Annals of Medicine
ISSN: 0785-3890 (Print) 1365-2060 (Online) Journal homepage: http://www.tandfonline.com/loi/iann20
Approaches to Gene Therapy of Complex Multigenic Diseases: Cancer as a Model and Implications for Cardiovascular Disease and Diabetes Theodore Friedmann To cite this article: Theodore Friedmann (1992) Approaches to Gene Therapy of Complex Multigenic Diseases: Cancer as a Model and Implications for Cardiovascular Disease and Diabetes, Annals of Medicine, 24:5, 411-417, DOI: 10.3109/07853899209147847 To link to this article: http://dx.doi.org/10.3109/07853899209147847
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 8
View related articles
Citing articles: 1 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iann20 Download by: [Monash University Library]
Date: 10 April 2016, At: 05:03
Special Section: Molecular Genetics and Genetic Epidemiology of Cardiovascular Disease and Diabetes
Approaches to Gene Therapy of Complex Multigenic Diseases: Cancer as a Model and lmplicationsfor Cardiovascular Disease and Diabetes
Downloaded by [Monash University Library] at 05:03 10 April 2016
Theodore Friedmann
The general concept of gene therapy is now well established and accepted by the medical, scientific and public policy communities, and is rapidly being implemented in human experimental studies. In addition to the Initial models of single gene defects, target diseases have now come to include multigenic and multifactorial diseases such as human cancer, neurodegenerative diseases such as Parkinson’s disease and firms of cardiovascular disease. While many conceptual and technical obstacles must still be overcome before therapy for disorders such as coronary artery disease and dlabetes mellitus will easily be approached at the genetic level, the early results with several multigenic disease models gives some cause for optimism that gene therapies for even those complicated disorders will eventually become available. Key words: gene therapy; multigenic; cancer; neurodegenerative; tumour suppressor; immunotherapy. (Annals of Medicine 2 4 411 -417,1992)
Introduction: Complex Disease Most cardiovascular diseases and types 1 and I diabetes mellitus represent some of the least understood and most complex kinds of human multifactorial and multigenic diseases. There is little doubt that important genetic components to these common and burdensome diseases interact in complex ways with each other and with environmental factors. Most of these factors, both environmental and genetic, are poorly understood. Furthermore, because treatment is usually aimed at superficial intermediate phenotypic effects of the underlying genetic defects, it is phenomenological and at best only partly effective. During the past several decades, a conceptually new approach has been suggested for genetic disorders, that of gene therapy. This approach to therapy is based on the concept of specific correlation of underlying diseaseassociated mutations in human cells to complement genetic defects and correct disease phenotypes. The simplest disease models for gene therapy have been the genetically simple, single gene disorders in which the From the Center for Molecular Genetics, UCSD School of Medicine, California, U.S.A. Address and reprint requests: Theodore Friedmann, M.D., Center for Molecular Genetics, Department of Pediatrics, UCSD School of Medicine, La Jolla, CA 92093, U.S.A.
nature of the genetic defect and the biochemical and metabolic consequences are well understood. The absence of definitive information on the nature and role of the genetic factors in the pathogenesis of complex disorders such as types I and 11 diabetes mellitus and most forms of atherosclerosis would seem to suggest that specific gene therapy would be difficult or impossible to achieve. However, there are ways in which even these kinds of diseases might be approached at the genetic level, and some of these genetic techniques are now being applied to other more complex disease models, including cancer and disorders of the central nervous system.
Gene Therapy The concept of gene therapy arose in very general terms during the late 1960s (12), catalysed particularly by the discovery that some kinds of tumour viruses cause neoplastic changes in cells by integrating their foreign genetic information into the infected cell genome in a way that permits long-term and stable gene expression of some of the transduced viral genes. In 1972, before the advent of the recombinant DNA era, it became obvious that genetically engineering viruses could be made to act as efficient gene transfer vectors for the Ann
mnd 94
Downloaded by [Monash University Library] at 05:03 10 April 2016
412
Friedmann
introduction of potentially therapeutic genes into defective human cells (19). To make such tools applicable to the treatment of human disease, it has been necessary to develop vectors for the efficient, stable and safe transfer of genes into human and other mammalian cells, to identify accessible and relevant target cells involved in genetic disease, and to test the complementation of genetic defects and the correction of disease phenotypes in vitro and in vivo by expression of foreign genes. Many of those requirements have now been met so well that the concepts and tools of gene therapy have progressed from being highly improbable and technically out of'reach to being imminent and even in progress. The earliest disease models for which gene therapy was envisioned were genetically simple single gene defects such as sickle cell anaemia and the thalassaemias, the Lesch Nyhan syndrome, adenosine deaminase deficiency. Because the more complex disorders, multigenic and multifactorial diseases, such as most degenerative defects, atherosclerosis and other cardiovascular diseases and cancer were thought to be the result of many interacting genes often witti strong environmental influences, direct genetic correlation for these disorders has generally seemed much less feasible. However, recent advances in the understanding of the pathogenesis of cancer has given some cause for optimism that even multigenetic disorders may provide suitable targets for gene therapy in human patients. Relatively uncomplicated genetic modifications have abrogated or suppressed features of the cancer phenotype, and it Seems likely that other disorders that result from the expression of several cooperating genetic defects, almost certainly of the kind found in most forms of atherosclerosis and many other disorders of the cardiovascular system, will also become accessible to a genetic attack. The manipulations described illustrate conceptually simple genetic approaches to the treatment of complex multigenic and multifactorialdiseases.
Oncogenes and Cancer During the recent decade or so, several kinds of cancerrelated genes have been described. Transforming oncogenes are generally mutated forms of cellular 'proto-oncogenes' and represent gain of function changes that cause tumour development by de-regulating cell replication. In contrast, tumour development by the tumour suppressor genes occurs through loss of heterozygosity of genes that encode functions that affect cell cycle events and gene transcription. But it seems very likely that neoplasia in humans is most frequently the result of disruption of normal cell growth regulation resulting from an accumulation of multiple mutations in proto-oncogenes and tumour suppressor genes, some germ-line and inherited in Mendelian fashion, most somatically acquired. Each of the mutations contributes incrementally to the deregulation of cell growth, with each mutational step producing alterations in the phenotype that finally, with the last few pivotal mutations, cause the cell to become tumourigenic. Some of the mutations
are particularly crucial and produce especially damaging effects in the cells. Because of their ubiquity in human cancers and their role in regulating the cell cycle, it seems likely that tumour suppressor genes, in particular p53, play such central roles in tumourigenesis in human cells. It also seems clear, however, that the accumulated mutations are most likely to produce progressive tumours in cases where the immune system is unable to recognize them and mount an effective immune response. In most cases, an intact human immune system is thought to recognize differences on the cancer cells and to respond to their appearance or to the inflammatory effects of tumour development with an induction of mechanisms of cellular immunity such as those mediated by cytotoxic T cells (CTL), natural killer (NK) cells and tumour infiltrating lymphocytes (TIL). New approaches to cancer therapy will, therefore, in all probability take advantage of genetic as well as immunological manipulations to restore growth regulation to cancer cells and allow effective immune responses to remove them from the body. The best current approach to an understanding of human cancer requires that it be viewed as a genetic disease whose progression involves deficiencies in immunological defence mechanisms of the host. If that view is accurate, it is likely that the development of truly effective therapy will in most cases require the addition of multi-modal genetic and immune approaches to therapy to augment the more traditional pharmacological treatments. The previous concept that there are many forms of cancer with many etiologies, some metabolic, some environmental, some genetic and that the various forms of the disease will require many different, metabolically based forms of therapy must now be strengthened by an understanding of the potential for genetic and immune reconstruction approaches. We are now beginning to understand the mechanism of action of several classes of immune modulatory genes as well as transforming oncogenes and tumour suppressor genes whose expression or lack of expression defines the altered regulation of cell growth and replication characteristic of human and other turnours (4, 5,27,28,32,37, 38, 54). These aberrant genes must now become the target for therapy. With currently available concepts and reagents, a number of major approaches might be envisioned for gene therapy for cancer; enhancement of host immune response to cancer cells, restoration or enhancement of tumour suppressor gene function and interruption of the action of oncogenes.
