to testify before Congress. With this initial feedback, Smith and DeLisi began to pull the bureaucratic levers in Washington.
In a note to Smith, DeLisi outlined an approach to garner support from the scientific community, from his superiors at DOE, and from Congress. In a return note to DeLisi dated December 30, 1985, Smith mentioned previous discussions of sequencing the human genome at a Gordon conference and at a meeting at the University of California the previous summer but said he did not know what had come of these efforts. He also anticipated the criticisms that would plague the DOE proposal for some time to come: that it was not science but technical drudgery, that directed research was less efficient than letting small groups decide what was important and then do it, and that effort should be concentrated on genes of interest rather than global sequencing. In a reply the next day, DeLisi contended that “regarding the grind, grind, grind argument there will be some grind; what we are discussing is whether the grinding should be spread out over 30 years or compressed into 10.” He presciently noted that “we are talking about $100-150 million per year spread out over somewhat more than a decade ” and further asserted that such a project would certainly rate as more important than the lower 1 percent of grants that funding of this magnitude would displace. He suggested that the political effort should focus not on whether it would displace other work but instead on how to gain support for new funding.
In January 1986, DeLisi discussed the idea with his superior, Alvin Trivelpiece, who as director of the Office of Energy Research reported directly to the secretary of energy (then Herrington). Trivelpiece supported the project and charged the DOE biological sciences advisory committee (the Health and Environmental Research Advisory Commit-
tee, or HERAC) to report back to him about the idea. This action by Trivelpiece followed several discussions with DeLisi about the possibility of doing a genome project in DOE. The two men had discussed why the agency did not have the same high stature in biology that it had in high-energy physics, and both aspired to change that situation by providing a project that would propel DOE to the forefront of biology. As part of the outreach to the scientific community, Los Alamos was asked to convene a workshop (1) to find out if there was consensus that the project was feasible and should be started, (2) to delineate medical and scientific benefits and to outline a scientific strategy, and (3) to discuss international cooperation, especially with the Soviet Union.
During 1986, the wheels continued to turn. A workshop was held at Santa Fe on March 3 and 4, with “a rare and impassioned esprit,” according to Bitensky's memo that summarized it. Discussions at the workshop resulted in a clear emphasis on physical mapping by ordering clone libraries as a crucial first step (collected papers from 1986 Santa Fe Workshop, DOE, not published; Bitensky, 1986). In letters back to the conference organizer, Mark Bitensky, there was consensus on the importance of a new project, a fair degree of agreement on what should be done next, and a wide range of opinions about how to organize the effort. Anthony Carrano and Elbert Branscomb from Lawrence Livermore National Laboratory stressed the importance of clone maps and warned that “a program whose announced purpose was simply to ‘sequence the human genome' might unnecessarily and incorrectly arouse fears of territorial and financial usurpation in the biomedical research community.” They were certainly right in that regard. In contrast, David Comings was rather far off the mark when he averred that the whole physical mapping component might be funded “without any stirring up of any congressmen or other related creatures.” The creatures were not so docile; indeed, they proved downright ornery.
By May, DeLisi had produced an internal planning memo to carry the request for a line-item budget. The memo was transmitted to Trivelpiece and from there up through the DOE bureaucracy. By the time the memo was prepared, the project had been broken into two phases. Phase I had three components. The first, physical mapping of the human chromosomes, to last five or six years, took up much of the first phase. The other two components in Phase I were development of high-speed automated DNA sequencing and a research program to improve computer analysis of sequence information. DeLisi's background in computational biology came to the fore here. Phase II, which was contingent on success in Phase I, entailed sequencing the banks of DNA clones put together in a physical map of the chromosomes.
In his memo to Trivelpiece dated May 6, DeLisi spoke of a project
analogous to a space program, but requiring the efforts of many agencies and a more distributed work structure, with “one agency playing the lead, managerial role DOE is a natural organization to play the lead management role.” In a separate memo DeLisi requested a budget of $5, $10, $19, $22, $22, and $22 million dollars for fiscal years 1987-1992. The plans survived internal DOE review, and a series of meetings were scheduled in late 1986 with Judy Bostock, the DOE life sciences budget officer in the White House Office of Management and Budget (OMB), in conjunction with planning for fiscal year 1988 and beyond. Bostock was a physicist from the Massachusetts Institute of Technology, with a strong interest in biology, especially in improving the speed and efficiency of biological research. The budget briefing documents for DeLisi's Office of Health and Environmental Research/OMB meetings included a budget projection for fiscal years 1987-1990 of $5.64, $11.55, $18, and $22 million. In the DOE copy “$22 million ” for 1990 is scratched out and replaced with “$23.5,” and there is a handwritten note that the changes resulted from discussions with OMB. The document's cover sheet specifies a four-year project beginning October 1, 1987, extending to September 30, 1991, and costing $95 million. By simple arithmetic, this suggests there was an agreement for a fiscal year 1991 budget of $40 to $45 million. Decisions about a Phase II budget were to be made in 1990 and 1991. Bostock confirmed that there had been minor revisions, but essentially the proposal worked out by DOE and OMB in fall 1986 became the basis for a multiyear program agreement.
The DOE HERAC endorsed the plan for a DOE genome initiative in a report from its special ad hoc subcommittee. The subcommittee was composed of 14 scientists, only one of whom was from a national laboratory. It was a blue-ribbon scientific group chaired by Ignacio Tinoco, a highly respected chemist from the University of California at Berkeley (then on sabbatical for a year at the University of Colorado in Boulder). The report urged a budget of $200 million per year, and made a case for DOE leadership of the effort. A few observations must be made about this advisory process, however. First, the subcommittee's budget projections were not at all connected to the multiyear DOE-OMB budget agreement discussed above. The subcommittee first considered budget projections on February 5 and 6, 1987, at a meeting in the Denver Stouffer's Hotel (see the discussion of costs below). The DOE-OMB agreement is dated one and a half months earlier, December 18, 1986. DeLisi had briefed OMB earlier, on September 5, and received tentative agreement (Hall, 1988). Clearly, DeLisi was willing to listen to the subcommittee's advice, but it is equally clear that the commitment to go ahead with a project, including a multiyear budget, was made long
before DeLisi knew what the subcommittee would say. Second, at its final meeting to draft the report, the subcommittee did not discuss which agency should lead the effort. This deficit was pointed out to HERAC when it met to consider the subcommittee's report in March; by April, when the report was released, Tinoco as subcommittee chairman and Mort Mendelsohn, a member of the subcommittee and chairman of HERAC, had canvassed members to gain support for language in favor of DOE leadership. Later interviews with members of the subcommittee revealed that at least 7 of the 14 had reservations about giving DOE a blank check; they agreed to the suggested language, however, because they perceived a lack of action on the part of NIH and thought the project so important that it should be done no matter which agency did it.
In the waning days of 1985, DeLisi and Smith forged a plan that propelled the human genome project onto the public agenda. It is clear from memos and personal notes that they did this deliberately and with the purpose of establishing a new mission for the DOE-supported laboratories centered on sequencing the human genome. The process for obtaining funding included successful transit of the DOE bureaucracy and agreement from a highly involved OMB budget officer.
Yet despite the go-ahead from the bureaucracy, the job was not complete: now came the two-step congressional process. Here DeLisi was less adept, although he managed it. Any new action of the federal government requires congressional authorization and appropriation. These twin processes are interdependent but distinct. Authorization falls to a pair of committees, one each in the House and Senate. Which of the authorization committees handles a particular science agency is determined by an intricate set of jurisdictional rules negotiated over the years by the committees. The authorizing committee structure is not exactly parallel between the House and the Senate because the two houses have different boundaries, drawn in part to accommodate the individual interests of past and current committee chairmen. The appropriations process, in contrast, is a parallel process with a relatively stable annual routine.
The President's budget proposal is submitted in January each year and then goes to the appropriations committees. Except in unusual circumstances (as occurred once during the Reagan years, violating the spirit, if not the letter, of the Constitution), the House takes action first, and the Senate works from the House figures. If there are new programs under consideration, appropriations are theoretically, and in most cases actually, contingent on prior passage of an authorization statute. The appropriations committees are not to legislate but rather to fund activities under rules set forth by other committees. The interpretation of this proviso can be liberal or strict, depending on the circumstances.
To get the genome program started, DeLisi took $4.5 million in funds from the preexisting fiscal year 1987 budget and reallocated them to the genome effort. Such limited “reprogramming” is common practice, permitted by the appropriation and authorization committees within reasonable limits with written justification. For 1988 and later budgets, however, DOE needed support from its authorization committees and funding from the appropriations committees. DeLisi had noted the need for congressional action in his December 1985 note to David Smith, and he had held some meetings with congressional staff in 1986. There was little problem in the Senate, as DOE had the strong support of Senator Pete Domenici and tacit approval of Senator Wendell Ford, the key figures on the authorization committee. Domenici also sat on the appropriations and budget committees and could be counted on for support there. The problem was in the House.
Staff of the relevant DOE authorization subcommittee in the House were getting mixed signals about the DOE genome initiative. They had read the generally negative response to it in Science magazine, and a few calls to contacts in the molecular biology field elicited both support and opposition. Eileen Lee was the committee 's resident biologist, and was understandably uncertain about the tack the committee should take. The problem was further complicated by the politics of DeLisi's other biology programs. The committee staff of the majority party were generally disposed to support initiatives coming from DOE staff, who were, after all, paid to do just such planning, but DeLisi's relations with the committee were problematic and it was unclear to the staff whether they should expend the political capital to defend him on the genome initiative. Claudine Schneider, ranking Republican on the committee, was dissatisfied with DOE's record on research into environmental health hazards, and her staff director, Eric Erdheim, was rumored to have called the Delegation for Biomedical Research to ask James Watson to testify against the DOE genome program. (Erdheim did, indeed, speak with Bradie Metheny, lobbyist for the delegation, but nothing came of it; Erdheim was not so much opposed to the genome project as suspicious of any new proposal coming from DeLisi. Watson stated in an interview a few months later that he was never asked to testify. DeLisi had a backup if he ran into trouble in the subcommittee, however, because Manuel Lujan, ranking Republican on the full committee, came from New Mexico and was well known as a national laboratory supporter. If Schneider had wanted to change DOE's direction, she would at least have had to notify him.)
Eileen Lee arranged for Leroy Hood to testify before the committee, after calling OTA and several other contacts for suggestions. Hood agreed, oblivious to the political maelstrom swirling around him, and on
March 17, 1987, projected a passionate vision of the genome project (U.S. Congress, House, 1987a). Hood strongly supported a new genome initiative and proposed a role for DOE, NIH, and NSF, thus deftly ducking the troublesome question of which agency should hold the reins. Schneider's latent distrust broke the surface in a series of questions about DOE reports on health effects of radiation on submarine workers, Hiroshima and Nagasaki survivors, and nuclear plant workers and reports on least-cost energy, but the genome program glided through the hearings unscathed. The appropriations process was less troublesome than authorization and presented no major obstacles once the genome project had OMB approval.
The DOE budget process for fiscal years 1988 and 1989 held true to the initial agreement with OMB—the agency sought $12 million and $18 million, respectively, for the two years. It exceeded the initial agreement only in 1990, when it sought $28 million instead of $22 million.
The seeds that Charles DeLisi planted found fertile soil in the U.S. Senate, but for very different reasons. Senator Pete Domenici was a staunch supporter of the national laboratories in New Mexico, although he had long believed that they produced far fewer long-term benefits for the local economy of his state than they should. He convened a panel to discuss the future of the national laboratories one Saturday morning, May 2, 1987, in the U.S. Capitol. Domenici's staff brought together an impressive group of people, including former Senator Barber Conable, now head of the World Bank; Donald Fredrickson, former director of NIH; Ed Zschau, former California congressman and successful entrepreneur; and the directors of several national laboratories. In the middle of the meeting, Domenici posed the question, “What happens if peace breaks out?” This question was of great concern because the vast bulk of work supported at the two laboratories in New Mexico was focused on nuclear weapons production and defense-related research and development. Domenici wanted to know how the immense research resources of the national laboratories could be better integrated into local economies. He also sought a new mission for the labs that did not depend on cold war rhetoric and that might move them into the growth areas of science, which clearly included biology. There was no way that Domenici could have foreseen the events of late 1989 and the transformation of Eastern Europe, but it did seem likely that sooner or later the Reagan defense spending juggernaut would lose steam.
Donald Fredrickson, then president of HHMI, suggested that the national laboratories might play a role in the human genome project. Jack McConnell, director of advanced technologies for Johnson & Johnson, took hold of the ideas discussed at the meeting and worked with
Domenici's staff to draft legislation that resulted in Senate Bill No. 1480, giving DOE the mandate to mount a genome project. (The bill also legislated issues relating to technology transfer in semiconductors and military research and included some patent policy directives.) By that time, Los Alamos was already in the thick of beginning its genome program, but this show of strong support from the Senate secured its future at a time of potential vulnerability.
It was clear from the outset that the DOE initiative would include national laboratory centers for genome research. Taking advantage of national laboratory resources was one of two principal justifications for DOE assuring the lead role in the project (the other being the areas of related research that were part of the DOE mission, particularly mutation detection). It was also clear that Los Alamos National Laboratory would be one such center, because it housed groups working on the cloning techniques, was the site for GenBank (the DNA sequence data base), and was in Domenici's home state (Domenici sat on both the authorization and appropriations committees for DOE and was also a ranking member of the budget committee). Most people assumed that Lawrence Livermore also would be designated a center, because it housed the other group engaged in large-scale cloning and DNA library-ordering efforts.
There were several problems, however. First, both Los Alamos and Livermore were weapons laboratories. This fact placed them in a different administrative category in DOE and made them somewhat less subject to direction by the Office of Energy Research, and the Office of Health and Environmental Research within it, because the laboratories answer to those parts of DOE concerned with national security policies. This administrative aspect also makes technology transfer issues (e.g., negotiation of patent agreements and personnel exchanges) more difficult because such transfers are covered by a different set of contracts with the University of California. (DOE-funded laboratories are operated by contractors. The University of California operates Los Alamos, Livermore, and Berkeley.) Finally, Livermore is sited on a flat, dry, windswept California plain whereas its sister laboratory, Lawrence Berkeley, nestles among the eucalyptus groves in the foothills overlooking beautiful San Francisco Bay. It would be much easier to recruit new researchers to a beautiful spot near one of the world 's most prestigious universities than to an isolated steppe known best for the birth of the fusion bomb.
Early in the DOE initiative, the various national laboratories were invited to submit proposals for review, and many did so, including both Berkeley and Livermore. The Berkeley proposal was initially rejected by DOE staff, which did not surprise those who had worked on the
proposal in Berkeley, who knew it was in serious trouble. But in September 1987, press reports of an American Association for the Advancement of Science meeting announced that Berkeley had been designated a center. This decision was apparently the result of discussions between DeLisi, Bostock, and other DOE and OMB higher-ups. Livermore, on the other hand, retained its project to map chromosome 19 and several other ongoing efforts, but was not given center status for several years.
The Berkeley center began with a first-year budget of roughly $3 million but without an approved proposal or a site visit. The designation was made with the expectation that Charles Cantor would become the center's director, but negotiations with him were still going on when the announcement was made. Clearly, Berkeley's designation as a center was a judgment—which required the agreement of upper management at DOE and OMB that establishing a new resource center at Berkeley was a better course of action than having to solve all the bureaucratic problems associated with supporting the ongoing work at Livermore.
DeLisi and Smith's anticipation of some arguments that would be made for and against the program was excellent. But what was missing from their thoughts proved to be just as important—competition with NIH and acceptance among molecular biologists and human geneticists. DeLisi remarked later that “moving unilaterally was not my preference, nor did I consider it optimal.” One source of great enthusiasm was Vincent DeVita, director of the National Cancer Institute, where DeLisi had worked before. The problem, from DeLisi's perspective, was that the National Institute of General Medical Sciences supported the fields of science most relevant to the genome project. DeLisi saw a hole, put his head down, and ran. He put the genome project on the public agenda, but not without getting tackled.
