NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Strachan T, Read AP. Human Molecular Genetics. 2nd edition. New York: Wiley-Liss; 1999.

Cover of Human Molecular Genetics

Human Molecular Genetics. 2nd edition.

Show details

Chapter 22Gene therapy and other molecular genetic-based therapeutic approaches

22.1. Principles of molecular genetic-based therapies and treatment with recombinant proteins or genetically engineered vaccines

22.1.1. Principles of molecular genetic-based approaches to treating disease

Once a human disease gene has been characterized, molecular genetic tools can be used to dissect gene function and explore the biological processes involved in the normal and pathogenic states. The resulting information can be used to design novel therapies using conventional drug-based approaches. In addition, molecular genetic technologies have recently provided a variety of novel therapeutic approaches that can be categorized into two broad groups, depending on whether the therapeutic agent is a gene product/vaccine or genetic material.

Recombinant proteins and genetically engineered vaccines

Here the therapy is to deliver proteins or vaccines which have been produced by genetic engineeering instead of traditional methods. Methods involve:

  • expression cloning of normal gene products — cloned genes are expressed in microorganisms or transgenic livestock in order to make large amounts of a medically valuable gene product;
  • production of genetically engineered antibodies — antibody genes are manipulated so as to make novel antibodies, including partially or fully humanized antibodies, for use as therapeutic agents;
  • production of genetically engineered vaccines — includes novel cancer vaccines and vaccines against infectious agents.

Gene therapy

The term gene therapy describes any procedure intended to treat or alleviate disease by genetically modifying the cells of a patient. It encompasses many different strategies and the material transferred into patient cells may be genes, gene segments or oligonucleotides. The genetic material may be transferred directly into cells within a patient (in vivo gene therapy), or cells may be removed from the patient and the genetic material inserted into them in vitro, prior to transplanting the modified cells back into the patient (ex vivo gene therapy). Because the molecular basis of diseases can vary widely, some gene therapy strategies are particularly suited to certain types of disorder, and some to others. Major disease classes include:

  • infectious diseases (as a result of infection by a virus or bacterial pathogen);
  • cancers (inappropriate continuation of cell division and cell proliferation as a result of activation of an oncogene or inactivation of a tumor suppressor gene or an apoptosis gene — see Chapter 18);
  • inherited disorders (genetic deficiency of an individual gene product or genetically determined inappropriate expression of a gene);
  • immune system disorders (includes allergies, inflammations and also autoimmune diseases, in which body cells are inappropriately destroyed by immune system cells).

A major motivation for gene therapy has been the need to develop novel treaments for diseases for which there is no effective conventional treatment. Gene therapy has the potential to treat all of the above classes of disorder. Depending on the basis of pathogenesis, different gene therapy strategies can be considered (Box 22.1 and Figure 22.1). One, rather arbitrary, subdivision of gene therapy approaches is as follows:

Box Icon

Box 22.1

General gene therapy strategies (see also Figure 22.1). For diseases caused by loss of function of a gene, introducing extra copies of the normal gene may increase the amount of normal gene product to a level where the normal phenotype is restored (see (more...)

Figure 22.1. Five approaches to gene therapy.

Figure 22.1

Five approaches to gene therapy. Of the five illustrated approaches, four have been used in clinical trials. Gene augmentation therapy by simple addition of functional alleles has been used to treat several inherited disorders caused by genetic deficiency (more...)

  • Classical gene therapy. The rationale of this type of approach is to deliver genes to appropriate target cells with the aim of obtaining optimal expression of the introduced genes. Once inside the desired cells in the patient, the expressed genes are intended to do one of the following:
    1. produce a product that the patient lacks;
    2. kill diseased cells directly, e.g. by producing a toxin which kills the cells;
    3. activate cells of the immune system so as to aid killing of diseased cells.
  • Nonclassical gene therapy. The idea here is to inhibit the expression of genes associated with the pathogenesis, or to correct a genetic defect and so restore normal gene expression.

Current gene therapy is exclusively somatic gene therapy, the introduction of genes into somatic cells of an affected individual. The prospect of human germline gene therapy raises a number of ethical concerns, and is currently not sanctioned (see Section 22.6.1).

22.1.2. Recombinant pharmaceuticals can be produced by expression cloning in microorganisms or transgenic livestock

Advantages of obtaining medically valuable reagents by expression cloning

Once a human gene has been cloned, large amounts of the purified product can be obtained by expression cloning (Section 4.4.2). Often the desired gene is expressed in bacterial cells, which have the advantage that they can be cultured easily in large volumes. Using this approach, large amounts of recombinant pharmaceuticals can be generated. Expression cloning may provide the only product-based therapeutic route in those cases where biochemical purification of the product from a human or animal source is difficult or impossible. Safety risks are minimal, unlike those associated with products obtained from conventional human or animal sources (Box 22.2).

Box Icon

Box 22.2

Treatment using conventional animal or human products can be hazardous. Animal products have been prepared by standard biochemical purification techniques and used to treat human patients. In many cases the treatment is based on the animal product having (more...)

Recombinant human insulin was first marketed in 1982 and, subsequently, a number of other cloned human gene products of medical interest have been produced commercially (see Table 22.1). Treatment with the products of cloned genes is not free from risks, however. For example, patients who completely lack a normal product may mount a vigorous immune response to the administered pharmaceutical product as in the case of some patients with severe hemophilia A who have been treated with recombinant factor VIII.

Table 22.1. Examples of pharmaceutical products obtained by expression cloning.

Table 22.1

Examples of pharmaceutical products obtained by expression cloning.

Transgenic livestock as a source of medically valuable products

Expression cloning often involves the use of microorganisms, but this approach may not always be suitable. For example, expression of a human gene in a bacterial cell can give a product that shows differences from the normal human product: the polypeptide may have the same sequence of amino acids but patterns of glycosylation may be different. This may mean that the gene product is not particularly stable in a human environment, or it may provoke an immune response, or its biological function may be less effective than desired. In addition, the cost of producing purified recombinant pharmaceuticals may be rather high.

The disadvantages of microbial expression cloning has prompted consideration of alternatives. In particular, increasing attention has been paid to constructing transgenic livestock (see Lubon, 1998), where the post-translational processing systems are more similar to analogous human systems. For example, a cloned human gene can be fused to a sheep gene specifying a milk protein then inserted into the genome of the sheep germline. The resultant transgenic sheep can secrete large quantities of the fusion protein in its milk. Transgenic pigs have also been designed to express human proteins in their milk, as in the case of human factor VIII (Paleyanda et al., 1997).

22.1.3. Genetically engineered antibodies and vaccines have great therapeutic promise

Antibody engineering

Antibodies are natural therapeutic agents which are produced by B lymphocytes. In each B-cell precursor, a cell-specific rearrangement of antibody gene segments occurs so that individual B cells produce different antibodies (Section 8.6.2). Additional diversity is provided by other mechanisms, including frequent somatic mutation events. As a result, each one of us has a population of B cells which collectively ensures a huge repertoire of different antibodies as a defense system against a diverse array of foreign antigens. The antibody may be thought of as an adaptor molecule: it contains binding sites for foreign antigen at the variable (V) end, and at the constant (C) end it has binding sites for effector molecules. Binding of an antibody may by itself be sufficient to neutralize some toxins and viruses, but it is more common for the antibody to trigger the complement system and cell-mediated killing.

Artificially produced therapeutic antibodies are designed to be monospecific (they recognize a single type of antigenic site) and can recognize specific disease-associated antigens, leading to killing of the disease cells (see Berkower, 1996). Notable targets for such therapy are cancers (especially lymphomas and leukemias); infectious disease (using antibodies raised against antigens of the relevant pathogen); and autoimmune disorders (where antibodies recognize inappropriately expressed host cell antigens). A favorite way of producing immortal monospecific antibodies is to fuse individual antibody-producing B lymphocytes from an immunized mouse or rat with cells derived from an immortal mouse B-lymphocyte tumor. From the fusion products, hybridomas, a heterogenous mixture of hybrid cells which have the ability to make a particular antibody and to multiply indefinitely in culture, are selected. The hybridomas are propagated as individual clones, each of which can provide a permanent and stable source of a single type of monoclonal antibody (mAb).

Until recently, the therapeutic antibody approach was not straightforward. Although rodent monoclonal antibodies (mAbs) can be created against human pathogens and cells, they normally have limited use in the clinic. This is because rodent mAbs have a short half-life in human serum and they can elicit an unwanted immune response in patients (producing human antirodent antibodies). In addition, only some of the different classes can trigger human effector functions. The generation of human mAbs would avoid these problems but has been difficult to achieve using standard hybridoma technology. Once immunoglobulin genes had been cloned, however, the possibility of designing artificial combinations of immunoglobulin gene segments arose (antibody engineering). Because different exons encode different domains of an antibody molecule, domain swapping could be done easily at the DNA level by artificially shuffling exons between different antibody genes (see Winter and Harris, 1993).

Chimeric and humanized antibodies

One immediate goal of antibody engineering was the production of chimeric and humanized antibodies, which are rodent-human recombinant antibodies (Winter and Harris, 1993; Figure 22.2). Humanizing of rodent antibodies could allow access to a large pool of wellcharacterized rodent mAbs for therapy, including those with specificities against human antigens that are difficult to elicit from a human immune response. Early versions contained the variable domains of a rodent antibody attached to the constant domains of a human antibody, a so-called chimeric (V/C) antibody. The immunogenicity of the rodent mAb is reduced, while allowing the effector functions to be selected for the therapeutic application. A further stage of humanizing antibodies is possible. The essential antigen-binding site is a subset of the variable region characterized by hypervariable sequences, the complementarity-determining regions (CDRs). Accordingly, second generation humanized antibodies were CDR-grafted antibodies: the hypervariable antigen-binding loops of the rodent antibody were built into a human antibody, creating a humanized antibody. Chimeric V/C antibodies and CDR-engrafted antibodies have been constructed against a wide range of microbial pathogens and against human cell surface markers, including tumor cell antigens. Their clinical potential is considerable (see Table 22.2).

Figure 22.2. Genetically engineered antibodies.

Figure 22.2

Genetically engineered antibodies. Antibodies consist of two light (L) chains, each containing a variable domain (VL) and a constant domain (CL), plus two heavy chains, containing a variable domain (VH) and three constant domains (CH1, CH2 and CH3). The (more...)

Table 22.2. Examples of the clinical potential of humanized antibodies.

Table 22.2

Examples of the clinical potential of humanized antibodies.