Genetic Approaches to Modification of Immunological Functions The earliest approaches to the genetic treatment of cancer have taken advantage more of the immunological response to tumours than of an understanding of the molecular genetic mechanisms of tumourigenesis and the role of oncogenes and tumour suppressor genes. Initial therapeutic studies have centered around attempts
Downloaded by [Monash University Library] at 05:03 10 April 2016
Gene Therapy of Multigenic Diseases to enhance cellular immune mechanisms and to circumvent the apparently ineffective cytokine production arm of the immune response by providing surrogate cytokine-producing cells to the tumours. Several investigators have demonstrated that the introduction of one of several cytokine genes into tumour cells results in important changes in the tumour phenotype. In several animal tumour models, cells infected with retroviral vectors expressing IL-2, IL-4 or other cytokine genes lose their ability to produce tumours when re-introduced into recipient animals. But much more important in a therapeutic setting is the fact that, when such genetically modified tumour cells are transplanted subcutaneously into host animals, the animals become protected from tumour development by tumour cells subsequently grafted into the recipient animals (14, 15, 21). The animals are successfully vaccinated by the cytokineproducing tumour cells! The mechanisms for this protection seems to involve the induction of a host anti-tumour CTL response through the expression of cytokines in the context of the putative tumour-specific antigens on the tumour cells. The genetically modified tumours are thought to be acting as surrogates for the ineffective helper T cells in the host. The induced CTL response recognizes and destroys not only the vaccinating tumour cells but also tumour cells elsewhere in the body. Consistent with this interpretation is the fact that, in one study, existing tumours have been found to regress after vaccination with cytokine-expressing tumour cells (23). Another approach to genetically based cancer immunotherapy involves the use of TIL cells, those lymphocytes that accumulate in some solid tumours presumably to carry out an anti-tumour function (18, 39, 51). If such cells could be modified genetically to express potent anti-tumour functions, they might be returned to a tumour-bearing host in a way that could provide a locally high concentration of an anti-tumour produce directly to a tumour. In this way, factors such as tumour necrosis factor (TNF) that are exceedingly toxic when given systematically might be targeted specifically to tumours to provide effective therapy and obviate systemic toxicity. Such clinical studies with TNF-expressing TIL cells are already underway at the NIH (51). Still another immunological approach to cancer gene therapy is based on the presumption that the tumour cells can be made to appear ‘non-self’ to the host immune system through the introduction to tumour cells in vivo of class I MHC genes not ordinarily expressed by the tumour cells. Toward this end, permission has recently been given by U.S.federal regulatory agencies for human studies in which HLA B-27 genes will be introduced into melanoma in vivo by liposome-mediated gene transfer in an attempt to induce an immune attack on the tumour cells. The results of these kinds of immunotherapy studies, both published and unpublished, with a variety of immuno-modulatory genes are very promising, and it seems likely that immune-based genetic manipulations will become useful in therapy of some cancers, either by immunization or by physically targeting the immunomodulatory genes to tumours in vivo via TIL or other tumour-trophic cells.
413
Restoration of Tumour Suppressor Function In the case of cancers arising from cells made tumourigenic at least partly through the loss of heterozygosity for a tumour suppressor gene, it has been demonstrated that the tumour phenotype of such cells can be suppressed by restoration of a state of heterozygosity of the relevant tumour suppressor gene through the introduction of a single wild-type allele of the gene into the cancer cells. We have suggested elsewhere that this phenomenon might eventually serve as the basis for specific genetic therapy of some cancers (11). Of course, it seems obvious that cells harbouring multiple mutations and corrected only at the site of a single tumour suppressor will continue to be highly susceptible to development of a tumourigenic phenotype by subsequent mutations of the therapeutic transgene or in other critical growth regulatory genes. But that assumption must be tested. If found to be true, it may be possible to circumvent that problem by the addition of multiple copies of the therapeutic tumour suppressor gene to minimize the likelihood of knockout of a single allele of a ‘linchpin’ tumour suppressor function. Model studies have demonstrated that restoration of tumour suppressor function in a variety of human tumours can suppress cell growth and tumour induction. Tumour suppressor genes are normal cellular genes that encode functions vital for the orderly and regulated progression of cells through the cell cycle. In the case of loss of function of both alleles of these genes, controlled growth and replication of cells is disrupted and the deregulated growth properties of the cells results in tumour development. A number of such genes encoding tumour suppressor functions have been identified and characterized, including the retinoblastoma gene (Rb) (20, 29), several Wilrn’s tumour-associated genes (35), NF1 (2) and the p53 gene (30) and others. Mutations or losses of p53 have been found associated with a very high percentage of human tumours, suggesting very strongly that it plays a particularly important role in human tumourigenesis. The most compelling current models of human carcinogenesis indicate clearly that most human cancers results from the development of multiple cooperating and interacting mutations. For that reason, it may be unrealistic to expect that the genetic correction of only one of the many defects in a tumour could result in a therapeutic phenotypic change. Nevertheless, initial results from our laboratories (25) and more recent results from a number of investigators (1, 9, 11, 41, 47, 52) have shown that restoration of a state of functional heterozygosity for Rb or p53 in a number of human tumour cells frequently causes a suppression of tumourigenicity after implantation of the genetically modified cells into immuno-compromised nude mice. For instance, we have recently studied the effects of the restoration of p53 expression by retroviral gene transfer on human T-cell leukaemia cells deficient in the p53 tumour suppressor gene. After the restoration of a state of functional heterozygosity for p53 expression in those
Downloaded by [Monash University Library] at 05:03 10 April 2016
414
Friedmann
cells, they lose their ability to produce colonies in anchorage-independent conditions and to produce tumours in immunodeficient nude mice (11). Similar in vivo tumour suppression as now been found in a number of other p5bdeficient tumour cells, including osteosarcoma, adenocarcinoma of the breast, hepatocellular carcinoma and glioblastomas (Yee, Runnebaum and Friedman, unpublished results). Because of difficulties in achieving efficient and targeted gene delivery in vivo, most approaches to gene therapy to date have taken advantage of genetic modification of target cells in vitro followed by transplantation of the genetically altered cells into an appropriate organ or tissue in the recipient animal (17). Using this approach, we and others have shown that introduction of a single wild-type allele of a tumour suppressor gene into a tumour cell line deficient in that gene frequently results in the loss of the ability of that cell line to form tumours in nude mice (11, 25). While these studies represent proofs of principle for this genetic approach to cancer gene therapy, they do not present practical approaches to real clinical situations for the following reasons. It seems likely that the majority of the cells in an existing tumour will have to be infected with the transducing vector and genetically modified to have a significant effect on tumour growth. Of course, it is known that there is a threshold in tumourigenesis studies for the number of cells that must be implanted in order for a tumour to develop, a fact that suggests the possihility that the reduction in total number of tumourigenic cells to some number below that threshold, or a 'bystander' effect through which the presence of a certain number of genetically corrected cells can affect the tumourigenic potential of the unmodifiedtumour cells. Furthermore, in cells whose tumourigenic properties result from the interactions between many cooperating mutations, the genetic correlation of only one mutation is likely to leave the cells poised on the brink of tumourigenicity. A single subsequent mutation in the transgene can result in renewed unregulatory cell growth. Obviously, many major advances toward the solution of these problems, toward targeting and efficiency of gene transfer, and toward a greater understanding of the mechanisms of action of tumour suppressor gene biology provided by such model studies can be translated into therapy for human cancer.
Interruption of Oncogene Function In the case of cells made tumourigenic through the activation of proto-oncogenes to transforming oncogenes, it would be necessary to interfere with the expression of the responsible oncogene. A number of methods toward that end have already been suggested, including the introduction into tumour cells of anti-sense sequences (8, 40, 554, ribozymes (10, 42), transdominant negative mutations of oncogenes (7, 24), introduction of non-allelic genes such as rev (6)and others. Until now, these approaches to cancer gene therapy have remained much less well developed, conceptually
or technically, than the immunotherapy or tumour suppressor approaches. The major technical problems facing clinical applications for the concepts of gene therapy, whether it be for cancer or other diseases, include the efficient introduction of a therapeutic gene into target cells, the persistence of genetically modified cells in vivo and the maintenance of suitably high levels of transgene expression in the modified cells. Approaches to these problems require study of both the vector carrying the gene of interest as well as of the means to deliver that vector. This task will require systematic studies of vector delivery to target tissues in vivo as well as the characterization of a large number of different viral vectors carrying marker genes as well as genes relevant to cancer therapy. Although in vitro transduction of tumour cells may not be appropriate for the majority of applications requiring the introduction of tumour suppressor genes, this approach is suitable for some types of cancer gene therapy, such as those based on vaccination with cytokine-producing cancer cells or targeting of therapeutic molecules to cancers by means of TIL cells. In these cases, it may not be necessary to infect a large proportion of the cancer cells or even to achieve stable expression of the foreign gene. Rather, suitably transduced cells can be selected in vitro and high-level, short-term expression of immuno-modulatory genes may be sufficient to produce a significant anti-cancer effect. Thus, the different conceptual approaches to cancer gene therapy, i.e. modification of oncogene expression and immunomodulatory therapy, may require different methods of gene delivery.
