Gene therapy in the treatment of disease

Abstract

Gene therapy or transfer is increasingly used as a therapeutic modality in diseases that span the medical spectrum. The American Academy of Allergy and Immunology and the National Institute of Allergy and Infectious Diseases jointly sponsored the Symposium, “Gene Therapy in the Treatment of Disease,” which was held December 6, 1993 at the National Institutes of Health (NIH). The program committee consisted of Lawrence Lichtenstein, MD, PhD, (Baltimore) Chairman; Philip Fireman, MD, (Pittsburgh); Howard B. Dickler, MD, (Bethesda); William Paul, MD, (Bethesda); and Raif Geha, MD, (Boston). Eight outstanding presentations on current and potential uses of gene therapy were provided by speakers who are leading researchers in this area of broad interest and unusually promising clinical application. This report summarizes the highlights of those presentations. Dr. Harry L. Malech (NIH) began the meeting with an overview of human gene therapy. He noted that the history of gene therapy is relatively short. In 1988, the Recombinant DNA Advisory Committee of the NIH approved the first guidelines for clinical trials. The first clinical studies were done in 1989 and involved the introduction of gene markers into patients with melanoma. In 1990, the first therapeutic trials began with the introduction of the adenosine deaminase (ADA) gene into the T cells of a patient with immunodeficiency. Today, there are over 60 approved protocols.1Anderson WF. Human gene therapy.Science. 1992; 256: 808-813Crossref PubMed Scopus (702) Google Scholar Gene therapy broadly involves the delivery of DNA encoding many different types of proteins. In addition to enabling the production of natural protein products as therapy, gene therapy also enables the production of engineered proteins (e.g., fusion proteins containing toxins, immunoregulatory proteins, receptors or their ligands); produces cell types resistant to chemotherapy, infection, or immune rejection; delivers ribozymes, decoy DNAs, or DNA binding proteins to prevent viral infection, such as human immunodeficiency virus (HIV)-1; and immunizes patients with the introduction of genes, including the injection of naked DNA. Several essential factors are needed as a foundation before human gene therapy trials should be considered. First, the disease gene or therapeutic gene must be cloned and thus, the sequence known. Second, the functional properties of the gene product must be determined. Third, the introduction of the gene should be tested in an in vitro or laboratory model of the disease. If possible, gene therapy in an animal model of the disease should also be examined. Human trials should then be considered for studying the safety and efficacy of the gene transfer technology.2Wivel NA Walters L. Germ-line gene modification and disease prevention: some medical and ethical perspectives.Science. 1993; 262: 533-538Crossref PubMed Scopus (70) Google Scholar An ideal gene transfer delivery system would ensure high rates of gene transfer, target the gene to specific cell types, regulate when in cell differentiation the product would appear, express the correct amount of gene product, and allow temporary or permanent expression. The delivery system should have no pathogenic effects itself and induce no unwanted immune response. Three main vector systems are in use. First, naked DNA can be injected alone, with cationic lipids, or attached to gold particles. The advantages of this system are: (1) it is chemically defined, and (2) high transient expression of the gene is obtained. The disadvantage is poor insertion of the gene into the genome. Second, retroviral vectors, after deletion of the viral proteins and insertion of the gene of interest, have been used. Efficient insertion of the gene into the genome is obtained, but mitotic cycling of the targeted cells is required for the uptake of the DNA. Another limitation is that high titers of the virus cannot be obtained for use because of the fragile physical nature of the virus. In addition, there is a risk of transfer of helper virus with this system. Third, adenovirus is being investigated as a vector. The advantages of this virus are: (1) it can be obtained in high titers, (2) high levels of expression of protein are possible, and (3) cell replication of the targeted cell is not required. However, the expression is transient with low rates of genomic insertion. Vector systems in development include adeno-associated virus. This virus allows efficient insertion into the genome and may not require replication of the targeted cells, but there has been variability in the titers of virus that can be produced. Herpes virus, which is neurotropic, should be effective in targeting the central nervous system, but the development of this virus as a vector is still in the early stages, and pathogenic effects have been observed. Development of special packaging cell lines is necessary for production of all replication-defective viruses. Gene therapy can be administered in several ways. First, the altered vectors or the DNA itself can be directly injected, or the vectors can be introduced into the respiratory tract. Second, an artificial secretory organ, that is a collagen matrix or semipermeable bag containing cells secreting the product, can be implanted. Third, cells can be removed from the patient, altered in the laboratory to contain the gene of interest, and given back to the patient.3Mulligan RC. The basic science of gene therapy.Science. 