Fully human antibodies

Two approaches have been taken towards the construction of fully human antibodies (Vaughan et al., 1998):

  • Phage display technology. This technology bypasses hybridoma technology, and even immunization. Instead, antibodies are made in vitro by mimicking the selection strategies of the immune system (see Section 20.4.3).
  • Transgenic mice. One powerful strategy involves transferring yeast artificial chromosomes containing large segments of the human heavy and light chain immunoglobulin loci into mouse embryonic stem cells, and subsequent production of transgenic mice. For example, Mendez et al. 1997 report the construction of transgenic mice containing 1.02 Mb human Ig heavy chain and 0.8 Mb human Ig light chain loci (transloci; see Figure 21.3). Such mice contain a very considerable portion of the human V gene segment repertoire and the human immunoglobulin transloci are able to undergo the normal program of rearrangement and hypermutation to generate human antibodies.

Genetically engineered vaccines

Recombinant DNA technology is also being applied to the construction of novel vaccines. Several different strategies are being used (Liu, 1998; Pardoll, 1998):

  • Nucleic acid vaccines. These are typically bacterial plasmids containing genes encoding pathogen or tumor antigens which are delivered in saline solution by direct intramuscular injection. They normally carry a strong viral promoter which drives the expression of the gene of interest directly in the injected host. An alternative method of gene transfer uses a ‘gene gun’ to deliver gold beads onto which the DNA has been precipitated (see also Section 22.2.3). In the last few years, abundant evidence has been obtained that DNA vaccines can be effective (see Donnelly et al., 1997).
  • Genetic modification of antigen. This can be achieved, for example, by fusion with a cytokine gene to increase antigenicity.
  • Genetic modification of viruses. Viral vector systems have been used to deliver the genes of heterologous pathogens. DNA vectors based on alphaviruses (which have single-stranded RNA genomes) have been the focus of much recent attention.
  • Genetic modification of microorganisms. One way is to disable an organism, e.g. by removing genes required for pathogenesis or survival. This is a genetic method of attenuation so that a live vaccine can be used without undue risk. Another approach involves inserting an exogenous gene that will be expressed in bacteria or parasites.

Note that some gene therapy approaches, such as adoptive immunotherapy, are effectively forms of genetically engineered vaccination (Section 22.5.2).

22.2. The technology of classical gene therapy

22.2.1. Genes can be inserted into the cells of patients by direct and indirect routes, and the inserted genes can integrate into the chromosomes or remain extrachromosomal

An essential component of classical gene therapy is that cloned genes have to be introduced and expressed in the cells of a patient in order to overcome the disease. Practically, this usually involves targeting the cells of diseased tissues. However, deliberate targeting of unaffected cells may be preferred in some approaches:

  • Immune system-mediated cell killing. In many gene therapies the target cells are healthy immune system cells, and the idea is to enhance immune responses to cancer cells or infectious agents (Section 22.5.1).
  • Delivery of gene products from cells at a remote location. Genes may be targeted initially to one type of tissue while the gene products may be delivered to a remote location. For example, the myonuclei in muscle fibers have the advantage of being very long lived. Genetically engineered myoblasts therefore have the potential to ameliorate some nonmuscle diseases through long-term expression of exogenous genes which encode a product secreted into the blood stream (see for example, Jiao et al., 1993).

Two major general approaches are used in the transfer of genes for gene therapy: transfer of genes into patient cells outside of the body (ex vivo) or inside the body (in vivo).

Ex vivo gene transfer

This initially involves transfer of cloned genes into cells grown in culture. Those cells which have been transformed successfully are selected, expanded by cell culture in vitro, then introduced into the patient. To avoid immune system rejection of the introduced cells, autologous cells are normally used: the cells are collected initially from the patient to be treated and grown in culture before being reintroduced into the same individual (see Figure 22.3). Clearly, this approach is only applicable to tissues that can be removed from the body, altered genetically and returned to the patient where they will engraft and survive for a long period of time (e.g. cells of the hematopoietic system and skin cells). Note that this type of gene therapy involves transplantation of autologous genetically modified cells and so can be considered a modified form of cell therapy (see Box 22.3).

Figure 22.3. In vivo and ex vivo gene therapy.

Figure 22.3

In vivo and ex vivo gene therapy. In vivo gene therapy (blue arrow) entails the genetic modification of the cells of a patient in situ. Ex vivo gene therapy (black arrows) means that cells are modified outside the body before being implanted into the (more...)

Box Icon

Box 22.3

Cell therapy. Unlike gene therapy, cell therapy is a well-established form of treating disorders where patients can be treated with cells from different sources (Gage, 1998). Often treatment is performed by transplanting cells from another individual (allotransplantation). (more...)

In vivo gene transfer

Here the cloned genes are transferred directly into the tissues of the patient. This may be the only possible option in tissues where individual cells cannot be cultured in vitro in sufficient numbers (e.g. brain cells) and/or where cultured cells cannot be re-implanted efficiently in patients. Liposomes and certain viral vectors are increasingly being employed for this purpose. In the latter case, it is often convenient to implant vector-producing cells (VPCs), cultured cells which have been infected by the recombinant retrovirus in vitro: in this case the VPCs transfer the gene to surrounding disease cells. As there is no way of selecting and amplifying cells that have taken up and expressed the foreign gene, the success of this approach is crucially dependent on the general efficiency of gene transfer and expression.

Principles of gene transfer

Classical gene therapies normally require efficient transfer of cloned genes into disease cells so that the introduced genes are expressed at suitably high levels. In principle, there are numerous different physicochemical and biological methods that can be used to transfer exogenous genes into human cells. The size of DNA fragments that can be transferred is in most cases comparatively very limited, and so often the transferred gene is not a conventional gene. Instead, an artificial minigene may be used: a cDNA sequence containing the complete coding DNA sequence is engineered to be flanked by appropriate regulatory sequences for ensuring high level expression, such as a powerful viral promoter. Following gene transfer, the inserted genes may integrate into the chromosomes of the cell, or remain as extrachromosomal genetic elements (episomes).

Genes integrated into chromosomes

The advantage of integrating into a chromosome is that the gene can be perpetuated by chromosomal replication following cell division (Figure 22.4). As progeny cells also contain the introduced genes, long-term stable expression may be obtained. As a result, gene therapy using this approach may provide the possibility of a cure for some disorders. For example, in tissues composed of actively dividing cells, the key cells to target are stem cells (a minority population of undifferentiated precursor cells which gives rise to the mature differentiated cells of the tissue; see Section 2.2.2). This is so because stem cells not only give rise to the mature tissue cells, but during this procedure they also renew themselves. As a result, they are an immortal population of cells from which all other cells of the tissue are derived. High efficiency gene transfer into stem cells, and subsequent stable high level expression of a suitable introduced gene, can therefore provide the possibility of curing a genetic disorder.

Figure 22.4. Exogenous genes that integrate into chromosomes can be stably transmitted to all daughter cells, unlike episomal (extrachromosomal) genes.

Figure 22.4

Exogenous genes that integrate into chromosomes can be stably transmitted to all daughter cells, unlike episomal (extrachromosomal) genes. The figure illustrates two possible fates of genes that have been transferred into nucleated cells. If the cells (more...)

Chromosomal integration has its disadvantages, however, because normally the insertion occurs almost randomly. As a result, the location of the inserted genes can vary enormously from cell to cell. In many cases, inserted genes may not be expressed, e.g. they may have integrated into a highly condensed heterochromatic region. In some cases the integration event can result in death of the host cell (the insertion may occur within a crucially important gene, thereby inactivating it). Such an event has consequences only for the single cell in which the integration occurred. A greater concern is the possibility of cancer: an integration event in one of the many cells that are targeted could disturb the normal expression patterns of genes that control cell division or cell proliferation. For example, the integration can cause activation of an oncogene or it could inactivate a tumor suppressor gene or a gene involved in apoptosis (programmed cell death). Ex vivo gene therapy at least offers the opportunity for selecting cells where integration has been successful, amplifying them in cell culture and then checking the phenotypes for any obvious evidence of neoplastic transformation, prior to transferring the cells back into the patient.

Nonintegrated genes

Some gene transfer systems are designed to insert genes into cells where they remain as extrachromosomal elements and may be expressed at high levels (see Table 22.3). If the cells are actively dividing, the introduced gene may not segregate equally to daughter cells and so long-term expression may be a problem. As a result, the possibility of a cure for a genetic disorder may be remote: instead, repeated treatments involving gene transfer will be necessary. In some cases, however, there may be no need for stable long-term expression. For example, cancer gene therapies often involve transfer and expression of genes into cancer cells with a view to killing the cells. Once the malignancy has been eliminated, the therapeutic gene may no longer be needed.

Table 22.3. Properties of major methods of gene transfer used in gene therapy and their applications.

Table 22.3

Properties of major methods of gene transfer used in gene therapy and their applications.

22.2.2. Most gene therapy protocols have used mammalian viral vectors because of their high efficiency of gene transfer

The method chosen for gene transfer depends on the nature of the target tissue and whether transfer is to cultured cells ex vivo or to the cells of the patient in vivo. No one gene transfer system is ideal; each has its limitations and advantages. However, mammalian virus vectors have been the preferred vehicle for gene transfer because of their high efficiency of transduction into human cells (Anderson, 1998).

Oncoretroviral vectors

Retroviruses are RNA viruses which possess a reverse transcriptase function, enabling them to synthesize a complementary DNA form. Following infection (transduction), retroviruses deliver a nucleoprotein complex (preintegration complex) into the cytoplasm of infected cells. This complex reverse transcribes the viral RNA genome and then integrates the resulting DNA copy into a single site in the host cell chromosomes (Figure 18.2). Retroviruses are very efficient at transferring DNA into cells, and the integrated DNA can be stably propagated, offering the possibility of a permanent cure for a disease. Because of these properties, retroviruses were considered the most promising vehicles for gene delivery and currently about 60% of all approved clinical protocols utilize retroviral vectors.

The retrovirus vectors that have traditionally been used in gene therapy are derived from simple retroviruses (oncoretroviruses), notably murine leukemia virus. Unlike adenoviruses they can only be produced at relatively low titers and so it is not possible to get a large number of vector particles to the desired cell type in vivo. Since all the viral genes are removed from the vector, the viruses cannot replicate by themselves. They can accept inserts of up to 8 kb of exogenous DNA and require a variety of packaging systems to enclose the viral genome within viral particles (simple injection of retroviral vectors is usually inappropriate for in vivo gene therapy because they can generally be killed by human complement).