The well-known NIGMS response was that if it were to be done, they should do it, but it should not be done One of my choices was to use the NIH style of cautious consensus building. At times, perhaps most of the time, that is the best procedure; but in my judgment, this was not such a time. I made a deliberate decision to move vigorously forward with the best scientific advice we could muster (HERAC). I am quite willing to take the criticism, rational or not, that such movement provokes I would have been far more timid about subjecting myself to criticisms if I saw my future career path confined to government. (C. DeLisi, personal communication, March 1990)
Several technical elements are also remarkable by their absence from early consideration. The DOE proposals for the project contain very little discussion of genetic linkage mapping—the first and arguably the most important step in making the project useful to the research
community—and scant attention to the study of nonhuman organisms as either pilot projects or even scientifically important topics. One could argue that these areas were outside the range of biology research at DOE, but this is straining the argument because genetic linkage mapping is highly mathematical, requires systematic repetitive searches for DNA markers, and thus presents a great opportunity for just the sort of large group effort advocated by DOE. The omissions were undoubtedly in part also premised on a tacit understanding that the requisite work would of course get done. The omissions were, however, noted by biologists.
DOE strongly emphasized bacterial genetics immediately after World War II, and continued to support many groups working on nonhuman biology. Whatever the justifications, the neglect of genetic linkage mapping and nonhuman genetics drove a wedge between DOE and much of the biomedical research community. The enthusiasm driving the DOE human genome proposal proved sufficient to keep it going, but it was a rough ride.
On March 7, 1986, as many of the Santa Fe workshop participants were returning to their laboratories, Science magazine published an article by Renato Dulbecco (1986). Dulbecco, a Nobel laureate and president of the Salk Institute, was highly respected for his quiet demeanor and careful approach to science. The article thus brought wide attention and generated a wave of discussion in the laboratories of universities and research centers throughout the world. Dulbecco argued that the early emphasis in cancer had been on exogenous factors—viruses, chemical mutagens, and their mechanisms of action. Cancer research was at a turning point, he said, so “ if we wish to learn more about cancer, we must now concentrate on the cellular genome.” The nature of the connection to research specifically on cancer was imprecise; but scientists took note of the proposal, coming as it did from a scientist of Dulbecco's stature. Like Sinsheimer, Dulbecco came to the idea of the genome project deliberately thinking big. He had been preparing a review paper on the genetic approach to cancer. Although cancer clearly is not a purely genetic disease except in rare cases, it is equally clear that the steps leading to uncontrolled cellular growth involve changes in DNA.
Dulbecco further explained his rationale in a January 1987 interview. His argument for sequencing was that genetic techniques were among the most powerful in biology and extensive sequencing information would be a tool of immense utility in the study of cancer. He saw the sequence as a reference standard against which to measure the changes in DNA that take place in cancer. He argued that some such standard was needed because of human genetic variation. In other
species, such as the mouse, for which there are some 150 well-characterized, genetically homogeneous strains, controlled breeding is possible; in humans it is not. He saw the human gene sequence information as itself generating new biological hypotheses for experimental testing. Dulbecco believed that sequence information would be an intimate part of understanding some of the most fundamental problems of biology —cancer, chronic disease, evolution, and development. He noted the need for biology to encompass some collective enterprises of use to all, in addition to its extremely successful agenda of mounting small, narrowly focused inquiries. When asked what he thought about big science in biology, he smiled and said perhaps that would take care of itself over time.
THE SCIENTIFIC COMMUNITY RESPONDS
By the summer of 1986, the rumor networks of molecular biology were abuzz with talk of the DOE human genome proposal. News of the Santa Fe workshop had been disseminated by the participants and those in the mainstream of molecular biology were beginning to take the idea seriously. As is so often the case, Cold Spring Harbor Laboratory became the focal point of the debate. In June 1986 the lab sponsored a large meeting titled “The Molecular Biology of Homo sapiens.” The symposium brought together the giants of human genetics and molecular biology, with 123 speakers and an audience of 311 (Watson, 1986). The genome project was the hottest topic of discussion. Walter Bodmer, a British human geneticist familiar with both molecular methods and mathematical analysis, was the keynote speaker. Well known for his broad view of the field, he emphasized the importance of gene maps and the advantages of a DNA reference dictionary. Bodmer argued that the project was “enormously worthwhile, has no defense implications, and generates no case for competition between laboratories and nations. ” Moreover, it was better than big science in physics or space because it was “no good getting a man a third or a quarter of the way to Mars However, a quarter or a third of the total human genome sequence could already provide a most valuable yield of applications” (Bodmer, 1986a). He concluded his talk by urging a commitment to systematic mapping and sequencing, as “a revolutionary step forward.”
Victor McKusick, dean of human genetics and keeper of Mendelian Inheritance in Man, the gold standard human genetic disease data base, was the next speaker. He summarized the status of the gene map and finished his talk by urging a dedicated effort to genomic mapping and sequencing (McKusick, 1986). He argued that “complete mapping of the human genome and complete sequencing are one and the same thing,”
and explained the intricate interdependence of genetic linkage maps, physical maps, and DNA sequence data. He urged the audience to get on with the work and pointed to the future importance of managing the massive flood of data to come from human genetics.
The issue came to a head at an evening session not originally on the program. Paul Berg was unaware of the discussions in DOE and Santa Cruz but he had read Dulbecco's article and suggested to Watson that it might be useful to have an informal discussion of a genome sequencing effort at the Cold Spring Harbor meeting (P. Berg, personal communication, March 1990). Watson called Gilbert to find out if he would chair such a session with Berg (W. Gilbert and J. Watson, personal communication, June 1989). Berg thus arrived at the symposium to find that Watson had scheduled the session and that he and Gilbert would chair it.
The purpose of the session was to discuss the proposals for a genome project. Berg led off by trying to have the symposium participants discuss the scientific merits of mapping and sequencing and what technical approaches might make the effort feasible. Gilbert then got up and briefly described the Santa Cruz and Santa Fe meetings, following with the essentials covered in his earlier letter to Sinsheimer. He noted that the DNA sequence was accumulating at the rate of 2 million base pairs per year, in which case there would be no reference sequence for the genome for a thousand years. Gilbert thought that figure could be reduced to 100 years with no special effort but that a dedicated effort involving 30,000 person-years, on the scale of the space shuttle project, would produce a dramatic acceleration with enormous benefits. He began to write down numbers—large numbers that evoked great interest from the audience. The numbers on costs, however, provoked the most vigorous reaction. Gilbert estimated that, at $1 per base pair, there could be a reference sequence for approximately $3 billion dollars.
Berg called for discussion about whether it would be worthwhile to have the DNA sequence of the human genome. David Botstein rose to the podium and asked that scientists “not go forward under the flag of Asilomar,” meaning presumably that they should be aware of the political elements of their endeavor from the outset. He noted that if Lewis and Clark had followed a similar approach to mapping the American West, a millimeter at a time, they would still be somewhere in North Dakota; he also cautioned that scientists were amateur politicians and should be wary of making grand political proposals. Botstein voiced concern that researchers would become “indentured” to a mindless sequencing project and closed by pleading that molecular biologists “maybe accept the goal, but not give away our ability to decide what is im-
portant because we have decided on the space shuttle” (D. Botstein, remarks recorded by C. Thomas Caskey). This speech set loose the dammed-up energy of the assembly, and infectious applause spread through the audience. Several speakers followed, including Maxine Singer, Leonard Lerman, Peter Pearson, and Giorgio Bernardi, as well as other well-known molecular biologists. Many reiterated Botstein's sentiments; others supported the notion of an appealing proposal that could attract public support but were ambivalent about its impact on science. David Smith of DOE spoke midway through the session on the focused nature of the DOE proposal, but he was clearly on the defensive, even admitting in response to one question that perhaps DOE should not lead such an effort. His comments were largely lost in the wash. (Smith noted later, however, that many people in the audience had come forward privately to indicate their support.) Berg struggled from time to time to refocus the meeting on the technical and scientific aspects of the proposals, but his cause was lost. As the session ended, it was clear that molecular biologists were not enthused by the DOE human genome project, perceiving it as a misguided bureaucratic initiative and, more important, as a direct threat to their own research funding.
The dispute was covered by Roger Lewin of Science magazine, and his news articles were the first signals of the coming debate that encompassed many in science and government (Lewin, 1986a,b). The first step, however, was to shift the scene of the action from the quiet scientific mecca of Cold Spring Harbor to Washington, D.C.
A gala event held on the NIH campus, July 23, 1986, was sponsored by HHMI. Donald Fredrickson, former director of NIH and then president of HHMI, introduced and closed the meeting chaired by Walter Bodmer, which became something of a celebration for a redefined genome project. There were several brief presentations about the technologies and about what was going on in U.S. agencies and in other parts of the world. Mainly, however, it was a show of power—a battleship summit for molecular biology.
The HHMI interest can be traced along several paths. HHMI staff credit Ray Gesteland and Charles Scriver as the people principally responsible for interesting the institute in gene mapping. Gesteland was a student in the Watson laboratory in the mid-1960s and became a Hughes Investigator at the University of Utah. He suggested to George Cahill, HHMI vice president for scientific training and development, that HHMI support research on some ideas for systematic genetic mapping proposed by a group in Massachusetts, in particular, Ray White. Cahill recruited White to go to Utah, a feat accomplished in part by the attraction of the incredibly rich and detailed Mormon pedigrees kept by the university that might be useful for clinical genetic studies. (It
also helped that White liked to ski.) Thus, in November 1980, White began to construct a genetic linkage map.
Charles Scriver, a Canadian and human geneticist of international reputation, had served on the HHMI Medical Advisory Board from the late 1970s to the mid-1980s. (During this period the institute's annual funding for biomedical research increased from just over $10 million to more than $200 million, following the death of Howard Hughes in 1976.) Scriver was fascinated by the prospect of a human genome project, thinking primarily of the immense impact systematic mapping could have on clinical genetics and the patients he saw each day in his Montreal genetics clinic. Yet he called the decision to fund genetic linkage mapping, for which he became a champion on the Medical Advisory Board, “a close thing” on the part of HHMI.
Scriver became convinced, mainly through conversations with Francis Ruddle and later with White, that support of genetics data bases was essential, and he worked to persuade the other members of the Medical Advisory Board. As a result, HHMI began to support human gene mapping workshops held every two years, the Human Gene Mapping Library at its facility near Yale University, and the computerized version of McKusick's Mendelian Inheritance in Man data base (Pines, 1986).
A special meeting to discuss the HHMI human genetic resources effort was convened in Coconut Grove, Florida, on February 15, 1986. The meeting focused on how to manage the massive increase in information about genetic marker maps, locations determined by somatic cell genetics, and new DNA probes. It took place two weeks before the DOE meeting in Santa Fe on sequencing the genome but covered a completely different set of issues that only later took refuge under the umbrella of the human genome project. Watson met with Fredrickson on April 1 to indicate his strong support for an HHMI presence in genome research.
A July meeting at NIH was scheduled to gain information for a meeting of the Hughes trustees in August. In the wake of Cold Spring Harbor, the focus on data bases converged with the dispute about the DOE proposal to generate a tension surrounding the HHMI forum at NIH. The forum thus became another turning point in the debate, but the new direction was not entirely clear. Roger Lewin opened his report of the meeting in Science with the observation, “The drive to initiate a Big Science project to sequence the entire human genome is running out of steam” (Lewin, 1986c). Hood, whose views on sequencing were taken quite seriously, asserted that massive sequencing was premature and that the focus instead should be on improving the technologies. But the meeting was far from an outright rejection of the genome project—it merely rechanneled its energies.
Those attending the meeting agreed that the time was ripe to mount a special initiative in gene mapping. Bodmer could not contain himself when David Smith presented an outline of the DOE genome initiative, and Bodmer interjected that the proposal did not acknowledge the importance of genetic mapping. While Smith continued, Sydney Brenner, seated at the meeting table, conspicuously passed a note to Gilbert and Watson that was read by those around them: “This is a retreat. ” DOE was on hostile turf there in the NIH homeland, and the meeting was another event in a several-year period of tension between NIH and DOE regarding management strategies. Indeed, which agency would lead the effort became the dominant topic of discussion surrounding the genome project until well into 1988. The importance of the HHMI forum, however, lay in the fact that the question had imperceptibly shifted from whether to start a genome project to what it encompassed, how best to do it, and who should lead it.
HHMI was presumed to be a neutral party in the dispute, a philanthropy with international reach and a commitment to shared informational resources in human genetics. The institute was seen as a small partner to the federal agencies, rapidly responding when federal agencies could not, and filling in niches left vacant by the NIH behemoth. The shape of the HHMI program was becoming clear. A special presentation on the genome project was scheduled for the HHMI trustees in August. Maya Pines, a highly respected science writer, was commissioned to write a background piece on gene mapping and sequencing as background for deciding on continued support of basic genetics and a multiyear funding initiative for genomic data bases (Pines, 1986). The proposal was approved, with George Cahill (later, Max Cowan) and Diane Hinton of HHMI assigned principal administrative responsibility.
Before and after the Cold Spring Harbor meeting, James Watson was busy behind the scenes, trying to put together the pieces of a project that would be to his liking. Watson had been a power broker in molecular biology, since soon after he and Francis Crick discovered the double helical structure of DNA, and he had a well-deserved reputation for speaking his mind. Within the scientific community, many found him distasteful, but almost every molecular biologist learned to respect his biological intuition, his ability to frame important questions, and his talent for creating an environment in which bright people could contend with the best in the world. First at Harvard and later at Cold Spring Harbor Laboratory, he was an impresario of topnotch molecular biology, and he used his status as “the father of DNA” to get what he thought was needed.
He thought the nation needed a genome project—but not the one offered by DOE. Consequently, he enticed Berg and Gilbert to hold the
Cold Spring Harbor rump session on the genome project, and he began to agitate for involvement of the National Academy of Sciences (NAS), supporting its efforts to mount a study quickly. His position was quite simple, and he stated it publicly at the HHMI meeting: “I am for the project, although everyone I talk to at Cold Spring Harbor is against it.” He was following his intuition, again, against the stream.
CALLS FOR EVALUATION
NAS was a logical place to go for an assessment of research strategy. Devising a national plan for mapping and sequencing clearly involved substantial scientific and technical issues, for which the National Research Council (NRC) at the Academy was created. Furthermore, the Academy process ensured a systematic assessment often absent from open-ended debate, and its reports carried special weight in Congress and in the executive agencies. Convening a panel on the genome project was a considerable risk to genome supporters at the time, however, because sentiments were largely against the DOE proposal, which had dominated discussion to that point. An Academy report that equivocated or came out against a genome project would likely kill the idea for several years at least. A positive report would not guarantee its success, particularly if it required extra funding, but a negative report would be an almost insurmountable obstacle.
Plans to involve NAS were hatched around the time of the Cold Spring Harbor meeting in June 1986. On July 3, John Burris, executive director of the Academy's Board on Basic Biology, wrote a short proposal to fund a small group meeting to discuss the genome project in August. A discussion of the options was placed on the agenda of the board 's meeting in Woods Hole, Massachusetts, on August 5. The meeting included DeLisi, Wyngaarden, Watson, Cantor, Gilbert, Hood, White, Ruddle, Kingsbury, Ruth Kirschstein (director of NIGMS), Frank Press (president of the Academy), and several Academy staff. The board noted its support for physical mapping and expressly withheld its support from a massive sequencing program. It suggested that the NRC might wish to organize a study to formulate whether any special project made technical sense and if so, what its goals should be. A proposal was prepared, and Watson directed Burris to Michael Witunsky of the James S. McDonnell Foundation for funding of the study. Within a week, McDonnell had sent a check to the Academy. Bruce Alberts, known to be generally opposed to large targeted research efforts in biology because of an editorial he had written for Cell (Alberts, 1985), was selected to chair the project because he would be seen as neutral. Furthermore, his experience in writing a major textbook gave him the tools
to make sure a report was written quickly and well. The original hope was to complete the study in six months, or at least by mid-summer 1987.
Several people identified as skeptics were appointed to the panel, notably Botstein and Shirley Tilghman. The committee was also peppered with Nobel laureates: Gilbert, Watson, and Daniel Nathans from Johns Hopkins. Sydney Brenner was invited to represent the views of British mappers and sequencers, and John Tooze from the European Molecular Biology Organization was asked to speak for the Europeans as a group. An effort to secure a Japanese scientist was unsuccessful, although at least one was invited. Cantor, Hood, and Ruddle represented different technical backgrounds, and McKusick, Leon Rosenberg (dean of Yale Medical School), and Stuart Orkin (whose laboratory had done seminal work on chronic granulomatous disease and many other conditions) represented the field of human genetics.
