The possibilities for applying the emerging techniques of molecular biology to human gene therapy were appreciated by a handful of people as early as the late 1960s, although at that time no one had the slightest idea of how human genes might be isolated. The first human gene would not even be cloned for another dozen or so years. But there were tantalizing clues of what might lie ahead. In the 1960s, a physician-scientist named Stanfield Rogers at the Oak Ridge National Laboratories made an extraordinary observation with the common wart-causing Shope papilloma virus. One of the genes carried by this virus encodes an enzyme called arginase, which degrades excess amounts of the amino acid arginine. In addition to its role as a building block of protein, arginine also plays a role of its own as an intermediate in the processing of nitrogenous wastes in the body. Rogers noted that in rabbits infected with the Shope virus, the levels of arginine in the blood were unusually low. This did not particularly bother the rabbits; in fact high levels of arginine can be toxic. But what struck Rogers was that a viral gene appeared to be carrying out an important metabolic function in a rabbit. Rogers became curious as to whether any of the laboratory personnel handling the infected rabbits had themselves become infected with the virus, a not uncommon occurrence in the laboratory. Upon removing blood samples from people in the laboratory, he found out that not only had some of them become infected (not a serious event in itself; the immune system can easily control papilloma infections), but they also had extremely low levels of arginine in their bloodstreams.

This was a profound and profoundly interesting finding. It showed that genes carried as part of a viral genome could alter normal physiological processes in a human being. Rogers immediately recognized the potential of viruses for delivering genes to human beings. A few years later, he came upon an opportunity to test his thinking in a very direct way on two young patients who were suffering from the toxic effects of excess arginine in their systems. He deliberately exposed them to the Shope virus in an attempt to reduce their arginine levels. Hoping to err on the side of caution, he used amounts too low to be effective, and neither good nor harm came of his experiment. But Rogers himself was severely criticized for taking what some saw as an unwarranted risk based on premature and less than compelling data. Nevertheless, the overall lesson about the ability of viruses to carry genes into human beings would not be forgotten.

The reality of gene therapy would be brought a step closer a decade later, with another experiment that would again be roundly criticized by the scientific community. The first human gene to be isolated and cloned (in 1977) was that for b-globin, one of the subunits of hemoglobin, the oxygen-carrying molecule of red blood cells. Defects in the b-globin gene can cause serious diseases such as sickle-cell anemia and thalassemia. Just two years after the b-globin gene had been isolated, a physician at UCLA, working with a basic scientist at his institution, tried introducing the gene into mouse bone marrow cells using a viral vector. The experiment worked; moreover, when the altered bone marrow cells were put back into mice, there was evidence that they survived and that the added gene was functioning in vivo. The researchers then applied to their university’s Human Subjects Protection Committee for permission to try the same thing in human patients with thalassemia. Unable to get permission from his home institution, the physician member of the team proceeded to recruit patients suffering from thalassemia in institutions outside the U.S. Several patients had samples of their bone marrow removed, exposed to the vector containing the b-globin gene, and reinfused into their bloodstream. The experiment failed to alter the course of their disease, although the patients appeared to suffer no harm from the procedure. The physician was taken severely to task by his institution, by the federal government, and by the international scientific community. The results of this first attempt at gene therapy using cloned DNA, and the overwhelmingly negative response it engendered, gave caution to those eager to seize the day and be the first to make the dream of molecular medicine come true. But it also became abundantly clear to everyone who was watching that the technology for gene therapy was now in hand, and that it would only be a matter of time before genuine clinical trials for gene therapy would begin.

Clinical trials represent the transition phase between highly promising basic laboratory research on a new drug or treatment method, and the general release of that drug or treatment to the larger medical community for use in standard therapy. Just a few decades ago, clinical trials were generally informally organized studies carried out by physicians, usually in collaboration with drug companies, to test a new drug, medical device or clinical procedure. These studies were often not standardized, lacked important controls, and did not pay attention to proper statistical analysis of data. As a result, many clinical trials produced data of little real value. Yet patients were put at risk during such trials, and patients subsequently treated by drugs or devices approved for general use as a result of faulty clinical trials were unknowingly also at risk.