Oncoretroviruses can only transduce cells that divide shortly after infection: the preintegration complex is excluded from the nucleus and can only reach the host cell chromosomes when the nuclear membrane is fragmented during cell division. This therefore limits potential target cells. Only certain blood cells (but not stem cells) and the cells lining the gastrointestinal tract are continually in division; other cell types undergo division but not continually and some important cell targets never divide (e.g. mature neurons). The property of transducing only dividing cells can, however, be beneficial to gene therapy for cancers of tissues that normally have nondividing cells: the actively dividing cancer cells can be selectively infected and killed without major risk to the nondividing cells of the normal tissue (see Section 22.5.3).

Adenovirus vectors

Adenoviruses are DNA viruses that produce infections of the upper respiratory tract and have a natural tropism for respiratory epithelium, the cornea and the gastrointestinal tract. Adenovirus vectors have been the second most popular delivery system in gene therapy (with extensive applications in gene therapy for cystic fibrosis and certain types of cancer) and have several advantages as gene delivery vectors. They are human viruses which can be produced at very high titers in culture, and they are able to infect a large number of different human cell types including nondividing cells. Entry into cells occurs by receptor-mediated endocytosis (Figure 22.5; see also below) and transduction efficiency is very high (often approaching 100% in vitro). They are large viruses and so have the potential for accepting large inserts.

Figure 22.5. Adenoviruses enter cells by receptor-mediated endocytosis.

Figure 22.5

Adenoviruses enter cells by receptor-mediated endocytosis. Binding of viral coat protein to a specific receptor on the plasma membrane of cells is followed by endocytosis, a process in which the plasma membrane invaginates and then pinches off to form (more...)

They also have some major disadvantages. The inserted DNA does not integrate, and so expression of inserted genes can be sustained over short periods only. The first generation recombinant adenoviruses used in cystic fibrosis gene therapy trials showed that transgene expression declined after about 2 weeks and was negligible after only 4 weeks (see below). Because they can infect virtually all human cells, adenovirus vectors may conceivably pose a risk in some therapies that are designed to kill cancer cells without causing toxicity to normal surrounding cells. Most importantly, first generation adenovirus vectors can generate unwanted immune responses, causing chronic inflammation.

Many of these difficulties have been addressed in the construction of second generation adenovirus vectors. For example, all of the adenovirus genes have been deleted from some newer adenovirus vectors (‘gutless vectors’) which then require the assistance of a helper virus. Such a virus provides certain viral functions in trans (e.g. enzymes involved in viral DNA replication etc.) which are essential for productive infection (including viral DNA replication, viral assembly and infection of new cells) by certain natural viruses, such as AAV, or artificially disabled viruses.

The risk of immune response to these vectors is negligible. This is an important consideration given the need to administer treatment frequently (because of the inability of adenovirus to integrate into chromosomal DNA). They also have the advantage that they can accept much larger inserts (up to 35 kb). Unfortunately, however, deletion of adenoviral genes can also be counterproductive. Deletion of the E3 region removes the capacity to encode a protein that protects the virus from immune surveillance mechanisms in the host. In addition, fully disabled adenoviral vectors have much lower transduction efficiencies.

Adeno-associated virus vectors

Adeno-associated viruses (AAVs) are a group of small, single-stranded DNA viruses which cannot usually undergo productive infection without co-infection by a helper virus, such as an adenovirus or herpes simplex virus. In the absence of co-infection by a helper virus, unmodified human AAV integrates into chromosomal DNA, usually at a specific site on 19q13.3-qter. Subsequent superinfection with an adenovirus can activate the integrated virus DNA, resulting in progeny virions. AAV vectors can only accommodate inserts up to 4.5 kb, but they have the advantage of providing the possibility of long-term gene expression because they integrate into chromosomal DNA. They also provide a high degree of safety: because 96% of the parental AAV genome has been deleted, the AAV vectors lack any viral genes and recombinant AAV vectors only contain the gene of interest.

Herpes simplex virus vectors

HSV vectors are tropic for the central nervous system (CNS) and can establish lifelong latent infections in neurons. They have a comparatively large insert size capacity (>20 kb) but are nonintegrating and so long-term expression of transferred genes is not possible. Their major applications are expected to be in delivering genes into neurons for the treatment of neurological diseases, such as Parkinson's disease, and for treating CNS tumors.


The lentivirus family, which includes HIV (human immunodeficiency virus), are complex retroviruses that infect macrophages and lymphocytes. Unlike oncoretroviruses, lentiviruses are able to transduce nondividing cells. In the case of HIV, for example, the preintegration complex contains nuclear localization signals that permit its active transport through nuclear pores into the nucleus during interphase. Because of their ability to infect nondividing cells and to integrate into host cell chromosomes, considerable efforts are now being devoted to making lentivirus vectors for gene therapy (Naldini, 1998).

22.2.3. Concerns over the safety of recombinant viruses have prompted increasing interest in nonviral vector systems for gene therapy

Increasingly, concern has been expressed regarding the safety of viral vector systems. The recombinant viruses which are used for ex vivo gene therapy are designed to be disabled: typically some viral genes required for viral replication are deleted, and the therapeutic genes that are to be transferred are inserted in their place. The resulting replication-incompetent viruses are then intended to infect individual cells. In the case of retrovirus vectors, chromosomal integration is still possible but, like other replication-incompetent virus vectors, they should not be able to undergo a productive infection in which they replicate, assemble new virions and infect new cells. However, there is the remote possibility that the introduced viruses can recombine with endogenous retroviruses, resulting in recombinant progeny that can undergo productive infection. Additionally, adenoviruses are generally nonintegrating and the repeated injections that may be required may provoke severe inflammatory responses to the recombinant adenoviruses as has happened recently in a gene therapy trial for cystic fibrosis. Increasingly, therefore, attention has been focused toward studying alternative methods of gene transfer (Kay et al., 1997).


Liposomes are spherical vesicles composed of synthetic lipid bilayers which mimic the structure of biological membranes. The DNA to be transferred is packaged in vitro with the liposomes and used directly for transferring the DNA to a suitable target tissue in vivo (Figure 22.6). The lipid coating allows the DNA to survive in vivo, bind to cells and be endocytosed into the cells. Cationic liposomes (where the positive charge on liposomes stabilize binding of negatively charged DNA), have become popular vehicles for gene transfer in in vivo gene therapy (see Huang and Li, 1997 for references). Unlike viral vectors, the DNA/lipid complexes are easy to prepare and there is no limit to the size of DNA that is transferred. However, the efficiency of gene transfer is low, and the introduced DNA is not designed to integrate into chromosomal DNA. As a result, expression of the inserted genes is transient.

Figure 22.6. In vivo liposome gene delivery.

Figure 22.6

In vivo liposome gene delivery. (A) and (B) Structure of liposomes. Liposomes are synthetic vesicles which can form spontaneously in aqueous solution following artificial mixing of lipid molecules. In some cases, a phospholipid bilayer is formed, with hydrophilic (more...)

Direct injection/particle bombardment

In some cases, DNA can be injected directly with a syringe and needle into a specific tissue, such as muscle. This approach has been considered, for example, in the case of DMD, where early studies investigated intramuscular injection of a dystrophin minigene into a mouse model, mdx (Acsadi et al., 1991). An alternative direct injection approach uses particle bombardment (‘gene gun’) techniques: DNA is coated on to metal pellets and fired from a special gun into cells. Successful gene transfer into a number of different tissues has been obtained using this approach. Such direct injection techniques are simple and comparatively safe. However, there is poor efficiency of gene transfer, and a low level of stable integration of the injected DNA. The latter property is particularly disadvantageous in the case of proliferating cells, and would necessitate repeated injections. It may be less of a problem in tissues such as muscle which do not regularly proliferate, and in which the injected DNA may continue to be expressed for several months.

Receptor-mediated endocytosis

The DNA is coupled to a targeting molecule that can bind to a specific cell surface receptor, inducing endocytosis and transfer of the DNA into cells. Coupling is normally achieved by covalently linking polylysine to the receptor molecule and then arranging for (reversible) binding of the negatively charged DNA to the positively charged polylysine component. For example, hepatocytes are distinguished by the presence on the cell surface of asialoglycoprotein receptors which clear asialoglycoproteins from the serum. Coupling of DNA to an asialoglycoprotein via a polycation such as polylysine can target the transfer of exogenous DNA into liver cells. The complexes can be infused into the liver either via the biliary tract or vascular bed, whereupon they are taken up by hepatocytes.

A more general approach utilizes the transferrin receptor which is expressed in many cell types, but is relatively enriched in proliferating cells and hemopoietic cells (Figure 22.7). Gene transfer efficiency may be high but the method is not designed to allow integration of the transferred genes. A further problem has been that the protein— DNA complexes are not particularly stable in serum. Additionally, the DNA conjugates may be entrapped in endosomes and degraded in lysosomes, unless previously co-transferred with, or physically linked to, an adenovirus molecule (see Figure 22.5).

Figure 22.7. Gene transfer via the receptor-mediated endocytosis pathway.

Figure 22.7

Gene transfer via the receptor-mediated endocytosis pathway. The negatively charged plasmid DNA can bind reversibly to the positively charged polylysine attached to the transferrin molecule. During this process, the DNA is condensed into a compact circular (more...)

22.3. Therapeutics based on targeted inhibition of gene expression and mutation correction in vivo

22.3.1. Principles and applications of therapy based on targeted inhibition of gene expression in vivo

One way of treating certain human disorders is to selectively inhibit the expression of a predetermined gene in vivo. In principle, this general approach is particularly suited to treating cancers and infectious diseases, and some immunological disorders. In these cases, the basis of the therapy is to knock out the expression of a specific gene that allows disease cells to flourish, without interfering with normal cell function. For example, attention could be focused on selectively inhibiting the expression of a particular viral gene that is necessary for viral replication, or an inappropriately activated oncogene.

In addition to the above, targeted inhibition of gene expression may offer the possibility of treating certain dominantly inherited disorders. If a dominantly inherited disorder is the result of a loss-of-function mutation, treatment may be difficult using conventional gene augmentation therapy: given that heterozygotes with 50% of normal gene product can be severely affected, very efficient expression of the introduced genes would be required for the gene therapy to be successful. However, dominantly inherited disorders which arise because of a gain-of-function mutation may not be amenable to simple addition of normal genes. Instead, it may be possible, in some cases, to inhibit specifically the expression of the mutant gene, while maintaining expression of the normal allele. Such allele-specific inhibition of gene expression would be facilitated if the pathogenic mutation results in a significant sequence difference between the alleles.

