Gene Therapy: 
A New Frontier 
in the Treatment of Disease

In the middle of the 19th century, Gregor Mendel indicated that certain traits can be inherited. Subsequently, these discrete units were localized on chromosomes and termed genes. In this century, genes have been identified whose malfunction contributes to many diseases, such as diabetes, Alzheimer's disease, and some forms of cancer. In theory, gene therapy may be defined as a technique in which an absent or defective gene is replaced by a working gene so that the body can make the correct product, thereby restoring normal body function.

 

The practical use of gene therapy relates to the emerging relationship between the development of molecular technology, the characterization of gene structure and function, the ability to manipulate the genetic material of a cell to essentially create new organisms, and the understanding of defects that underlie a variety of human genetic disorders. In this manner, it is possible to begin to devise new strategies that will exponentially increase our potential for preventive and palliative health care.

 

The understanding of nucleic acid structure, genetic organization, and gene function provides the fundamental basis for the potential use of gene therapy in patient care. For that reason, a short discussion of nucleic acid chemistry, DNA (deoxyribonucleic acid) and RNA (ribonucleic acid) structure, and gene isolation is provided. This is intended both as a review and orientation for the reader. As such, the material and concepts discussed in this section are an a priori requirement for a consideration of the principles and utility of gene therapy. For a more detailed discussion, the reader should consult a general textbook on biochemistry.

 

Sugar moieties in nucleic acids provide a significant, albeit unappreciated, component of nucleic acid structure. First, bases are connected to the sugar at the 1´-carbon. In RNA, the ribose sugar contains an OH group at both the 2´- and 3´-carbons. Water breaks the ribose sugar in a nonenzymatic reaction. In DNA, the sugar, termed a deoxyribose (ie, without oxygen), contains a 3´-OH group but only a hydrogen (H) at the 2´-carbon. This lack of an oxygen, another small change in structure, produces a great change in properties. In this case, it results in the inability of water to split the deoxyribose sugar. This confers on DNA a pronounced stability in vivo in that our bodies are composed of 75% water. Phosphate groups are bound at two sites on the sugar molecule, the 5´-carbon and the 3´-carbon. In this manner the sugar moiety provides a bridge between two phosphate groups as well as its binding to the individual base.

 

DNA is arranged in a double-helical secondary structure termed the double helix. The two strands of DNA may be envisioned as two circular staircases interwound with each other. This is an antiparallel structure in that one strand is oriented in a 5´ --> 3´ direction while the other strand is oriented in a 3´ --> 5´ direction. The sugar-phosphate portion of the molecule is located on the outside of the double helix and is termed the DNA backbone. The DNA bases are located close to each other in the interior of the double helix. They are attracted to each other by hydrogen bonding. This would be similar to a magnet in which opposite poles attract. Hydrogen bonding occurs between extracyclic amino (NH₂) and keto (C=O) groups as well as ring nitrogens with hydrogen atoms. The hydrogens on the amino group have a net positive charge, while the oxygen on the keto group or the ring nitrogen has a net negative charge. These are based on electronegativity, a basic chemical property in covalent linkages. Structural analysis demonstrates that adenine forms two hydrogen bonds with thymine (AT base pair), while guanine forms three hydrogen bonds with cytosine (GC base pair). As such, normal base pairing involves a purine base bonding with a pyrimidine base. This large base ­ small base pairing results in uniform size within the interior of the double helix.

 

The biological significance of DNA structure was identified through chromosomal and genetic analysis. This involved the localization of chromosomes within the nucleus and the discovery that they contained the hereditary information of the cell. Further study demonstrated that genes, located on chromosomes, determined our physical characteristics. Composed of DNA, the physical location of a gene on a chromosome was highly specific.

 

A significant advance in our understanding of nucleic acid structure and function was the "one gene ­ one enzyme" hypothesis. This theory suggested that each gene controls the expression of one enzyme. Gene 1 provides for enzyme 1; gene 2 for enzyme 2; gene 1000 for enzyme 1000; etc. An individual gene is solely responsible for its own enzyme; ie, gene 2 cannot provide enzyme 1 and vice versa. Thus, should something occur to destroy the ability of gene 1 to provide enzyme 1, the cell would lose that enzyme. More specifically, the gene that provides the information for the synthesis of insulin cannot be used to produce adrenaline and vice versa.