1993; 260: 926-932Crossref PubMed Scopus (1511) Google Scholar Studies on development of gene therapy for chronic granulomatous disease illustrate the steps in the development of gene therapy for an inherited deficiency disease. Chronic granulomatous disease is a rare inherited immunodeficiency, affecting two to four persons per million with life-threatening pyogenic infections and granuloma formation. The basic cellular defect is an abnormality in the production of hydrogen peroxide and superoxide by phagocytic cells. With the cloning of the genes responsible, it became clear that chronic granulomatous disease is not one disease but four distinct diseases, each involving a separate gene defect. The oxidase consists of three cytoplasmic proteins and a heterodimeric flavocytochrome in the membrane. The gene for each of these proteins has been cloned. When pathogens enter the phagosome, the proteins are activated and the complex is formed; then the reduced form of nicotinamide-adenine dinucleotide phosphate transfers electrons to oxygen to produce superoxide and then hydrogen peroxide. Bone marrow transplantation can be curative, and carrier mothers, who are normal and have no problem with infection, may have as little as 8% normal phagocytes. This suggests that correction of only a small percentage of the patient's cells may have a marked effect on the clinical course. Each of the component genes has been put separately into replication-deficient retroviruses. These have been used to attempt to correct in vitro the defect in the patient's blood progenitor cells with a defect in that gene. In each case the introduction of the vector carrying a normal copy of the patient's defective gene has resulted in correction of the defect in phagocytes derived from the gene-corrected progenitors. The percentage of cells carrying the normal gene varies and suggests that these cells may need to be selected before reintroduction into the patient. This disease provides a model of the steps before beginning human trials. Dr. Richard A. Morgan (NIH) reviewed the progress in developing gene therapy for the treatment of acquired immunodeficiency syndrome (AIDS). There are three potential advantages of gene transfer therapy for HIV infection. First, the ability to deliver the treatment directly to the cell that is the target of HIV would minimize toxicity and side effects. Second, the gene product could be produced continuously by the cell, obviating the need for repeated treatment. Third, the production of the treatment by the cell could be responsive to HIV's regulatory signals, such as the tat and rev proteins. Four general issues must be approached in development of gene therapy to treat HIV infection. First is the general and specific safety of the procedure. Culturing cells from HIV-infected patients requires safeguards to prevent the spread of the virus. The effect of the treatment on the normal function of the patients' cells must be considered. Second, the specific cells to target for gene transfer must be selected. The CD4+ peripheral blood lymphocytes and other CD4+ cells are usually selected because they are easy to obtain and grow and represent a well-defined population of cells. However, introduction of genes into stem cells may be more appropriate for the regeneration of the patient's immune system with a protective block to HIV infection. Third, a delivery system for the gene transfer must be selected. Viral delivery systems have been used most frequently, but other nonviral delivery systems may be considered. Fourth, the specific strategies for the inhibition of the HIV infection must be developed.4Morgan RA Anderson WF. Human gene therapy.Annu Rev Biochem. 