The expression of a selected gene might be inhibited by a variety of different strategies. One possible type of approach involves specific in vivo mutagenesis of that gene, altering it to a form that is no longer functional. Gene targeting by homologous recombination offers the possibility of site-specific mutagenesis to inactivate a gene (Section 21.2.3). However, this technique has only very recently become feasible with normal diploid somatic cells and is still very inefficient. Instead, methods of blocking the expression of a gene without mutating it have been preferred. In principle, this can be accomplished at different levels: at the DNA level (by blocking transcription); at the RNA level (by blocking post-transcriptional processing, mRNA transport or engagement of the mRNA with the ribosomes); or at the protein level (by blocking post-translational processing, protein export or other steps that are crucial to the function of the protein).

Various techniques for selectively inhibiting expression of a specific gene have been devised, and include examples where expression is inhibited at all three major levels (see Figure 22.8):

Figure 22.8. Targeted inhibition of gene expression in vivo.

Figure 22.8

Targeted inhibition of gene expression in vivo. Gene therapy based on selective inhibition of a predetermined gene in vivo can be achieved at several levels. In principle, it is possible to mutate the gene via homologous recombination-mediated gene targeting (more...)

  • Targeted inhibition of expression at the DNA level. Under certain conditions, DNA can form triple-stranded structures, as occurs naturally in the case of a portion of the mitochondrial genome (Section 7.1.1). The rationale of triple helix therapeutics is to design a gene-specific oligonucleotide that will have a high chance of base-pairing with a defined double-stranded DNA sequence of a specific target gene in order to inhibit transcription of that gene (Vasquez and Wilson, 1998). Binding of the single-stranded oligonucleotide to a pre-existing double helix occurs by Hoogsteen hydrogen bonds and certain bases are preferred. The most stable of such bonds are formed by a G binding to the G of a GC base pair and a T binding to the A of an AT base pair.
  • Targeted inhibition of expression at the RNA level. Antisense therapeutics involves binding of gene-specific oligonucleotides or polynucleotides to the RNA; in some cases, the binding agent may be a specifically engineered ribozyme, a catalytic RNA molecule that can cleave the RNA transcript (Section 22.3.2).
  • Targeted inhibition of expression at the protein level. Oligonucleotide aptamers and intracellular antibodies can be designed to specifically bind to and inactivate a selected polypeptide/protein (Section 22.3.3).

22.3.2. Antisense oligonucleotides or polynucleotides can bind to a specific mRNA, inhibiting its translation and, in some cases, ensuring its destruction

During transcription, only one of the two DNA strands in a DNA duplex, the template strand (or antisense strand), serves as a template for making a complementary RNA molecule. As a result, the base sequence of the single-stranded RNA transcript is identical to the other DNA strand (the sense strand), except that U replaces T. Any oligonucleotide or polynucleotide which is complementary in sequence to an mRNA sequence can therefore be considered to be an antisense sequence.

Binding of an antisense sequence to the corresponding mRNA sequence would be expected to interfere with translation, and thereby inhibit polypeptide synthesis. Indeed, naturally occurring antisense RNA is known to provide a way of regulating the expression of genes in some plant and animal cells, as well as in some microbes. Synthetic oligonucleotides can be designed to be complementary in sequence to a specific mRNA and, when transferred into cells, show evidence of inhibition of expression of the corresponding gene, which can occur at different levels not just translation. Antisense therapeutics is the application of antisense technology to block expression of a disease-causing gene such as a viral gene or a cancer gene with the aim of combating disease.

Antisense oligodeoxynucleotides

The use of artificial antisense oligodeoxynucleotides is often favored, simply because they can be synthesized so simply. They can be transferred efficiently into the cytoplasm of cells using liposomes, and can migrate rapidly to the nucleus by passive diffusion through the pores of the nuclear envelope. Note that even although antisense oligonucleotides migrate to the nucleus, they do not bind the double-stranded DNA because they are specifically designed to have a low chance of binding to double-stranded DNA, unlike the oligonucleotides used in triple helix technologies which are deliberately designed for this purpose.

Antisense oligodeoxyribonucleotides (ODNs) are preferred as they are generally less vulnerable to nuclease attack than oligoribonucleotides. Nevertheless, to protect against degradation by cellular exonucleases it is still usual to modify the oligonucleotides at their 3′ or 5′ ends e.g. by introducing more resistant phosphorothioate bonds where sulfur atoms are linked to phosphate groups instead of the normal oxygen atoms. Antisense ODNs are also preferred because they have the additional advantage of inducing the destruction of an mRNA to which they bind. This is so because an ODN-mRNA hybrid, like all DNA-RNA hybrids, is vulnerable to attack by a specific class of intracellular ribonuclease, RNase H which selectively cleaves the RNA strand.

General optimism about the power of antisense technology has been sufficient for it to be used in several clinical trials (Wagner, 1994 and Table 22.7). However, formidable technical challenges remain to be overcome, including unexpected ‘nonantisense effects’. Folding of target RNAs and/or their association with specific proteins in the cell often means that the antisense molecule is unable to bind to its target (see Branch, 1998).

Table 22.7. Examples of cancer gene therapy trials.

Table 22.7

Examples of cancer gene therapy trials.

Peptide nucleic acids

Peptide nucleic acids (PNAs) are artificially constructed by attaching the bases found in nucleic acids to a pseudopeptide backbone. The normal phosphodiester backbone is entirely replaced with a polyamide (peptide) backbone composed of 2-aminoethyl glycine units. As a result, PNAs have improved flexibility compared to DNA or RNA, which permits more stable hybridization to DNA or RNA (by Watson-Crick hydrogen bonding). They are also more resistant to nuclease attack and may therefore be useful alternatives to conventional antisense oligonucleotides (see Corey, 1997).


Some RNA molecules are able to lower the activation energy for specific biochemical reactions, and so effectively function as enzymes (ribozymes). They contain two essential components: target recognition sequences (which base-pair with complementary sequences on target RNA molecules), and a catalytic component which cleaves the target RNA molecule while the base-pairing holds it in place. The cleavage leads to inactivation of the RNA, presumably because of subsequent recognition by intracellular nucleases of the two unnatural ends. Examples include human ribonuclease P and various ribozymes obtained from plant viroids (virus-like particles).

Genetic engineering can be employed to custom design the recognition sequence so that the ribozyme now contains sequences that are complementary to a specific mRNA molecule, and appropriately modified hammerhead ribozymes may be useful in gene therapy (Figure 22.9). A particularly promising application has been evisaged in treating dominant genetic disorders by specifically targeting the mutant allele (Phylactou et al., 1998). Ribozymes such as the group I intron ribozyme can also be used to repair mutant RNA molecules (see Section 22.3.4)

Figure 22.9. Genetically engineered hammerhead ribozymes.

Figure 22.9

Genetically engineered hammerhead ribozymes. The hammerhead ribozyme is a constituent of some plant viroids (virus-like particles) and is so called because of the shape of its catalytic component. It is trans-acting and cleaves specific target RNA molecules, (more...)

22.3.3. Artificially designed intracellular antibodies (intrabodies), oligonucleotide aptamers and mutant proteins can inhibit the function of a specific polypeptide

Intracellular antibodies (intrabodies)

Antibody function is normally conducted extracellularly. Once synthesized, they are normally secreted into the extracellular fluid, or are transported to the surface of the B cell to act as an antigen receptor. Recently, however, it has been possible to design genes encoding intracellular antibodies, or intrabodies. Intrabodies are engineered to have a single chain by coupling the variable domain of the heavy chain to the variable domain of the light chain through a peptide linker, thereby preserving the affinity of the parent antibody. Intrabodies can be directed to a particular cell compartment where they can bind to and inactivate a specific cell molecule such as a disease-causing protein, and so they have been envisaged to have potential for treating certain diseases, such as infectious diseases (Rondon and Marasco, 1997).

Oligonucleotide aptamers

Oligonucleotide aptamers are oligonucleotides which can bind to a specific protein sequence of interest. A general method of identifying aptamers is to start with partially degenerate oligonucleotides, and then simultaneously screen the many thousands of oligonucleotides for the ability to bind to a desired protein. The bound oligonucleotide can be eluted from the protein and sequenced to identify the specific recognition sequence. Transfer of large amounts of a chemically stabilized aptamer into cells can result in specific binding to a predermined polypeptide, thereby blocking its function. Currently, the therapeutic potential of this technology has yet to be realized (Osborne et al., 1997).

Mutant proteins

Naturally occurring gain-of-function mutations can involve the production of a mutant polypeptide that binds to the wild-type protein, inhibiting its function. In many such cases, the wild-type polypeptides naturally associate to form multimers, and incorporation of a mutant protein inhibits this process (see Section 16.5). In some cases, gene therapy may be possible by designing genes to encode a mutant protein that can specifically bind to and inhibit a predetermined protein, such as a protein essential for the life-cycle of a pathogen. For example, one form of gene therapy for AIDS involves artificial production of a mutant HIV-1 protein in an attempt to inhibit multimerization of the viral core proteins (see Section 22.5.4).

22.3.4. Artificial correction of a pathogenic mutation in vivo is possible, in principle, but is very inefficient and not readily amenable to clinical applications

Certain disorders are not easy targets for conventional gene therapy. For example, dominantly inherited disorders where a simple mutation results in a pathogenic gain of function cannot be treated by gene augmentation therapy, and targeted inhibition of gene expression may be difficult to achieve. An alternative to conventional gene therapy involves repair of a mutant sequence in vivo. In principle, this can be done by a variety of different experimental strategies at both the level of the mutant gene or its transcript (Woolf, 1998).

Therapeutic repair at the DNA level

One possible approach is to achieve correction of the genetic defect by therapeutic gene targeting. However, while there have been substantial advances in our understanding of homologous recombination-based gene targeting in the cells of complex eukaryotes, the efficiency of homologous recombination remains extraordinarily low in such systems and there are formidable challenges in applying this technology to in vivo gene therapy. An alternative, recently developed gene targeting method, which uses chimeric RNA/DNA oligonucleotides has, however, been claimed to be a much more efficient way of inducing site-directed mutagenesis in vivo (Kren et al., 1998). Other possibilities for therapeutic DNA repair utilize triple helix formation and peptide nucleic acids (see Section 22.3.2 and Woolf, 1998).