Alberts and Burris hatched a strategy intended to build consensus slowly, if that proved possible. The first meeting on December 5, 1986, was intended to give the committee a sense of the general lay of the land, with presentations from the U.S. organizations with special genomerelated activities (NIH, DOE, NSF, HHMI, and OTA) followed by a survey of activities in Europe. The remainder of the day centered on deciding what a final report should cover and what information needed to be gathered for it. The committee elected to focus early meetings almost exclusively on technical background and to postpone discussion of policy options and funding until the technical stakes were clear. To that end, they brought in individuals with “hands-on ” experience in the fields under discussion. A January meeting had three technical sessions: (1) genetic linkage mapping, somatic cell hybrid mapping, and physical mapping; (2) large-scale sequencing; and (3) data bases related to protein and DNA structure.
In the meantime, the dynamics of the committee took an interesting turn. Walter Gilbert announced plans to form the Genome Corporation to map and sequence the genome as a private company and consequently he resigned from the NRC committee when he did so to avoid a conflict of interest. Gilbert had been a strong proponent of a fast-track genome project. Several committee members felt he was such a strong champion that it was becoming difficult to reach any consensus because his assertiveness elicited a backlash from several other members. His resignation paradoxically made it possible for those who were highly skeptical of the project to gracefully redefine it and shift to its support.
The next meeting, in March, began with a discussion of genetic linkage mapping in greater detail, with Donis-Keller, Gusella, and White all making presentations. The afternoon was an attempt to assess the
political context, with presentations from OTA and Wyngaarden. Gilbert 's slot had been filled by Maynard Olson, whose work in physical mapping and large DNA fragment cloning was in the thick of the technologies under discussion. Indeed, Bruce Alberts viewed “my major contribution to the NRC as the appointment of Maynard Olson to replace Gilbert ” (personal communication, October 12, 1990). Olson also brought a philosophical approach well suited to forging consensus. It was he who noted the importance of having enough genetic linkage markers to assemble a physical map, thus cementing the “marriage” of those twin objectives, and he also clearly articulated the distinctive feature of genome research that would set it apart from other genetic studies: a focus on projects of increasing scale (size of DNA to be handled or mapped, degree of map resolution, speed, cost, accuracy, or other factors). He argued that support should be given to those projects that promised to increase these scale factors by three-to tenfold. The second day of the March meeting opened with Cahill describing the HHMI interests. The rest of the day was devoted to trying to sort through the first drafts of several technical background chapters and beginning the process of deciding what would be said about policy. It became clear that the skeptics had been converted by the project's redefinition.
In effect, the NRC panel was a microcosm of biomedical research. Its deliberations for the first time systematically assessed the arguments for and against a dedicated genome project and surveyed the various technical components necessary to bring it together. Alberts called the NRC committee “the most fun of any committee I have worked on” because of the talented people on it, the rapid learning process it entailed, the uncertainty of its outcome, and its direct impact on policy (personal communication, August 18, 1988). The NRC report succeeded to a remarkable degree in setting a scientific agenda —the critical missing element from 1986 to early 1988.
The report, however, had one critical weakness—its recommendations about how the project should be organized. The scientists on the committee made little attempt to survey what the agencies were doing, and there was a great deal of activity going on in NIH, in DOE, and in Congress. The committee commissioned only one paper, by Eric Juengst and Albert Jonsen, on the ethical implications of the research. The committee members had some informal contacts, principally with NIH, but there was no systematic attempt to gather information critical to making a policy recommendation. The federal bureaucracies are highly complex, and the political process is unpredictable; having an impact requires extensive knowledge about the backgrounds of large bureaucracies, jurisdictional boundaries in Congress, and the histories of pivotal figures. Such knowledge is necessary to make credible recommen-
dations, or at least formulate options, that do not seem naive or counterproductive. One of the reviewers who received the penultimate draft of the NRC report had great familiarity with the organization of science agencies and was appalled by the organizational options. This response and others provoked a rewrite of the section and many last-minute changes. It is clear from subsequent interviews that the committee did not have enough data on which to base a recommendation but felt it had to do so anyway to fulfill its responsibility. There had been no meeting to discuss this topic, and the phone calls that were conducted in its stead did not permit a solid consensus to form.
The report was released recommending that there be one lead agency but failing to specify which one (NRC, 1988). The option preferred by the NRC required Congress to decide whether NIH or DOE should lead the genome project. Clearly, the preferable organizational structure would have been to develop a program de novo from within one agency, but this was historically not how it happened. It was too late to make the program fit the ideal. The NRC committee ignored the fact that by 1988 each agency had multimillion-dollar budgets, advisory committees, planning documents, and, just as important, expectant constituencies and congressional patrons. In addition, it was not clear what was meant by a “lead agency.” If it meant that one agency should have a formal mandate, with funding coming from several sources, then it would have been politically feasible but effectively meaningless. How would NIH as “lead” agency decide how DOE should spend its funds? If it meant all the funding should come from one place, then it required a dismantling of either the NIH or the DOE programs, a politically hopeless task, particularly when congressional interests were taken into account. The main reason for fumbling the administrative recommendation was ambivalence about both NIH and DOE. NRC committee members were disappointed by what they saw as a faint-hearted commitment from Wyngaarden and near hostility from Kirschstein. Yet DOE undermined its credibility by asserting it wanted to do the project to detect mutations and monitor human exposure to radiation and environmental toxins. This was seen as asking to buy a sledgehammer to put in thumbtacks. There was also suspicion of DOE peer review, fomented by the process by which the genome center had been established at Lawrence Berkeley Laboratory.
Congress was following the debate and independently taking steps to gather the information needed to make policy choices. McKusick presented the arguments for mapping the human genome at a 1986 meeting of a biotechnology advisory panel at OTA, at the invitation of OTA staff Gary Ellis and Kathi Hanna. When the news from Cold Spring Harbor was reported, it attracted the attention of several Hill staff. I
wrote a memo in mid-July to OTA upper management urging that the agency undertake a study because the issues were highly technical and complex and because DOE and NIH were on a collision course. A short discussion with Lesley Russell, science advisor to John Dingell, chairman of the Energy and Commerce Committee, led to a request for a proposal from Dingell to which OTA responded. By pure happenstance, the OTA and the NAS projects were approved in the same hour on September 30, 1986.
With separate governmental departments (in the form of NIH and DOE) vying for position, there were only two places to resolve the issue —in the White House or in Congress. Science policy in the Reagan administration was largely dictated by the budget process. Until the final year of the administration the ritual was to propose unrealistically low NIH budgets, leaving room for increased funding in other areas. NIH was one of the most popular executive agencies in Congress because of its medical research mission and a reputation for being well run and “clean,” if a bit stodgy and paranoid. For the first seven years under Reagan, Congress increased the NIH budget far above the requested amount, a course of action that effectively gave power over the NIH budget, especially new budget items, to the appropriations committees in the House and Senate. This process contrasted starkly with that of the DOE budget, for which the President's request was much more likely to be cut than augmented. For DOE, the “inside game” that DeLisi played, going through formal budget review in DOE and OMB, was much more important than for NIH. NIH's budget, on the other hand, was an “outside game,” played in the public arena of congressional politics—hearings, press reports, and Capitol Hill meetings.
Tracking the policy aspects of the genome project was left to OTA. Patricia Hoben kept abreast of technical developments and wrote a clear, well-illustrated introduction to the technologies. Jacqueline Courteau gathered information about data bases and repositories and sought information about foreign genome plans. Other staff collected information about what U.S. agencies and research institutions were doing in the field of genome research. Papers on Japan and Europe were commissioned, as was this history of the biology and of analogous periods of development in physics. OTA commissioned two papers to assess the ethical implications of the project, and several papers were commissioned and dozens of letters sent to gain technical background (a process unnecessary for the NRC committee). The OTA process was relatively open, with four panel meetings and workshops, each of which was attended by more than 100 people. The drafts were circulated to roughly 200 people, and served as an informal communication link among those following genome activities.
In contrast to the NRC report, which lays out a clear scientific strategy, OTA had virtually no role in setting the scientific agenda because it was not positioned to render a scientific judgment. Yet consensus on the strategy was a necessary precondition for the political decision about how to fund a program; thus, the NRC and OTA reports complemented one another. NRC performed the most important function, namely, articulating a scientific program that captured the insights of those who saw the need for collective resources and focused efforts. OTA systematically gathered information about government programs and acted as a well-informed but neutral observer, expert in science policy but not about the science itself. It fell to OTA to propose the options for coordinating the NIH and DOE efforts.
The OTA report was used mainly in two hearings on the NIH and DOE programs, held in April and June 1988. Senator Lawton Chiles (D-Fla.) had introduced a bill, when the Domenici legislation met with resistance, to create an interagency task force to tackle the genome. The new bill was passed by the Senate by a margin of 88-1 and sent to the House. NIH and DOE were faced with two options: conspicuous cooperation or a strong likelihood of legislation mandating a specific framework not entirely to their liking. They opted to sign a memorandum of understanding in hopes of staving off House action on the bill, having reached a tacit agreement with staff from both the Science, Space, and Technology and the Energy and Commerce committees. The content of the memorandum was immaterial, but the process was important for each agency in facing the reality that they would both have genome programs for the foreseeable future and that Congress would be quite sensitive to interagency bickering.
Until the release of the NRC and OTA reports, and, indeed, for a few months afterward, staff from both agencies appeared to believe that Congress would somehow designate their agency the lead organization. They expressed disappointment that NRC and OTA had not made a tough call, but in fact there was no call to make. The existence of twin genome programs was in the cards as soon as DOE pushed its first formal authorization and appropriation through Congress. The only strategy that could have prevented the birth of two programs was either the death of both in a fit of internecine warfare or a preemptive strike by NIH in spring or summer 1987. DOE could not have stopped NIH from mounting an effort once leaders at NIH concurred and the NRC committee reached a consensus, and NIH could have stopped DOE only before it cleared its authorization hurdles for the 1988 budget.
THE NATIONAL INSTITUTES OF HEALTH AND CONGRESS RESPOND
The critical figure in the effort to secure funding for an NIH genome program was James Wyngaarden. From Duke University Medical School, where he had been chairman of its largest department, medicine, for 15 years, Wyngaarden became the director of NIH after being nominated by President Reagan in spring 1982. He was highly respected as a clinician and human geneticist, and he accepted the job with some reluctance, stating this openly. In his confirmation hearings before the Senate, he noted, “I did not actively seek the post my acceptance of that honor is out of a sense of obligation based on an awareness of the vital role of NIH in biomedical research ” and went on to emphasize the importance of basic biomedical research as the best long-term strategy to solve the nation's health needs (U.S. Congress, Senate, 1982). In a 1988 interview, he said he accepted the position because of considerable worry about what might happen to NIH if a caretaker were nominated instead of a person thoroughly familiar with the biomedical research process.
Wyngaarden first heard about the genome project in London at a meeting of the European Medical Research Council in June 1986, when someone asked him what he thought about a DOE plan to spend $3.5 billion sequencing the human genome. Shocked, Wyngaarden said this idea seemed to him “like the National Bureau of Standards proposing to build the B-2 bomber.” At this same time, Ruth Kirschstein, director of NIGMS, began to get feedback from the March workshop in Santa Fe.
DeLisi of DOE had invited an NIH representative to the Santa Fe meeting in March, but the invitation got lost in the deluge of mail that enters the NIH director's office. He had sent background materials about the meeting afterward, as preparation for a meeting of himself, Wyngaarden, and Norman Anderson. When he returned from London, Wyngaarden asked Ruth Kirschstein to bring together an NIH panel to decide what might be done in response to the DOE foray. Kirschstein summarized the June 27 meeting of that group in a memo dated July 2 to Wyngaarden, noting that “first and foremost, while it is clear that the Department of Energy has taken, and will continue to have, the lead role in this endeavor, the NIH must and should play an important part.” The group recommended that Wyngaarden focus the upcoming Director's Advisory Committee meeting in October on the genome project, in time to make plans for the fiscal year 1988 budget. They also noted the need for increased support of GenBank, the data base that stored DNA sequence information.
The October 16-17, 1986, meeting of the Director's Advisory Committee had another all-star cast that included Nobel lights and Nobel
aspirants. The meeting was more structured than the HHMI forum of several months before, and the policy issues were becoming more evident. The meeting's main conclusions were that (1) NIH should eschew “big science” or a crash program, (2) the study of nonhuman organisms was just as important as the study of humans, (3) it would soon be feasible to sequence the human genome, and (4) information handling was already a problem. An NIH working group was appointed after this meeting, to be chaired by Wyngaarden and including Kirschstein, Duane Alexander (director of the National Institute of Child Health and Human Development), Betty Pickett (director of the Division of Research Resources), Donald Lindberg (director of the National Library of Medicine), and Jay Moskowitz (Program Planning and Evaluation), with George Palade (Nobel laureate from Yale) as the lone outsider. Rachel Levinson was named executive secretary. The working group met in November and December and produced recommendations for enhanced support of data bases as well as two new research program announcements.
Wyngaarden's early concern was to ensure that NIH had a major role in any large genome program that went forward but without making any long-term commitments. He was in favor of the concept of the genome project “from the very start,” but resisted the impetus to go too far out on a limb when there was so much dissension among NIH-supported researchers. In a revealing analogy, he likened his position on the genome project to Lincoln 's waiting for success at Antietam to announce the Emancipation Proclamation, so as not to lose Union support from Europe. His second analogy was to Roosevelt's delay in pushing the Lend-Lease Act until public sentiment supported the course he had already chosen. However, Wyngaarden did support the genome project, and strongly, where it counted the most —in the appropriations process.
In his summary statement to the House and Senate appropriation committees for fiscal year 1988 (in February and March 1987), Wyngaarden gave gene mapping a high profile. He mentioned NIH's centennial and the urgency of acquired immune deficiency syndrome (AIDS) research, and then flagged the genome project. A straight reading of his text would suggest it had second priority to AIDS. The NIH appropriation for genome research did not require a special authorization, as such research clearly fell within the bounds of NIH's biomedical research mission. Unless someone in Congress objected, much could be done through appropriations alone. Fiscal year 1988 was one of the years in which the NIH budget dance was played by ignoring the administration proposals, and Congressman David Obey said as much. Because the NIH director is part of the administration, Wyngaarden had to toe
the administration line in supporting budget requests to Congress. Any testimony before legislative or appropriations committees is reviewed by officials in the Department of Health and Human Services and in the Office of Management and Budget. But the bureaucracy cannot interfere with Congress's authority to ask whatever questions it likes, and interfering with honest answers is a violation of federal whistleblowing laws. The appropriations committees thus had a simple way to ascertain NIH's true priorities, as opposed to those in the administration's fictitious request. Each year, they merely asked the NIH director what he would do with sums of money in addition to those requested, in $100 million increments.
In the House appropriations hearing, Congressman David Obey, who had read an article in the Washington Post about the genome project, tapped into what had become the major issue related to genome research by asking several questions about gene mapping. He wanted to know why DOE was proposing to lead such a project. Wyngaarden replied that the agency had legitimate interests in detecting mutations but that NIH was outspending DOE by a hundred to one in the relevant fields and so NIH should—He was about to finish his policy recommendation when Obey interrupted, asking for further clarification of DOE's interest. Wyngaarden never finished his recommendation but said to Nature magazine several weeks later that he thought it was presumptuous of DOE to claim leadership when it was spending less than $10 million a year in the area (Palca, 1987). Again he was not pressed on what NIH should do about it. Leslie Roberts of Science magazine opened the “Research News” section of the September 18 issue with a depiction of interagency squabbling (Roberts, 1987) that captured the confused positions of scientists and administrators during this formative period.
David Kingsbury of NSF, who emerged as one of the mediating forces, attempted to channel the conflict, first through the Biotechnology Science Coordinating Committee (formed principally by the White House Office of Science and Technology Policy to deal with interagency disagreements over the release of genetically altered organisms into the environment) and then the Domestic Policy Council (a cabinet-level group). Kingsbury's decision to mediate meant that NSF had to stay out of the competition, and NSF's policy position was quite clear for several years—it had no genome program per se, although its support for instrumentation and nonhuman biology was directly relevant. (This was clearly a position crafted in that bureaucratic netherworld where truth wears gray.)
Kingsbury's political situation deteriorated quickly, however, when he was implicated in a conflict-of-interest investigation related to his
financial connections with Porton, a company that grew out of the chemical warfare establishment in England and that had aspirations in biotechnology. NSF thus was removed from contention for several years, reentering only in 1989 with its instrumentation centers and proposals for a plant genome program focused on Arabidopsis thaliana, a weed with a conveniently small genome.