1993; 62: 191-217Crossref PubMed Scopus (276) Google Scholar HIV is a standard retrovirus that binds to the cell surface and enters the cell. Once inside, its genomic RNA is reverse-transcribed into double-stranded DNA, which is integrated into the host cell's genome and becomes a permanent part of the host cell's chromosomes. HIV initially produces mRNA that codes for a series of regulatory proteins. The two most important of these are rev and tat. Tat increases the production of the HIV's mRNAs when it binds to the TAR, a specific sequence near the 5′ terminus of the HIV RNA. Rev is responsible for inducing the processing and transport of the mRNA for the late genes of HIV, which code for its structural proteins. Rev binds to an internal region of the HIV RNA called the rev responsive element. Strategies designed to inhibit HIV infection fall into three categories. The first are schemes to induce intracellular immunity. In this case cells that are CD4+ and thus targets for HIV are programmed to inhibit HIV infection intracellularly. One approach is for the cell to produce nucleic acid decoys for the binding of the regulatory proteins, tat and rev. In addition, some mutated HIV proteins, either regulatory or structural, can act transdominantly to prevent infection. These mutant proteins probably interfere with the formation or function of necessary protein complexes. Also, cells can be programmed to produce a suicide gene product only when infected with HIV. The second category is that of cell-independent protection schemes. These schemes could protect infected cells themselves and potentially their neighboring cells. The classic example is the use of soluble CD4 or of CD4 chimeric proteins to prevent binding of HIV to the cell. Cells could also be engineered to make HIV antigens, which could elicit a secondary immune response to infected cells. Cells could be programmed to produce generalized antivirals, such as interferon, or appropriate cytokines, which could upregulate the immune system. The last category includes schemes that result in the release of defective virus particles. The initial infection by HIV would not be inhibited, but the virus would not be able to spread to other cells. Examples include mutations in the envelope gene or mutations in the p17 matrix protein.5Yu M Poeschlo E Wong-Staal F. Progress towards gene therapy for HIV infection.Gene Therapy. 1994; 1: 13-26PubMed Google Scholar Dr. Morgan described the use of the retroviral delivery system to transfer several different anti-HIV genes into peripheral blood lymphocytes (PBLs). These cells were then challenged with both laboratory and clinical isolates of HIV to determine the ability of various strategies to inhibit HIV infection. The efficiency of transduction of PBLs without selection was poor, but with the addition of selection using neomycin resistance, more than 50% of the cells carried the engineered gene. Cells producing soluble CD4 or a CD4-IgG fusion protein inhibited the amount of HIV made in the cells, but on challenge with the clinical isolate, did not completely prevent spreading of the virus in the culture. The kinetics of production of HIV were shifted, but eventually the culture produced the same levels of virus as the control culture in this protein decoy system.6Vanden Driessche T Chuah MKL Morgan RA. Gene therapy for acquired immune deficiency syndrome.AIDS Updates. 1994; 7: 1-14Google Scholar Genes for antiviral or cytotoxic products, such as interferon, diphtheria toxin, or cytosine deaminase, can be engineered to be induced by HIV infection. The interferon-α gene was engineered to be controlled by the tat and rev proteins of HIV. Both proteins were needed to get good induction of interferon-α in these cells. These cells showed a greater than 2 log reduction in the amount of HIV made on challenge compared with control cells in this HIV-inducible protein system. One disadvantage of this construct is that the cells should be filled with defective immature HIV particles, so the cell may not function normally. Mutation of certain HIV proteins can have a transdominant effect and inhibit the production of HIV. The introduction of a mutant gene for the rev regulatory protein with either a point mutation or a frameshift mutation resulted in a decrease in the production of HIV in PBLs on challenge with either a laboratory or clinical isolate. This is thought to be due to a transdominant effect of the mutant protein on the production of the functional complex. In addition to the introduction of protein genes, the introduction of nucleic acids can interfere with the production of HIV. One example is the introduction of anti-sense of the TAR region. A cell transduced with a retrovirus carrying a fusion of tRNA and TAR anti-sense produced less HIV than control cells when challenged with a clinical isolate. One exciting extension of the above-mentioned studies is the combination of multiple vectors in an attempt to interfere with HIV infection at multiple stages. A cell making soluble CD4 and mutant transdominant rev did not show any increase in the inhibition of HIV above that of the rev transdominant mutant alone. However, soluble CD4 may have a beneficial effect for the patient above the inhibitory effect on the production of HIV itself. Another combination of a rev transdominant mutant and anti-sense TAR was very effective in inhibiting production of HIV in PBLs. These studies are an exciting start to the long-term goal of developing an effective anti-HIV therapy. Dr. Alan H. Beggs (Children's Hospital, Boston, Mass.) gave an overview of gene therapy for the treatment of muscular dystrophy, in particular Duchenne's and Becker's muscular dystrophy. Duchenne's