 

Although initially promulgated as the one gene ­ one enzyme hypothesis, it is clear that this term should be modified to the "one gene ­ one product" maxim. This is because of our understanding that not all proteins are enzymes but may have other functions (ie, structural, hormonal, wound healing, etc). However, it should be noted that the fundamental concept linking the nucleic acid structure and sequence within the genome and the use of that information synthesizing a protein remains constant.

 

This transfer of genetic information has been termed the central dogma. Briefly, it states that this transmission involves the conveyance of genetic information from nucleotide sequences in DNA to comparable sequences in RNA. Subsequently, the RNA, termed a messenger RNA (mRNA), is used to produce a protein that reflects the information in the DNA gene. This involves a nuclear to cytoplasmic transmission and is subject to the modifications indicated above with respect to the one gene ­ one enzyme hypothesis. In addition, the central dogma may be reversed in specific situations with information transfer from RNA to DNA. This is characteristic of RNA tumor viruses catalyzed by the enzyme reverse transcriptase and directly involved in HIV.

 

The genetic alphabet that is used in this information transfer is a four-letter DNA alphabet: A, C, G, and T. The genetic code using this alphabet is a series of triplet RNA bases, each of which is complementary to the DNA base. The complementary base is that which hydrogen bonds to the respective DNA base. Thus, if a DNA sequence is AGCTAGCT, the mRNA sequence is UCGAUCGA. The enzyme RNA polymerase catalyzes the synthesis of mRNA. The triplet RNA genetic code specifies each of the 20 naturally occurring amino acids. For example, the triplet UUU specifies the amino acid phenylalanine. The triplet AAA codes for the amino acid lysine. With a four-letter alphabet, there is the possibility of 64 such codes, termed codons. Each amino acid may be specified by more than one codon. In addition, a single codon is used as a start signal (AUG) and three codons may be used as a stop signal (UAA, UGA, UAG) thereby delineating the initiation and termination of the protein product.

 

This genetic code is contained within the mRNA, which is synthesized in the nucleus using DNA as the template. Following mRNA transport to ribosomes in the cytoplasm, its information is translated into protein. This process involves another RNA species, transfer RNA (tRNA). This molecule, which has a cloverleaf structure, serves two functions. First, it binds specifically to the codon using three bases within its sequence. This is termed the anticodon and is present in what is termed the anticodon loop. This binding is determined by hydrogen bonding demonstrating again the use of electronegativity as a basic driving force in gene expression. Second, at the other end of the molecule, termed the acceptor stem, tRNA binds specifically to the amino acid specified by that codon. Thus, if the triplet code within the mRNA is UUU, the anticodon region in tRNA is AAA, and the amino acid selected by the tRNA is phenylalanine. There are 20 different tRNA species, each of which is distinct for an individual amino acid.

 

The cellular factory in which proteins are made is termed the ribosome, which binds to the mRNA. To make a protein, one may imagine each of the tRNA molecules bound to its respective amino acid arranged in a specific order defined by the triplet code within the mRNA molecule. As such, at the other end, the amino acids will also be arranged in a chain yet unconnected. At this point, the enzyme peptidyltransferase catalyzes the formation of peptide bonds between the amino acids, thereby synthesizing the particular protein. The amino acid sequence within the protein will be defined by the sequence of codons in the mRNA. That sequence determines the order in which tRNA molecules bind to the mRNA. Consequently, the chain of amino acids within the protein will be specified by that sequence. En toto, this elaborate and complicated process results in the use of the information contained in the nucleotide sequences in DNA to produce a protein with a defined sequence of amino acids. As such, the structure and function of that protein will be dependent upon the order of amino acids within the protein. In this manner, cells are able to use their genetic information in vivo.

 

 

Our ability to establish gene therapy treatment regimens is, and will be, a direct consequence of the development and growth of molecular technology. In particular, the capacity to isolate individual genes and then manipulate their expression provides the foundation for gene therapy in health care. As briefly described in this section, several approaches were used in the structure-function analysis of gene organization.

 

Three of these protocols are illustrated in Figure 1. Primarily, each depends on the isolation of the protein of interest. Historically, the production of monoclonal or polyclonal antibodies provided the first step in this process. These antibodies were used to screen cDNA libraries established in bacteria (prepared as described below) for individual cells that express an immunoreactive protein. As related in the previous section, the one gene ­ one product hypothesis defines that a positive result would indicate that the gene specifying that protein is contained within that cell's genome. Following isolation of the respective cell colony, usually by limited dilution, the individual cDNA specifying that product could be isolated and characterized. Its nucleotide sequence would then be determined by gel electrophoretic protocols using dideoxynucleotides in a DNA polymerase reaction. Again, the reader is referred to a general biochemistry textbook for details of these protocols.