Vectors and Efficient Gene Transfer The development of techniques for the introduction of potentially therapeutic genes in vivo is one of the major goals of cancer gene therapy. Aside from immunomodulatory techniques, it is only through some form of in vivo targeting of therapeutic tumour suppressor genes or genes encoding toxic products to the cells of most tumours, especially solid tumours, that it will be possible to infect a substantial proportion of tumour cells and achieve a significant anti-tumour response. Furthermore, if the introduction of immuno-modulatory genes directly at the site of a tumour can stimulate an effective antitumour immune response, this approach to immunotherapy will certainly be more direct, more universally useful, more applicable to many solid tumours whose cells cannot be grown easily in culture and, in principle, much less expensive than methods that rely on in vitro modificationof cultured tumour cells. An ideal vehicle for in vivo gene delivery would be a targetable and tissue-specific, high titre viral vector that would infect or express transgenes only in cancer cells. When produced in sufficiently high titres, such a vector could be delivered to the desired site through the general circulation. Unfortunately, techniques for the development of such tissue-specific or targetable
Downloaded by [Monash University Library] at 05:03 10 April 2016
Gene Therapy of Multigenic Diseases vectors for the development of such tissue-specific or targetable vectors have not yet been developed. Some studies from our laboratory and other groups have shown that retroviruses can be constructed in which the retroviral env gene product can be replaced by other proteins, such as the VSV G protein to produce a vector with broadened cell range (34),or by CD4 to produce a vector that, in principle, is trophic for HIV-infected cells (58). Alternatively, vectors might also be ’targeted’ physically to specific tissues by introduction directly into the mass of a tumour or specifically into its blood supply by interventional radiology, much as other chemotherapeutic agents are currently being delivered to some tumours. Approaches to targeting genes to specific cell types such as hepatocytes using receptor-mediated uptake, for instance, of asioalyglycoprotein-DNA conjugates have also been developed (56, 57). Unfortunately, tissue-specific gene delivery techniques do not exist for the majority of cells types. Nor are there any methods that specifically target tumour cells over their untransformed counterparts. Which vectors would be most useful for such direct gene delivery in vivo? A number of gene transfer methods and viral vectors have become available that might allow direct, in vivo genetic modification of tumour cells that could enhance their immunogenicity, suppress their growth or make them susceptible to the toxic effects of pharmacological agents. The greatest experience has accumulated with the retroviruses, the most common of which are those derived from murine retroviruses (43,46,48,49,53).However, the utility of these vectors for in vivo gene delivery is severely limited by the relatively low titres of approximately < lo6 infectious units/ml that can be achieved in most laboratories. A fundamental property of retroviral vectors is that they will infect only actively dividing cells, again rendering them unsuitable for efficient in vivo infection of many types of cancers in very highly proliferating tissues of tumours with low mitotic indices. On the other band, other vectors, such as herpesbased (13,16,22,36,44) and adenoviral vectors (3, 31, 45, 50)can be prepared to much higher titres, up to and even exceeding 1010 infectious units/ml. The availability of very large amounts of high titre virus is one of the requirements for testing the feasibility of direct in vivo delivery of gene transfer vectors for gene therapy purposes. Such vectors have come to be developed over the past several years, and while they are still less well understood than the retroviruses and still display undesirable features such as some persistant cytotoxicity, those characteristics may not be a serious impediment to their application in some forms of cancer gene therapy. Herpes vectors can be produced to very high titres, can infect a very broad range of cells in vivo and can express some transgenes at very high levels (33).Most current forms of herpes vectors also retain some degree of cytotoxicity (26),but this may not represent a serious disadvantage for the delivery of immuno-modulatory genes to tumours, provided that gene expression can persist long enough to induce a host immune response. Furthermore, adenoviral vectors can also be produced at
41 5
very high titres and are also capable of infecting most tissues, even relatively quiescent cells of the liver, after direct delivery in vlvo (33). An alternative to the direct use of viral vectors is the introduction into tumours of producer cells that are generating retroviral vectors in the midst of the target tumour. If the surrounding tissue consists largely of nonreplicating cells, as in the cases of the central nervous system or even an organ like the liver, only the replicating cells of a tumour would readily be susceptible to infection with locally produced virus. After allowing a sufficiently long time for the infection of many of the replicating tumour cells, they and the producer cells can then be killed by systemic administration of potentially toxic substrates of the herpes TK gene, such as ganciclovir.
Summary and Implications for Genetically Complex Diseases The concept of gene therapy for a number of human diseases has now come to be widely accepted by the medical, molecular genetic, public policy and ethics communities in the United States. Cancer constitutes one of the most important target diseases for this conceptually new kind of treatment, and many of the initial gene therapy studies currently being undertaken are directed at forms of cancer. With good reason. The medical urgencies are compelling and the techniques are rapidly becoming sufficient to the task. There is a good reason to be optimistic that the next several years will see the dawning of a new era in cancer therapy. Is that same optimixm warranted in the case of other complex disorders such as diabetes and cardiovascular disease? Certainly, for those rare diseases with single major genetic causes, such as in familial hypercholesterolaemia resulting from mutations in the LDL receptor gene, there is reason to expect effective gene therapy approaches. For disorders in which specific mutations have not been identified, genetic approaches toward modification of intermediate phenotypes might still be envisioned. For instance, genetic methods for increasing the levels of HDL, even in the absence of specific metabolic defects in HDL or its constituent lipoproteins, may protect against atherosclerosis. The genetic characterization of types 1 and II diabetes is currently limited to correlations with HLA genotypes and does not present any obvious approaches to gene therapy. The present understanding of the relationship of these genotypes with insulin deficiency, insulin resistance and the serum lipid disorders associated with diabetes mellitus does not lend itself easily to direct genetic forms of therapy. Nevertheless, one might imagine that constitutive expression of low levels of insulin from genetically modified autologous cells might reduce or stabilize insulin requirements. One might envision genetic methods of protecting pancreatic islet cells from toxic or apoptotic cell loss by providing trophic factors required by the islet cells. Obviously, these approaches are not yet with us, but it does seem justified to encourage imaginative approaches to therapy even of such
41 6
Friedmann
complex disorders at the genetic level. With the continued characterization of genotypic associations and the eventual identification of the biochemical consequences and their interactions to produce the 'Itimate disease phenotype, genetic approaches to the interruption of the pathoaenesis cascade will become possible.
References
Downloaded by [Monash University Library] at 05:03 10 April 2016
1. Baker SJ, Markowltz S, Fearon ER, Willson JK, VOgel-
steln B. Suppression of human colorectal carcinoma cell growth by wild-type p53. Science 1990; 249: 912-5. 2. Ballester R, Marchuk D. Boguskl M, Saullno A, Letcher R, Wigler M, Collins F. The NF1 locus encodes a protein functionally related to mammalian GAP and yeast IRA proteins. Cell 1990; 63: 851-9. 3. Berkner KL. Development of adenovirus vectors for the expression of heterologous genes. BioTechniques1988: 6: 616-29. 4. Bishop JM. Cancer genes come of age. Cell 1983; 32: 1018-20. 5. Bishop JM. Molecular themes in oncogenesis. Cell 1991; 64: 235-48. 6. BBranger F, Goud B, Tavltlan A, De Gunzburg J. Association of the Ras-antagonistic Rap1/Krev-1 proteins with the Golgi complex. Proc Natl Acad Sci 1991: 88: 1606-10. 7. Cal H, Szeberhyl J, Cooper GM. Effect of a dominant inhibitory Ha-ras mutation on mitogenic signal transduction in NIH 3T3 cells. Mol Cell Biol 1990; 10: 5314-23. .8. Calabretta, B. Inhibition of protooncogene expression by antisense oligodeoxynucleotldes: biological and therapeutic implications. Cancer Res 1991; 51:4505-1 0. 9. Casey G, Lo-Hsueh M, Lopez ME, Vogelrtein B, Stanbridge EJ. Growth suppression of human breast cancer cells by the introduction of a wild-type p53 gene. Oncogene 1991: 6:1791-7. 10. Cech TR. Ribozymes and their medical implications. JAMA 1988; 260: 327-68. 11. Cheng J, Yee J-K, Yeargln J, Friedman T, Haas M. Suppression of acute lymphoblastic leukemia by the human wild-type p53 gene. Cancer Res 1992; 52: 222-6. 12. Davis BD. Prospects for genetic intervention in Man. Science 1970; 170: 1279-83. 13. Dobson AT, Margolls TP, Sedaratl F, Stevens JG, Feldman LT. A latent, nonpathogenic HSV-1-derived vector stably expresses b-galactosidase in mouse neurons. Neuron 1990; 5: 353-60. 14. Esuml N, Hunt B, ltaya T, Frost P. Reduced tumorigenicity of murine tumor cells secreting gamma-interferon is due to non-specific host responses and is unrelated to class I major histocompatibility complex expression. Cancer Res 1991; 51: 1185-9. 15. Fearon ER, Pardoll DM, ltaya T, Golumbek P, et al. Interleukin-2 production by tumor cells bypasses T helper function in the generation of an antitumor response. Cell 1990; 60:397-403. 16. Fink DJ, Sternberg LR, Weber PC, Mata M, Golns WG, Glorloso JC. In vivo expression of b-galactosidase in hippocampal neurons by HSV-mediated gene transfer. Hum Gene TheraDy 1992: 3: 11-1 9. 17. Frledmann T. Piogress toward human gene therapy. Science 1989; 244: 1275-81. 18. Frledmann T. Genetically modified tumor-infiltrating lymphocytes for cancer therapy. Cancer Cells 1991; 3: 271-4. 19. Frledmann T, Roblln R. Gene therapy for human genetic disease? Science 1972; 175: 949-55. 20. Friend SH, Bernards R, Rogelj S, et at. A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma. Nature 1986: 323: 643-6.