muscular dystrophy is characterized by recurrent necrosis and regeneration of muscle fibers, leading to weakness evident at 3 to 5 years of age. It is severely progressive, with loss of ambulation between 11 and 13 years of age. This X-linked disease is the most common neuromuscular disease affecting 1 in 3500 boys. Becker's muscular dystrophy is a milder disease with a variable age of onset between 3 and 65 years. The clinical course is also variable, and some patients without muscular weakness have been described. The incidence of this disease is one tenth that of Duchenne's muscular dystrophy. Both these diseases are due to defects in the dystrophin gene. Dystrophin is a large cytoskeletal protein with homology to the spectrin superfamily. It consists of an actin-binding domain at the amino terminus, a rod-shaped domain of 24 helical units, and a carboxy terminus that binds a complex of integral and peripheral membrane proteins. In addition to all muscle types (skeletal, smooth, and cardiac), full-length dystrophin is also found in the brain. In muscle, dystrophin forms a network on the inner surface of the plasma membrane and is thought to give strength and flexibility to the membrane similar to the function of spectrin in the red blood cell. The gene for this protein is also large (2500 kb) and includes at least 79 exons. Deletions in this gene account for most of the mutations in muscular dystrophy. In Duchenne's muscular dystrophy, the mutations lead to a frameshift and thus a complete absence of functional protein. In Becker's muscular dystrophy, generally the deletions are in frame, and an internally deleted smaller protein is produced. This protein is thought to have partial function, which leads to the milder phenotype in these patients.7Ahn AH Kunkel LM. The structural and functional diversity of dystrophin.Nat Genet. 1993; 3: 283-291Crossref PubMed Scopus (568) Google Scholar Several unique features of skeletal muscle are important in designing a rational therapy for muscular dystrophy. Muscle is a large syncytium with thousands of nuclei, and the protein produced from mRNA from each nucleus has a restricted distribution in the myofiber. Thus the transgene must be targeted to most of the nuclei in a given muscle. Because muscle is not an actively dividing tissue, vectors that require replication for integration may not be very effective in treatment of this disease. Several features of the pathogenesis of muscular dystrophy are also important for the design of gene therapy for this disease. First, even though dystrophin is absent from the time of conception, patients do not show weakness until 3 to 4 years of age. This suggests that the lack of dystrophin itself is not pathogenic; rather, successive rounds of degeneration and regeneration lead to eventual muscle wasting, which leads to weakness. A therapeutic effect is not likely after muscle wasting has occurred, suggesting that the timing of gene therapy will be crucial. Second, replacement of dystrophin in the skeletal muscle may not correct all of the defects found in these patients. Cognitive defects found in some of the patients are thought to be due to the lack of dystrophin in the brain. Many patients also have a dilated cardiomyopathy, which is thought to be due to the lack of dystrophin in cardiac muscle. Neither of these defects will be improved with replacement of dystrophin in skeletal muscle.8Karpati G Acsadi G. The potential for gene therapy in Duchenne muscular dystrophy and other genetic muscle diseases.Muscle Nerve. 1993; 16: 1141-1153Crossref PubMed Scopus (52) Google Scholar Because of the large size of the dystrophin gene and the limit on the size of the gene that can be transferred in many gene transfer vectors, it may not be possible to transfer the complete dystrophin gene. Minidystrophin, which results from a deletion of exons 17-48, was found originally in a family with very mild Becker's muscular dystrophy. Several groups are currently experimenting with the use of this gene for therapy. Another aspect of this disease that may confound the ability to interpret the results of gene therapy is the relatively frequent occurrence of revertant myofibers. Both in the mdx mouse, an animal model of muscular dystrophy, and in patients with Duchenne's muscular dystrophy, a second mutation in some myofibers leads to deletion of the original mutation, converting the defect into an in-frame mutation, with production of a partially functional protein. Thus interpretation of biochemical assays for dystrophin in patients treated with gene therapy must account for the possible effect of somatic reversion. Recently, transgenic mice expressing the full-length dystrophin cDNA under the control of a muscle-specific promotor and enhancer were produced. These animals showed an increase in the levels of dystrophin protein and functional improvement in contraction of the diaphragm. One interesting aspect of these animals was that gross overexpression of dystrophin in muscle cells did not lead to toxic effects.9Vincent N Ragot T Gilgenkrantz H et al.Long-term correction of mouse dystrophic degeneration by adenovirusmediated transfer of a minidystrophin gene.Nat Genet. 