Starting in the 1970s, the federal government began formulating specific sets of guidelines for clinical trials, and today all new drugs and invasive medical devices are subjected to rigorously controlled clinical tests before they are made available for general clinical use. Clinical trials in the U. S. are now overseen by the Food and Drug Administration, and the FDA has final authority for approving a new drug or invasive medical device for manufacture, marketing and general use by the medical community at large. Clinical trials are most often carried out in university medical centers under the guidance of physicians who also have strong basic science backgrounds, or have basic science consultants as part of the overall clinical trial team. In most cases, a potential manufacturer or marketer of a new drug or procedure will be an active partner in clinical trials, providing the drug itself and any other materials needed for the trials, and generally underwriting their costs.

All institutions sponsoring clinical trials must have an internal Institutional Review Board (IRB) to review clinical trial proposals before they are even submitted for FDA approval. It is the job of the IRB (called in some institutions the Human Subjects Protection Committee) to carry out an initial assessment of the scientific soundness of the proposal, and to be sure that the data collection and analysis procedures are valid and meaningful. It is also the IRB’s responsibility to determine that the proposed patient population is appropriate for the aims of the trial, and that proper patient informed consent procedures will be followed (see below.)

Although the human genes proposed for use in gene therapy might not seem to fit the category of a new drug, the fact that they will be delivered in viral or chemical vectors, that they will eventually be prepared for use by standard manufacturing procedures, and that we know little about the longterm effects of introducing extraneous DNA into human beings, suggests that the same caution should be exercised with DNA as with any other new drug until such time as the safety and efficacy of this procedure have been established beyond reasonable doubt.

Before a clinical trial can begin, the FDA must see compelling evidence from laboratory studies that a proposed new drug or procedure can actually do what its creators hope to achieve in humans. This preclinical phase of testing generally involves laboratory experiments with human cells grown outside the body, to gain insight into potential toxicity and to be sure that the drug will actually work on human cells. The next step is to test the drug or procedure in animals. This work often begins with rats and mice, for reasons of economy and because of the large backlog of experience with these animals and knowledge of how their physiology compares with humans. A very useful animal model that finds increasing use in drug testing is the so-called nude mouse. Nude mice have a genetic defect that prevents them from immunologically rejecting human cells and tissues. (A closely linked defect prevents them from developing fur; hence the "nude" designation.) It is thus possible to transplant into nude mice a small piece of the human tissue a new drug is supposed to affect, to inject that drug into the mice, and to monitor the drug’s effect on the human tissue under "in vivo" conditions. In some cases it may be appropriate to test the drug further on a larger animal before testing in humans, but increasingly the nude mouse model has been able to satisfy federal regulators.

Ultimately, of course, any new drug or procedure must be tested on human beings to be absolutely certain that it is safe, and that it has the effect intended by its developers. The FDA has developed very strict guidelines for conducting human clinical trials. The first principle of any clinical trial is fully informed consent of the human subjects who will participate in the trial. Patients (or their families, in those cases where patients are gravely ill) must clearly understand the experimental nature of the procedures they will undergo, the possible dangers they may face, and how the information gathered will be used. They must not be misled about potential benefits or improvements to their underlying disease. Confidentiality of the information gathered during the trial must be assured. A copy of information provided to each patient is reviewed by the FDA as part of the overall approval process for any new clinical trial.

Clinical trials are typically divided into four phases. Each phase is reviewed while it proceeds, and each phase must be completed and approved before the next phase can begin. Although the exact description of each phase may be slightly different for each new drug or procedure, the following general guidelines apply to the majority of trials conducted.