Figure 1.  Strategies for Gene Isolation

Improvements in technology permitted the synthesis of oligonucleotide sequences containing the DNA bases that encoded the N-terminal amino acid sequence of the protein. These sequences, radiolabeled with (³²P), were then used in hybridization studies to determine the presence of the respective gene sequence in the cDNA library. This represented a significant technological improvement for genetic studies. The antibody studies described above required the synthesis of immunoreactive proteins. This limited the number of colonies detected, as expression of the gene sequence was required. The use of radiolabeled oligonucleotides eliminated that requirement. It was now possible to detect a genetic sequence without its mRNA transcription and translation into a protein. This protocol also obviated the need for antibody production, the extended interval for their production and characterization, the animal and laboratory facilities required for such an undertaking, as well as the financial resources for these costly experimental protocols.

 

A further improvement in gene isolation was the technological development of the polymerase chain reaction (PCR). This protocol, using Thermus aquaticus DNA polymerase and specific 5´ and 3´ oligonucleotide primers, permitted the extensive synthesis and resynthesis of a defined DNA segment. This amplification of DNA allowed the use of a small amount of a DNA sample, which yielded an extensive quantity of product. Reverse transcriptase (RT)-PCR represented a further technological advancement. This procedure obviated the need for a DNA sample using RT to synthesize a DNA copy from an mRNA sample.

 

The a priori experimental protocol underlying these studies is the ability to digest DNA at a specific sequence and thus to isolate defined DNA segments. This procedure was actually developed prior to and concurrent with the development of antibody strategies to isolate individual genes. Intriguingly, it evolved from very basic studies in bacteriophage genetics. These investigations, begun in the 1940s, addressed very fundamental biochemical and genetic questions using bacteria and the viruses that infect them (ie, bacteriophages) as the experimental paradigm.

 

Successful bacteriophage infection requires the injection of phage DNA into the bacteria and its subsequent use of the cell's transcriptional and translational machinery to conquer the bacteria. These studies revealed that bacteria contain very unique enzymes, which function as an intracellular defense mechanism. These proteins, termed restriction endonucleases, digest bacteriophage DNA, thereby preventing successful infection. They do not recognize bacterial DNA, which has specific modifications that prevent its digestion. Furthermore, each individual restriction enzyme displays an exquisite sequence specificity. Hypothetical examples are illustrated in Figure 2. Restriction enzymes 1 and 2 digest DNA at the indicated sequences. The origin of the DNA is immaterial; only the sequence matters. Thus, in vitro, these enzymes can be used on any DNA apart from host DNA. The significance of restriction endonucleases for the study of genetic organization resides in the observation that, literally, a plethora of such bacterial enzymes exists. Each has a different specificity (termed a restriction site) and thus can be used in sequence to analyze DNA. These enzymes have been extensively purified and characterized. They are readily available commercially and are routinely used as reagents for DNA analysis. A model is presented in Figure 3, in which a DNA sequence contains sites for both hypothetical restriction enzymes 1 and 2. This illustration demonstrates how each enzyme can be used to digest the DNA into segments that may then be isolated by electrophoretic protocols based on their size.

 

These enzymes were used to create the cDNA libraries referred to above. Using a heterogeneous mRNA sample, RT was used to make DNA copies (termed complementary DNA or cDNA). This cDNA could be randomly digested with a specific restriction enzyme. The cut DNA, comprised of different sizes, was then ligated into a bacteriophage that contained the identical restriction site. In this manner, a hybrid DNA was formed that contained bacteriophage DNA combined with DNA from another organism. Termed a recombinant plasmid (ie, DNA from diverse organisms has been combined), this newly created molecule was used to infect a bacterial host. Initially, bacteriophage l was used as its genetic map was well known. Of equal significance, it could kill bacteria through lytic infection or transform bacteria through lysogenic infection. The latter condition permitted the replication of the l genome as well as the cDNA now contained within it.

 

These new technologies allowed the manipulation of the cellular genome to an extraordinary degree. An analogy which may be used is that cell DNA may be viewed either as a living Lego or Erector set. In both instances, individual pieces may be rearranged in different sequences or combined to create new structures. Likewise, using restriction endonucleases, a gene from one organism may be removed and inserted into a second entity. The only restriction is the viability of the newly created DNA.