Ann Med 24
21. Gansbacher B, Zier K, Daniels 8, Cronin K, BannerjI R, Gilboa E. lnterleukin 2 gene transfer into tumor cells abrogates tumorigenicity and induces protective immunity. J Exp Med 1990; 172: 1217-24. 22. Geller At, Freese A. Infection of cultured central nervous system neurons with a defective herpes simplex virus 1 vector results in stable expression of Escherichia coli b-galactosidase. Proc Natl Acad Sci USA 1990; 87: 1149-53. 23. Golumbeck PT, Lazenby AJ, Levltsky HI, et al. Treatment of established renal cancer by tumour cells engineered to Secrete interleukin-4. Science 1991; 254: 713-1 6. 24. Herskowltz 1. Functional inactivation of genes by dominant negative mutations. Nature 1987; 329: 219-22. 25. Huang H-JS, Yee J-K, Shew J-Y, et al. Suppression of the neoplastic phenotype by replacement of the RB gene in human cancer cells. Science 1988; 242: 1563-6. 26. Johnson PA, Yoshlda K, Gage FH, Frledrnann 1.Effects of gene transfer into cultured CNS neurons with a replication defective herpes simplex virus type 1 vector. Mol. Brain Res 1992; 12: 95- 102. 27. Kleln G. The approaching era of the tumor suppressor genes. Science 238: 1539-45. 28. Lee W-H, Booksteln R, Hong F, Young LJ, Shew JY, Lee EY-HP. Human retinoblastoma susceptibility gene: cloning, identification, and sequence. Science 1987; 235: 1394-9. 29. Lee WH, Shew JY, Hong FD, et al. The retinoblastomasusceptibility gene product, ppAB110, is a nuclear phos[phoprotein associatedwith DNA binding activity. Nature 1988. 30. Levlne AJ, Momand J, Flnlay CA. The p53 tumour suppressor gene. Nature 1991; 351: 453-6. 31. Levero M, Barban V, Manteca S, et al. Defective and nondefective adenovirus vectors for expression foreign genes in vitro and in vivo. Gene 1991; 101: 195-202. 32. Marshall CJ. Tumor suppressor genes. Cell 1991; 64: 313-26. 33. Mlyanohara A, Johnson PA, Elam RL, Daf Y, Wltztum JL, Verma IM, Frledmann T. Direct gene transfer to the liver with herpes simplex virus type 1 vectors: Transient production of physiologically relevant levels of circulating factor IX. New Biol1992 (in press). 34. Nobuhlko E, Frledmann T, Yee J-K. Pseudotype formation of murine leukemia virus with the G of vesicular stomatitis virus. J Virol1991; 65: 1202-7. 35. Pelletler J, Bruenlng W, Kashtan CE, et at. Germline mutations in the Wilms' tumor suppressor gene are associated with abnormal urogenital development in DenysDrash syndrome. Cell 1991; 67: 437-47. 36. Roemer K, Johnson PA, Frledmann T. Recombination between a herpes simplex virus type 1 vector deleted for immediate-early gene 3 and the infected cell genome. J Gen Virool 1992; 73 (in press). 37. Rosenberg SA. The development of new immunotherapies for the treatment of cancer using interleukin-2: a review. Ann Surg 1988; 208: 121-35. 38. Rosenberg SA. lmmunotherapy and gene therapy of cancer. Cancer Res 1991;51: 5074s-5079s. 39. Rosenberg SA, Aebersoid P, Cornetia K, et al. Gene transfer into humans: immunotherapy of patients with advanced melanoma. using tumor-infiltrating lymphocytes modified by retroviral gene transduction. N Engl J Med 1990; 323: 570-8. 40. RUden TV, Gllboa E. Inhibition of human T-cell leukemia virus type I replication in primary human T cells that express antisense RNA. J Virol1989; 63: 677-82. 41. SchOfer R, lyer J, lten E, Nlrkko AC. Partial reversion of the transformed phenotype in HRAS-transfected tumorigenic cells by transfer of a human gene. Proc Natl Acad Sci USA 1988; 85: 1590-4. 42. Shepard AA, Tolentino P, DeLuca NA. trans-dominant inhibition of herpes simplex virus transcriptional regulatory protein ICP4 by heterodimer formation. J Virol 1990; 64: 3916-26.
Downloaded by [Monash University Library] at 05:03 10 April 2016
Gene Therapy of Multigenic Diseases 43. Shimotohno K, Temin HM. Formation of infectious progeny virus after insertion of herpes simplex thymidine kinase gene into DNA of an avian retrovirus. Cell 1981; 26: 67-77. 44. Spaete RR, Frenkel N. The herpes simplex virus amplicon: analyses of cis-acting replication functions. Proc Natl Acad Sci USA 1985; 82: 694-8. 45. Stratford-Perricaudet L, Levrero M, Chasse J-F, Perricaudet M, Brland P. Evaluation of the transfer and expression in mice of an enzyme-encoding gene using a human adenovirus vetor. Hum Gene Therapy 1990; 1: 241-56. 46. Tabin CJ, Hoffmann JW, Soff SP, Weinberg RA. Adaptation of a retrovirus as a eucaryotic vector transmitting the herpes simplex thymidine kinase gene. Mol Cell Biol 1982; 2: 426-36. 47. Tanaka K, Oshimura M, Kikuchi R, Seki M, Hayashi T, Miyaki M. Suppression of tumorigenicity in human colon carcinoma cells by introduction of normal chromosomes 5 or 18. Nature 1991; 349: 340-2. 48. Temin HM. Retrovirus vectors for gene transfer: Efficient integration into and expression of exogenous DNA in vertebrate cell genomes. In: R. Kucherlapati (ed.), Gene transfer. Plenum Press, New York, 1986; pp. 149-87. 49. Temin HM. Safety considerations in somatic gene therapy of human disease with retroviral vectors. Hum Gene Therapy 1990; 1: 111-23. 50. Thummel C, Tjian R, Grodzicker T. Construction of adenovirus expression vectors by site-directed in-vivo
417
recombination. J Mol Appl Genet 1982; 1: 435-46. 51. Topalian SL, Solomon D, Rosenberg SA. Tumor-specific cytolysis by lymphocytes infiltrating human melanomas. J lmmunoll989; 142: 3714-25. 52. Uzvolgyl E, Classon M, Henriksson M, Huang HJ, Szekeiy L, Lee W-H, Sumegi J. Reintroduction of a normal retinoblastoma gene into retinoblastoma and osteosarcoma cells inhibits the replication-associated function of SV40 large T antigen. Cell Growth Diff 1991; 2: 297-303. 53. Wei C, Gibson M, Spear PG, Scolnick EM. Construction and isolation of a transmissible retrovirus containing the src gene from Harvey Murine Sarcoma Virus and the thymidine kinase gene from herpes simplex virus type 1. J Virol 1981; 39: 935-44. 54. Weinberg RA. Tumor suppressor genes. Science 1991; 254: 1138-46. 55. Whitesell L, Rosolen A, Neckers LM. Anti-sense supression of N-myc expression inhibits the transdifferentiation of neuroectodermal tumor cell lines. Prog Clin Biol Res 1991; 366: 45-54. 56. Wu GY, Wu CH. Gene delivery systems for gene therapy. Biotech 1991; 3: 87-95. 57. Wu GY, Wu CH. Receptor-mediated gene delivery and expression in vivo. J Biol Chem 1988; 263: 14621-4. 58. Young JAT, Bates P, Willert K, Varmus HE. Efficient incorporation of human CD4 protein into avian leukosis virus particles. Science 1990; 250: 1421-3.