Phase I. The principal purpose of a Phase I clinical trial is to gather information on safety of the proposed drug or procedure in human beings. In the case of a new drug (including DNA used for gene therapy) investigators will look at things like how long the drug remains in the system, whether its properties change once it is inside a human body, and whether it causes any measurable side effects, either as reported by the subject or as detected in laboratory tests. Usually a range of doses of the drug will be tested, guided by previous toxicity tests in animals. However, in most Phase I trials patients will only receive a single administration of an experimental drug. The clinical tests may be carried out on persons with the disease for which the drug is intended, or occasionally they may be carried out on healthy volunteers. If the drug is to be tested on persons with active disease, it is usually the case that patients with very advanced disease, who have failed to respond to standard current therapies for the disease, are enrolled in the trials first. The number of patients involved in Phase I studies is small, no more than the number required to get statistically believable data for the points under study (typically a few to a few dozen individuals.)

Phase II. While the effectiveness of a new drug on the condition it is intended to treat (its efficacy) is monitored during a Phase I trial, efficacy is really the focus of Phase II trials. Dosage and toxicity limits derived from Phase I trials are used to design a larger- scale trial to begin assessing the value of the new drug or treatment as compared to existing treatments. Patients with less advanced stages of disease may be entered into Phase II trials, and may receive multiple administrations of the drug. Patients are still closely monitored for toxicity or side effects of any kind. (Healthy volunteers do not participate in clinical trials beyond Phase I.) Somewhat larger numbers of patients may be involved (in the range of a few dozen to a hundred or so.) Numerous controls are usually built into Phase II trials, including placebo treatments for some patient groups; patients with no previous treatment for their disease; patients receiving various combinations of standard therapies; etc.

Phase III. If drug efficacy with acceptable side effects can be established in Phase I and II trials, larger numbers of patients (hundreds to thousands) are enrolled in Phase III trials, usually in a variety of clinical settings (community hospitals as well as university medical centers.) Additional data on the interaction of the new drug with existing drugs used to treat the disease are gathered. Information gathered in Phase III will eventually be used to instruct physicians about use of the new drug; toxicity is still closely monitored. The drug developer will usually apply for formal approval from the FDA to market the new drug after successful conclusion of Phase III trials.

Phase IV. Occasionally, the FDA will approve a new drug for general use, but require the manufacturer to continue monitoring the effects of the drug for a limited period after general release. In Phase IV trials, the drug may be extended to slightly different patient populations than those studied in earlier trials, or dosages may be altered, or the drug tested in combinations with other, previously approved drugs. The drug may also be extended for use in related conditions not specified in the original trials.

Although viewed by some as unnecessarily rigid and time-consuming, there is no question that clinical trials carried out in the U.S. today produce sound, scientifically meaningful information about proposed new drugs, devices or procedures, and protect the patients involved in the trials. By the time human gene therapy was ready for clinical trials, the steps necessary for approval were fairly well defined. But as we will see shortly, gene therapy introduced some new complications and concerns that at least initially made the approval process even more difficult than it had been in the past.

The involvement of ADA deficiency in one form of SCID had been appreciated from the early 1970s, and the disease was recognized as a single-gene defect almost as soon as it was described. Moreover, long before the gene was cloned and sequenced, it was also realized that the different severities of ADA-SCID observed clinically would likely be due to different mutations in a single, common gene.

The gene encoding ADA was one of the earlier human genes to be isolated and cloned for study. Three separate laboratories published the gene sequence in 1983. By that time several strategies had been developed for going after a gene that had never been isolated before. It is never easy; the degree of difficulty depends upon how much information is available about the gene or its product before starting. For example, in the 1950s scientists began isolating and sequencing proteins. As alluded to in Chapter 3, getting the complete amino acid sequence of a protein is a daunting task, and can take years. Nevertheless, by the time restriction nucleases became available in the 1970s, which made isolating individual genes possible, the complete amino acid sequences of quite a few proteins were known, and partial sequences were available for a good many more. Knowing an amino acid sequence, and using the genetic code, it is possible to work backwards and predict the nucleic acid sequence of the corresponding gene. That information can then be used to chemically synthesize a complementary DNA probe that will detect the gene on a southern blot.