Therapeutic repair at the RNA level

An alternative approach to gene targeting is to repair the genetic defect at the RNA level. One possibility is to use a therapeutic ribozyme (Rossi, 1998). One method envisages using a class of ribozyme known as group I introns, which are distinguished by their ability to fold into a very specific shape, capable of both cutting and splicing RNA. If a transcript has, for example, a nonsense or a missense mutation, it may be possible to design specific ribozymes that can cut the RNA upstream of the mutation and then splice in a corrected transcript, a form of trans-splicing (see Figure 22.10). Thus far, this technology is in its infancy, and catalytic efficiency needs to be improved.

Figure 22.10. Some ribozymes also have the potential of repairing mutations in mRNA.

Figure 22.10

Some ribozymes also have the potential of repairing mutations in mRNA. Group I introns are a class of self-splicing intron (see Box 14.3). The RNA transcript acts as a ribozyme by catalyzing the cleavage of the RNA and subsequent splicing. They possibly (more...)

Another possibility is therapeutic RNA editing. This involves using a complementary RNA oligonucleotide to bind specifically to a mutant transcript at the sequence containing the pathogenic point mutation, and an RNA editing enzyme, such as double-stranded RNA adenosine deaminase, to direct the desired base modification. Again this technology is in its infancy and formidable technical difficulties need to be overcome before clinical applications can be envisaged.

22.4. Gene therapy for inherited disorders

Over the last two decades molecular genetic technologies have been spectacularly successful in identifying and characterizing novel disease genes, and in devising novel diagnostic tests for inherited disorders. In contrast, the dream of successfully applying molecular genetic technologies on a large scale to curing, or even treating disease has remained unfulfilled. A new era was heralded when the first gene therapy trial for an inherited disease began in 1990, but exciting though this prospect was, reviews of clinical trials have shown that the initial enthusiams were misplaced (Ross et al., 1996; Knoell and Yiu, 1998). Even now, gene therapy has not cured any patient and there is precious little evidence for any significant clinical benefit in the trials that have been conducted thus far: any amelioration of the diseases that have been treated have been modest and very short-lived. Instead, there is now widespread recognition of the limitations of the current technologies and the need for safer and more efficient gene delivery systems (Verma and Somia, 1997; Anderson, 1998).

While recognizing that current gene delivery methods are not very effective, there remains considerable optimism that this is a temporary difficulty that can be overcome by future technological improvements. However, some genetic disorders may not be so easy to treat as others. Common nonmendelian genetic diseases may involve a complex interplay between different genetic loci and/or environmental factors, and so possible gene therapy approaches may not be straightforward. Single gene disorders where individuals are severely affected and where there is no effective treatment, are more obvious candidates for gene therapy. Within the single gene disorder category, however, differing pathogeneses means that certain single gene disorders will be more amenable to gene therapy approaches than others (Table 22.4).

Table 22.4. Factors governing the amenability of single gene disorders to gene therapy approaches.

Table 22.4

Factors governing the amenability of single gene disorders to gene therapy approaches.

22.4.1. Recessively inherited disorders are conceptually the easiest inherited disorders to treat by gene therapy

Those disorders where the disease results from a simple deficiency of a specific gene product are generally the most amenable to treatment: high level expression of an introduced normal allele should be sufficient to overcome the genetic deficiency. Recessively inherited disorders have been of particular interest as candidates for gene therapy because the mutations are almost always simple loss-of-function mutations. Affected individuals have deficient expression from both alleles and so the disease phenotype is due to complete or almost complete absence of normal gene expression. Heterozygotes, however, have about 50% of the normal gene product and are normally asymptomatic. Additionally, there is, in at least some cases, wide variation in the normal levels of gene expression, so that a comparatively small percentage of the average normal amount of gene product may be sufficient to restore the normal phenotype. It is also often observed that the severity of the phenotype of recessive disorders is inversely related to the amount of product that is expressed (see Table 16.4). As a result, even if the efficiency of gene transfer is low, modest expression levels for an introduced gene may make a substantial difference. This is quite unlike dominantly inherited disorders where heterozygotes with loss-of-function mutations have 50% of the normal gene product and may yet be severely affected.

Although recessively inherited disorders are, in principle, amenable to gene augmentation therapy, certain disorders are less amenable than others. In addition to the question of accessibility of the disease tissue, some disorders may be difficult to treat for other reasons. A good example is provided by β-thalassemia which results from mutations in the β-globin gene, HBB. This is a severe disorder affecting hundreds of thousands of people worldwide, and superficially would appear to be an excellent candidate for gene therapy: the gene is very small and has been characterized extensively, the disorder is recessively inherited and affects blood cells. An initial attempt at gene therapy for this disorder in 1980 failed, largely because of inefficient gene transfer and poor expression of the introduced β-globin genes. Even though we now know much about how this gene is expressed, there have been no subsequent gene therapy attempts. This is due to the problem of the very tight control of gene expression required following insertion of a normal β-globin gene into the desired cells: the amount of β-globin product made must be equal to the amount of α-globin. If too much β-globin were to be made, the imbalance between β-globin and α-globin chains would result in an α-thalassemia phenotype.

22.4.2. The first gene therapy trial for an inherited disease was initiated in 1990

The first gene therapy trial for an inherited disorder was initiated on 14 September 1990. The patient, Ashanthi DeSilva, was just 4 years old and was suffering from a very rare recessively inherited disorder, adenosine deaminase (ADA) deficiency. ADA is involved in the purine salvage pathway of nucleic acid degradation, and is a housekeeping enzyme which is synthesized in many different types of cell. An inherited deficiency of this enzyme has, however, particularly severe consequences in the case of T lymphocytes, one of the major classes of immune system cells. As a result, ADA- patients suffer from severe combined immunodeficiency. This severe disorder was particularly amenable to gene therapy for a variety of reasons: the ADA gene is small, and had previously been cloned and extensively studied; the target cells are T cells which are easily accessible and easy to culture, enabling ex vivo gene therapy; the disorder is recessively inherited and, importantly, gene expression is not tightly controlled (enzyme levels in the normal population show huge differences between healthy individuals). The observation that allogeneic bone marrow transplantation can cure the disorder suggested that engraftment of T cells alone may be sufficient, and transfer of normal ADA genes into ADA- T cells was noted to result in restoration of the normal phenotype.

Alternative treatments for ADA deficiency do exist. Indeed, the treatment of choice is bone marrow transplantation from a perfectly HLA-matched sibling donor, which provides a cure in about 80% of cases. For children where this is not an option, an alternative is enzyme replacement therapy, consisting of weekly intramuscular injections of ADA conjugated to polyethylene glycol (PEG). PEG stabilizes the ADA enzyme, allowing it to survive and function in the body for days. Inevitably, however, enzyme replacement therapy does not provide full immune reconstitution and so life expectancy is still likely to be shortened (T cells are required for mounting effective immune responses against invading microorganisms, and in preventing cancer).

The novel ADA gene therapy approach involved essentially four steps:


cloning a normal ADA gene into a retroviral vector;


transfecting the ADA recombinant into cultured ADA- T lymphocytes from the patient;


identifying the resulting ADA+ T cells and expanding them in culture;


re-implanting these cells in the patient (see Figure 22.11).

Figure 22.11. Ex vivo gene augmentation therapy for adenosine deaminase (ADA) deficiency.

Figure 22.11

Ex vivo gene augmentation therapy for adenosine deaminase (ADA) deficiency. Note that identification of suitably transformed cells is helped by having an appropriate selectable marker in the retrovirus vector, such as a neoR gene which confers resistance (more...)

This approach was never going to be a cure; instead it was designed to be a form of treatment which would need to be repeated on many occasions. Successful treatment would require high efficiency gene transfer into bone marrow stem cells and high levels of expression. The trouble here is that human bone marrow stem cells are very difficult to isolate and insertion of retroviral vectors into such cells is very inefficient. Enrichment for such cells is possible using the monoclonal antibody CD34 which selectively binds a population of cells that includes totipotent stem cells and subsequent gene therapy trials conducted on neonates used retroviral transduction of selected CD34+ umbilical cord blood cells.

All patients in the ADA gene therapy trials were treated in parallel by conventional enzyme replacement therapy using PEG-ADA. The combined gene therapy plus enzyme replacement therapy appeared to give initially promising results (as assayed by various measures of antibody and T-cell function, and a dramatic decrease in infections compared with the incidence before treatment). However, cessation of the parallel PEG-ADA treatment led to a decline in immune function despite the persistence of ADA+ T lymphocytes (Kohn et al., 1998). The inescapable conclusion is that improved gene transfer and expression will be needed before ADA gene therapy can be successful.

22.4.3. Since the pioneering work on ADA deficiency, gene therapy trials have been initiated for a few inherited disorders

Gene therapy has been initiated for only a comparatively few inherited disorders in addition to ADA deficiency (see Table 22.5 for examples). Different recessively inherited disorders have been targets for in vivo or ex vivo gene augmentation therapy and, in the one case where a dominantly inherited disorder has been treated, familial hypercholesterolaemia, the patients had the homozygous form of the disease. The following examples are simply illustrative of current progress and difficulties.

Table 22.5. Examples of gene therapy trials for inherited disorders.

Table 22.5

Examples of gene therapy trials for inherited disorders.

Familial hypercholesterolemia (FH)

This disorder is caused by a dominantly inherited deficiency of low density lipoprotein (LDL) receptors, which are normally synthesized in the liver, and is characterized by premature coronary artery disease. About 50% of heterozygous affected males die by 60 years of age, unless treated. Because FH is such a common single gene disorder, homozygotes are occasionally seen. They suffer precocious onset of disease and increased severity, with death from myocardial infarction commonly occurring in late childhood. The first, and only, gene therapy for FH was initiated on the first of five patients with the homozygous form of the disease in 1992. The liver, being a solid internal organ, may not seem to be an ideal choice for targeting gene therapy, and its major cell population, the differentiated hepatocyte, is refractory to infection with retroviruses, the most widely used vector system. However, hepatocytes can be cultured in vitro and, under such conditions, are susceptible to retroviral infection.

Ex vivo gene therapy became a possibility when animal experiments showed that cultured hepatocytes could be injected via the portal venous system - the veins which drain from the intestine directly into the liver - after which they appear to seed in the liver. The gene therapy involved surgical removal of a sizeable portion of the left lobe of the patient's liver, disaggregation of the liver cells and plating in cell culture prior to infection with retroviruses containing a normal human LDLR gene (Grossman et al., 1994). The genetically modified cells were infused back into the patient through a catheter implanted into a branch of the portal venous system. The patient's LDL/high density lipoprotein (HDL) ratio subsequently declined from 10–13 before gene therapy to 5–8, and such improvement was maintained over a long period. However, because of the invasiveness of the procedure and limited effectiveness of therapy, subsequent gene therapies are not intended until there is a significant advance in gene transfer efficiency.