A n n AAnd 34
ISSN: 0785-3890 (Print) 1365-2060 (Online) Journal homepage: http://www.tandfonline.com/loi/iann20
Approaches to Gene Therapy of Complex Multigenic Diseases: Cancer as a Model and Implications for Cardiovascular Disease and Diabetes Theodore Friedmann To cite this article: Theodore Friedmann (1992) Approaches to Gene Therapy of Complex Multigenic Diseases: Cancer as a Model and Implications for Cardiovascular Disease and Diabetes, Annals of Medicine, 24:5, 411-417, DOI: 10.3109/07853899209147847 To link to this article: http://dx.doi.org/10.3109/07853899209147847
Published online: 08 Jul 2009.
Submit your article to this journal
Article views: 8
View related articles
Citing articles: 1 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=iann20 Download by: [Monash University Library]
Date: 10 April 2016, At: 05:03
Special Section: Molecular Genetics and Genetic Epidemiology of Cardiovascular Disease and Diabetes
Approaches to Gene Therapy of Complex Multigenic Diseases: Cancer as a Model and lmplicationsfor Cardiovascular Disease and Diabetes
Downloaded by [Monash University Library] at 05:03 10 April 2016
Theodore Friedmann
The general concept of gene therapy is now well established and accepted by the medical, scientific and public policy communities, and is rapidly being implemented in human experimental studies. In addition to the Initial models of single gene defects, target diseases have now come to include multigenic and multifactorial diseases such as human cancer, neurodegenerative diseases such as Parkinson’s disease and firms of cardiovascular disease. While many conceptual and technical obstacles must still be overcome before therapy for disorders such as coronary artery disease and dlabetes mellitus will easily be approached at the genetic level, the early results with several multigenic disease models gives some cause for optimism that gene therapies for even those complicated disorders will eventually become available. Key words: gene therapy; multigenic; cancer; neurodegenerative; tumour suppressor; immunotherapy. (Annals of Medicine 2 4 411 -417,1992)
Introduction: Complex Disease Most cardiovascular diseases and types 1 and I diabetes mellitus represent some of the least understood and most complex kinds of human multifactorial and multigenic diseases. There is little doubt that important genetic components to these common and burdensome diseases interact in complex ways with each other and with environmental factors. Most of these factors, both environmental and genetic, are poorly understood. Furthermore, because treatment is usually aimed at superficial intermediate phenotypic effects of the underlying genetic defects, it is phenomenological and at best only partly effective. During the past several decades, a conceptually new approach has been suggested for genetic disorders, that of gene therapy. This approach to therapy is based on the concept of specific correlation of underlying diseaseassociated mutations in human cells to complement genetic defects and correct disease phenotypes. The simplest disease models for gene therapy have been the genetically simple, single gene disorders in which the From the Center for Molecular Genetics, UCSD School of Medicine, California, U.S.A. Address and reprint requests: Theodore Friedmann, M.D., Center for Molecular Genetics, Department of Pediatrics, UCSD School of Medicine, La Jolla, CA 92093, U.S.A.
nature of the genetic defect and the biochemical and metabolic consequences are well understood. The absence of definitive information on the nature and role of the genetic factors in the pathogenesis of complex disorders such as types I and 11 diabetes mellitus and most forms of atherosclerosis would seem to suggest that specific gene therapy would be difficult or impossible to achieve. However, there are ways in which even these kinds of diseases might be approached at the genetic level, and some of these genetic techniques are now being applied to other more complex disease models, including cancer and disorders of the central nervous system.
Gene Therapy The concept of gene therapy arose in very general terms during the late 1960s (12), catalysed particularly by the discovery that some kinds of tumour viruses cause neoplastic changes in cells by integrating their foreign genetic information into the infected cell genome in a way that permits long-term and stable gene expression of some of the transduced viral genes. In 1972, before the advent of the recombinant DNA era, it became obvious that genetically engineering viruses could be made to act as efficient gene transfer vectors for the Ann
mnd 94
Downloaded by [Monash University Library] at 05:03 10 April 2016
412
Friedmann
introduction of potentially therapeutic genes into defective human cells (19). To make such tools applicable to the treatment of human disease, it has been necessary to develop vectors for the efficient, stable and safe transfer of genes into human and other mammalian cells, to identify accessible and relevant target cells involved in genetic disease, and to test the complementation of genetic defects and the correction of disease phenotypes in vitro and in vivo by expression of foreign genes. Many of those requirements have now been met so well that the concepts and tools of gene therapy have progressed from being highly improbable and technically out of'reach to being imminent and even in progress. The earliest disease models for which gene therapy was envisioned were genetically simple single gene defects such as sickle cell anaemia and the thalassaemias, the Lesch Nyhan syndrome, adenosine deaminase deficiency. Because the more complex disorders, multigenic and multifactorial diseases, such as most degenerative defects, atherosclerosis and other cardiovascular diseases and cancer were thought to be the result of many interacting genes often witti strong environmental influences, direct genetic correlation for these disorders has generally seemed much less feasible. However, recent advances in the understanding of the pathogenesis of cancer has given some cause for optimism that even multigenetic disorders may provide suitable targets for gene therapy in human patients. Relatively uncomplicated genetic modifications have abrogated or suppressed features of the cancer phenotype, and it Seems likely that other disorders that result from the expression of several cooperating genetic defects, almost certainly of the kind found in most forms of atherosclerosis and many other disorders of the cardiovascular system, will also become accessible to a genetic attack. The manipulations described illustrate conceptually simple genetic approaches to the treatment of complex multigenic and multifactorialdiseases.
Oncogenes and Cancer During the recent decade or so, several kinds of cancerrelated genes have been described. Transforming oncogenes are generally mutated forms of cellular 'proto-oncogenes' and represent gain of function changes that cause tumour development by de-regulating cell replication. In contrast, tumour development by the tumour suppressor genes occurs through loss of heterozygosity of genes that encode functions that affect cell cycle events and gene transcription. But it seems very likely that neoplasia in humans is most frequently the result of disruption of normal cell growth regulation resulting from an accumulation of multiple mutations in proto-oncogenes and tumour suppressor genes, some germ-line and inherited in Mendelian fashion, most somatically acquired. Each of the mutations contributes incrementally to the deregulation of cell growth, with each mutational step producing alterations in the phenotype that finally, with the last few pivotal mutations, cause the cell to become tumourigenic. Some of the mutations
are particularly crucial and produce especially damaging effects in the cells. Because of their ubiquity in human cancers and their role in regulating the cell cycle, it seems likely that tumour suppressor genes, in particular p53, play such central roles in tumourigenesis in human cells. It also seems clear, however, that the accumulated mutations are most likely to produce progressive tumours in cases where the immune system is unable to recognize them and mount an effective immune response. In most cases, an intact human immune system is thought to recognize differences on the cancer cells and to respond to their appearance or to the inflammatory effects of tumour development with an induction of mechanisms of cellular immunity such as those mediated by cytotoxic T cells (CTL), natural killer (NK) cells and tumour infiltrating lymphocytes (TIL). New approaches to cancer therapy will, therefore, in all probability take advantage of genetic as well as immunological manipulations to restore growth regulation to cancer cells and allow effective immune responses to remove them from the body. The best current approach to an understanding of human cancer requires that it be viewed as a genetic disease whose progression involves deficiencies in immunological defence mechanisms of the host. If that view is accurate, it is likely that the development of truly effective therapy will in most cases require the addition of multi-modal genetic and immune approaches to therapy to augment the more traditional pharmacological treatments. The previous concept that there are many forms of cancer with many etiologies, some metabolic, some environmental, some genetic and that the various forms of the disease will require many different, metabolically based forms of therapy must now be strengthened by an understanding of the potential for genetic and immune reconstruction approaches. We are now beginning to understand the mechanism of action of several classes of immune modulatory genes as well as transforming oncogenes and tumour suppressor genes whose expression or lack of expression defines the altered regulation of cell growth and replication characteristic of human and other turnours (4, 5,27,28,32,37, 38, 54). These aberrant genes must now become the target for therapy. With currently available concepts and reagents, a number of major approaches might be envisioned for gene therapy for cancer; enhancement of host immune response to cancer cells, restoration or enhancement of tumour suppressor gene function and interruption of the action of oncogenes.