That was one of the approaches used to identify the gene for ADA. One of the laboratories had isolated the human ADA protein in the late 1970s, and by the early 1980s had obtained enough amino acid sequence data on the purified protein to predict a region of the DNA sequence that seemed likely to be useful as a probe (Figure 6-1). A stretch of DNA consisting of 17 nucleic acids (a "17-mer") was selected based on minimal predicted codon ambiguity (the confusion that can arise from the fact that more than one codon can specify a give amino acid). Even with a relative minimum of ambiguous codons, however, a total of 64 different 17-mer probes had to be synthesized to cover all possible combinations. These probes were used to screen a cDNA library made from human T cells, where the ADA gene was known to be expressed in high concentrations. The clones reactive with the probe mixture were then expanded, and the cDNA inserts isolated and sequenced. (Once a cDNA has actually been isolated, the cDNA itself becomes the probe of choice for future work.)

At the time all of this was done, the possibility of using cloned DNA for gene therapy still seemed to most researchers a distant goal; the main purpose of cloning the ADA gene was to get more information on the relationship of genetic mutations to protein function. The usefulness of cloned portions of the gene as probes for genetic screening of at-risk populations was, however, immediately recognized. (Genetic screening will be discussed in detail in Chapter 12.) But within a short time, the possibility of using the newly isolated ADA gene for the first human gene therapy trials was under serious consideration. One of the first to make a formal proposal to do so was W. French Anderson, a physician-scientist at the National Institutes of Health. Anderson had been among the pioneers in pushing other physicians as well as research scientists to pursue the possibilities of applying recombinant DNA technology to the treatment of human genetic disorders. He had trained with Marshall Nirenberg, and had been part of the team that isolated the second human gene, a-globin, in 1977. Anderson recognized that since ADA-SCID can be cured by a bone marrow transplant, the only critical locus for the ADA gene in this disease must be in a bone marrow cell, or in a cell derived from bone marrow. Moreover, since heterozygotes are perfectly normal, a single copy of a "good" ADA gene within the critical cell type must be sufficient to correct the underlying defect. ADA-SCID seemed made for gene therapy.

By 1987 Anderson had prepared an initial proposal for clinical trials using the cloned human ADA gene to treat patients with ADA-SCID. But first, Anderson had to clear his proposal through one of the most feared governmental regulatory agencies: the Recombinant DNA Advisory Committee. This committee, which oversees all federally funded research involving recombinant DNA, is referred to by all who go before it simply by its acronym: "the RAC".

The intrusion of the RAC into clinical trials involving recombinant DNA grew out of a defining moment in the history of molecular biology. Scientists began using restriction nucleases almost as soon as they became available in the early 1970s to make recombinant DNA molecules. Initial experiments involved nothing more than working out the conditions for cutting and stitching together pieces of DNA from various sources in test tubes. But before long scientists were eager to move on to more interesting possibilities. Some began carrying out experiments involving intact genomes removed from microorganisms such as bacteria and viruses. The first biologically functional recombinant DNA molecule was made by Herb Boyer and Stanley Cohen in 1973. They had placed a toad gene in a bacterial plasmid, and introduced the plasmid into a bacterial host strain; the bacteria promptly began making the corresponding toad protein. Paul Berg, a colleague of Boyer and Cohen and a pioneer in the field of molecular biology, also carried out experiments recombining DNA from the genomes of different bacterial viruses (phage) that infect the common bacterium E. coli. He had additionally begun recombining phage genomes with portions of the genome of a virus called SV-40, which causes cancer in monkeys.