The basis for gene therapy derives from, first, the ability to isolate individual genes as described above; second, to use plasmid technology to introduce them into a living cell; and, third, to use the machinery of that cell to transcribe an mRNA from that gene, which is then translated into the respective protein. As indicated in Figure 4, two such protocols have important pharmacologic ramifications. For example, on the left-hand side, the isolated gene (ie, human insulin) is incorporated into a bacterial plasmid, which is then used to infect a bacterial cell. This cell is not killed (ie, lysed) but instead is permanently transformed so that as it grows it expresses the genes encoded by the plasmid. In this case, the protein specified by the cloned gene is produced in large quantities. It may be purified by conventional biochemical chromatography to provide a reliable source for the indicated protein. It is through this approach that we have developed a new source for insulin as a treatment for diabetes. The human insulin gene was cloned, inserted into a plasmid that was used to transform a recipient bacteria. The bacteria produces large quantities of human insulin, which is purified and available for therapeutic use. Termed recombinant insulin, it is now available for patient care and represents a far more reliable source than that used previously (ie, purification of insulin from animal tissue). Thus, in this instance, bacteria are used as a living factory for the production of a human protein that is then prescribed for the patient.

 

Gene therapy represents a similar strategy for the use of cloned genes in patient care (Figure 4, right-hand side). However, in this case, the isolated gene is incorporated into a recombinant plasmid, termed a vector, which is capable of infecting a human (ie, patient) cell. This allows for the use of human cells to produce this protein directly. This would be done in the patient on a continual basis. This obviates the need for bacterial expression, protein purification, and administration to the patient.

 

There are three a priori requirements for the use of cloned genes in patient treatment. These are the development of recombinant plasmids, methods for infection of patient cells, and determination of the expression of the respective gene in the target cells. All are currently under intensive investigation and will provide the technical basis for effective gene therapy.

 

It is axiomatic in life that everything has advantages and disadvantages. For example, the AIDS virus, HIV, represents a clear and present danger to human health. A basic HIV mechanism is the infection of target cells followed by the incorporation of the HIV genome into host DNA by reverse transcription. In this manner, HIV is able to use the cell's biosynthetic machinery to transcribe its own genes, synthesize its own proteins, and thus produce its respective pathology.

 

Recombinant DNA technology provides an opportunity to turn this malevolent situation into an advantage for treatment. In this approach, restriction endonucleases are used to rearrange the retrovirus genetic material analogous to the Erector or Lego model. In this strategy, genes are retained that provide for the ability of the retrovirus to enter (infect) the host cell. However, "deleterious" viral genes are removed, thereby preventing the disease. The gene of choice (ie, human insulin) is inserted, so that it is under the control of the retrovirus regulatory elements. En toto, through this procedure, a new recombinant organism will be produced. It will be capable of infecting a human cell, of transferring its genetic information into the human cell genome, and, similar to the retroviral genome, it will be replicated during cell division and its information transferred to daughter cells.

 

Two main approaches may be used to "infect" human cells with the recombinant retroviral plasmid. The first is termed the ex vivo approach. In this protocol, cells are removed from the patient, treated in vitro, then transferred back to the individual. It may be reasonable to suggest that this presents the most feasible approach at the current time. With modifications, a person's own lymphocytes may be removed, treated in vitro, then reinjected into the individual using protocols similar to those used in blood transfusions. From a theoretical standpoint, considering that lymphocytes are an HIV target, this approach could present a high likelihood of success.

 

The second approach would be what is termed the in vivo method. In these protocols, individuals are treated with the recombinant plasmid directly. The plasmid is introduced into the target tissue by mechanical means. There are obvious inherent difficulties in the use of this protocol. However, it may be necessary if tissue-specific expression of the respective gene is required for its therapeutic function.

 

 

Prime candidates for gene therapy span the gamut of preventive and palliative health care. These include replacement of a nonfunctional protein to restore a normal cell function, pharmacologic modulation of enzyme activity to increase therapeutic efficacy, stimulation of the immune response by augmenting antibody production, and detoxification of deleterious agents by increasing the concentration of enzymes through which they are degraded. In each case, recombinant DNA technology would provide the essential tools for the manipulation of the human genome.