Interagency disagreements at the strategic policy level had little material impact on those individuals who were administering grants and sponsoring activities in NIH and DOE, or among those researchers receiving grants from the agencies. If anything, there were special efforts to work jointly because of the intense public scrutiny, at least among readers of Science and Nature. Indeed, the degree of disruptive battling between NIH and DOE was less than for other high-stakes turf disputes within the Public Health Service—for example, among the National Institute on Aging, the National Institute of Neurological and Communicative Disorders and Stroke, and the National Institute of Mental Health regarding Alzheimer' s disease funding in the late 1970s to mid-1980s; or among the National Cancer Institute, the National Institute of Allergy and Infectious Diseases, and the Centers for Disease Control over AIDS research beginning in the mid-1980s.
Squabbling over the genome nonetheless reached directly into Congress in the form of legislation. Senator Pete Domenici introduced S. 1480 early in 1987, the bill crafted by Jack McConnell and Domenici's staff to promote technology transfer from DOE-funded national laboratories. A section was devoted to the genome project. Domenici's bill gave a mandate to the secretary of energy to map the human genome and directed the secretary to establish and head a consortium dedicated to this purpose. Coordination of research from other agencies was to occur through a National Policy Board on the Human Genome, chaired by the secretary and including the NIH director, the NSF director, the secretary of agriculture, and other officials. Domenici attempted to add the bill as an amendment to the trade bill under active consideration in spring 1987, and his staff began to call other Senate and House committees with jurisdiction. Key to this effort were Senator Chiles, chairman of the NIH appropriations subcommittee and of the entire Budget Committee and generally accepting of NIH initiatives in biotechnology, and Senator Edward Kennedy, chairman of the NIH authorization committee.
By an irony typical of congressional politics, the genome project was linked to orange groves in Florida. Chiles's interest in biotechnology stemmed from a 1982 or 1983 meeting with one of his constituents, Francis Aloysius Wood, dean of the School of Agriculture at the University of Florida. Wood caught the senator's attention by describing how gene
manipulation could move the frost belt 60 miles north, which meant more land could be devoted to cultivating a large crop plant of immense importance to Florida. Wood had found a graphic way to explain how the use of so-called ice-minus bacteria might delete some of the genes that cause ice crystals to form on fruit, thus lowering the temperature at which the fruit sustained damage. Changing the temperature at which fruit becomes damaged would reduce the annual worries of Florida's orange growers and effectively expand the territory acceptable for planting.
When the Senate became Democratic in the 1986 election, Chiles became the chairman of the appropriations subcommittee for NIH. His interests at NIH focused on biotechnology policy. The genome project became linked to biotechnology through Domenici's bill and through the language used by scientists to justify the project. When Domenici's bill first came to his attention, Chiles spoke with his legislative aide Rand Snell in a brief conversation en route from the Senate floor after a vote. The DOE element didn't seem quite right; it didn't seem fair to NIH. Later, Patricia Hoben from OTA happened to be meeting with Snell from Senator Chiles 's staff on another matter, the competitiveness of U.S. biotechnology. When she heard about the proposal, she asked whether there had been outside consultation with university researchers. She suggested that Snell call Bruce Alberts, chairman of the NRC study committee. Alberts was noncommittal but indicated that there was, indeed, ambivalence about DOE leadership and a strong feeling among some on the committee that NIH should be the lead agency. Chiles refused to bite on Domenici 's bill and thus began a long process of negotiation that led to a Chiles-Kennedy-Domenici bill, S. 1966, which included a genome project provision modified from Domenici's to give NIH and DOE joint leadership.
Kennedy's staff also called their contacts. During the week, a storm of protest calls came into the offices of Domenici, Chiles, and Kennedy, and the idea of passing the Domenici bill as an amendment to the trade bill was dropped. The Industrial Biotechnology Association, a trade association for the larger biotechnology companies, began a survey of its members in response to the Domenici bill. The survey showed a strong consensus in favor of funding a genome project, but only under the aegis of NIH.
Domenici held a workshop in Santa Fe on August 31, 1987, to determine what should be done about the genome project. It was Charles DeLisi 's last day on the job at DOE: he was leaving to head a department of mathematical biology at Mt. Sinai Medical Center in New York City. Domenici urged his strong support for a DOE role in genome research. Norman Anderson pulled out all the stops in a moment of zeal:
I think so far as the man in the street is concerned to say that here is the possibility at one shot of finding the cause of some 2,500 human diseases is really stunning. A century from now, as history books are written, the big projects that were important in this century are the genome project, and after it possibly space and then the atomic bomb (the order of those, I don't know). But the man who first proposes to do the genome project in the United States Congress is in history. (U.S. Congress, Senate, 1987b)
It was a good way to get Domenici's attention.
Back in Washington at hearings on Domenici's bill on September 17, 1987, Wyngaarden articulated his desire for what one might paraphrase “the mission and the money, but not the management.” This came during an interchange with Domenici in the question-and-answer session following Wyngaarden's testimony:
Domenici: If you were assured that it was not the intention of the legislation to in any way denigrate or detract from your ongoing activities, would you recommend that the United States of America have a policy of mapping the human genome as expeditiously as possible?
Wyngaarden: Yes, sir. Unequivocally, yes.
Domenici (several exchanges later): If Congress wants to do it, how do we do it? Just give the NIH more money under their existing program and give DOE some more money
Wyngaarden: I think that is a very good way to do it.
Domenici: And would it get done?
Domenici: Without any changes in the law?
Wyngaarden: I think so.
James Decker, representing DOE, concurred with Wyngaarden.
Domenici went on: I love you both and I think you are great. But I absolutely do not believe you. I believe it would get done. But I am quite sure that it would not get done in the most expeditious manner, because I do not think you would be charged with doing that. I do not think you would send up any requests of a priority nature with reference to it, because you do not have enough money to do what you are doing. And if you tried to send up the request, it would be thrown in the waste basket at OMB (U.S. Congress, Senate, 1987a)
Wyngaarden and Domenici locked horns for several minutes more
over definitions of what the other had meant, but it was clear that the basic issue was the mutual distrust between the legislative and executive branches of government. Congress, in the person of Domenici, did not trust the agencies to act quickly, and the agencies, principally in the person of Wyngaarden, did not want to have Congress tying internal priority-setting and budgeting processes in knots. Neither side could score a decisive win, and the policy process in this case was typical in that it unfolded over many months of thrusts and parries.
Within NIH, and among the power brokers in molecular biology, there was a division of opinion about the NIH role. Kirschstein articulated one position strongly. She was particularly concerned that the genome project not become a political juggernaut that could endanger small-group pursuit of basic genetic knowledge, for which NIGMS was the largest source of funding in the world. NIGMS issued two new announcements for grants in mapping and computation in May 1987, to demonstrate a special willingness to support such work, but did not formally commit dedicated funds for this purpose. These two grant announcements were the main product of the genome working group set up after the NIH Directors Advisory Committee the previous October.
Earlier, Kirschstein had canvassed all the NIH institutes to find out how much was being spent on grants that involved gene mapping or DNA sequencing and had produced a figure of $313 million in fiscal year 1987, of which $90 million was for work on humans. The grant officers of her own institute spent several days poring over their portfolios to come up with the figures, revealing the energy with which Kirschstein worked to support her position that NIH was already acting aggressively. Kirschstein argued that the NIGMS announcements were “not exactly business as usual, but not highly targeted either. ” Rachel Levinson, a member of Wyngaarden's staff who worked on the genome policies, agreed with this viewpoint and maintained that there was no need “for a concerted effort because it is not new. Every institute has work related to mapping and sequencing ” (Roberts, 1987). This position was no doubt intended to assuage fears of a major shift in policy that could threaten investigator-initiated research, but it backfired. The message heard by the opinion leaders in molecular biology, including many Kirschstein supporters, was that NIH thought it was doing all it needed to do.
Many scientists, however, saw this contention as failure to appreciate the collective and dedicated efforts needed to finish maps and develop new technologies. NIH's neglect of dedicated genetic linkage mapping and DNA sequencing instrumentation was cited as symptomatic of a deficiency in NIH's resource planning. In interviews with dozens of molecular biologists, including Berg, David Baltimore, Botstein, Watson,
Gilbert, Hood, and others, NIH's official position was decried for missing the point of the genome project—to fill a need for concerted and focused efforts to create common resources. Whether Kirschstein took this position because she saw it as essential, or whether she was attempting to placate those worried about the genome project's impact on other basic genetic research and defend NIH against incursions from DOE, cannot be determined now, but it set the NIH genome program on a course that made separation from NIGMS inevitable.
NIH's first dedicated funding for genome research—$17.2 million—came in December 1987, when President Reagan signed the 1988 appropriations law (two months into the fiscal year). To determine how to allocate the funds, Wyngaarden convened an ad hoc advisory committee on February 29 and March 1, 1988, in Reston, Virginia, only a few weeks after release of the NRC study, which recommended a vigorous $200 million annual genome effort. The committee was chaired by David Baltimore, a Nobel laureate who had written against the Big Science genome project approach (Baltimore, 1987). The ad hoc committee made recommendations that closely followed those contained in the NRC report. Watson urged that an esteemed scientist be appointed director of the NIH genome efforts. Later, he stated, “I did not realize that I could be perceived as arguing for my own subsequent appointment” (Watson, 1990).
Whether Watson's arguments contributed to his subsequent appointment is unclear; nevertheless Wyngaarden named Watson director of a new genome office in October 1988. Watson then hired Elke Jordan, former associate director under Kirschstein and erstwhile resident of Matthew Meselson 's laboratory when it was down the hall from Watson's in the 1960s. He also hired Mark Guyer, a bacterial geneticist who had worked at Genex Corporation before joining the NIGMS staff. Several people who were interviewed for this history thought Watson 's ascension to the position of genome director was due to sexism or power seeking. Watson did, indeed, want power—enough to get the genome project moving. It seems clear, however, that he did not initially contemplate running the project himself.
One pointed exchange took place between Kirschstein and Watson in August 1987 at an OTA workshop on the costs of the genome project. The meeting was chaired by Berg and was intended to ferret out strategies for genome mapping and sequencing by forcing a discussion of budget items that would be of concern to Congress. At one point, the discussion digressed to management issues. Kirschstein and Watson clashed over the need for assertive planning by NIH. Watson wanted powerful direction; Kirschstein argued for the wisdom of the investigator-initiated grant mechanism. Watson was interviewed after the
OTA meeting and asked if he were willing to be the “czar” that he thought necessary. He said no, saying “I can't think of a job I'd like less” (Roberts, 1987). He later called several other people to find someone willing to take the job, but none of those who were able were also willing.
In his actions regarding the genome project, Wyngaarden made a tough call between the advice provided by Watson and others in his camp and the advice offered by Kirschstein and others in hers. In the end, the NIH director decided a question of policy directly. To her credit, Kirschstein supported the project publicly even after it was moved away from NIGMS, calling it “an important part of what the Public Health Service is all about in the next century,” although allowing that it had politically “taken on a life of its own” (Jenks, 1989). Certainly, no one could doubt that.
THE PROJECT IS FUNDED
Wyngaarden was successful on another front—obtaining a genome budget at NIH. In his replies to the House Appropriations Committee for fiscal year 1988, he asked for $30 million in genome research funds in his fifth increment of $100 million above the administration request, and another $15 million in the eleventh increment (the penultimate of 12 such increments) (U.S. Congress, House, 1987b). After Wyngaarden testified in early spring 1987, David Baltimore and Watson met to brief members and staff of the House and Senate appropriations committees. They were invited to speak informally as part of a series of occasional meetings put together by Bradie Metheny on behalf of the Delegation for Basic Biomedical Research. Baltimore and Watson met briefly just before the session on May 1 to go over their remarks. The meeting included Congressmen William Natcher and Silvio Conte, chairman and ranking Republican of the NIH appropriations subcommittee, and also Congressman Joseph Early, a subcommittee member and staunch NIH supporter of many years. Senator Lowell Weicker, then chairman of the Senate appropriations subcommittee for NIH, was also present. The principal aim of the meeting was to promote funding for AIDS research. Watson, however, also supported adding $30 million to NIH's budget for genome research (Watson, 1990).
The House responded to Wyngaarden by appropriating $30 million for genome research. The Senate was less enthusiastic, inserting only $6 million. Maureen Byrnes, staff to Senator Weicker, recalled that he was not as enthusiastic about the genome project as the House delegation; other senators, such as Tom Harkin, were more enthusiastic but were also more junior and thus were unable to influence decisions as strongly. The House and Senate bills went to a conference committee for
resolution of differences. The usual response in such cases was to split the difference unless one house could convince the other. In this case, the arithmetic mean of $18 million emerged from the House-Senate conference, passed, and became law. Because of Gramm-Rudman-Hollings rescissions, NIH had a final appropriation of $17.2 million for genome research at NIGMS that year. In private conversations, NIGMS staff estimated that $5 million of this was diverted from existing funds and the rest was “new” money.
An additional $3.85 million found its way to NIH's coffers in the 1988 to fund a National Center for Biotechnology Information. The regents of the National Library of Medicine (NLM) identified molecular biology as an important area in which the NLM 's emerging expertise in electronic data basing would become increasingly important. An outside support organization, the Friends of the National Library of Medicine, took up the cause and drafted a bill for Congressman Claude Pepper to introduce. Pepper was an old friend of Fran Howard, who was a long-time NLM supporter—later an employee—as well as the sister of the late Hubert Humphrey and the widow of a prominent academic physician at Johns Hopkins. The bill was to establish an information management center to support biomedical research and biotechnology efforts in the United States, with annual budget authority rising to $10 million. Pepper held a moving hearing on the bill on March 6, 1987, but the hearing was under the auspices of his subcommittee on the Select Committee on Aging, which has no legislative authority. Pepper called members of the legislative committees, but the news did not reach staff of the Energy and Commerce Committee, which had legislative jurisdiction. (Staff for those committees learned of the hearing through OTA and NIH.) NIH, unlike many other agencies, is authorized for three-year intervals as a rule, and 1987 was not one of the years when such a bill was in Congress. There was thus no logical vehicle to which the NLM bill could be attached, and so it stood as a freestanding act. These factors delayed action on it, and eventually it was folded into the NIH authorization passed a year and a half later. However, the appropriations committee acted before then, appropriating $3.85 million for fiscal year 1988 with Pepper's full support, with the understanding that it was to be spent for the purposes specified in the languishing Pepper bill.
NIH appropriations for fiscal year 1989 were more or less routine, with NIGMS requesting $28 million for genome research in this, the final year of the Reagan administration. Congress and the President had agreed on a two-year budget plan the previous fall, in the wake of the October 1987 stock market crash, and the President's budget request held to this agreement. In fact, this was the one year under Reagan when the NIH request was taken seriously by the appropriations com-
mittees, and the requested amount was granted. There was one added feature by then, in that the NRC genome report had been made available for the appropriations hearing cycle. Representative Natcher led off his questioning of Wyngaarden by asking how the $28 million budget request from NIH fit with the $200 million recommended by the NRC committee. This gave Wyngaarden an opening to explain that there would be higher budget requests in future years.
Natcher also asked which agency should assume the leadership of the project. Wyngaarden was unequivocal and direct in his answer: “I think NIH is the appropriate agency” (U.S. Congress, House, 1988). The congressional hearing took place within weeks of the NIH ad hoc advisory committee meeting in Reston, Virginia. Wyngaarden's strong support for NIH leadership on the human genome project in Reston and now before the appropriations committee were the clear statements of purpose that had been eagerly awaited by opinion leaders in molecular biology. The NIH program began to pick up steam.
What led to Wyngaarden's assertion of leadership is instructive. He was beset by disagreement about the proper style for promoting genome research, with Kirschstein and Watson articulating incompatible options. He had to choose. To aid his decision, he scheduled the ad hoc planning meeting in Reston, and he had already met with Watson and Baltimore on December 17, 1987, to discuss AIDS research and the human genome. Watson expressed his views about how NIH had missed the boat on the genome project and was clear in his opposition to Kirschstein's approach. With the backing of an NRC report presenting a coherent approach and advocating a focused effort with a $200 million annual budget, Wyngaarden chose the high road.