The resulting recombinant genomes were, in effect, totally new life forms - never before seen among organisms that had evolved naturally over eons of time. The problem is that E. coli lives, among other places, in the human gut. How would these new life forms living inside them - the recombinant phage - behave? The new experiments with recombinant DNA were closely followed and widely known, and they were beginning to make some scientists slightly nervous. No one had ever before tampered with the genome of a living organism. When the resulting genome was associated with an organism that can infect human beings - and especially when it carried genes from a cancer-causing virus - it was felt that such tamperings needed a fuller discussion by the wider scientific community. There was no evidence that such experiments were in fact dangerous, but the suggestion was made that Berg and others should suspend further experimentation until everyone could get together for a talk.

The meeting to discuss the implications and possible risks of recombining the DNA of living organisms took place at the Asilomar Conference Center near Monterey, California, in 1974. All of the major laboratories working with recombinant DNA were invited, along with representatives of the federal agencies funding such research. To forestall any possible charges of a scientific elite meeting behind closed doors to decide the biological fate of humanity, the press was invited to participate. Many issues, some highly emotional as well as scientific, were aired in the formal sessions, but to an even larger extent over meals, in the lodge where people gathered in the evenings, and in strolls along the spectacular adjoining beaches. Would the creation of recombinant genomes interfere with normal evolutionary processes? Do human beings have the right to reach into nature and create new life? Is that interfering with divine purpose? Could these organisms eventually have any military - biological warfare - applications?

Out of this meeting came a proposal to ask the prestigious National Academy of Sciences to establish a committee to review the current status of recombinant DNA technology, and to advise the government whether and how such research using this technology might need to be regulated. Paul Berg himself, as a member of the Academy, chaired the resulting committee. Berg’s committee ultimately recommended that the National Institutes of Health establish a permanent Recombinant DNA Advisory Committee, which was formed almost immediately. Over the next two years, the RAC began formulating guidelines for carrying out recombinant DNA research funded by federal research grants. These guidelines were published by NIH in 1976, and were adopted immediately by virtually all laboratories in the United States carrying out such research, however funded, and were also eventually adopted in one form or another by most governments throughout the world. The press was invited to observe every stage of this process, and acted very responsibly in informing the public of what was taking place.

The early workings of the RAC dealt almost exclusively with safety issues for laboratory research. The major concern initially was that altered life forms would escape from laboratories and infect plants, animals or people on the outside. Guidelines for "containment" procedures, handling and storage of recombinant DNA, and other practical issues were disseminated to research laboratories throughout the country. Institutions sponsoring such research were required to establish Institutional Review Boards (IRBs) to ensure implementation of the RAC guidelines, and to assure that all applications for government research funding submitted to NIH met RAC standards. Mindful of the furor that had attended the earlier forays in the direction of human gene therapy, the RAC also established a permanent Human Gene Therapy Subcommittee to carry out initial reviews of all research involving human genes, whether intended for therapeutic purposes or not. The government later decided that any proposals to introduce DNA into human beings would also have to be subjected to clinical trials as defined by the Food and Drug Administration. Currently both the RAC and the FDA must approve clinical trials for gene therapy. There was a move on the part of some biotechnology companies and a few academics to have the RAC abolished, and to transfer sole authority for approving clinical trilas to the FDA. However, after a thorough review of the workings of the RAC it was decided that the unique role the RAC plays in assessing the quality and value of the basic science underlying, and likely to emerge from, clinical trials will be of considerable value for the foreseeable future.

The political and personal dramas surrounding the attempts to gain RAC and FDA approval for a human trial with the ADA gene have been admirably detailed in Jeffrey Lyons’ and Peter Gorner’s book Altered Fates. Initially, French Anderson had hoped to be able to introduce an ADA gene into hematopoietic stem cells from the bone marrow. Since these stem cells last for the life of an individual, and give rise to all cells of the blood - including the T cells crippled in SCID - correcting the gene defect in stem cells would mean permanent protection from disease. But try as he would, Anderson could never get stable transformation of human bone marrow stem cells using the retroviral vectors he had generated. The reason for his lack of success is now clear: retroviruses only infect rapidly dividing cells, and it turns out that the vast majority of stem cells from normal bone marrow are not actively dividing at any given time.