 

The replacement of a non-functional protein provides perhaps the most likely starting point for gene therapy. In these instances, a protein required for an essential cell process is either missing or, through gene mutation, is present but its activity is substantially diminished based on a defined change in its amino acid sequence. The former may be exemplified by hemophilia, a genetic disorder in which a specific clotting protein is not produced. This genetic condition is displayed at birth. It is well known and characterized at the population, biochemical, and molecular level. Another example, age-dependent diabetes, may be inherited but occurs at a later interval in life. In this instance, the normal ability to synthesize insulin may be lost. The latter may be exemplified by sickle cell anemia. In this genetic condition, a mutant hemoglobin is produced as a result of a genetic change in the sixth amino acid, ie, glutamic acid to valine. Again, population genetics, and biochemical and molecular analysis have well characterized this human disease.

 

Gene therapy would provide a means to correct each disorder. With respect to hemophilia, this would entail the incorporation of a normal clotting factor gene into target cells under retroviral regulation such that it is constitutively produced. A similar situation would exist with respect to sickle cell anemia. In either case, the respective genes have been cloned and await the appropriate technological developments for their utilization. Age-dependent diabetes presents a different problem due to its development later in life. Thus, early treatment is unnecessary. It is unclear what modifications of gene therapy would be needed as a function of the age of the patient.

 

The second area appropriate for gene therapy could be modulation of the therapeutic efficacy of diverse pharmacologic agents. This would be particularly significant for drugs that require host metabolism for their activity. In this instance, it may be useful to enhance the patient's ability to metabolize the drug in question. For example, 5-fluorouracil requires metabolism by salvage pathway enzymes, ie, uridine phosphorylase and uridine kinase, for its conversion to 5-fluorodeoxyuridine monophosphate. As the monophosphate it binds to thymidylate synthase as its major site of action. Augmentation of either activity through gene therapy could increase the therapeutic efficacy of the drug.

 

Alternatively, an effective strategy may be to prevent the degradation of the drug. In this instance, cell metabolic pathways diminish therapeutic efficacy by catabolism of the drug in question. In this instance, it would be necessary to decrease the activity of a specific enzyme. This would be comparable to the use of allopurinol as supportive therapy during the use of purine antagonists in cancer treatment. The resultant inhibition of xanthine oxidase prevents the degradation of the drug increasing its therapeutic efficacy. Accordingly, a strategy for gene therapy may be to introduce a gene whose product inhibits the respective enzyme or to introduce a mutant protein that in some way competes with the normal protein.

 

A third area ripe for gene therapy relates to the immune response. Antibody production is an a priori requirement for this defense mechanism. As our understanding of the genetics of antibody production increases, our ability to use gene therapy as a strategy becomes more attractive. Such an approach would be comparable to that observed now using bone marrow transplantation as an effective treatment in immune-compromised patients. Gene therapy for augmentation of the immune response could be attempted initially using circulating cells. In this instance, ex vivo treatment of lymphocytes would be an attractive first step.

 

The fourth area for potential gene therapy addresses the significant question of preventive health care. For example, a great concern, reflected in the consumption of nonprescription drugs, is the role of oxidative stress as a causative pathological agent. Gene therapy could provide a means to augment constitutively the antioxidant capacity of an individual. Such therapy could involve the introduction of genes whose protein products function as internal antioxidants. For example, the introduction of the superoxide dismutase gene could significantly augment the ability of an individual to degrade reactive oxygen species.

 

As indicated, several potential areas are described that may provide the initial focus for the use of gene therapy in patient care. Other uses may also be considered at present or, as our technologies develop and our understanding of basic cell functions advance, may be added to this nonexclusive list. A likely area is gene therapy of neurodegenerative disorders such as Alzheimer's disease and Parkinson's disease. Advances in the understanding of these as well as other neuronal diseases will impact dramatically on the utility of gene therapy in patient care.

As described in the previous section, gene therapy offers tremendous potential as a potent weapon in our therapeutic arsenal. However, certain problems, mostly technical in nature, need to be overcome before gene therapy becomes a routine treatment. Some of these relate to the delivery of the gene, the specificity for its expression, the pharmacokinetics of vector transfer, and, last, and perhaps most intriguing, the eternal nature of gene therapy.