NIH's appropriations for 1990 involved several complications. NIH forwarded a budget request to the Department of Health and Human Services that went on to OMB, with a final request of $62 million as the result. When the President's budget request came out, it asked for $100 million for genome research at NIH. (The $62 million apparently had been increased to $100 million by dividing up some excess monies left from the removal of other programs during OMB review [John Barry, free-lance writer, personal communication, May 1990].) The increase surprised NIH and signaled support for the NIH genome project high in OMB or elsewhere in the White House; but in the end it did not matter, as the appropriations committee staff used the initial request level from NIH as the basis for their deliberations. The final 1990 appropriation was $59.5 million after last-minute adjustments.
Now that the genome budget had become sufficiently large, Wyngaarden discussed with the appropriations committee staff the need to create a separate administrative center for the project for the 1990 fis-
cal year. The House agreed to allocate the 1990 budget request to a new center that the department would create by administrative fiat. The Senate, however, was looking for ways to fund new initiatives in health and human services, and, as a result, the 1990 NIH genome budget was subject to last-minute negotiations. One eleventh-hour proposal reduced the genome budget from $62 million to $50 million, with the $12 million added to funds taken from elsewhere in NIH to fund programs for the homeless. In the end, however, the Senate agreed to roughly the same budget figure as the House but left the funds in NIGMS. In conference, the report followed the House, creating a new center.
This process illustrates the twofold vulnerability of new programs at NIH. Activities that show a rapid percentage rise in funding from year to year are highlighted by the procedures used by appropriations staff to track budgets, and NIH takes up an increasingly high fraction of the discretionary funding in the Department of Health and Human Services. The department disburses more than $300 billion in funds each year, but the vast bulk goes to entitlement programs—Social Security, Medicare, and Medicaid—that are not subject to congressional appropriations or direct agency control. This makes NIH's $8 billion budget loom large as a potential source of funding “offsets” to support new initiatives for health and social services.
The budget history also illustrates the illusory dichotomy between “new” and “existing” monies. One of the most divisive debates within the biomedical research community has been miscast in these terms, with supporters of investigator-initiated small grants contending that the genome project was carved from their province, while staunch defenders of the genome project argue that the political attractiveness of the project has increased the size of the pie without in any way cutting into other efforts.
There is scant evidence for pure versions of either view. Would the $87 million 1990 genome budgets at NIH and DOE have been appropriated elsewhere for biomedical research if there had been no genome proposal? Only those who actually made the decisions for the appropriations committees could answer such a question, but they simply did not make the decisions on these terms—nor should they. The NRC report and the rhetoric supporting the genome initiative leaned heavily on the principle that the new initiative should come from “new” funds, but this kind of money did not exist. The genome budget was not given a great deal of attention in the appropriations process, and it was merely one of thousands of such decisions; in interviews with appropriations committee staff, it was clear that this was not a highly contentious part of the budget deliberations and did not generate enough controversy to leave strong memory traces.
There are hints that substantial pressures are being felt to reduce the overall NIH budget to leave more room for other health and social service spending, as in the case of the homeless funding. If this were the case, it would lend support to the contention that, without the dedicated genome line item and the new NIH administrative center, there would have been a lower overall NIH budget, and that the genome project provided a new justification for a budget increment. But the budget history makes clear that it was largely left to NIH, specifically Wyngaarden, to indicate internal NIH priorities, and the funds given to the genome project would probably have gone to NIH in any case. The argument thus is not truly about new or existing funding but about NIH priorities. Initial funding in 1988 was just under 2 percent of a budget supplement appropriated to NIH beyond the presidential request (amounting to 0.2 percent of the overall budget). Was Wyngaarden right in his decision to dedicate the funds to the genome project? The answer hinges on whether the genome project filled an unmet need. The NRC committee and leaders of the biomedical research community certainly identified a weakness in the pattern of NIH funding—a neglect of genome-scale mapping efforts, inattention to development of new technologies, and insufficient funding of data bases and shared resources.
The real question is whether addressing these needs merits just under 1 percent of the NIH budget now and up to 2 or so percent of the budget in the future. Is this field more important than the several hundred grants that could be funded for other biomedical research? The answer depends on whether one believes that society will benefit more from funding an additional 3 to 4 percent of investigator-initiated grants or from devoting attention to developing maps and technologies useful to all of molecular genetics. The debate is not about a funding mechanism—the funds will be dispersed by the same mechanisms used throughout NIH, although differently from tradition in basic genetics, which has been undirected research. The decision is analogous to deciding when a new territory is crowded enough to want to build roads and make rules about land and water use. The genome is largely virgin territory, but molecular biologists have begun to stake claims. When is it time to devote resources to planning and constructing projects for the common good?
Wyngaarden and Kirschstein placed the genome project on the NIH agenda, principally through the appropriations process. Kirschstein's reputation among appropriations committee staff as a solid administrator of great integrity was a necessary element. Wyngaarden first created the line item in response to a ritual question from Congress and then shepherded the budget request through its labyrinthine appropriations
process. The process of securing an NIH budget was much messier and more public than securing funds for DOE, a process that required input from a dozen or so individuals but the direct approval of only a handful. The NIH process did not take much longer, but it involved input from hundreds to thousands of individuals and the direct approval of scores of them.
The DOE process was initiated by a few individuals who sensed an opportunity; the NIH process entailed a cacophonous but productive discussion that redefined the project and flagged the policy issues it would raise. Because of its conspicuousness, the NIH process was also much harder on, but more exciting for, the individuals involved. During 1987 and well into 1989, there was an article almost every month in the news pages of Science or Nature, or both. The genome project was a way for one's name to become widely known—but not always with the “spin” one might want. The extensive coverage also meant that Wyngaarden, Kirschstein, DeLisi, Watson, Gilbert, and others often learned of personal criticisms first by reading about them, an injury that always leaves scars.
SOCIAL ISSUES EMERGE
The debate about the genome project changed substantially in 1989, moving from the question of which agency should lead it to issues of international scientific cooperation, economic competition, and concern about social implications of the research. These issues had always been in the background, but with successful joint planning by NIH and DOE and little controversy about budget levels, attention turned to them. In Europe, concern about the history of eugenics delayed by a year the approval of a 15 million ECU (European currency unit, slightly over a dollar in value) genome project. In the United States, there were two congressional hearings, one by Ralph Hall 's international scientific cooperation subcommittee in the House and the other by Senator Albert Gore's science subcommittee. International and ethical issues were foremost in both hearings. On October 19, Hall summarized his concern about equitable sharing of the research burden: “If you want to ride on the train, you've got to buy a ticket” (U.S. Congress, House of Representatives, 1989). Watson, testifying before him, concurred. George Cahill, representing the Human Genome Organization, acknowledged the problem but pointed to the destructive impact on science of imposing any unilateral restrictions on data flow. This issue arises from the two faces of science—at once the pursuit of pure knowledge, conforming to moral values that transcend national borders, and also an investment in the future of national economies, expected to produce technological
capacities that will yield new products, new jobs, and new wealth. As the rate of growth of the U.S. national economy lagged far behind that of Japan in the late 1970s and through the 1980s, and as Japan captured selected high-technology markets once the sole province of U.S. corporations, members of Congress became concerned that Japan was commercially exploiting the basic research results paid for by U.S. tax-payers. Economic competitiveness became the buzzword of the day; although there were few concrete policy options to address the concern, it would clearly persist as an issue for the genome project for years to come.
The Gore hearing, held November 9, 1989, touched on international data sharing but concentrated even more on the social implications of the project (U.S. Congress, Senate, 1989). These were not entirely new, having been raised by other work in human genetics, but the highly public debate about the genome project brought these issues to the fore. The project would clearly result in much greater knowledge about human genes and would produce technologies to make genetic tests faster, cheaper, and more accurate, as well as applicable to many more diseases. The issues of genetic discrimination in employment or insurance, and the prospects of backdoor racism through genetic screening and testing, were thus more urgent because of the genome project. Indeed, just as the genome project was being formulated, a run of books began on issues related to genetic screening, genetic testing and counseling, and related issues (Holtzman, 1989; Nelkin and Tancredi, 1989; Rothstein, 1989; U.S. Congress, OTA, 1988b). These issues bespoke a renewed public concern about how genetic tests would be used.
Watson saw the need to confront these issues early in the project and stated at the press conference announcing his appointment as associate director of NIH, head of a new Office of Human Genome Research, in October 1988, that he thought the NIH genome program should spend some money to discuss the ethical implications of the work. He elaborated these ideas further at a speech at UCLA in December 1988. Watson foresaw the importance of educating the public through courses, books, and public meetings, and of devising new means to think through the consequences of genome research and anticipate public policy needs. His argument was that, although the genome project was “completely correct” in pursuing gene maps and DNA sequence data as fast as possible, it was essential to be completely candid about how such information could be abused and to suggest laws to prevent such abuse, because, as he said “we certainly don't want to mislead Congress” (Watson, 1988).
As one of the next speakers, I first had to recover from my surprise. This was not the Watson I expected from reading The Double Helix.
Watson's commitment to the consideration of the ethical implications of the project was clarified at several subsequent meetings, and the NIH advisory committee agreed to devote 3 percent of the NIH genome budget to fund the activities of a working group chaired by Nancy Wexler of Columbia University's College of Physicians and Surgeons, and to support a research program.
This commitment was noted in Watson's opening statement before Senator Gore, and Gore commented on it favorably in his opening remarks and again after Watson spoke. Robert Wood, acting director of DOE's Office of Health and Environmental Research, spoke after Watson. As Wood was reading his prepared statement into the microphone, Gore turned to his staff and me, seated behind him, and asked if DOE had made a similar commitment of funds to study the ethical and social implications of the genome project. We were not sure but could not remember seeing any budget commitment in the prepared statement. Gore interrupted Wood to ask. Wood began a reply to the effect that he believed that the NIH effort would address the necessary issues and that DOE was quite concerned about them. Gore responded by asking specifically whether DOE had made a commitment similar to NIH's. Wood said no, and Gore stepped up to the plate. He suggested quite strongly that they do so and noted that there would be future hearings on the genome project at which this issue would come up. Gore's position was reiterated by Senator John Kerry. It was as clear a congressional signal as can be made (U.S. Congress, Senate, 1989).
Gore's interest in the implications of human genetics dated from the early 1980s, when he held a series of hearings on human gene therapy, in vitro fertilization, and biotechnology. In talks with constituents he had found that genetics was especially worrisome to the general public, and he shared some concerns. It was not an antiscience bias but rather an inchoate discomfiture with the prospects of meddling in something as fundamental as a person's genes. Gore had introduced legislation in 1983 that eventually led to the ill-fated Biomedical Ethics Advisory Committee (which quietly died in September 1989 without issuing a report), and his interests in the topic had continued unabated.
Gore was not alone among senators in these concerns. Edwin Froelich was a physician and staff person for Senator Orrin Hatch, ranking Republican on the Senate committee that authorizes NIH activities. Froelich called DeLisi to his office late in 1986, soon after he learned of the DOE plans for a genome project. He expressed grave concern about the project and urged that the research be scrutinized for its broader impact, particularly whether it would lead to more prenatal diagnosis and abortion. He likewise called Ruth Kirschstein when he heard of the NIH plans in 1987. Kirschstein and W. French Anderson then met
with him to reassure him that NIH was, indeed, concerned about these matters. (Froelich wanted some assurance that there would be explicit attention to such matters, or the program would be in jeopardy.) In several meetings late in 1988, Barbara Mikulski, the brusque senator from Maryland, also expressed concern to me and others about “go-go ” science that potentially might race far in advance of the policies developed to contain its adverse impacts on individuals and society. More pointedly, Congressman David Obey raised serious questions about how insurers and employers might use genetic information to discriminate unfairly against certain individuals. He urged strongly in the House appropriations hearings for the 1991 budget (held early in 1990) that NIH come up with a systematic plan to deal with such issues.
Human genetics was of special concern for several reasons. First, it appeared threatening because it studies the very stuff of life —not directly who we are but the “recipe book” that makes each of us possible. Second, human genetics offered technological options that were not available before. Before genetic testing was developed, individuals at risk for Huntington's disease or anxious about whether they were carriers of sickle cell or Tay-Sachs disease did not have to worry about such tests. Technology brought choice, sometimes agonizing choice. This situation was not different in principle from other medical advances, but it seemed to hit especially hard in the case of genetic disease, which is caused by factors utterly beyond the control of the person carrying the genes. Finally, genetics had been used as a tool for political abuse in the past.
Human genetics research labored in the shadow of eugenics and racial hygiene. A new spate of scholarship detailed the role of scientists and physicians in promoting the racist agenda of those movements in the first half of this century (Kevles, 1985; Lifton, 1986; Muller-Hill, 1988; Proctor, 1988; Reilly, 1977). The medical model of nondirective genetic counseling explicitly rejected the tenets of eugenics and racial hygiene, but the magnitude of the abuses left a strong legacy of distrust. Nonscientists were not about to give this trust automatically; scientists would have to earn it.
Several observers, both scientists and nonscientists, predicted that the research program and other activities to investigate the ethical, legal, and social implications of human genetics would be an important legacy of the genome project, perhaps even its most substantial one. Concomitantly with the National Center for Nursing Research at NIH, the National Center for Human Genome Research did, indeed, become a pioneer in offering NIH support for such work. Although bioethics had been supported intermittently by NIH in the past, it had not had the support of any ongoing program and lacked a dedicated budget.
The commitment to fund such work was a dramatic departure from past practices, an innovation in NIH policy that was likely to have deep and long-lasting impacts on NIH well beyond the genome center.
STILL, ORGANIZATIONAL ISSUES
The organization of the genome project remained in flux. Under threat of legislation, NIH and DOE signed a memorandum of understanding in fall 1988, choosing a joint agreement rather than a structure imposed by the Chiles-Kennedy-Domenici bill. The agreement ratified an existing informal arrangement but grew into substantially more, as bona fide cooperation began to seem advantageous to both agencies. Throughout 1989, staff from NIH and DOE met to discuss how to carry out the terms of the memorandum. They finally settled on a joint NIH-DOE advisory group, composed of members of the advisory panels for each agency's outside advisory group.
Watson was insistent on having a “serious” planning document. Rand Snell and Michael Hall, of Senator Chile 's staff, inserted language into the 1989 appropriations conference report, a document that accompanied the bill to explain congressional intent, expressing concern about interagency coordination and stipulating that NIH and DOE develop “the optimal strategy for mapping and sequencing the human genome” in time for the 1991 budget cycle (U.S. Congress, Senate, 1988). The research was given a big boost at a joint NIH-DOE planning retreat held at the Banbury Center, Cold Spring Harbor Laboratory, in late August 1989.
For many months, there had been an informal coordinating committee: Diane Hinton of HHMI, Mark Guyer of the genome office, Irene Eckstrand of NIGMS, John Wooley of NSF, and Ben Barnhart of DOE. Others attended occasionally. The loosely coordinated plans formed by that group began to gel when combined in a retreat setting with the powerhouses of genome research at Banbury. The advisory committees for both DOE and NIH were present, as well as invited experts from other laboratories engaged in genome research.
Going into the meeting, NIH and DOE staff were expecting to prepare a five-year plan. DOE's Barnhart and Norton Zinder, of Rockefeller University, thought that no specific planning draft would emerge from the retreat (Palca, 1989), but they proved themselves wrong. Zinder organized the discussion into task areas, and a format of specifying goals and the means of achieving them developed naturally out of this discussion. Much of the meeting focused on how to construct physical maps, sets of ordered, overlapping cloned DNA fragments. Maynard Olson and others mentioned the idea of using short stretches of DNA
sequence as unique “tags” that would serve as landmarks on the chromosomes that could be used by laboratories using different methods. The suggestion was seized upon quickly, and a group agreed to author a paper for Science (Olson et al., 1989; Roberts, 1989). Thus, the shape of the plan that the staff had hoped to develop became considerably clearer after the retreat; NIH and DOE staff agreed that the report on the five-year plan should follow the goal-oriented format (NIH/DOE, 1990).