Anderson had almost given up when his colleague Michael Blaese suggested another approach - why not just transform the T cells themselves? If it worked, the effect would only be transient, because T cells do not live very long unless stimulated by foreign antigen. It would probably be necessary to repeat the procedure every few months. But those T cells stimulated by antigen do become long-lived "memory" T cells; in most normal adults, the overwhelming bulk of an immune response is carried by the pool of previously generated memory cells. Over time, Blaese reasoned, a SCID patient in whom even a small proportion of T cells were kept alive long enough to be stimulated by antigen might build up a repertoire of T cells capable of responding to most environmental antigens.

Although initially sceptical, Anderson gradually warmed to the idea. The specific delivery vector proposed by Blaese, Anderson, and their colleague Ken Culver for the initial ADA-SCID gene therapy trials with T cells involved a retrovirus that causes leukemia in mice, called MoMLV. Most of the viral genes were removed from the provirus form of MoMLV DNA, and replaced with a ds-cDNA form of an unmutated copy of the human ADA gene. The passenger ADA gene was placed under the control of the retrovirus’s own promoter to facilitate transcription into a functional mRNA. This recombinant DNA was then introduced into retroviral packaging cells for conversion to an infectious (but, outside the packaging cells, non-replicative) retrovirus.

Before submitting an application for the first clinical trials, these recombinant viruses were subjected to a number of intensive laboratory tests. Using T cells isolated from ADA-SCID patients, it was shown that the ADA gene delivered by the retroviral vector could in fact produce ADA in human T cells. Moreover, the resulting enzyme could prevent the buildup of toxic deoxyadenosine, preventing the premature death of the T cells. Next, the vector was used to deliver the ADA gene to the T cells of mice and of monkeys, and the altered T cells were then injected into live animals to see if they would continue to function. The results were highly encouraging. Finally, in the late summer of 1990, the RAC and the FDA were sufficiently convinced by the preliminary laboratory data to approve the first human gene therapy trials using the MoMLV-based delivery vector.

Months before, Anderson and other physician-scientists on his team had selected two young girls with ADA-SCID as potential candidates for gene therapy. Ashanti DeSilva, whom we met earlier, would receive the first transfer of the gene. Ashanti was in advanced stages of her disease. Standard therapies, such as PEG-ADA, were not working. Even before final FDA approval had been obtained for the entire gene therapy procedure, samples of T cells were collected from her blood and transfected with the ADA vector in vitro. The cells were first triggered to start dividing, in order to enhance penetration by the retroviral vector. After exposure to the virus, the cells were grown in an incubator for a week or so to expand their total numbers. Final FDA approval was received on the morning of September 14, 1990; that afternoon, four-year-old Ashanti was infused with her own T cells containing the MoMLV-ADA vector, and became the first human being in history to undergo gene therapy for therapeutic purposes.

As discussed in Chapter Four, the procedure went smoothly. After four infusions over a four-month period, Ashi’s T cell counts were climbing toward normal. The second young patient to be treated with this procedure, Cynthia Cutshall, received an infusion of her own T cells transformed with the same MoMLV vector on January 31, 1991. One of the conditions imposed by the RAC was that these young girls and all subsequent patients receiving ADA gene therapy also be maintained on PEG-ADA. This drug (ADA protein complexed with polyethylene glycol) had been approved by the FDA as a standard treatment for ADA-SCID in 1990, just shortly before the RAC approved ADA gene therapy. In many children it caused a marked initial increase in the number of T cells, alleviating many of the complications of ADA-SCID. On the other hand, some children gain little or no sustainable benefit after a few administrations of the drug, and it is enormously expensive - upwards of $200,000 per year for the average patient. Because PEG-ADA does not correct the underlying defect, but simply alleviates its symptoms, it must be taken regularly for the life of the patient. Both Ashi and Cynthia were being treated with PEG-ADA at the time they began gene therapy. Although Ashi did not seem to be responding to the drug, it was considered inappropriate to discontinue its use for either patient. Thus evaluation of the efficacy of gene therapy in its first and longest lasting trial is complicated by the continued administration of a drug whose end effect can be the same as that of the gene therapy; an increase in the number of viable T cells.