1993; 5: 130-134Crossref PubMed Scopus (192) Google Scholar Several groups are testing myoblast transfer in patients with muscular dystrophy. Although not technically gene transfer therapy, development of this technology will be necessary for the transfer of myoblasts engineered in the future. In these patients myoblasts from normal donors, usually fathers, are obtained, grown in culture for several cell divisions, and then transplanted by intramuscular injection into the patients. The results have been very disappointing. Several factors may account for the disappointing results. First, the patients have been 6 to 10 years old and may already have had irreversible muscle wasting. Furthermore, because the stem cells for myoblasts are not known, the selection process may be deleting the very cells that are needed. The donors have been older, and the number of cell divisions that the transplants have undergone may be substantial. In addition, the transplanted cells do not show good dispersion throughout the muscle. The only encouraging results that have been reported are in uncontrolled trials in patients also treated with cyclosporine, which may have a beneficial effect of its own. The direct injection of naked engineered DNA is appealing for the treatment of muscle disease because the delivery is directly to the affected tissue. However, the results of these studies have shown poor integration of the DNA and poor dispersion in the muscle. Use of a retroviral vector has been tried with the minidystrophin gene, but in very early experiments only 6% of fibers were positive for the protein. Finally, adenoviral vectors have shown the most promise for treatment of this disease. A construct containing the minidystrophin gene was injected in mdx mice. In injected muscles large numbers of myofibers expressed dystrophin, and the proportion of centronucleated fibers was reduced, suggesting a functional improvement related to the therapy. Significantly, this effect was maintained for up to 6 months, so it may not be necessary to continuously inject this agent. The tools necessary for gene therapy of Duchenne's and Becker's muscular dystrophy are now available. Clearly, many more experiments on improving vectors, dystrophin constructs, and strategies for introduction at appropriate times need to be done. However, the work to date is encouraging in that no adverse reactions to dystrophin overexpression have been reported. Hopefully, the next 5 years will show significant progress in treatment of these disorders. Dr. R. Michael Blaese (NIH) reviewed the use of gene therapy for the treatment of severe combined immunodeficiency caused by the absence of the ADA gene. In developing the first clinical protocol for gene therapy, several criteria were thought to be important. The patients should have a life-threatening disease for which existing therapy was inadequate. The gene involved should have been isolated and characterized. Preferably, the gene should code for a single-chain protein, and tight regulation of levels of the protein should not be required. There should be no defective protein produced from the defective allele. In addition, there should be evidence that ex vivo therapy would be effective. Patients with ADA deficiency have severe defects in T- and B-cell function and usually die of overwhelming infection in the first few years of life. The gene for the missing enzyme, which is a single-chain protein, has been cloned and characterized. The level of this enzyme found in persons with normal immune function varies from 10% of the normal level to 50 times the normal amount. Bone marrow transplantation can be curative in these patients. Thus these patients met the criteria important for the treatment of a disease with gene therapy and were particularly well suited for the first clinical trials of gene therapy.10Blaese RM Anderson WF Culver KW. The ADA human gene therapy clinical protocol.Hum Gene Ther. 1990; 1: 327-362Crossref PubMed Scopus (105) Google Scholar A retroviral vector was used to transfer the ADA gene into the T cells of the patients. This vector is characterized by integration into the genome of the host with high efficiency. This integration requires DNA synthesis and thus the use of an actively dividing cell, such as the cells of the lymphoid or hematopoietic system. Stem cells would be a better site of insertion of the deficient gene. However, because these cells are not actively dividing, the efficiency of transfer into them is poor. One disadvantage of the retroviral vector is that it cannot be obtained in high titers. This means that treatment of patients with the virus in vivo is not feasible. T cells offer several advantages for treatment, because much is known about culturing and