Genetic Approaches to Modification of Immunological Functions The earliest approaches to the genetic treatment of cancer have taken advantage more of the immunological response to tumours than of an understanding of the molecular genetic mechanisms of tumourigenesis and the role of oncogenes and tumour suppressor genes. Initial therapeutic studies have centered around attempts
Downloaded by [Monash University Library] at 05:03 10 April 2016
Gene Therapy of Multigenic Diseases to enhance cellular immune mechanisms and to circumvent the apparently ineffective cytokine production arm of the immune response by providing surrogate cytokine-producing cells to the tumours. Several investigators have demonstrated that the introduction of one of several cytokine genes into tumour cells results in important changes in the tumour phenotype. In several animal tumour models, cells infected with retroviral vectors expressing IL-2, IL-4 or other cytokine genes lose their ability to produce tumours when re-introduced into recipient animals. But much more important in a therapeutic setting is the fact that, when such genetically modified tumour cells are transplanted subcutaneously into host animals, the animals become protected from tumour development by tumour cells subsequently grafted into the recipient animals (14, 15, 21). The animals are successfully vaccinated by the cytokineproducing tumour cells! The mechanisms for this protection seems to involve the induction of a host anti-tumour CTL response through the expression of cytokines in the context of the putative tumour-specific antigens on the tumour cells. The genetically modified tumours are thought to be acting as surrogates for the ineffective helper T cells in the host. The induced CTL response recognizes and destroys not only the vaccinating tumour cells but also tumour cells elsewhere in the body. Consistent with this interpretation is the fact that, in one study, existing tumours have been found to regress after vaccination with cytokine-expressing tumour cells (23). Another approach to genetically based cancer immunotherapy involves the use of TIL cells, those lymphocytes that accumulate in some solid tumours presumably to carry out an anti-tumour function (18, 39, 51). If such cells could be modified genetically to express potent anti-tumour functions, they might be returned to a tumour-bearing host in a way that could provide a locally high concentration of an anti-tumour produce directly to a tumour. In this way, factors such as tumour necrosis factor (TNF) that are exceedingly toxic when given systematically might be targeted specifically to tumours to provide effective therapy and obviate systemic toxicity. Such clinical studies with TNF-expressing TIL cells are already underway at the NIH (51). Still another immunological approach to cancer gene therapy is based on the presumption that the tumour cells can be made to appear ‘non-self’ to the host immune system through the introduction to tumour cells in vivo of class I MHC genes not ordinarily expressed by the tumour cells. Toward this end, permission has recently been given by U.S.federal regulatory agencies for human studies in which HLA B-27 genes will be introduced into melanoma in vivo by liposome-mediated gene transfer in an attempt to induce an immune attack on the tumour cells. The results of these kinds of immunotherapy studies, both published and unpublished, with a variety of immuno-modulatory genes are very promising, and it seems likely that immune-based genetic manipulations will become useful in therapy of some cancers, either by immunization or by physically targeting the immunomodulatory genes to tumours in vivo via TIL or other tumour-trophic cells.
413
Restoration of Tumour Suppressor Function In the case of cancers arising from cells made tumourigenic at least partly through the loss of heterozygosity for a tumour suppressor gene, it has been demonstrated that the tumour phenotype of such cells can be suppressed by restoration of a state of heterozygosity of the relevant tumour suppressor gene through the introduction of a single wild-type allele of the gene into the cancer cells. We have suggested elsewhere that this phenomenon might eventually serve as the basis for specific genetic therapy of some cancers (11). Of course, it seems obvious that cells harbouring multiple mutations and corrected only at the site of a single tumour suppressor will continue to be highly susceptible to development of a tumourigenic phenotype by subsequent mutations of the therapeutic transgene or in other critical growth regulatory genes. But that assumption must be tested. If found to be true, it may be possible to circumvent that problem by the addition of multiple copies of the therapeutic tumour suppressor gene to minimize the likelihood of knockout of a single allele of a ‘linchpin’ tumour suppressor function. Model studies have demonstrated that restoration of tumour suppressor function in a variety of human tumours can suppress cell growth and tumour induction. Tumour suppressor genes are normal cellular genes that encode functions vital for the orderly and regulated progression of cells through the cell cycle. In the case of loss of function of both alleles of these genes, controlled growth and replication of cells is disrupted and the deregulated growth properties of the cells results in tumour development. A number of such genes encoding tumour suppressor functions have been identified and characterized, including the retinoblastoma gene (Rb) (20, 29), several Wilrn’s tumour-associated genes (35), NF1 (2) and the p53 gene (30) and others. Mutations or losses of p53 have been found associated with a very high percentage of human tumours, suggesting very strongly that it plays a particularly important role in human tumourigenesis. The most compelling current models of human carcinogenesis indicate clearly that most human cancers results from the development of multiple cooperating and interacting mutations. For that reason, it may be unrealistic to expect that the genetic correction of only one of the many defects in a tumour could result in a therapeutic phenotypic change. Nevertheless, initial results from our laboratories (25) and more recent results from a number of investigators (1, 9, 11, 41, 47, 52) have shown that restoration of a state of functional heterozygosity for Rb or p53 in a number of human tumour cells frequently causes a suppression of tumourigenicity after implantation of the genetically modified cells into immuno-compromised nude mice. For instance, we have recently studied the effects of the restoration of p53 expression by retroviral gene transfer on human T-cell leukaemia cells deficient in the p53 tumour suppressor gene. After the restoration of a state of functional heterozygosity for p53 expression in those
Downloaded by [Monash University Library] at 05:03 10 April 2016
414
Friedmann
cells, they lose their ability to produce colonies in anchorage-independent conditions and to produce tumours in immunodeficient nude mice (11). Similar in vivo tumour suppression as now been found in a number of other p5bdeficient tumour cells, including osteosarcoma, adenocarcinoma of the breast, hepatocellular carcinoma and glioblastomas (Yee, Runnebaum and Friedman, unpublished results). Because of difficulties in achieving efficient and targeted gene delivery in vivo, most approaches to gene therapy to date have taken advantage of genetic modification of target cells in vitro followed by transplantation of the genetically altered cells into an appropriate organ or tissue in the recipient animal (17). Using this approach, we and others have shown that introduction of a single wild-type allele of a tumour suppressor gene into a tumour cell line deficient in that gene frequently results in the loss of the ability of that cell line to form tumours in nude mice (11, 25). While these studies represent proofs of principle for this genetic approach to cancer gene therapy, they do not present practical approaches to real clinical situations for the following reasons. It seems likely that the majority of the cells in an existing tumour will have to be infected with the transducing vector and genetically modified to have a significant effect on tumour growth. Of course, it is known that there is a threshold in tumourigenesis studies for the number of cells that must be implanted in order for a tumour to develop, a fact that suggests the possihility that the reduction in total number of tumourigenic cells to some number below that threshold, or a 'bystander' effect through which the presence of a certain number of genetically corrected cells can affect the tumourigenic potential of the unmodifiedtumour cells. Furthermore, in cells whose tumourigenic properties result from the interactions between many cooperating mutations, the genetic correlation of only one mutation is likely to leave the cells poised on the brink of tumourigenicity. A single subsequent mutation in the transgene can result in renewed unregulatory cell growth. Obviously, many major advances toward the solution of these problems, toward targeting and efficiency of gene transfer, and toward a greater understanding of the mechanisms of action of tumour suppressor gene biology provided by such model studies can be translated into therapy for human cancer.
Interruption of Oncogene Function In the case of cells made tumourigenic through the activation of proto-oncogenes to transforming oncogenes, it would be necessary to interfere with the expression of the responsible oncogene. A number of methods toward that end have already been suggested, including the introduction into tumour cells of anti-sense sequences (8, 40, 554, ribozymes (10, 42), transdominant negative mutations of oncogenes (7, 24), introduction of non-allelic genes such as rev (6)and others. Until now, these approaches to cancer gene therapy have remained much less well developed, conceptually
or technically, than the immunotherapy or tumour suppressor approaches. The major technical problems facing clinical applications for the concepts of gene therapy, whether it be for cancer or other diseases, include the efficient introduction of a therapeutic gene into target cells, the persistence of genetically modified cells in vivo and the maintenance of suitably high levels of transgene expression in the modified cells. Approaches to these problems require study of both the vector carrying the gene of interest as well as of the means to deliver that vector. This task will require systematic studies of vector delivery to target tissues in vivo as well as the characterization of a large number of different viral vectors carrying marker genes as well as genes relevant to cancer therapy. Although in vitro transduction of tumour cells may not be appropriate for the majority of applications requiring the introduction of tumour suppressor genes, this approach is suitable for some types of cancer gene therapy, such as those based on vaccination with cytokine-producing cancer cells or targeting of therapeutic molecules to cancers by means of TIL cells. In these cases, it may not be necessary to infect a large proportion of the cancer cells or even to achieve stable expression of the foreign gene. Rather, suitably transduced cells can be selected in vitro and high-level, short-term expression of immuno-modulatory genes may be sufficient to produce a significant anti-cancer effect. Thus, the different conceptual approaches to cancer gene therapy, i.e. modification of oncogene expression and immunomodulatory therapy, may require different methods of gene delivery.