Nevertheless, direct analysis of Ashi’s T cells has shown that nearly all of them now express the newly inserted ADA gene, and she no longer receives regular infusions of gene-altered T cells. While Cynthia is showing a lower level of expression of the ADA transgene in her T cells, she is definitely using her new gene and is showing a gradual increase in the ability of her T cells to function normally.

Some investigators feel that the PEG-ADA may actually be working against the effectiveness of the underlying gene therapy treatment. PEG-ADA helps keep all T cells alive, whether or not they have been transduced by the ADA gene. From a gene therapist’s point of view, PEG-ADA is helping "bad" (untransduced) as well as "good" T cells to survive. There is reason to believe that, in the absence of PEG-ADA, those T cells transduced with a healthy ADA gene would have a selective advantage for survival, and would eventually outgrow and displace the untransduced T cells. However, they have been unable to convince the RAC and FDA to allow them to wean this initial set of patients off the PEG-ADA in order to find out.

That PEG-ADA may indeed be having a less than helpful effect when administered to patients undergoing ADA gene therapy is suggested by more recent studies based on French Anderson’s original idea of transducing stem cells rather than T cells in ADA-SCID patients. Dr. Donald Kohn, a pediatrician who had studied with Michael Blaese at NIH before joining Children’s Hospital in Los Angeles, was presented with an unusual opportunity in early 1993. Three at-risk infants were identified in utero as having ADA-SCID, through a combination of amniocentesis and analysis of their DNA with ADA-specific probes. This is an extraordinary coincidence, given the rarity of this disease. Because of his unique background and training, Kohn was able to convince the RAC to allow him to remove "cord blood" from these three infants at birth, and transduce it with the retroviral ADA vector used for the initial ADA-SCID trials with T cells. The umbilical cord is known to be an unusually rich source of stem cells for the blood cell system; cord blood has actually been used in lieu of bone marrow in certain transplant situations. Moreover, procedures had recently been developed for enriching the concentration of stem cells, taking advantage of a surface molecule on stem cells called CD-34. The RAC agreed to the new trials but insisted that, as in the earlier trials, the infants also receive PEG-ADA as standard therapy. But importantly Kohn was able to convince the RAC to allow him to gradually wean these infants from PEG-ADA over time.

This approach appears to be working. Tests carried out shortly after infusion of the altered cord blood cells showed that about .01 to .10 percent of the T cells in these infants were expressing the transgene. Although low, these numbers were real, and represented the first demonstration that human hematopoietic stem cells can be transduced with a retroviral vector. As the PEG-ADA has been reduced, the overall proportion of T cells expressing the transgene has increased to between one and ten percent - a 100-1,000-fold increase! Kohn and his colleagues are moving slowly with the PEG-ADA reduction, but are confident their infants will eventually be able to do fine without it. If they are successful, these youngsters (now over four years old) may be the first to achieve the ultimate goal of all gene therapy: a permanent cure of their disease through genetic manipulation alone.

The form of SCID that affected David the Bubble-Boy was, as we have seen, genetically distinct from ADA-SCID. The gene that, when mutated, causes X-linked SCID had not yet been identified during David’s lifetime, so gene therapy was never an option for treating him. The biochemical nature of the defect causing X-SCID was not even known; there was no protein known to be involved that would have allowed working backward to pick up the gene. As recently as 1990 it semed that identification of the X-SCID gene was still a long way off. But the isolation and cloning of this gene was greatly hastened by one of those marvelous situations in which the threads of apparently unconnected inquiries can suddenly entwine in unexpected ways with dramatic consequences.