 

The first problem is that we have not yet perfected the delivery systems for gene therapy. Excellent systems exist in tissue culture for the transfection of human cells in culture with viral or nonviral vectors, the successful propagation of the transfected cells in culture, and their characterization at the genetic, cellular, and biochemical levels. These studies have been, and continue to be, very successful. In contrast, our ability to undertake similar investigations ex vivo or in vivo lags far behind. It may be reasonable to state that, with respect to these studies, we are still in our infancy. However, this is primarily a technical problem which should, with the explosion of work in gene therapy, be solved within the next decade.

 

The second problem in gene therapy relates to the particular cell that needs to be transfected to achieve therapeutic efficacy. Directing gene transfer and/or expression to the appropriate cells within the body could appreciably enhance gene therapy approaches. In normal individuals, insulin is a product of the pancreas, and detailed mechanisms exist for its transport and function in vivo. Considering the discussion above, it is conceivable that ex vivo protocols may be developed for the transfection of lymphocytes with the insulin gene. These cells could then be returned to the individual. However, the physiologic consequences of lymphocytes producing insulin in vivo are unknown. What effects will this new source of insulin have on homeostasis in vivo? What are the consequences for the body when insulin is constitutively synthesized by this mechanism? Similar studies would have to be undertaken with other genes who also would be prime candidates for therapy.

 

The third problem relates to the pharmacokinetics of the synthesized product. It is unknown what mechanisms exist with the body to transport, activate, or degrade the particular product. It may be that intrinsic mechanisms would treat the gene product similarly to that of the endogenously synthesized molecule. Alternatively, depending on the cell in which the product is synthesized, other mechanisms would be involved that could affect its therapeutic efficacy.

 

The last problem associated with gene therapy is curiously intriguing. It relates to the alteration of the genetic composition of human beings. By definition, gene therapy would modify the unique genetic character of that individual. If this was performed in somatic cells with a defined lifetime, human gene structure would revert to its original form. Alternatively, if this was undertaken in stem cells, it would represent an eternal change in that individual. This procedure would alter the most singular characteristic that marks a human being. That is a metaphysical problem that has the scenario of a brave new world. It would also be a more critical question if, for some reason, meiotic cells were involved in some manner.

 

A change in genetic information in stem cells would last the lifetime of the individual. What would happen if, for some reason, it was considered advisable to discontinue treatment? How would one stop this regimen? One cannot simply decide not to renew the prescription. Again, this would be a particular problem if stem cell transfer was involved. It would be much less of a problem with gene therapy using cells with a defined lifetime. Again, this would seem to target lymphocytes as the most appropriate initial vehicle for the first studies on gene therapy in humans.

 

The industrial age witnessed an explosion in technology and knowledge that formed the basis for the dramatic advances in patient care over the past century. The use of gene therapy represents a potential next wave for the accelerated improvement of both preventive and palliative health care. Over the next decade, it is a reasonable expectation that the technological barriers to the use of gene therapy will be overcome. Discoveries and treatments now in the exploratory stage in the laboratory will progress to clinical analysis. The success of these research efforts will present a significant impact on treatment. Furthermore, it is sensible to expect a similar dramatic effect with respect to the ability to devise preventive measures. As such, this progression will usher in new, more effective treatments to replace current regimens and will enable the initiation of protocols for use in conditions currently not amenable to therapy.

 

This enthusiasm needs to be tempered by both the technical difficulties that remain and which are formidable. The introduction of genetic material into a patient as a routine treatment requires significant advances beyond our capabilities at the present time. In addition, the ethical aspects of gene therapy need detailed discussion, and recommendations are required from oversight organizations. The latter is particularly critical for therapies that have the potential of altering meiotic cells. These guidelines may also be required for treatment in utero. Nevertheless, it remains a reasonable expectation that the next decade will witness significant advances in gene therapy protocols as well as the beginnings of its consideration and use in preventive and palliative health care.

Questions 16 through 19, select from the following choices:
(a) hemophilia
(b) sickle cell anemia
(c) 5-fluorouracil
(d) superoxide dismutase

Return to Introduction

References

Fielding AK, Ager S, Russell SJ. The future of haematology, molecular biology and gene therapy. Brit Med J. 1997;314: 1396-1399.

Friedmann T. Overcoming the obstacles to gene therapy. Scientific American 1997; 276:96-101.

Horton HR, Moran LA, Ochs RS, Rawn JD, Scrimegeour KG. Principles of Biochemistry. 2nd ed. Upper Saddle River, NJ: Prentice Hall; 1996.

Weichselbaum RR. Gene therapy of cancer. Lancet. 1997;349:sh10-sh12.