Cystic fibrosis

Cystic fibrosis is an autosomal recessive disorder that results in defective transport of chloride ions through epithelial cells, and results from mutations in a gene, CFTR, which encodes a cAMP-regulated chloride channel. The primary expression of the defect is in the lungs: a sticky mucus secretion accumulates which is prone to chronic infections. Because there are no methods to culture lung cells routinely in the laboratory, in vivo gene therapy approaches have been adopted. As respiratory epithelial cells are differentiated, retroviral vectors cannot be used. Instead, gene therapy trials have used adenovirus vectors or liposomes to transfer a suitably sized CFTR minigene, either through a bronchoscope or through the nasal cavity.

The first adenovirus-based protocol began in 1993 and, although preliminary data have confirmed gene transfer into respiratory epithelium in vivo, there have been major concerns regarding the safety of the procedure. The first patient to be treated with a high dose of recombinant adenovirus experienced transient pulmonary infiltrates and alterations in vital signs, before recovering uneventfully. This experience prompted recognition of the need to confirm the maximum tolerated adenovirus dose. The liposome-based gene therapy trials are regarded as safer procedures, but the efficiency of gene transfer is much lower. Despite an impressive amount of research, CF gene therapy remains ineffective (Boucher, 1999).

Duchenne muscular dystrophy

DMD is a severe X-linked recessive disorder: affected males suffer progressive muscle deterioration, are confined to a wheelchair in their teens and die usually by the third decade. The target tissue is skeletal muscle, and initial interest in treatment for this disorder focused on cell therapy because of the unique cell biology of muscle (Miller and Boyce, 1995). As well as muscle fibers (or myofibers - very long, post-mitotic, multinucleate cells), skeletal muscle contains mononucleate myoblasts which are normally quiescent but can divide and subsequently fuse with myofibers to repair muscle damage. Although implanting normal or genetically modified myoblasts into diseased muscles appeared attractive, difficulties have been evident with this approach in humans, despite promising pilot studies with myoblast transfer in mice.

Suitable gene therapy approaches have also been difficult to conceive, largely because of the lack of a suitable gene transfer system. Oncoretroviral vectors cannot be used because adult skeletal muscle fibers are postmitotic and hence not susceptible to oncoretroviral infection. Adenovirus vectors have been used to deliver genes to muscle fibers in vivo and, although the postmitotic state of muscle nuclei allows the expression to persist, the need for expression to continue over the course of a lifetime (which would be required for successful therapy) remains doubtful. A final problem is the sheer size of the dystrophin coding sequence (~14 kb), although a very large central segment appears not to be crucially important.

In addition to simple gene replacement strategies, alternative methods have been considered including up-regulation of genes encoding proteins that may have a compensatory function. The principle of therapeutic reactivation of fetal genes has been considered for β-thalassemia and sickle cell anemia. Here the idea is to offset genetic deficiencies in β-globin production by reactivation of other globin genes which are largely expressed during the fetal period, such as the γ-globin genes (β-thalassemia and sickle cell anemia patients may have a mild form of the disease if they produce unusually high levels of the HbF fetal haemoglobin; Olivieri and Weatherall, 1998). A similar strategy has been considered for Duchenne muscular dystrophy. Dystrophin has a close relative, utrophin, which is highly expressed during the fetal period and so there is the possibility that up-regulation of utrophin may confer a protective effect. Encouraging results have been obtained in mice where expression of utrophin transgenes in mice with dsytrophin deficiency leads to major improvements in muscle function (Deconinck et al., 1997). Now, considerable effort is being devoted to standard drug-finding approaches to identify small molecules that naturally upregulate the utrophin gene with a view to administering them as drugs in future treatments.

22.5. Gene therapy for neoplastic disorders and infectious disease

22.5.1. General principles of gene therapy for neoplastic disorders and infectious disease

Cancer gene therapies

Many different approaches can be used for cancer gene therapy (see Table 22.6) and, in marked contrast to the few gene therapy trials for inherited disorders, numerous cancer gene therapy trials are currently being conducted (Table 22.7). This reflects partly the severity of the disorders that are being treated and the considerable funding for cancer research, and partly reflects the comparative ease in applying treatments based on targeted killing of disease cells, by introducing genes that encode toxins, etc. or by provoking enhanced immune responses. In a few cases, the gene therapy approach has focused on targeting single genes, such as TP53 gene augmentation therapy and delivery of antisense KRAS genes in the case of some forms of non-small-cell lung cancer. In most cases, however, targeted killing of cancer cells has been conducted without knowing the molecular etiology of the cancer. Thus far, some significant advances have been made against local and metastatic tumor growth, but effective therapy awaits development of more effective methods to transfer and express transgenes or to induce antitumor responses (Hall et al., 1997).

Table 22.6. Potential applications of gene therapy for the treatment of cancer.

Table 22.6

Potential applications of gene therapy for the treatment of cancer.

Gene therapy for infectious disorders

The gene therapy approaches for treating infectious disorders are slightly different. In common with cancer gene therapy, strategies can involve provoking a specific immune response or specific killing of infected cells. An increasingly popular additional approach targets the life-cycle of the infectious agent, reducing its ability to undergo productive infection. Some infectious agents are genetically comparatively stable. Others, however, may be undergoing rapid evolution and (much as in the case of cancer cells) present problems for any general therapy. The classic example is AIDS, where the infectious agent, HIV-1, appears to mutate rapidly.

22.5.2. Ex vivo cancer gene therapies frequently involve attempts to recruit immune system cells to destroy the tumor cells

Gene transfer into tumor-infiltrating lymphocytes

One of the earliest gene therapy protocols used a population of immune system cells for specifically targeting a foreign protein to a tumor. The therapy could be considered to be a form of adoptive immunotherapy (see below) because a gene encoding a cytokine, tumor necrosis factor-α (TNF-α), was transferred into tumor-infiltrating lymphocytes (TILs) in an effort to increase their antitumor efficacy. The TIL population is a natural population of T lymphocytes which can seek out and infiltrate tumor deposits, such as metastatic melanomas. TNF-α is a protein naturally produced by T lymphocytes which, if infused in sufficient amounts in mice, can destroy tumors. However, it is a toxic substance and intravenous infusion of TNF has significant adverse side-effects in humans. An attractive alternative was to use TILs as cellular vectors for transferring the toxic protein directly to tumors. The gene therapy approach that was used, therefore, involved retroviral-mediated transfer of a TNF gene to a TIL population which had initially been obtained from an excised tumor and then grown in culture. Subsequent transfusion of the genetically modified TILs into a patient with metastatic melanoma was expected to result in the TILs ‘homing in’ on the melanomas, expression of the introduced TNF gene and tumor regresssion (Figure 22.12). However, the trial has been marked by comparatively poor efficiency of gene transfer into human TILs and a down-regulation of cytokine expression by the TILs.

Figure 22.12. Genetic modification of cultured tumor-infiltrating lymphocytes can be used to target therapeutic genes to a solid tumor.

Figure 22.12

Genetic modification of cultured tumor-infiltrating lymphocytes can be used to target therapeutic genes to a solid tumor. This approach has been used in an attempt at ex vivo gene therapy for metastatic melanoma. The tumor-infiltrating lymphocytes (TIL) (more...)

Adoptive immunotherapy by genetic modification of tumor cells

Animal studies in which murine tumor cells were genetically modified by the insertion of genes encoding various cytokines [several different interleukins (ILs), TNF-α, interferon (IFN)-γ, granulocyte—macrophage colonystimulating factor (GM-CSF)] and then re-implanted in mice gave cause for encouragement. In each case, the genetically altered tumor cells either never grew, or grew and then regressed. In addition, most of the treated mice were then systemically immune to reimplantation of nonmodified tumors. However, the results were much less satisfactory when animals with established, sizeable tumors were treated. Nevertheless, the idea of modifying a patient's own tumor cells for use as a vaccine (adoptive immunotherapy) caught on, and human gene therapy trials have been approved for the insertion of cytokine genes using retrovirus vectors for treating a wide variety of cancers (see Table 22.7 and Ockert et al., 1999).

In each case, the idea is to immunize the patients specifically against their own tumors by genetically modifying the tumor with one of a variety of genes that are expected to increase the host immune reactivity to the tumor. In addition to cytokine genes, other genes such as foreign HLA antigen genes have been transferred to tumors for the same general reason. Insertion of genes encoding HLA-B7 into tumors of patients lacking HLA-B7 is intended to provoke an immune response to the tumors as a consequence of the presence on the tumor cell surface of the effectively foreign HLA-B7 antigen (see Table 22.7 for some examples). Such a response is hoped to provide subsequent immunity against the same type of tumor even in the absence of the HLA-B7 antigen.

Adoptive immunotherapy by genetic modification of fibroblasts

One problem with ex vivo therapy for tumors is the difficulty in growing tumor cells in vitro: less than 50% of tumor cell lines grow in long-term culture. As an alternative, fibroblasts, which are much easier to adapt to long-term tissue culture, have been targeted in some cases. For example, transfer of genes encoding the cytokines IL-2 and IL-4 into skin fibroblasts grown in culture provides the basis of some clinical trials for treatment of breast cancer, colorectal cancer, melanoma and renal cell carcinoma. The IL-2- and IL-4-secreting fibroblasts are then mixed with irradiated autologous tumor cells and injected subcutaneously. In such cases, the hope is that the local production and secretion of cytokines by the transferred fibroblasts will induce a vigorous immune response to the nearby irradiated tumor cells and thereby result in a systemic anticancer immune response.

Other immunological approaches

Two other ex vivo gene therapy strategies use immunological approaches to tumor destruction. One involves transferring an antisense insulin-like growth factor-1 (IGF1) gene into tumor cells in order to block production of IGF-1 (Anthony et al., 1998). Animal studies have shown that when tumor cells modified in this way are reimplanted in vivo, they provoke an immune response which can lead to destruction of nonmodified tumors, but the basis of immunological destruction is not known. A second approach involves the insertion of a co-stimulatory molecule such as B7-1 or B7-2, molecules which are normally present on lymphocytes, being required for full T-lymphocyte activation (see Putzer et al., 1997).