Vectors and Efficient Gene Transfer The development of techniques for the introduction of potentially therapeutic genes in vivo is one of the major goals of cancer gene therapy. Aside from immunomodulatory techniques, it is only through some form of in vivo targeting of therapeutic tumour suppressor genes or genes encoding toxic products to the cells of most tumours, especially solid tumours, that it will be possible to infect a substantial proportion of tumour cells and achieve a significant anti-tumour response. Furthermore, if the introduction of immuno-modulatory genes directly at the site of a tumour can stimulate an effective antitumour immune response, this approach to immunotherapy will certainly be more direct, more universally useful, more applicable to many solid tumours whose cells cannot be grown easily in culture and, in principle, much less expensive than methods that rely on in vitro modificationof cultured tumour cells. An ideal vehicle for in vivo gene delivery would be a targetable and tissue-specific, high titre viral vector that would infect or express transgenes only in cancer cells. When produced in sufficiently high titres, such a vector could be delivered to the desired site through the general circulation. Unfortunately, techniques for the development of such tissue-specific or targetable
Downloaded by [Monash University Library] at 05:03 10 April 2016
Gene Therapy of Multigenic Diseases vectors for the development of such tissue-specific or targetable vectors have not yet been developed. Some studies from our laboratory and other groups have shown that retroviruses can be constructed in which the retroviral env gene product can be replaced by other proteins, such as the VSV G protein to produce a vector with broadened cell range (34),or by CD4 to produce a vector that, in principle, is trophic for HIV-infected cells (58). Alternatively, vectors might also be ’targeted’ physically to specific tissues by introduction directly into the mass of a tumour or specifically into its blood supply by interventional radiology, much as other chemotherapeutic agents are currently being delivered to some tumours. Approaches to targeting genes to specific cell types such as hepatocytes using receptor-mediated uptake, for instance, of asioalyglycoprotein-DNA conjugates have also been developed (56, 57). Unfortunately, tissue-specific gene delivery techniques do not exist for the majority of cells types. Nor are there any methods that specifically target tumour cells over their untransformed counterparts. Which vectors would be most useful for such direct gene delivery in vivo? A number of gene transfer methods and viral vectors have become available that might allow direct, in vivo genetic modification of tumour cells that could enhance their immunogenicity, suppress their growth or make them susceptible to the toxic effects of pharmacological agents. The greatest experience has accumulated with the retroviruses, the most common of which are those derived from murine retroviruses (43,46,48,49,53).However, the utility of these vectors for in vivo gene delivery is severely limited by the relatively low titres of approximately < lo6 infectious units/ml that can be achieved in most laboratories. A fundamental property of retroviral vectors is that they will infect only actively dividing cells, again rendering them unsuitable for efficient in vivo infection of many types of cancers in very highly proliferating tissues of tumours with low mitotic indices. On the other band, other vectors, such as herpesbased (13,16,22,36,44) and adenoviral vectors (3, 31, 45, 50)can be prepared to much higher titres, up to and even exceeding 1010 infectious units/ml. The availability of very large amounts of high titre virus is one of the requirements for testing the feasibility of direct in vivo delivery of gene transfer vectors for gene therapy purposes. Such vectors have come to be developed over the past several years, and while they are still less well understood than the retroviruses and still display undesirable features such as some persistant cytotoxicity, those characteristics may not be a serious impediment to their application in some forms of cancer gene therapy. Herpes vectors can be produced to very high titres, can infect a very broad range of cells in vivo and can express some transgenes at very high levels (33).Most current forms of herpes vectors also retain some degree of cytotoxicity (26),but this may not represent a serious disadvantage for the delivery of immuno-modulatory genes to tumours, provided that gene expression can persist long enough to induce a host immune response. Furthermore, adenoviral vectors can also be produced at
41 5
very high titres and are also capable of infecting most tissues, even relatively quiescent cells of the liver, after direct delivery in vlvo (33). An alternative to the direct use of viral vectors is the introduction into tumours of producer cells that are generating retroviral vectors in the midst of the target tumour. If the surrounding tissue consists largely of nonreplicating cells, as in the cases of the central nervous system or even an organ like the liver, only the replicating cells of a tumour would readily be susceptible to infection with locally produced virus. After allowing a sufficiently long time for the infection of many of the replicating tumour cells, they and the producer cells can then be killed by systemic administration of potentially toxic substrates of the herpes TK gene, such as ganciclovir.
Summary and Implications for Genetically Complex Diseases The concept of gene therapy for a number of human diseases has now come to be widely accepted by the medical, molecular genetic, public policy and ethics communities in the United States. Cancer constitutes one of the most important target diseases for this conceptually new kind of treatment, and many of the initial gene therapy studies currently being undertaken are directed at forms of cancer. With good reason. The medical urgencies are compelling and the techniques are rapidly becoming sufficient to the task. There is a good reason to be optimistic that the next several years will see the dawning of a new era in cancer therapy. Is that same optimixm warranted in the case of other complex disorders such as diabetes and cardiovascular disease? Certainly, for those rare diseases with single major genetic causes, such as in familial hypercholesterolaemia resulting from mutations in the LDL receptor gene, there is reason to expect effective gene therapy approaches. For disorders in which specific mutations have not been identified, genetic approaches toward modification of intermediate phenotypes might still be envisioned. For instance, genetic methods for increasing the levels of HDL, even in the absence of specific metabolic defects in HDL or its constituent lipoproteins, may protect against atherosclerosis. The genetic characterization of types 1 and II diabetes is currently limited to correlations with HLA genotypes and does not present any obvious approaches to gene therapy. The present understanding of the relationship of these genotypes with insulin deficiency, insulin resistance and the serum lipid disorders associated with diabetes mellitus does not lend itself easily to direct genetic forms of therapy. Nevertheless, one might imagine that constitutive expression of low levels of insulin from genetically modified autologous cells might reduce or stabilize insulin requirements. One might envision genetic methods of protecting pancreatic islet cells from toxic or apoptotic cell loss by providing trophic factors required by the islet cells. Obviously, these approaches are not yet with us, but it does seem justified to encourage imaginative approaches to therapy even of such
41 6
Friedmann
complex disorders at the genetic level. With the continued characterization of genotypic associations and the eventual identification of the biochemical consequences and their interactions to produce the 'Itimate disease phenotype, genetic approaches to the interruption of the pathoaenesis cascade will become possible.
References
Downloaded by [Monash University Library] at 05:03 10 April 2016
1. Baker SJ, Markowltz S, Fearon ER, Willson JK, VOgel-
steln B. Suppression of human colorectal carcinoma cell growth by wild-type p53. Science 1990; 249: 912-5. 2. Ballester R, Marchuk D. Boguskl M, Saullno A, Letcher R, Wigler M, Collins F. The NF1 locus encodes a protein functionally related to mammalian GAP and yeast IRA proteins. Cell 1990; 63: 851-9. 3. Berkner KL. Development of adenovirus vectors for the expression of heterologous genes. BioTechniques1988: 6: 616-29. 4. Bishop JM. Cancer genes come of age. Cell 1983; 32: 1018-20. 5. Bishop JM. Molecular themes in oncogenesis. Cell 1991; 64: 235-48. 6. BBranger F, Goud B, Tavltlan A, De Gunzburg J. Association of the Ras-antagonistic Rap1/Krev-1 proteins with the Golgi complex. Proc Natl Acad Sci 1991: 88: 1606-10. 7. Cal H, Szeberhyl J, Cooper GM. Effect of a dominant inhibitory Ha-ras mutation on mitogenic signal transduction in NIH 3T3 cells. Mol Cell Biol 1990; 10: 5314-23. .8. Calabretta, B. Inhibition of protooncogene expression by antisense oligodeoxynucleotldes: biological and therapeutic implications. Cancer Res 1991; 51:4505-1 0. 9. Casey G, Lo-Hsueh M, Lopez ME, Vogelrtein B, Stanbridge EJ. Growth suppression of human breast cancer cells by the introduction of a wild-type p53 gene. Oncogene 1991: 6:1791-7. 10. Cech TR. Ribozymes and their medical implications. JAMA 1988; 260: 327-68. 11. Cheng J, Yee J-K, Yeargln J, Friedman T, Haas M. Suppression of acute lymphoblastic leukemia by the human wild-type p53 gene. Cancer Res 1992; 52: 222-6. 12. Davis BD. Prospects for genetic intervention in Man. Science 1970; 170: 1279-83. 13. Dobson AT, Margolls TP, Sedaratl F, Stevens JG, Feldman LT. A latent, nonpathogenic HSV-1-derived vector stably expresses b-galactosidase in mouse neurons. Neuron 1990; 5: 353-60. 14. Esuml N, Hunt B, ltaya T, Frost P. Reduced tumorigenicity of murine tumor cells secreting gamma-interferon is due to non-specific host responses and is unrelated to class I major histocompatibility complex expression. Cancer Res 1991; 51: 1185-9. 15. Fearon ER, Pardoll DM, ltaya T, Golumbek P, et al. Interleukin-2 production by tumor cells bypasses T helper function in the generation of an antitumor response. Cell 1990; 60:397-403. 16. Fink DJ, Sternberg LR, Weber PC, Mata M, Golns WG, Glorloso JC. In vivo expression of b-galactosidase in hippocampal neurons by HSV-mediated gene transfer. Hum Gene TheraDy 1992: 3: 11-1 9. 17. Frledmann T. Piogress toward human gene therapy. Science 1989; 244: 1275-81. 18. Frledmann T. Genetically modified tumor-infiltrating lymphocytes for cancer therapy. Cancer Cells 1991; 3: 271-4. 19. Frledmann T, Roblln R. Gene therapy for human genetic disease? Science 1972; 175: 949-55. 20. Friend SH, Bernards R, Rogelj S, et at. A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma. Nature 1986: 323: 643-6.