The human genome project began to come of age in late 1989 and 1990. Secretary of Health and Human Services Louis Sullivan created the National Center for Human Genome Research in October 1989, giving the project administrative authority to spend federal funds, pending approval of an advisory council. Much staff time went into organizing chromosome-specific meetings, workshops on cloning large DNA inserts, DNA sequencing, informatics, and other topics. Upper management, particularly Elke Jordan and Mark Guyer at NIH and Ben Barnhart at DOE, focused heavily on preparing a joint five-year plan to present to Congress. The two oversight hearings in October and November 1989 gave way to appropriations hearings in both houses of Congress in spring 1990. The process for reviewing grants began to become more routine for the genome proposals, although there continued to be disagreement about the specific goals and scope of the project among grant reviewers. Steps were taken to establish standing review panels for genome grants in mid-1990, and the advisory council charter was approved. Watson declared that, because the first few years had been dedicated to getting organized, the genome project should officially begin with fiscal year 1991. And so it began.
The history of the genome project makes it clear that scientists played a crucial role in starting it, and they were the sources to which policymakers turned for advice along the way. The NRC was particularly influential, but there were many independent mechanisms as well. Hundreds of scientists attended the meetings that hatched the genome scheme. Scientists were also called upon as independent witnesses in hearings in addition to NRC and OTA staff. There were no disease constituencies who rose to champion the project, in contrast to cancer, heart disease, Alzheimer's disease, or AIDS. There was strong support from the Alliance for Aging Research and a few other groups in 1989; these sources did not constitute a wellspring of public sentiment but rather the support of organization members in leadership positions. Indeed, aside from scattered press reports, the public remained largely ignorant of the project even after it had been under way for three years.
Public policy on the genome project was formed in the cruel daylight of productive conflict, and key actors in federal agencies responsible for conducting research did most of the political spadework. It is impossible to judge now whether the genome project would have happened without DeLisi's efforts, but it certainly would not have happened as fast. Probably, there would have been considerable discussion about sequencing, about data bases, about genetic linkage maps, about physical mapping, and about computational biology, but these “abouts ” could conceivably have remained segregated in their scientific communities of origin. The human genome project need not have emerged so quickly from the collision of human genetics and molecular biology, and it need not have been projected as a major new initiative meriting serious high-level political attention. Large-scale mapping and sequencing efforts might well have emerged piecemeal and been resolved incrementally, but the proposals would have met substantial resistance within the scientific disciplines. The genome project provided a vehicle for biological projects on a larger scale with a greater focus on technological improvement per se.
The science agencies did not have to discover the human genome project as a way to package the road-building work of human genetics. Indeed, given the bureaucratic tendencies toward caution and narrow definitions of mission, they might well have resisted any new initiative but for the prodigious efforts of a few champions. DeLisi put the genome project on the public agenda. Once it was there, it provoked competition between DOE and NIH and forced the issue to the surface of science policy. Wyngaarden rose to the challenge, and Congress then had to make several decisions about budgets and agency leadership. Watson channeled his prodigious energies into the project, and gave it scientific credences it desperately needed.
Without strong impetus from DOE, NIH almost certainly would not have reacted as strongly, as quickly, or as systematically as it did. Without these agencies vying for leadership, neither might have fought quite so hard to assess its options, secure its budget, or influence the opinions of the scientific community. The extraordinary number of meetings on the genome project testify to the fact that, once the idea of the project was aired, it was immediately perceived as exciting and important. Norton Zinder noted the consistency with which the genome project was first greeted with skepticism and then accepted as inevitable (N. L. Zinder, Rockefeller University, personal communication, March 1990). Sinsheimer, Dulbecco, and DeLisi were the first to sense the importance of a new push for human genetics. But in science policy as in science, being first matters more than being absolutely right.
A few pivotal scientific figures—the scientists who took the trouble
to learn about the policy process and to interact with it—clearly had enormous influence. Watson was preeminent among these, but Hood, Gilbert, Bodmer, Baltimore, Berg, Dulbecco, Alberts, Cantor, Olson, and others had major effects at critical junctures. Many scientists from the national laboratories played decisive voices in steering DOE policies. Still, not every contribution was totally positive. Some scientific input was naive and some almost destructive. Scientists were rather poor at making the policy issues clear and were often quick to form opinions without fully informing themselves of the political consequences. The NRC recommendation of a lead agency was one such mistake.
Sometimes there were remarks that fanned the flames without focusing the heat. Robert Weinberg, for example, was quoted in New Scientist as saying, “I'm surprised consenting adults have been caught in public talking about it it makes no sense” (Joyce, 1987). This may well have been the case of a wayward pronoun, a loose “it” floating freely in that dangerous space between a reporter and his source. The context of this remark made it unclear whether Weinberg referred to the entire project or to the prospect of sequencing it alone. He later recalled referring to the latter (personal communication, December 12, 1990) and disavowed opposition to a concerted mapping effort, but at the time, the remark was cited by staff on Capitol Hill as opposition to the project as a whole. Even apart from losing whatever nuances surrounded the original arguments, politicians are well able to filter such isolated judgments. Politics is, after all, waged as a war of words, and politicians are accustomed to the rhetorical excesses of interest groups anxious to support a position.
Most of the policy statements made by scientists had little impact because those in Congress who were making decisions were insulated from them. Most members of Congress had a few regular sources of scientific information on which they relied, and they also listened to committee staff, the NRC, and OTA. Scientists, for their part, talked mainly to one another. Few took the time to visit with policymakers or to write for an audience outside science journals and science news publications. There was remarkably little effort to build a broad public constituency, a strategy that would instinctively have appeared important to an elected official. When reporters came, many scientists were eager to be quoted, but there were remarkably few attempts to articulate a broader vision of why the general public should support a genome project at taxpayer expense. Most of the debate was a narrow one regarding the unanswerable question of whose ox would be gored if the monies for the genome project were not actually “new.”
Gilbert, Watson, Hood, Cantor, and a few others were unusual in the degree to which they took public communication seriously. Ironi-
cally, this sensitivity cost them support within the scientific community, which saw these efforts as grandstanding. The issue recurs time and again in connection with public support of science, falling in the gray border zone between selling to and educating the public. Every politician knows that one needs support to get things done, and deliberately maintaining a high profile is a necessary component of policy formation in the world of modern media. DOE recognized this early on and sponsored a science writers' workshop at Brookhaven National Laboratories, which led to several articles. The Alliance for Aging Research, the American Medical Association, and DuPont sponsored a national conference with general media coverage in mind. A cover feature appeared in Time magazine a few weeks before the conference (Elmer-DeWitt et al., 1989; Jaroff et al., 1989), and there was a wave of articles afterward. The many national scientific meetings also generated press coverage. These activities, however, were hardly broad public education, relying as they did on snapshots of scientific opinion. There was no systematic assessment of what the public was worried about until the project was well along—it clearly was worried—and little attempt to identify issues of which the public was as yet unaware. The task of public education was largely left for the future.
Now, the future of the genome project depends on several factors: (1) whether it produces scientifically useful data, in particular, whether the systematic approach is quickly shown to be useful as a way to understand major diseases and illuminate fundamental biological questions; (2) whether it broadens its base of support among scientists and in the general public; and (3) whether it successfully confronts the broader social implications that emerge as human genetics advances.
Where Was OSTP?
The job of the White House's Office of Science and Technology Policy (OSTP) is to help make science policy. Its director is the President's science advisor. OSTP's purpose and position make it the logical place to resolve disputes such as those that arose between NIH and DOE during 1987 and 1988. Why, then, was it ineffective? The question cannot be answered completely, but a few observations are relevant. First, OSTP is severely limited by staff constraints. For most of the period in question, there were only two life scientists in OSTP, left to attend to AIDS, environmental issues, international science agreements, and the ceremonial and diplomatic duties inherent in any White House office. It was quite easy for urgent issues to displace those that
were merely important when there were so few staff covering broad expanses of federal policy.
Second, OSTP was not trusted by the agencies or OMB. If OSTP had been better staffed it might have been effective. But life sciences were given short shrift at an agency regarded as marginally effective even in its area of greatest expertise—military research. OSTP eventually began to hold monthly meetings about the genome project that included representatives from the Department of Agriculture, NIH, DOE, NSF, and HHMI, but the agency staff attending them, without exception, found them to be nonproductive.
OSTP has enormous power to convene high-level meetings, but for such meetings to succeed, either OSTP staff must have the power to enforce decisions—giving agencies a strong incentive to prepare their positions carefully—or they must know the issues and agencies' interests well and have the agencies' respect and an expectation of fair judgment. Because OSTP was irrelevant to the budget process, interagency science policy decisions were made elsewhere, that is, by OMB and Congress.
OSTP did have some effect in one area regarding the genome project: it nearly ruined an opportunity to create a platform for U.S.-Japanese cooperation on the program. OSTP was given principal responsibility for negotiating a U.S.-Japan science and technology agreement and participated in discussions about life sciences in Tokyo in April 1989. The United States came to the table with five proposals, most of them involving forestry and fermentation, areas of Japanese strength. Japan came with nine areas, two of which were related to genetics and another that was specifically focused on the human genome project. The OSTP representative dismissed the Japanese genome proposals as not showing promise for U.S.-Japan collaboration under the agreement because the United States was seen as so far ahead in these areas. This perception was true, in fact, but an opportunity was missed to leverage the Japanese government into much greater support for genome research, which might have avoided at least some of the controversy that later ensued over data sharing by U.S. and Japanese genome investigators.
The Japanese Ministry of Education (Monbusho, which supported most university-based research) had just received a report supporting a major commitment to genome research. Japan's Science and Technology Agency had translated into English another report that reached more or less the same conclusion. Apparently, Monbusho and the Science and Technology Agency were counting on U.S. support to approach their own Ministry of Finance to ask for substantial budget increases, but the OSTP statement in April scuttled this strategy. The cabinet-level Japanese science council issued a report supporting genome research in May, but the damage was done. Monbusho and the Science and
Technology Agency succeeded in each gaining two-year programs, but funding for them fell considerably short of their aspirations. It set them back until well into 1990.
In June 1989 a controversy with the United States over the level of research support in Japan began to heat up at an international genome meeting in Moscow and erupted publicly in October. The controversy had the effect of focusing renewed attention on the genome project in Japan, and scientists there became optimistic that more support might be forthcoming. As this volume goes to press, U.S. negotiations with Japan about the genome project are pending.
Can the Genome Project Keep Its Promises?
The genome project was “sold” on four main points: (1) it would create tools to combat human disease by expediting the process of biomedical research; (2) it would stimulate domestic economic development by keeping the United States in the lead in biotechnology; (3) it would enhance national prestige; and (4) it would stand as a cultural achievement to be hailed for centuries. Of these, the promise to promote understanding of health and disease seems secure. Because the main products of the genome project will be information (maps and sequence data) and methods of broad applicability (cloning, detecting, sequencing, and analyzing DNA), the genome project will certainly help elucidate the mechanisms of disease, and even when elucidation does not lead to cure, it may suggest means of prevention or amelioration.
Sickle cell anemia is a case in point. It has often been cited as an example of how molecular knowledge can fail to have clinical impact, but persuasive arguments can also be made for exactly the opposite position. Knowledge of a molecular defect in hemoglobin in this condition has been available since 1949, but this has not led to a cure. Morbidity and mortality rates for sickle cell anemia, however, are severalfold lower than four decades ago. Has molecular knowledge made a difference? Knowing which gene is involved has not led to direct gene therapy —at least not yet—but knowing the mechanism of disease has directed the development of new treatment strategies that have improved management of the disease by small increments over the years, and has provided a means for monitoring the effects of various treatments.
Noting that genome research will help in the study of disease does not answer the question about the optimal balance between the collective, concerted effort to map and sequence the genome and more decentralized, undirected research, or that done in pursuit of particular
diseases. It does mean that the money spent on the genome will not be wasted. If in 10 years researchers and policymakers generally agree that the money was spent wisely, then this efficiency criterion will also have been met. The continuing debate about whether the genome project is displacing better science hinges on three points, none of which is susceptible to rigorous analysis: (1) whether the genome project displaced other biomedical research, that is, whether the increase in NIH funding associated with the genome project would have gone to NIH in any event; (2) if such displacement occurred, whether the displaced science was of equal or greater merit (a notoriously difficult issue to judge); and (3) whether biomedical research was the best use of funds that might otherwise have been used to house the homeless or provide other social services (an equally impossible question).
The other three promises are more difficult to assess. The relationship of the genome project to biotechnology is murky and confused. Genome research will clearly push the limits of DNA-based methods, and these advances will have broad applicability, not only in research but also in medical diagnosis and treatment, agriculture, pharmaceutical development, and other industrial sectors. Even more important, those who learn genome research will be at the cutting edge of techniques, and this training is the most effective means of technology transfer.
If economic development were the principal goal of policy, however, rather than acceleration of biomedical research, would the genome project be the most direct route? The genome project is more likely than most other research to have industrial spinoffs, but direct funding of instrumentation, methods development, fermentation technologies, protein engineering, structural chemistry, and other targets is arguably at least as important because it would be focused on choke points of industrial development, as opposed to pure research. Industrial development of new technologies is widely acknowledged as the weakest point of U.S. technology policy, while basic science is its greatest strength (U.S. Congress, OTA, 1984). The genome project is one step in the direction of technological development because three- to fivefold increases in speed, scale, or cost reduction are explicit goals of its technology development component. Nevertheless, technologies more directly related to commercial application might be even more efficient in improving economic returns on public biotechnology investments. Several other countries are pursuing such commercially targeted policies; thus, time will tell whether the U.S. genome project was, as some scientists asserted, the best way to retain biotechnological supremacy.
The arguments regarding national prestige and cultural endowment are undecidable. If the United States has the largest program and the program is judged successful, then it will add to our national
prestige. So do Olympic gold medals, but the U.S. government does not fund the efforts leading to them. Further, arguing that a science program adds to prestige would have to satisfy the condition that it does so more quickly, surely, and cheaply than alternatives to the same end. This condition seems a heavy burden for long-term scientific or technical projects to carry. Such an argument may have worked for the Apollo project in the 1960s, when the U.S. economy was overwhelmingly the healthiest on the planet, but it is unlikely to carry the day in the 1990s. National prestige is a tenuous basis on which to form any policy, and it is especially risky as a foundation for science and technology funding.
If the genome project is as successful, and has as few adverse social side effects, as the optimists predict, it will undoubtedly be hailed as a major cultural advance. If the project fails technically, or if the information derived from it and the methods developed as part of it are transformed into tools for genetic discrimination or racism, then it will be judged a disaster. The degree to which scientists and health professionals take responsibility for seeing that the fruits of their labor are used to promote freedom, rather than circumscribe it, will be critical to the ultimate judgment about the benefits of the human genome project. Enthusiastic pronouncements, such as those voiced by Norman Anderson in Santa Fe, fill the pages of history books—sometimes as examples of prescient insight and other times as examples of egregious folly. History's verdict on such predictions is harsh and depends entirely on whether they are right.
Is Cost Wobble a Serious Problem?
There were many attempts to formulate budgets for the human genome project. The earliest were performed in conjunction with the Santa Cruz meeting, at the first Santa Fe meeting, and at meetings held thereafter. There were many different strategic approaches with diverse component parts that yielded wildly different projections. The early estimates are chronicled in Appendix B of the OTA report (U.S. Congress, OTA, 1988a). The generation of cost figures played a pivotal role in forcing consideration of the technical options. Cost figures also became a focus of policymaking, as those funding the project wanted to know how much they would be expected to spend. Scientists who were engaged in the genome debate quickly became aware of this expectation, and the cost projections of the three advisory committees were remarkably similar. The reasons for this concurrence are instructive.
Since DOE was the first federal program to start, it was also the
first to begin discussion of budget projections. This took place at two levels, within the DOE bureaucracy and among scientists promoting the genome project from the outside. Charles DeLisi's internal process is noted in the text. The external process began at the first Santa Fe conference in March 1986, where the budget numbers were extremely diverse and generally focused only on one or two components. By the second Santa Fe conference in January 1987, planning had become more serious. Several of the invited participants met over lunch at that conference to discuss what the budget should be. David Padwa, who had previously been involved with founding the agricultural biotechnology company Agrigenetics, noted the political constraints on the budget. It had to be large enough to command congressional attention, so it would have to be at least $50 million to $100 million per year, but it could not be so large that it threatened other research interests.