22.5.3. In vivo gene therapy may be the only feasible approach for some cancers

Currently, a variety of different gene therapy approaches are being used involving genetic modification of tumor cells in vivo. In some cases, adoptive immunotherapy approaches are being employed, as in the case of increasing the immunogenicity of melanoma, colorectal tumors and a variety of solid tumors by the direct injection of liposomes containing a gene which encodes HLA-B7. The tumor cells take up the liposomes by phagocytosis and express the foreign HLA-B7 antigen transiently on their cells. More recent modifications include the additional insertion of a gene encoding the conserved light chain of HLA antigens, β2-microglobulin.

A second approach has been the use of retrovirus-mediated transfer of a gene encoding a prodrug, a reagent that confers sensitivity to cell killing following subsequent administration of a suitable drug. In one recent example, the target cells were brain tumor cells, notably recurrent glioblastoma multiforme, and the retroviruses were provided in the form of murine fibroblasts that are producing retroviral vectors (retroviral vector-producing cells or VPCs). The cells were directly implanted into multiple areas within growing tumors using stereotactic injections guided by magnetic resonance imaging (Figure 22.13). Once injected, the VPCs continuously produce retroviral particles within the tumor mass, transferring genes into surrounding tumor cells. Although retroviruses are not normally used for in vivo gene therapy because of their sensitivity to serum complement, they are comparatively stable in this special environment and have the advantage that, since they only infect actively dividing cells, the tumor cells are a target, but not nearby brain cells (which are usually terminally differentiated).

Figure 22.13. In vivo gene therapy for brain tumors.

Figure 22.13

In vivo gene therapy for brain tumors. This example shows a strategy for treating glioblastoma multiforme in situ using a delivery method based on magnetic resonance imaging-guided stereotactic implantation of retrovirus vector-producing cells (VPCs). (more...)

The prodrug gene that was transferred is a HSV gene which encodes thymidine kinase (HSV-tk). HSV-tk confers sensitivity to the drug gancyclovir by phosphorylating it within the cell to form gancyclovir monophosphate which is subsequently converted by cellular kinases to gancyclovir triphosphate. This compound inhibits DNA polymerase and causes cell death (see Figure 22.13). Such therapy appears to benefit from a phenomenon known as the bystander effect: adjacent tumor cells that have not taken up the HSV-tk gene may still be destroyed. This is thought to be due to diffusion of the gancyclovir triphosphate from cells which have taken up the HSV-tk gene, perhaps via gap junctions.

22.5.4. Gene therapy for infectious disorders is often aimed at selectively interfering with the life-cycle of the infectious agent

Current gene therapy trials for infectious disorders are conspicuously targeted at treating AIDS patients. The infectious agent for this usually fatal disorder is a class of retrovirus known as HIV-1 which can infect helper T lymphocytes, a crucially important subset of immune system cells (see Figure 22.14). Two features of HIV-1 make it especially deadly: it eventually kills the helper T cells (thereby rendering patients susceptible to other infections), and the provirus tends to persist in a latent state before being suddenly activated (the lack of virus production during the latent state complicates antiviral drug treatment). A major problem is that the HIV genome is mutating at a very high rate.

Figure 22.14. The HIV-1 virus life-cycle.

Figure 22.14

The HIV-1 virus life-cycle. The HIV-1 virus is a retrovirus which contains two identical single-stranded viral RNA molecules and various viral proteins within a viral protein core, which itself is contained within an outer envelope. The latter contains (more...)

In principle, a variety of gene therapy strategies can be envisaged for treating AIDS. As in the case of cancer gene therapy, infected cells can be killed directly (by insertion of a gene encoding a toxin or a prodrug; see above) or indirectly, by enhancing an immune response against them. For example, this can involve transferring a gene that encodes an HIV-1 antigen, such as the envelope protein gp120, and expressing it in the patient in order to provoke an immune response against the HIV-1 virus, or the patient's immune system can be boosted by transfer and expression of a gene encoding a cytokine, such as an interferon. Another general approach, which is applicable to all disorders caused by infectious agents, is to find a means of interfering with the life-cycle of the infectious agent.

Gene therapy strategies designed to interfere with the HIV-1 life-cycle

A wide variety of such strategies are available (Gilboa and Smith, 1994). Inhibition has been envisaged at three major levels:

  • Blocking HIV-1 infection. HIV-1 normally infects T lymphocytes by binding of the viral gp120 envelope protein to the CD4 receptor on the cell membrane. Transfer of a gene encoding a soluble form of the CD4 antigen (sCD4) into T lymphocytes or hemopoietic cells and subsequent expression will result in circulating sCD4. If the levels of circulating sCD4 are sufficiently high, binding of sCD4 to the gp120 protein of HIV-1 viruses could be imagined to inhibit infection of T-lymphocytes without compromising T lymphocyte function.
  • Inhibition at the RNA level. The production of HIV-1 RNA can be selectively inhibited by standard antisense/ribozyme approaches (see Section 22.3.2), and also by the use of RNA decoys. The latter strategy exploits unique regulatory circuits which operate during HIV replication. Two key HIV regulatory gene products are tat and rev which bind to specific regions of the nascent viral RNA, known as TAR and RRE respectively (Figure 22.14). Artificial expression of short RNA sequences corresponding to TAR or RRE will generate a source of decoy sequences which can compete for binding of tat and rev, and possibly thereby inhibit binding of these proteins to their physiological target sequences.
  • Inhibition at the protein level. There are numerous different strategies. One strategy involves designing intracellular antibodies (see Section 22.3.3), against HIV-1 proteins, such as the envelope proteins. Another involves introducing genes that encode dominant-negative mutant HIV proteins which can bind to and inactivate HIV proteins (transdominant proteins). For example, transdominant mutant forms of the gag proteins have been shown to be effective in limiting HIV-1 replication, possibly by interfering with multimerization and assembly of the viral core (Gilboa and Smith, 1994).

22.6. The ethics of human gene therapy

All current gene therapy trials involve treatment for somatic tissues (somatic gene therapy). Somatic gene therapy, in principle, has not raised many ethical concerns. Clearly, every effort must be made to ensure the safety of the patients, especially since the technologies being used for somatic gene therapy are still at an undeveloped stage. However, confining the treatment to somatic cells means that the consequences of the treatment are restricted to the individual patient who has consented to this procedure. Many, therefore, view the ethics of somatic gene therapy to be at least as acceptable as, say, organ transplantation, and feel that ethical approval is appropriate for carefully assessed proposals. Patients who are selected for such treatments have severely debilitating, and often life-threatening, disease for which no effective conventional therapy is available. As a result, despite the obvious imperfections of the technology, it may even be considered to be unethical to refuse such treatment. The same technology has the potential, of course, to alter phenotypic characters that are not associated with disease, such as height for instance. Such genetic enhancement, although not currently considered, can be expected to pose greater ethical problems; attempts to produce genetically enhanced animals have not been a success and in some cases have been spectacular failures (Gordon, 1999).

Germline gene therapy, involving the genetic modification of germline cells (e.g. in the early zygote), is considered to be entirely different. It has been successfully practised on animals (e.g. to correct β-thalassemia in mice). However, thus far, it has not been sanctioned for the treatment of human disorders, and approval is unlikely to be given in the near future, if ever (see next section).

22.6.1. Human germline gene therapy has not been practised because of ethical concerns and limitations of the technology for germline manipulation

The lack of enthusiasm for the practice of germline gene therapy can be ascribed to three major reasons.

The imperfect technology for genetic modification of the germline

Germline gene therapy requires modification of the genetic material of chromosomes, but vector systems for accomplishing this do not allow accurate control over the integration site or event. In somatic gene therapy, the only major concern about lack of control over the fate of the transferred genes is the prospect that one or more cells undergoes neoplastic transformation. However, in germline gene therapy, genetic modification has implications not just for a single cell: accidental insertion of an introduced gene or DNA fragment could result in a novel inherited pathogenic mutation.

The questionable ethics of germline modification

Genetic modification of human germline cells may have consequences not just for the individual whose cells were originally altered, but also for all individuals who inherit the genetic modification in subsequent generations. Germline gene therapy would inevitably mean denial of the rights of these individuals to any choice about whether their genetic constitution should have been modified in the first place (Wivel and Walters, 1993). Some ethicists, however, have considered that the technology of germline modification will inevitably improve in the future to an acceptably high level and, provided there are adequate regulations and safeguards, there should then be no ethical objections (see, for example, Zimmerman, 1991). At a recent scientific research meeting in the USA some scientists have also come out in support of such a development (Wadman, 1998).

From the ethical point of view, an important consideration is to what extent technologies developed in an attempt to engineer the human germline could subsequently be used not to treat disease but in genetic enhancement. There are powerful arguments as to why germline gene therapy is pointless (see next section). There are serious concerns, therefore, that a hidden motive for germline gene therapy is to enable research to be done on germline manipulation with the ultimate aim of germline-based genetic enhancement. The latter could result in positive eugenics programs, whereby planned genetic modification of the germline could involve artificial selection for genes that are thought to confer advantageous traits.

The implications of human genetic enhancement are enormous. Future technological developments may make it possible to make very large alterations to the human germline by, for example, adding many novel genes using human artificial chromosomes (Grimes and Cooke, 1998). Some people consider that this could advance human evolution, possibly paving the way for a new species, homo sapientissimus. To have any impact on evolution, however, genetic enhancement would need to be operated on an unfeasibly large scale (Gordon, 1999).

Even if positive eugenics programs were judged to be acceptable in principle and genetic enhancement were to be practised on a small scale, there are extremely serious ethical concerns. Who decides what traits are advantageous? Who decides how such programs will be carried out? Will the people selected to have their germlines altered be chosen on their ability to pay? How can we ensure that it will not lead to discrimination against individuals? Previous negative eugenics programs serve as a cautionary reminder. In the recent past, for example, there have been horrifying eugenics programs in Nazi Germany, and also in many states of the USA where compulsory sterilization of individuals adjudged to be feeble-minded was practised well into the present century.

The questionable need for germline gene therapy

Germline genetic modification may be considered as a possible way of avoiding what would otherwise be the certain inheritance of a known harmful mutation. However, how often does this situation arise and how easy would it be to intervene? A 100% chance of inheriting a harmful mutation could most likely occur in two ways. One is when an affected woman is homoplasmic for a harmful mutation in the mitochondrial genome (see Section 16.6.4) and wishes to have a child. The trouble here is that, because of the multiple mitochondrial DNA molecules involved, gene therapy for such disorders is difficult to devise.