Ann Med 24
21. Gansbacher B, Zier K, Daniels 8, Cronin K, BannerjI R, Gilboa E. lnterleukin 2 gene transfer into tumor cells abrogates tumorigenicity and induces protective immunity. J Exp Med 1990; 172: 1217-24. 22. Geller At, Freese A. Infection of cultured central nervous system neurons with a defective herpes simplex virus 1 vector results in stable expression of Escherichia coli b-galactosidase. Proc Natl Acad Sci USA 1990; 87: 1149-53. 23. Golumbeck PT, Lazenby AJ, Levltsky HI, et al. Treatment of established renal cancer by tumour cells engineered to Secrete interleukin-4. Science 1991; 254: 713-1 6. 24. Herskowltz 1. Functional inactivation of genes by dominant negative mutations. Nature 1987; 329: 219-22. 25. Huang H-JS, Yee J-K, Shew J-Y, et al. Suppression of the neoplastic phenotype by replacement of the RB gene in human cancer cells. Science 1988; 242: 1563-6. 26. Johnson PA, Yoshlda K, Gage FH, Frledrnann 1.Effects of gene transfer into cultured CNS neurons with a replication defective herpes simplex virus type 1 vector. Mol. Brain Res 1992; 12: 95- 102. 27. Kleln G. The approaching era of the tumor suppressor genes. Science 238: 1539-45. 28. Lee W-H, Booksteln R, Hong F, Young LJ, Shew JY, Lee EY-HP. Human retinoblastoma susceptibility gene: cloning, identification, and sequence. Science 1987; 235: 1394-9. 29. Lee WH, Shew JY, Hong FD, et al. The retinoblastomasusceptibility gene product, ppAB110, is a nuclear phos[phoprotein associatedwith DNA binding activity. Nature 1988. 30. Levlne AJ, Momand J, Flnlay CA. The p53 tumour suppressor gene. Nature 1991; 351: 453-6. 31. Levero M, Barban V, Manteca S, et al. Defective and nondefective adenovirus vectors for expression foreign genes in vitro and in vivo. Gene 1991; 101: 195-202. 32. Marshall CJ. Tumor suppressor genes. Cell 1991; 64: 313-26. 33. Mlyanohara A, Johnson PA, Elam RL, Daf Y, Wltztum JL, Verma IM, Frledmann T. Direct gene transfer to the liver with herpes simplex virus type 1 vectors: Transient production of physiologically relevant levels of circulating factor IX. New Biol1992 (in press). 34. Nobuhlko E, Frledmann T, Yee J-K. Pseudotype formation of murine leukemia virus with the G of vesicular stomatitis virus. J Virol1991; 65: 1202-7. 35. Pelletler J, Bruenlng W, Kashtan CE, et at. Germline mutations in the Wilms' tumor suppressor gene are associated with abnormal urogenital development in DenysDrash syndrome. Cell 1991; 67: 437-47. 36. Roemer K, Johnson PA, Frledmann T. Recombination between a herpes simplex virus type 1 vector deleted for immediate-early gene 3 and the infected cell genome. J Gen Virool 1992; 73 (in press). 37. Rosenberg SA. The development of new immunotherapies for the treatment of cancer using interleukin-2: a review. Ann Surg 1988; 208: 121-35. 38. Rosenberg SA. lmmunotherapy and gene therapy of cancer. Cancer Res 1991;51: 5074s-5079s. 39. Rosenberg SA, Aebersoid P, Cornetia K, et al. Gene transfer into humans: immunotherapy of patients with advanced melanoma. using tumor-infiltrating lymphocytes modified by retroviral gene transduction. N Engl J Med 1990; 323: 570-8. 40. RUden TV, Gllboa E. Inhibition of human T-cell leukemia virus type I replication in primary human T cells that express antisense RNA. J Virol1989; 63: 677-82. 41. SchOfer R, lyer J, lten E, Nlrkko AC. Partial reversion of the transformed phenotype in HRAS-transfected tumorigenic cells by transfer of a human gene. Proc Natl Acad Sci USA 1988; 85: 1590-4. 42. Shepard AA, Tolentino P, DeLuca NA. trans-dominant inhibition of herpes simplex virus transcriptional regulatory protein ICP4 by heterodimer formation. J Virol 1990; 64: 3916-26.
Downloaded by [Monash University Library] at 05:03 10 April 2016
Gene Therapy of Multigenic Diseases 43. Shimotohno K, Temin HM. Formation of infectious progeny virus after insertion of herpes simplex thymidine kinase gene into DNA of an avian retrovirus. Cell 1981; 26: 67-77. 44. Spaete RR, Frenkel N. The herpes simplex virus amplicon: analyses of cis-acting replication functions. Proc Natl Acad Sci USA 1985; 82: 694-8. 45. Stratford-Perricaudet L, Levrero M, Chasse J-F, Perricaudet M, Brland P. Evaluation of the transfer and expression in mice of an enzyme-encoding gene using a human adenovirus vetor. Hum Gene Therapy 1990; 1: 241-56. 46. Tabin CJ, Hoffmann JW, Soff SP, Weinberg RA. Adaptation of a retrovirus as a eucaryotic vector transmitting the herpes simplex thymidine kinase gene. Mol Cell Biol 1982; 2: 426-36. 47. Tanaka K, Oshimura M, Kikuchi R, Seki M, Hayashi T, Miyaki M. Suppression of tumorigenicity in human colon carcinoma cells by introduction of normal chromosomes 5 or 18. Nature 1991; 349: 340-2. 48. Temin HM. Retrovirus vectors for gene transfer: Efficient integration into and expression of exogenous DNA in vertebrate cell genomes. In: R. Kucherlapati (ed.), Gene transfer. Plenum Press, New York, 1986; pp. 149-87. 49. Temin HM. Safety considerations in somatic gene therapy of human disease with retroviral vectors. Hum Gene Therapy 1990; 1: 111-23. 50. Thummel C, Tjian R, Grodzicker T. Construction of adenovirus expression vectors by site-directed in-vivo
417
recombination. J Mol Appl Genet 1982; 1: 435-46. 51. Topalian SL, Solomon D, Rosenberg SA. Tumor-specific cytolysis by lymphocytes infiltrating human melanomas. J lmmunoll989; 142: 3714-25. 52. Uzvolgyl E, Classon M, Henriksson M, Huang HJ, Szekeiy L, Lee W-H, Sumegi J. Reintroduction of a normal retinoblastoma gene into retinoblastoma and osteosarcoma cells inhibits the replication-associated function of SV40 large T antigen. Cell Growth Diff 1991; 2: 297-303. 53. Wei C, Gibson M, Spear PG, Scolnick EM. Construction and isolation of a transmissible retrovirus containing the src gene from Harvey Murine Sarcoma Virus and the thymidine kinase gene from herpes simplex virus type 1. J Virol 1981; 39: 935-44. 54. Weinberg RA. Tumor suppressor genes. Science 1991; 254: 1138-46. 55. Whitesell L, Rosolen A, Neckers LM. Anti-sense supression of N-myc expression inhibits the transdifferentiation of neuroectodermal tumor cell lines. Prog Clin Biol Res 1991; 366: 45-54. 56. Wu GY, Wu CH. Gene delivery systems for gene therapy. Biotech 1991; 3: 87-95. 57. Wu GY, Wu CH. Receptor-mediated gene delivery and expression in vivo. J Biol Chem 1988; 263: 14621-4. 58. Young JAT, Bates P, Willert K, Varmus HE. Efficient incorporation of human CD4 protein into avian leukosis virus particles. Science 1990; 250: 1421-3.
A n n AAnd 34