The budget discussions continued a month later, at a meeting of the scientific advisory committee assembled to render advice on DOE' s genome project. The DOE HERAC subcommittee met to discuss costs on February 5 and 6, 1987, a month before its report was to be considered by the full committee. The subcommittee met for an afternoon session, during which there was some discussion of costs, but generating cost estimates had been delegated to Lee Hood, who was not scheduled to appear until the next morning. The meeting started at 9 a.m. on the 6th, but Hood's plane had been delayed, so the group began to discuss what could be done within the range of budgets it was thought reasonable for the Office of Health and Environmental Research to request. There was some discussion of how much physical mapping and sequencing could be done with $20 million to $40 million, the maximum the subcommittee considered politically feasible. Hood entered the meeting at 10 a.m., armed with some handwritten notes that included a menu of necessary technologies and attached costs. The proposal included technology development, physical mapping of the human genome, mapping and sequencing of model organisms (yeast and bacteria), and regional sequencing of interesting chromosomal regions (e.g., those packed with genes). Hood's estimates were $200 to $300 million per year for a full program. A shocked silence settled over the room. Someone asked if that was at all possible, being a full order of magnitude higher than what had been discussed before. Hood did not wait for or provide an answer but instead asked passionately whether the budget would drive the vision or the vision drive the budget. Swept away, the group niggled over technical details of how to make the projections and settled on a figure of $200 million, thus exceeding Padwa's threshold. Those present at the Santa Fe discussion, including Charles Cantor, then endorsed the importance of having the budget reflect what was
needed to begin a realistic scientific program, and not to let perceived budget ceilings force the group into making promises they could not keep.
The NRC committee process for setting a budget was somewhat different. A subgroup was tasked to produce cost options. Botstein spear-headed this effort and came up with three options that would enable various amounts of genetic linkage mapping, physical mapping, sequencing, and other activities. He suggested annual budgets of $50 million, $100 million, and $200 million, with completion dates sooner for the higher figures (the year 2000 for $200 million versus 2025 for $50 million). The estimates were based in part on previous technical presentations on mapping and sequencing methods but mainly on how many people in how many laboratories could be funded at the different budget levels. Watson objected to the range of options, noting that it would naturally incline the committee members to seem reasonable by choosing the middle option. He therefore suggested a $500 million-per-year crash program as a fourth option. (Because Botstein had already deemed his top option the crash program, Watson's was variously dubbed the crash-crash or crash-boom.)
Comments began on Botstein's right and went counterclockwise around the table. One by one, the members supported some special effort, although there was no convergence on any one figure until a second go round the table. Then, there seemed to be a general acceptance of something near the $200 million figure, with Botstein responsible for reviewing the figures again after the discussion. The committee ultimately projected a need for $200 million a year for 15 years: $60 million for 10 centers, $60 million for grants and technology development, $55 million per year in early years for construction and capital costs, and $25 million per year for administration, a stock center, a data management center, quality control, and peer review (NRC, 1988).
The OTA budget projections were based on a two-stage process. A workshop was convened in August 1987, chaired by Paul Berg. The participants included representatives from the major funding agencies and scientific groups engaged in mapping and sequencing, as well as others familiar with quality control, data bases, and costs of materials distribution and handling. After some discussion, the panel agreed on cost figures for genetic linkage mapping and physical mapping. When the discussion turned to sequencing, however, it became contentious, lapsing into several disagreements about what strategy should be followed, how much reagents would cost, and how much automation could save. Subsequent estimates of the costs of storing DNA clones were shockingly high, and the last hour of the meeting was dominated by
debate about how the project should be managed and which agency should lead it.
Beyond genetic linkage and physical mapping estimates, the exercise failed to produce a cost table. The process did force the alternative strategies to the surface, however, and these were sent along with alternative cost estimates to all the participants and an additional 30 or 40 people with technical or science administrative background. This procedure yielded estimates in a much narrower range, which were summarized in an appendix to the OTA report. One clear decision was made at OTA: attempting to project more than a five-year budget was impossible because of technical uncertainties, and even estimates beyond two years were highly suspect. OTA estimated costs beginning at $47 million the first year and increasing to $228 million over five years.
How did all the estimates fall into such a narrow range? There are two explanations. First, there was really only one way to project costs—to estimate how many people could be productively engaged in the various component tasks. With genetic linkage mapping, the wobble came from disagreement about how high the resolution needed to be. The costs per marker were fairly certain because there were several years of experience in finding markers and mapping them. In the case of physical mapping, there was some experience with nematodes and yeast and a sense that new methods to clone larger fragments would reduce costs. For the other components, the wobble overwhelmed the axis of rotation. At bottom, all of these estimates were highly subjective intuitions of what would be needed.
The question of estimates based on intuition raises the second factor ensuring some convergence. The group of people making these estimates was fairly small, and they talked to one another frequently, if nowhere else than at that week's genome meeting. Hood, for example, was on all three advisory committees, for DOE, NRC, and OTA.
A policy problem is thus apparent in placing the authority to make budget projections, clearly one of the most important considerations in genome project planning, in the hands of such a limited number of people. Yet there are few alternatives when dealing with cutting-edge technologies. The group of people consulted was small because the number of experts who ran relevant programs was small, and there was no way to avoid this problem. Further, the process was not as narrow as it might seem from the discussion above: at least the OTA cost workshop was well attended (by more than 100 observers, in addition to the participants) and covered prominently by science journals (including a feature piece in Science [Roberts, 1987]), which elicited unsolicited comments from many quarters. Nevertheless, the fact remains that the basic cost estimates came from the handful of scientists who had direct experience with the technologies.
What Is “Technically Feasible” in the Policy Context?
When discussions about the genome project began in 1985, scientists claimed that only recently had it become technically feasible to sequence the entire human genome. It is not clear what was meant by this statement, but this and similar language cropped up in dozens of letters, articles, meetings, and private discussions. It is not technically feasible to sequence the human genome in 1990, despite a half decade of impressive gains in power, speed, and simplicity of component steps (e.g., cloning larger DNA fragments, ordering clone libraries, DNA sequencing, and analyzing map and sequence data). Many scientists might more accurately have said that the sort of problems posed by the project were of a kind that seemed likely to be solved without conceptual breakthroughs or revolutionary technological innovations. Carefully qualified statements, however, fit poorly into media sound bites and can be as inimical to communication with policymakers as deliberate obfuscation.
This type of communication is a problem endemic to any scientific enterprise requiring long-term federal support. It is an artifact of the time-scale mismatch between policy formulation and scientific or technical achievement, and the communication style mismatch between science and science policy. The medium of science is written, usually in the passive voice, carefully qualified, and driven by data. Communication about policy is verbal, oriented toward action, and driven by issues. Arguments must be made in favor of a scientific agenda many years before that agenda can be met, in order to ensure that the budget and infrastructure are in place so work can go forward if the intervening technical obstacles are overcome. Visionary scientists leap over considerable technical obstacles because they assume obstacles can be overcome; the money has to be ready and waiting when the science arrives. This process is related to the issue of promise keeping noted above, except that it is focused on how the technical objectives of a project can be met, rather than on whether those objectives relate to a larger mission such as promoting health, combating disease, or generating wealth through biotechnology.
Science has a healthy skepticism about glib pronouncements of what will be possible—but then so do politicians. Skepticism about claims is a natural reflex, without which no politician can long remain in office. Contending with interest groups is the very stuff of political life, but in crafting science policy, scientists have conflicting roles as objective observers and stakeholders.
The predictions made in 1987 that there would be a complete physical map of the human genome in a “few years” and that DNA sequenc-
ing would cost “only a few pennies a base” were clearly wrong. On the other hand, some parts of the X chromosome spanning more than 2 million base pairs are contained in ordered clone libraries; more than 10 million base pairs of the nematode genome have been contiguously mapped by overlapping clones. And although the longest continuous sequence to date is still only 200,000 base pairs, the hprt gene region on chromosome X has been sequenced, and large expanses of chromosome 4 and the T-cell receptor region are now being sequenced —projects unthinkable even in 1987.
There are no easy answers here. It does not seem that Congress has been seriously misled about the technical feasibility of mapping the genome. The current five-year plan stipulates specific goals and presents these openly to Congress, a remarkably forthright strategy. By producing this document, NIH and DOE have provided Congress and OMB with the tools by which to measure their progress in future years. Success or failure of the effort will thus be much easier to assess. It is a bold strategy: if a technical obstacle suddenly appears, support for the entire enterprise might evaporate, although this is unlikely. In the worst-case scenario technically (leaving aside for the moment the possible social abuses of genetic information), there will be a great deal of useful sequence data even if only 5 percent or so of the genome is sequenced, because investigators will attack the most clearly interesting regions first. Genetic and physical maps will be useful no matter how incomplete, but the value of the genome project depends on their completeness. The current NIH-DOE joint planning strategy is critically dependent on setting the right goals, requiring a stretch but not a break, and marshaling the resources necessary to meet them.
Alberts, B. M. 1985. Limits to growth: In biology, small science is good science. Cell 41:337-338.
Anderson, N. G., and N. L. Anderson. 1985. A policy and program for biotechnology. American Biotechnology Laboratory September/October:1-3.
Avery, O. T., C. M. MacLeod, and M. McCarty. 1944. Induction of transformation by a deoxyribonucleic acid fraction isolated from Pneumococcus Type III. Journal of Experimental Medicine 79:137-158.
Baltimore, D. 1987. Genome sequencing: A small-scale approach. Issues in Science and Technology 3:48-50.
Beam, A., and J. O'C. Hamilton. 1987. A grand plan to map the gene code. Business Week April 27:116-117.
Begley, S., with S. E. Katz and L. Drew. 1987. The genome initiative. Newsweek August 31:58-60.
Bitensky, M. 1986. Sequencing the Human Genome. Santa Fe, N.M.: Office of Health and Environmental Research, U.S. Department of Energy (published by the University of California under contract W 7405-ENG-36, Los Alamos National Laboratory, N.M.).
Bodmer, W. F. 1986a. Human genetics: The molecular challenge. Pp. 1-3 in Molecular Biology of Homo sapiens. Cold Spring Harbor Symposia on Quantitative Biology, Vol. 51. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory.
Bodmer, W. F. 1986b. Two cheers for genome sequencing. The Scientist October 20:11-12.
Botstein, D. 1980. Construction of a genetic linkage map in man using restriction fragment length polymorphisms. American Journal of Human Genetics 32:314-331.
Cook-Deegan, R. M. 1989. The Alta summit, December 1984. Genomics 5:661-663.
Coulson, A., J. Sulston, S. Brenner, et al. 1986. Toward a physical map of the genome of the nematode Caenorhabditis elegans. Proceedings of the National Academy of Sciences (USA) 83:7821-7825.
del Guercio, G. 1987. Designer genes. Boston Magazine August:79-87.
DeLisi, C. 1988. The human genome project. American Scientist 76:488-493.
Donis-Keller, H., P. Green, C. Helms, S. Cartinhour, B. Weiffenbach, K. Stephens, T. P. Keith, D. W. Bowden, D. R. Smith, E. S. Lander, et al. 1987. A genetic linkage map of the human genome. Cell 51:319-337.
Dulbecco, R. 1986. A turning point in cancer research: Sequencing the human genome. Science 231:1055-1056.
Elmer-DeWitt, P., with A. Dorfman and J. M. Nash. 1989. The perils of treading on heredity. Time March 20:70-71.
Friend, S. H., R. Bernards, S. Rogelj, R. A. Weinberg, J. M. Rapaport, D. M. Albert, and T. P. Dryja. 1986. A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma. Nature 323:643-646.
Gilbert, W. 1986. Two cheers for human gene sequencing. The Scientist October 20:11.
Gilbert, W. 1987. Genome sequencing: Creating a new biology for the twenty-first century Issues in Science and Technology 3:26-35.
Gusella, J. F., N. S. Wexler, P. M. Conneally, S. L. Naylor, M. A. Anderson, R. E. Tanzi, P. C. Watkind, K. Ottina, M. R. Wallace, A. Y. Sakaguchi, A. M. Young, I. Shoulson, E. Bonilla, and J. B. Martin. 1983. A polymorphic DNA marker genetically linked to Huntington's disease. Nature 306:234-238.
Hall, S. S. 1988. Genesis: The sequel. California (July):62-69.
Holtzman, N. A. 1989. Proceed with Caution. Baltimore, Md.: Johns Hopkins University Press.
Holzman, D. 1987. Mapping the genes, inside and out. Insight May 11:52-54.
Hood, L., and L. Smith. 1987. Genome sequencing: How to proceed. Issues in Science and Technology 3:36-46.
Ingram, V. M. 1957. Gene mutation in human haemoglobin: The chemical difference between normal and sickle cell haemoglobin. Nature 180:326-328.
Jaroff, L., with J. M. Nash and D. Thompson. 1989. The gene hunt. Time March 20:62-67.
Jenks, S. 1989. Gene map budget given a boost. Medical World News February 13:94.
Joyce, C. 1987. The race to map the human genome. New Scientist March 5:35-40.
Judson, H. F. 1979. The Eighth Day of Creation: The Makers of the Revolution in Biology New York: Simon and Schuster.
Judson, H. F. 1987. Mapping the Human Genome: Historical Background (Mapping Our Genes contractor reports, Vol. 1, NTIS Order No. PB 88-160-783/AS). Office of Technology Assessment, U.S. Congress.
Kan, Y. W., and A. M. Dozy. 1978. Polymorphism of DNA sequence adjacent to human beta-globin structural gene: Relationship to sickle mutation. Proceedings of the National Academy of Sciences (USA) 75:5631-5635.
Kanigel, R. 1987. The genome project. New York Times Magazine December 13:44, 98-101, 106.
Kerem, B.-S., J. M. Rommens, J. A. Buchanan, D. Markiewicz, T. K. Cox, A. Chakravarti, M. Buchwald, and L.-C. Tsui. 1989. Identification of the cystic fibrosis gene: Genetic analysis. Science 245:1073-1080.
Kevles, D. J. 1985. In the Name of Eugenics. Berkeley, Calif.: University of California Press.
Koenig, M., E. P. Hoffman, C. J. Bertelson, A. P. Monaco, C. Feener, and L. M. Kunkel. 1987. Complete cloning of the Duchenne muscular dystrophy (DMD) cDNA and preliminary genomic organization of the DMD gene in normal and affected individuals. Cell 50:509-517.
Kolata, G. B. 1980. The 1980 Nobel prize in chemistry. Science 210:887-889.
Lee, W. H., R. Bookstein, F. Hong, L. J. Young, J. Y. Shew, and E. Y. Lee. 1987. Human retinoblastoma gene: Cloning, identification, and sequence. Science 235:1394-1399.
Lewin, R. 1986a. Molecular biology of Homo sapiens. Science 233 July 11:157-160.
Lewin, R. 1986b. Proposal to sequence the human genome stirs debate. Science 232:1598-1600.
Lewin, R. 1986c. Shifting sentiments over sequencing the human genome. Science 233:620-621.
Lifton, R. J. 1986. The Nazi Doctors. New York: Basic Books.
Maxam, A. M., and W. Gilbert. 1977. A new method for sequencing DNA. Proceedings of the National Academy of Sciences (USA) 74:560-564.
McAuliffe, K. 1987. Reading the human blueprint. U.S. News and World Report (December 28, 1987/January 4, 1988):92-93.
McKusick, V. A. 1986. The gene map of Homo sapiens: Status and prospectus. Pp. 15-27 in Molecular Biology of Homo sapiens. Cold Spring Harbor Symposia on Quantitative Biology, Vol. 51. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory.
McKusick, V. A. 1988. The morbid anatomy of the human genome: A review of gene mapping in clinical medicine. Bethesda, Md.: Howard Hughes Medical Institute.
Muller-Hill, B. 1988. Murderous Science. New York: Oxford University Press.
National Institutes of Health and Department of Energy. 1990. Understanding Our Genetic Inheritance. The US Human Genome Project: The First Five Years, FY 1991-1995. Document DOE/ER-0452P. Washington, D.C.: National Institutes of Health, U.S. Department of Health and Human Services; and U.S. Department of Energy.
National Research Council, Committee on Mapping and Sequencing The Human Genome. 1988. Mapping and Sequencing the Human Genome. Washington, D.C.: National Academy Press.
Neel, J. W. 1949. The inheritance of sickle cell anemia. Science 110:64-66.
Nelkin, D., and L. Tancredi. 1989. Dangerous Diagnostics: The Social Power of Biological Information New York: Basic Books.
Olson, M. V., J. E. Dutchik, and M. Y. Graham. 1986. Random-clone strategy for genomic restriction mapping in yeast. Proceedings of the National Academy of Sciences (USA) 83:7826-7830.