A second situation concerns inheritance of mutations in the nuclear genome. To have a 100% risk of inheriting a harmful mutation would require mating between a man and a woman both of whom have the same recessively inherited disease, an extremely rare occurrence. Instead, the vast majority of mutations in the nuclear genome are inherited with at most a 50% risk (for dominantly inherited disorders) or a 25% risk (for recessively inherited disorders). In vitro fertilization provides the most accessible way of modifying the germline. However, if the chance that any one zygote is normal is as high as 50 or 75%, gene transfer into an unscreened fertilized egg which may well be normal would be unacceptable: the procedure would inevitably carry some risk, even if the safety of the techniques for germline gene transfer improves markedly in the future. Thus, screening using sensitive PCR-based techniques would be required to identify a fertilized egg with the harmful mutation. Inevitably, the same procedure can be used to identify fertilized eggs that lack the harmful mutation. Since in vitro fertilization generally involves the production of several fertilized eggs, it would be much simpler to screen for normal eggs and select these for implantation, rather than to attempt genetic modification of fertilized eggs identified as carrying the harmful mutation.

Further reading

  1. Lemoine N, Cooper D (1998) Gene Therapy. BIOS Scientific Publishers, Oxford.
  2. Thomas A (ed.) (1998) Therapeutic horizons. Nature, 392, (suppl.), 1–35.


  1. Acsadi G, Dickson G, Love D R. Human dystrophin expression in MDX mice after intramuscular injection of DNA constructs. Nature. (1991);352:815–818. [PubMed: 1881437]
  2. Anderson W F. Human gene therapy. Nature. (1998); 392 (Suppl.):25–30. [PubMed: 9579858]
  3. Anthony D D, Pan Y X, Wu S G, Shen F, Guo Y J. Ex vivo and in vivo IGF-I antisense RNA strategies for treatment of cancer in humans. Adv. Exp. Med. Biol. (1998);451:27–34. [PubMed: 10026846]
  4. Berkower I. The promise and pitfalls of monoclonal antibody therapeutics. Curr. Opin. Biotechnol. (1996);7:622–628. [PubMed: 8939645]
  5. Boucher R C. Status of gene therapy for cystic fibrosis lung disease. J. Clin. Invest. (1999);103:441–445. [PMC free article: PMC408111] [PubMed: 10021450]
  6. Branch A D. A good antisense molecule is hard to find. Trends Biochem. Sci. (1998);23:45–50. [PubMed: 9538686]
  7. Cech T. Group I introns: new molecular mechanisms for mRNA repair. Biotechnology. (1995);13:323–326. [PubMed: 9634773]
  8. Corey D R. Peptide nucleic acids: expanding the scope of nucleic acid recognition. Trends Biotechnol. (1997);15:224–229. [PubMed: 9183865]
  9. Culver K W, Blaese R M. Gene therapy for cancer. Trends Genet. (1994);10:174–178. [PubMed: 8036720]
  10. Curiel DT (1994) High-efficiency gene transfer mediated by adenovirus—polylysine—DNA complexes. In: Annals of the New York Academy of Sciences (Huber and Lazo, eds), vol. 716, pp. 36–41. The New York Academy of Sciences, New York. [PubMed: 7912914]
  11. Deconinck N, Tinsley J M, De-Backer F, Fisher R, Kahn D, Phelps S, Davies K E, Gillis J M. Expression of truncated utrophin leads to major functional improvements in dystrophin-deficient muscles of mice. Nature Med. (1997);3:1216–1221. [PubMed: 9359695]
  12. Donnelly J J, Ulmer J B, Shiver J W, Liu M A. DNA vaccines. Annu. Rev. Immunol. (1997);15:617–648. [PubMed: 9143702]
  13. Gage F H. Cell therapy. Nature. (1998);392 (Suppl.):18–24. [PubMed: 9579857]
  14. Gilboa E, Smith C. Gene therapy for infectious diseases: the AIDS model. Trends Genet. (1994);10:139–144. [PubMed: 7913251]
  15. Gordon J W. Genetic enhancement in humans. Science. (1999);283:2023–2024. [PubMed: 10206908]
  16. Grimes B, Cooke H. Engineering mammalian chromosomes. Hum. Mol. Genet. (1998);7:1635–1640. [PubMed: 9735385]
  17. Grossman M, Raper S E, Kozarsky K, Stein E A, Engelhardt J F, Muller D, Lupien P J, Wilson J M. Successful ex vivo gene therapy directed to liver in a patient with familial hypercholesterolaemia. Nature Genet. (1994);6:335–341. [PubMed: 8054972]
  18. Hall S J, Chen S -H, Woo S L C. The promise and reality of cancer gene therapy. Am. J. Hum. Genet. (1997);61:785–789. [PMC free article: PMC1716009] [PubMed: 9382085]
  19. Huang L, Li S. Liposomal gene delivery: a complex package. Nat. Biotechnol. (1997);15:620–621. [PubMed: 9219259]
  20. Jiao S, Gurevich V, Wolff J A. Long-term correction of rat model of Parkinsons-disease by gene-therapy. Nature. (1993);362:450–453. [PubMed: 8096625]
  21. Kay M A, Liu D, Hoogerbrugge P M. Gene therapy. Proc. Natl. Acad. Sci. USA. (1997);94:12744–12746. [PMC free article: PMC34169] [PubMed: 9398067]
  22. Knoell D L, Yiu I M. Human gene therapy for hereditary diseases: a review of trials. Am. J. Health Syst. Pharm. (1998);55:899–904. [PubMed: 9588249]
  23. Kohn D B, Hershfield M S, Carbonaro D. T lymphocytes with a normal ADA gene accumulate after transplantation of transduced autologous umbilical cord blood CD34+ cells on ADA-deficient SCID neonates. Nature Med. (1998);4:775–780. [PMC free article: PMC3777239] [PubMed: 9662367]
  24. Kren B T, Bandyopahyay P, Steer C L. In vivo site-directed mutagenesis of the factor IX gene in rat liver and in isolated hepatocytes by chimeric RNA/DNA oligonucleotides. Nature Med. (1998);4:285–290. [PubMed: 9500600]
  25. Liu M A. Vaccine developments. Nature Med. (Suppl.) (1998);4:515–519.
  26. Lubon H. Transgenic animal bioreactors in biotechnology and production of blood proteins. Biotechnol. Annu. Rev. (1998);4:1–54. [PubMed: 9890137]
  27. Mendez M J, Green L L, Corvalan J R F, Jia X C, Maynard-Currie C E, Yang X -D. Functional transplant of megabase human immunoglobulin loci recapitulates human antibody response in mice. Nature Genet. (1997);15:146–156. [PubMed: 9020839]
  28. Miller J B, Boyce F M. Gene therapy by and for muscle cells. Trends Genet. (1995);11:163–165. [PubMed: 7785070]
  29. Naldini L. Lentiviruses as gene transfer agents for delivery to non-dividing cells. Curr. Opin. Biotechnol. (1998);9:457–463. [PubMed: 9821272]
  30. Ockert D, Schmitz M, Hampl M, Rieber E P. Advances in cancer immunotherapy. Immunol. Today. (1999);20:63–65. [PubMed: 10098323]
  31. Olivieri N F, Weatherall D J. The therapeutic reactivation of fetal hemoglobin. Hum. Mol. Genet. (1998);7:1655–1658. [PubMed: 9735388]
  32. Osborne S E, Matsumura I, Elington A D. Aptamers as therapeutic and diagnostic reagents: problems and prospects. Curr. Opin. Chem. Biol. (1997);1:5–9. [PubMed: 9667829]
  33. Paleyanda R K, Velander W H, Lee T K. Transgenic pigs produce functional human factor VIII milk. Nat. Biotechnol. (1997);15:971–975. [PubMed: 9335047]
  34. Pardoll D M. Cancer vaccines. Nat. Med. (1998);4 (suppl.):525–531. [PubMed: 9585204]
  35. Phylactou L A, Kilpatrick M W, Wood M J A. Ribozymes as therapeutic tools for genetic disease. Hum. Mol. Genet. (1998);7:1649–1653. [PubMed: 9735387]
  36. Putzer B M, Hitt M, Muller W J, Emtage P, Gauldie J, Graham F L. Interleukin 12 and B7-1 costimulatory molecule expressed by an adenovirus vector act synergistically to facilitate tumor regression. Proc. Natl Acad. Sci. USA. (1997);94:10889–10894. [PMC free article: PMC23520] [PubMed: 9380730]
  37. Rondon I J, Marasco W A. Intracellular antibodies (intrabodies) for gene therapy of infectious disease. Annu. Rev. Microbiol. (1997);51:257–283. [PubMed: 9343351]
  38. Ross G, Erickson R, Knorr D. Gene therapy in the United States: a five year report. Hum. Gene Ther. (1996);7:1781–1790. [PubMed: 8886849]
  39. Rossi J J. Ribozymes to the rescue: repairing genetically defective mRNAs. Trends Genet. (1998);14:295–298. [PubMed: 9724958]
  40. Smith A. Cell therapy: in search of pluripotency. Curr. Biol. (1998);8:R802–R804. [PubMed: 9811599]
  41. Vasquez K M, Wilson J H. Triplex-directed modification of genes and gene activity. Trends Biochem. Sci. (1998);23:4–9. [PubMed: 9478127]
  42. Vaughan T J, Osbourn J K, Tempest P R. Human antibodies by design. Nat. Biotechnol. (1998);16:535–539. [PubMed: 9624683]
  43. Verma I M, Somia N. Gene therapy - promises, problems and prospects. Nature. (1997);389:239–242. [PubMed: 9305836]
  44. Wadmann Germline gene therapy ‘must be spared excessive regulation’ Nature. (1998);392:317. [PubMed: 9537311]
  45. Wagner R W. Gene inhibition using antisense oligodeoxynucleotides. Nature. (1994);372:333–335. [PubMed: 7969490]
  46. Winter G, Harris W. Humanized antibodies. Immunol. Today. (1993);14:243–246. [PubMed: 8397764]
  47. Woolf T D. Therapeutic repair of mutated nucleic acid sequences. Nature Biotechnol. (1998);16:341–344. [PubMed: 9555723]
  48. Wivel N A, Walters L. Germ-line gene modification and disease prevention: some medical and ethical perspectives. Science. (1993);262:533–538. [PubMed: 8211180]
  49. Zimmerman B. Human germ-line therapy: the case for its development. J. Med. Philos. (1991);16:593–612. [PubMed: 1787389]
Copyright © 1999, Garland Science.
Bookshelf ID: NBK7569
PubReader format: click here to try


Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to pubmed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...