The pursuit of the X-SCID gene began the hard way, with a detailed study of how the gene segregates in affected families, in order to pin down its chromosomal location as closely as possible. The gross chromosomal association - the X chromosome - was immediately obvious from the fact that this disease affects only males. But further refinements of its location on the X chromosome were slow to emerge because SCID is such a rare disease: there simply were not many families available for study of X-linked SCID inheritance, and individuals affected by the disease do not usually survive very long. Comparison of the inheritance pattern of the X-SCID gene with known X chromosomal markers had narrowed its location to Xq13. But what appears as a dot on a chromosomal map can be millions of nucleotides, much too large a region of DNA to clone and sequence, even if it could be isolated.

While these studies were in progress, other researchers, in what appeared to be a completely unrelated line of investigation, were busy trying to understand the structure of a molecule found on the surface of T cells called the IL-2 receptor (IL-2R.) T cells make and release a small hormone-like molecule called interleukin-2 (IL-2), which is used by other cells to help them respond to an infection. Interestingly, T cells themselves use the IL-2 they secrete. During an immune response, T cells responding to a particular antigenic challenge will release IL-2 into their immediate vicinity, and then use some of this IL-2 to help them proliferate and produce more of their own kind to help fight the infection (Figure 4-1). The IL-2 receptor allows them to pick up IL-2 and bring it inside. (A given T cell can use its own IL-2, or IL-2 produced by a neighboring T cell.) T cells also rely heavily on IL-2 during their development in the thymus. As the cells mature, they release IL-2 which they immediately take up through their IL-2R and use to help them undergo normal cell division. Without the ability to internalize exogenous IL-2, normal T cell development is aborted shortly after it starts.

The T-cell IL-2 receptor was initially thought to consist of two protein chains, referred to as IL-2Ra and IL-2Rb. But certain aspects of IL-2R function seemed inconsistent with this two-chain model, so researchers went back for a closer look. A third chain, called IL-2Rg, was indeed found. Enough of the g-chain amino acid sequence was determined to allow synthesis of a nucleic acid probe, which was then used to screen a cDNA library from a cell line known to be making IL-2R. The corresponding cDNA was isolated and sequenced; the results were published by a Japanese research team in the journal Science in the summer of 1992.

The description and subsequent characterization of the IL-2Rg chain explained many of the puzzles relating to IL-2R function. But it remained for a research team centered at the NIH, which had independently cloned the IL-2Rg gene, to make the connection of this gene with X-SCID. By hybridizing it directly with intact chromosomes, they were able to show that it bound only to the X chromosome, and specifically to the q13 locus. Moreover, they showed that DNA from three out of three X-SCID patients (including a DNA sample preserved from David the Bubble Boy) had mutations in the IL-2Rg gene; no mutations were found in ten normal individuals. These results were confirmed a few months later by one of the groups involved in the original chromosome mapping studies. Within a very short time it was clear that the IL-2Rg gene is the gene which, in mutant form, causes X-linked SCID.

So now we also have the gene for X-SCID, and through knowledge of the nature of its protein product we also know exactly how the disease is caused (Figure 6-2). As T cells begin to mature in the thymus, they soon reach a point where they need IL-2 to complete their development. In X-SCID children, there is plenty of IL-2 in the thymus, but the emerging T cells are unable to use it because they lack a functional IL-2 receptor to bring the IL-2 inside. As a result, X-SCID males are born without functional T cells, and hence with virtually no immune protection. The failure to display a fully competent IL-2R on other cells of the immune system further compounds the primary defect.

Portions of the X-SCID gene are already being used as a genetic screening probe to determine, through Southern blot tests on small blood samples, which daughters in an X-SCID family are carriers, and whether male fetuses or newborns in at-risk families have inherited a mutant form of the gene. Pre-clinical experiments with the IL-2g gene inserted into a retroviral vector have shown that this gene can be delivered to mouse bone marrow cells to B cells from X-SCID patients, and to CD-34-enriched cord cells. Clinical trial protocols involving gene therapy for X-SCID have been submitted for RAC and FDA approval, and may reasonably be expected to get underway sometime in 1977. There will be no more "bubble boys."