Olson, M. V., L. Hood, C. Cantor, and D. Botstein. 1989. A common language for physical mapping of the human genome. Science 245:1434-1435.
Palca, J. 1987. Human genome sequencing plan wins unanimous approval in US. Nature 326:429.
Palca, J. 1989. Gene mappers meet on strategy. Science 245:1036.
Pauling, L., H. A. Itano, S. J. Singer, and I. C. Wells. 1949. Sickle cell anemia: A molecular disease. Science 110:543-548.
Pines, M. 1986. Shall We Grasp the Opportunity to Map and Sequence All Human Genes and Create a ‘Human Gene Dictionary.' Bethesda, Md.: Howard Hughes Medical Institute.
Proctor, R. 1988. Racial Hygiene. Cambridge, Mass.: Harvard University Press.
Reeders, S. T., M. H. Breunig, K. E. Davies, R. D. Nicholls, A. P. Jarman, D. R. Higgs, P. C. Pearson, and D. J. Weatherall. 1985. A highly polymorphic DNA marker linked to adult polycystic kidney disease on chromosome 16. Nature 317:542-544.
Reilly, P. 1977. Genetics, Law, and Social Policy. Cambridge, Mass.: Harvard University Press.
Riordan, J. R., J. M. Rommens, B.-S. Kerem, N. Alon, R. Rozmahel, Z. Grzelczak, J. Zielenski, S. Lok, N. Plavsic, J.-L. Chou, M. L. Drumm, M. C. Iannuzzi, F. S. Collins, and L.-C. Tsui. 1989. Identification of the cystic fibrosis gene: Cloning and characterization of complementary DNA. Science 245:1066-1072.
Roberts, L. 1987. Human genome: Questions of cost. Science 237:1411-1412.
Roberts, L. 1989. New game plan for genome mapping. Science 245:1438-1440.
Rommens, J. M., M. C. Iannuzzi, B.-S. Kerem, M. L. Drumm, G. Melmer, M. Dean, R. Rozmahel, J. L. Cole, D. Kennedy, H. Hidaka, M. Zsiga, M. Buchwald, J. R. Riordan, L.-C. Tsui, and F. S. Collins. 1989. Identification of the cystic fibrosis gene: Chromosome walking and jumping. Science 245:1059-1065.
Rothstein, M. A. 1989. Medical screening and the employee health cost crisis. Washington, D.C.: Bureau of National Affairs.
Royer, B., L. Kunkel, A. Monaco, S. Goff, P. Newburger, R. Baehner, F. Cole, J. Curnutte, and S. Orkin. 1987. Cloning the gene for an inherited human disorder—chronic granulomatous disease—on the basis of its chromosomal location. Nature 322:32-38.
Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences (USA) 74:5463-5468.
Schwartz, D. C., and C. R. Cantor. 1984. Separation of yeast chromosome-sized DNAs by pulsed field gel electrophoresis Cell 37:67-75.
Sinsheimer, R. 1989. The Santa Cruz workshop, May 1985. Genomics 5:954-956.
Solomon, E., and W. F. Bodmer. 1979. Evolution of sickle cell variant gene [letter]. Lancet 1:923.
U.S. Congress, House. 1987a. Fiscal Year 1988 DOE Budget Authorization: Environmental Research and Development (No. 58). Subcommittee on Natural Resources, Agriculture Research, and Environment of the Committee on Science, Space and Technology, U.S. House of Representatives.
U.S. Congress, House. 1987b. Departments of Labor, Health and Human Services, Education, and Related Agencies Appropriations for 1987 (Part 4A). Subcommittee of the Committee on Appropriations, U.S. House of Representatives
U.S. Congress, House. 1988. Departments of Labor, Health and Human Services, Education, and Related Agencies Appropriations for 1989 (Part 4A). Subcommittee of the Committee on Appropriations, U.S. House of Representatives
U.S. Congress, House. 1989. Hearing on International Cooperation in Mapping the Human Genome. October 19, 1989, 2325 Rayburn House Office Building. Subcommittee on International Scientific Cooperation, Committee on Science, Space, and Technology.
U.S. Congress, Office of Technology Assessment. 1984. Commercial Biotechnology: An International Analysis. OTA-BA-218. Washington, D.C.: U.S. Government Printing Office.
U.S. Congress, Office of Technology Assessment. 1986. Technologies for Detecting Heritable Mutations in Human Beings. Washington, D.C.: U.S. Government Printing Office.
U.S. Congress, Office of Technology Assessment. 1988a. Mapping Our Genes—Genome Projects: How Big? How Fast? OTA-BA-373, Washington, D.C.: U.S. Government Printing Office.
U.S. Congress, Office of Technology Assessment. 1988b. Medical Testing and Health Insurance. Washington, D.C.: U.S. Government Printing Office.
U.S. Congress, Senate. 1982. Nominations. Committee on Labor and Human Resources.
U.S. Congress, Senate. 1987a. Department of Energy National Laboratory Cooperative Research Initiatives Act (S. Hrg. 100-602, Pt. 1). Subcommittee on Energy Research and Development of the Committee on Energy and Natural Resources. Washington, D.C.: U.S. Government Printing Office.
U.S. Congress, Senate. 1987b. Workshop on Human Gene Mapping (100-71). Committee on Energy and Natural Resources. Washington, D.C.: U.S. Government Printing Office.
U.S. Congress, Senate. 1988. Departments of Labor, Health and Human Services, and Education and Related Agencies Appropriation Bill, 1989, Report (Report 100-399, pp. 83-84). Senate Committee on Appropriations.
U.S. Congress, Senate. 1989. The Human Genome Project and the Future of Biotechnology. Subcommittee on Science, Technology, and Space, Committee on Commerce, Science, and Transportation. (S. Hrg. 101-528). Senate Committee on Commerce, Science and Transportation. Washington, D.C.: U.S. Government Printing Office.
Watson, J. D. 1986. Foreword. Pp. xv-xvi in Molecular Biology of Homo sapiens. Cold Spring Harbor Symposia on Quantitative Biology, Vol. 51. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory.
Watson, J. D. 1988. The NIH Genome Initiative. Los Angeles: Molecular Biology Institute, University of California.
Watson, J. D. 1990. The human genome project: Past, present, and future. Science 248:44-49.
Watson, J. D., and F. H. C. Crick. 1953. Genetical implications of the structure of deo xyribonucleic acid Nature 171:737-738.
This case study abounds with interesting opportunities for both retrospective and prospective analyses of science policymaking. Moreover, the study 's chronicle of the personalities, events, and decisions that led to the creation of the genome project provides a valuable record for evaluating the wisdom and effectiveness of the actions taken. Cook-Deegan's analysis also points out those policy decisions whose validity will have to be judged by events yet to come. Thus, this case study fulfills the purpose for which it was commissioned. It provides a record that is worth scrutinizing to determine how the policy and funding decisions were made and whether there are general and applicable lessons to be learned for advancing other biomedical programs.
Any new initiative in science funding needs a highly visible, easily understandable goal, and champions who can articulate that goal persuasively in the offices of influence and power. By almost anyone's criteria, mapping and solving the human genome's sequence was viewed as a bold and exceedingly ambitious scientific and technical challenge, but one that would very likely be expensive in terms of resources, personnel, and funding. Aside from its initial influential group of proponents —Sinsheimer, DeLisi, Gilbert, Dulbecco, Hood, and Cantor—the plan to sequence the entire human genome, when it
Paul Berg is the Willson Professor of Biochemistry and Director of the Stanford University Medical School's Beckman Center for Molecular and Genetic Medicine.
became known, was met with considerable consternation and resistance. After some vacillation, Watson emerged as the genome project's chief advocate and proceeded to orchestrate support from several practicing and committed scientists. Concomitantly, DeLisi and Wyngaarden managed the legislative and budgetary complexities characteristic of governmental bodies. Senator Domenici, concerned for his DOE and state constituencies, and Senator Chiles, protective of the NIH, dominated the congressional debate and its outcome.
Another influential participant was the NRC Committee on Mapping and Sequencing the Human Genome. Their report expressing unanimous support (including that of previously skeptical or opposing members) carried a great deal of weight with the scientific community and helped persuade Wyngaarden, the OMB, and the congressional bodies dealing with the proposal. The report prepared by OTA (Mapping Our Genes—Genome Projects: How Big? How Fast?; 1988a) also helped move the genome proposal toward acceptance by supporting its feasibility and emphasizing the wisdom of a phased program. The report also provided a more reasoned estimate of its costs and the ways they could be apportioned. However, the NRC and OTA reports diverged in their recommendations as to who should manage the project. The NRC chose not to express a preference for the managing agency, while OTA preferred a joint NIH-DOE management structure —a solution that may become a stumbling block in later stages of the project. However, the recently published NIH-DOE combined plan for the first five years of the project, harbors well for a cooperative collaborative research effort. This arrangement should also alleviate the often-voiced suspicions about the quality of DOE research and its peer review systems. But competition for funding and recognition between DOE's national laboratories and NIH-sponsored genome centers could threaten the presently well intentioned cooperation. Mutual understanding of research activities and a strong spirit of team play will be important ingredients of that partnership.
A puzzling feature of the discussions concerning management of the genome project is why the NSF never emerged as a contender, especially as the project is indisputably science and technology based. Was this because of NSF's lack of interest or lack of competence to manage such a project? Or was it because of the project's decidedly biomedical slant?
One aspect of the scientific debate is worth noting. As initially conceived, the human genome project's goal was to obtain the sequence of the 3 billion base pairs comprising the haploid genetic complement. But the debate among scientists broadened the scope of the project in several significant ways. One was to include a moderate resolution map of linked restriction fragment length polymorphism (RFLP) markers for use in locating disease genes. A second modification was to obtain a physical map of cloned DNA segments spanning the entire genome. Moreover, it soon became apparent
that the sequencing efforts would have to await the completion of the first two objectives, as well as the development of faster, more accurate, and far less expensive sequencing technologies. Such a reformulation of the project was inevitable once fine scientific minds turned to developing a coherent and workable strategy.
Another major modification of the original plan was to increase the number of genomes to be included in the project. This modification was not intended to add funds to an already large project, but was a result of the recognition of the close correspondence in genetic structures and functions between even distantly related organisms. Furthermore, it was evident that such relatedness would inform and speed the work on the human genome. Additionally, research with yeast, Drosophila, nematode, and mouse genomes provides experimental models with which hypotheses and technologies can be tested without resorting to human experimentation.
Cook-Deegan points up several innovations in government-supported science. Besides the somewhat novel joint and coordinated sponsorship of genome research by NIH and DOE, there is the acknowledgment that ethical and cultural values need to be reconciled with the program 's objectives and applications. Including public education in that function would be helpful because one of the ingredients lacking from the debate so far has been any evidence of significant public awareness of the human genome initiative's costs and implications. The amount of funding for education and ethical studies is not as important as the fact that an ethics panel exists and that congressional support is contingent on attention to such issues. Whether ethical acceptability becomes a significant factor in judging the permissibility of initiating certain lines of basic research remains to be seen. Judging the ethical value of basic research prospectively and preemptively would be a considerable departure from current practices.
Cook-Deegan's record of the project's prenatal history identifies the contending forces that shaped the debate. These include the competing research interests and priorities among the “power” and “fringe” scientists, the similar tensions in the executive agencies governing biomedical research, and the jurisdictional sensibilities of congressional committees. Even though the project gained official sanction, debate over it lingers and threatens to flare up and polarize the constituencies needed for the project's support.
One of the most disturbing threats stems from the biomedical science community's still divided views concerning the project's merit, particularly in view of its perceived impact on traditional ways of supporting science; the simplistic view of this challenge is “Big Science versus Little Science.” This debate is exacerbated by the current dismal research funding situation and the bleak prospects for any remedies. The present and next year's genome budgets can hardly be responsible for the current crisis. We can anticipate that unless the virtues and benefits of the early-phase mapping studies, the
availability of clone banks, and the development of improved sequence technology are seen to benefit all scientists, the sniping will continue. Moreover, if the genome budget rises to its projected steady state —$200 million per year—while the research grant crisis worsens, the project's continuation will be jeopardized. Alternatively, funding for the genome project could be stretched out, as was done in the space program.
The case study does not indicate the extent to which the scientist activists, various advisory panels, or government agencies seriously considered the consequences of a fiscal crisis on Congress's ability or willingness to support the genome initiative. Perhaps this is because throughout the debate it was assumed that the project would be funded by supplemental (incremental) appropriations to NIH and DOE. Even though explicit assurances of the validity of those assumptions were never made, discussions and negotiations seemed to proceed as if the project's funding would not detract from support of investigator-initiated research in areas unrelated to the genome project.
The current funding crisis for projects unrelated to the genome project and the subsequent reactions of scientists and Congress indicate that neglecting the possibility and consequences of a funding shortfall was a serious deficiency in the planning. Perhaps that oversight stems from the pace with which the project was launched. Hindsight suggests that a more deliberate review, particularly by “disinterested ” participants, could have led to contingency planning to contend with present or future difficulties in biomedical science funding.
Ernest R. May
This story begins in the mid-1980s. More or less at once, a number of geneticists and molecular biologists saw the possibility of mapping and sequencing the entire human genome. A few developed visions of an immense and expensive project, making an analogy with the space program. The analogy the case brought to my mind was the opening of Africa in the nineteenth century and the response of Victorian imperialists such a Cromer, Curzon, and Rhodes.
For these visions to become reality, several things were needed. One was wide approval among knowledgeable scientists. Another was evidence that something appropriate could be done both bureaucratically and politically. Could any agency do it? Would the taxpayers countenance it?
The first requirement was satisfied. Despite reserved commentary in Science and Nature, the visionaries succeeded in winning wide backing from their peers. An NAS panel, together with advisory committees in both DOE and NIH, helped measurably.
Meanwhile, DOE-NIH competition helped both agencies come to readiness to do something on a large scale. OTA, abetted by or abetting key congressional staffers, worked out means such that the competition could end in cooperation rather than paralysis.
In the larger world, local interests of New Mexico, Florida, and California enabled congressional leaders to push the project. Appeals to the sense of
Ernest R. May is Charles Warren Professor of History at Harvard University.
what might make a legislator's name live in history also helped. Watson's inspired decision to commit a percentage of project budgets to the study of ethical issues defused possible opposition.
If this is the story—more or less—what are its lessons? Henry Kissinger used to say that no proposition was interesting unless one could imagine an intelligent person arguing the exact opposite. One way of asking about the lessons of the genome case is to ask what interesting propositions it speaks to. I see three.
The first is the proposition that there should be consensus among scientists before efforts are made to construct science policy. The affirmative brief would say that otherwise, policy could be constructed on erroneous premises and/or be vulnerable in the public arena because of evidence of expert differences. The negative brief would argue, among other things, that policy may never get made if it waits for scientific consensus and perhaps science policy is too important to be left to scientists.
The genome case seems to support this latter brief. The development of policy proceeded while scientists were still actively disagreeing. DOE/Office of Energy Research and OMB defined objectives, mechanisms, and tentative funding levels at a time when, according to the case, a poll of attendees at the Cold Spring Harbor conference would have voted against any early undertaking. Of course, the policy choices defined in mid-1986 were later to change, but the later effort would have been slower and different without this spadework.
The second proposition is that scientists, in the process of developing a large, pathbreaking project, should think about the bureaucratics of the project as well as about the science. Affirmative: the key question is not just what to do but how, in what sequence, and by whom. Negative: what to do is a large enough question, and it is the only one that most scientists are competent to consider.
Here, too, the genome case seems evidence for the negative. The efforts of the NAS panel to deal with bureaucratic implementation were, says the case, appallingly naive. It didn't matter. The bureaucracy, OTA, and congressional committees understood implementation issues. They dealt with them.
The third proposition is that, in developing such a project, scientists should think about politics—that is, about the types of citizen concerns that might surface once commitment of public money enters open debate. The affirmative case is summarized in the charter for this Institute of Medicine committee. The negative case is rather like that for the second proposition—too complicated, not scientists' strong suit, someone else (more expert) should carry the can.
On this third proposition, the genome case testifies for the affirmative as well as the negative. DeLisi got as far as he did in part because he recognized,
and could take advantage of, New Mexicans' concern about the future of Los Alamos. Watson clearly pushed a hurdle out of his way by manifesting awareness of fears that the genome project might awaken. But the case, as written, does not show that thought about politics needed to be a major component of debate among scientists. So long as a few key figures kept the public in mind, that sufficed.