manipulating them in the laboratory. These cells can be readily isolated from the patients, transduced with the retrovirus, and grown in cell culture. The cultured cells can be cloned and tested for activity of both the transfected agent and normal cell function before reinfusion. In clinical studies the cells were cloned immediately after transduction with retrovirus. Approximately 20% of the cells had taken up the gene. The amount of enzyme made by each of these cells was quite variable, most likely because of the effect of the site of integration of the retrovirus in the genome. The enzyme was functional and deaminated deoxyadenosine. Thus the gene could be put into the T cells of these patients and could correct the metabolic defect in these cells. In addition, the cultures of T cells with the transferred ADA gene lived as long as cultures of normal cells, whereas cultures of ADA-deficient cells had a shortened life span. This suggests that the corrected cells would have a survival advantage in the patients.11Blaese RM Culver KW. Gene therapy for primary immunodeficiency.Immunodeficiency Rev. 1992; 3: 329-349PubMed Google Scholar The patients were treated repeatedly with engineered cells. Although not stem cells, T cells can be long-lived. However, only those T cells that are collected, transfected with the gene, and reinfused are corrected. Because the T-cell repertoire is changing over time, the T cells that are isolated at one time may not be the same as those isolated at another time. Thus to correct cells from multiple parts of the T-cell repertoire, the isolation, correction, and reinfusion of T cells must be repeated. Two girls were treated in this first protocol. The first has achieved 15% to 20% of the normal level of ADA, and her T-cell counts have been normalized. Her corrected cells were shown to have a survival advantage. Initially, the CD8+ cells were disproportionately grown out in culture, corrected, and reinfused into the patient. A shift in the proportion of CD4 to CD8 cells was seen in this patient. In later treatments her CD4 cells were specifically selected for correction and reinfusion. Both patients have had delayed-type hypersensitivity responses and normal cellular responses, and both are producing antibodies. One of the children was homebound but now attends public school with no major infections. The other child's cells have, for unknown reasons, been much harder to transduce with the gene. The proportion of cells containing the gene in this patient range from 0.1% to 1%. However, she has shown significant increases in her immune function.12Blaese RM. Development of gene therapy for immunodeficiency: adenosine deaminase deficiency.Pediatr Res. 1993; 33: S49-S55PubMed Google Scholar Because the ideal is to correct the defect in the stem cell, for the last year, Dr. Blaese's laboratory has concentrated on developing stem-cell gene therapy. Stem cells can be isolated from the bone marrow, mobilized from peripheral blood by granulocyte colony-stimulating factor, or isolated from umbilical cord blood, in which they form a high proportion of the cells. One of the original girls had been treated with corrected mobilized peripheral blood stem cells. More recently, stem cells were isolated from the cord blood of three infants known to be ADA-deficient. These cells were transduced with retrovirus carrying the normal gene and reinfused. The production of ADA by lymphocytes in these patients will be evaluated in the near future. This offers an exciting new approach to the correction of this defect. Dr. Malcolm Brenner (St. Jude Children's Research Hospital, Memphis, Tenn.) reviewed the use of gene transfer and therapy in the treatment of cancer. Gene therapy for cancer treatment can be divided into modification of the tumor or modification of the host. One form of host modification involves transferring genes that confer resistance to cytotoxic drugs into hemopoietic progenitors to reduce the marrow toxicity of the treatment agent. This should allow use of higher doses and potentially a higher tumor response rate. There are two major potential problems with this approach. First, if malignant cells are present in the marrow, they would also become resistant, and the differential benefit would be lost. Second, toxicity to other organ systems may occur at higher doses, necessitating transfer of resistance to these tissues, which will in turn require much higher efficiency gene transfer than is currently obtainable. A second form of host modification involves the immune system. If an anti-tumor response (cytotoxic T cells or activated killer cells) already exists, it may be possible to augment anti-tumor effects by introduction of cytokine genes. This approach has been taken by Dr. Steven Rosenberg's group at the NIH. Alternatively, T lymphocytes could be transduced with chimeric receptor molecules, which both confer target cell specificity and generate T-lymphocyte effector function. Modification of the