Why commercialization of gene therapy stalled; examining the life cycles of gene therapy technologies.

This report examines the commercialization of gene therapy in the context of innovation theories that posit a relationship between the maturation of a technology through its life cycle and prospects for successful product development. We show that the field of gene therapy has matured steadily since the 1980s, with the congruent accumulation of >35 000 papers, >16 000 US patents, >1800 clinical trials and >$4.3 billion in capital investment in gene therapy companies. Gene therapy technologies comprise a series of dissimilar approaches for gene delivery, each of which has introduced a distinct product architecture. Using bibliometric methods, we quantify the maturation of each technology through a characteristic life cycle S-curve, from a Nascent stage, through a Growing stage of exponential advance, toward an Established stage and projected limit. Capital investment in gene therapy is shown to have occurred predominantly in Nascent stage technologies and to be negatively correlated with maturity. Gene therapy technologies are now achieving the level of maturity that innovation research and biotechnology experience suggest may be requisite for efficient product development. Asynchrony between the maturation of gene therapy technologies and capital investment in development-focused business models may have stalled the commercialization of gene therapy.

Gene transfer and expression of human phenylalanine hydroxylase.

Phenylketonuria (PKU) is caused by a genetic deficiency of the enzyme phenylalanine hydroxylase (PAH). A full-length complementary DNA clone of human PAH was inserted into a eukaryotic expression vector and transferred into mouse NIH3T3 cells which do not normally express PAH. The transformed mouse cells expressed PAH messenger RNA, immunoreactive protein, and enzymatic activity that are characteristic of the normal human liver products, demonstrating that a single gene contains all of the necessary genetic information to code for functional PAH. These results support the use of the human PAH probe in prenatal diagnosis and detection of carriers, to provide new opportunities for the biochemical characterization of normal and mutant enzymes, and in the investigation of alternative genetic therapies for PKU.

Differential diagnosis of mut and cbl methylmalonic aciduria by DNA-mediated gene transfer in primary fibroblasts.

Methylmalonic aciduria can be caused by mutations in the gene encoding the methylmalonyl coenzyme A mutase apoenzyme (mut) or genes required for the provision of cofactor B12 (cbl). The mut and cbl forms are classically differentiated by somatic cell complementation. We describe a novel method for differential diagnosis of mut and cbl methylmalonic aciduria using DNA-mediated gene transfer of a methylmalonyl CoA mutase cDNA clone. Gene transfer of a functional methylmalonyl CoA mutase cDNA clone into mut fibroblasts reconstitutes holoenzyme activity measured by metabolism of [14C]-propionate in culture. Identical gene transfers into cbl fibroblasts have no effect. This method is used for the differential diagnosis of mut and cbl genotypes in cells from patients with a clinical diagnosis of methylmalonic aciduria and is shown to be a facile, sensitive, and specific method for genetic diagnosis. This work establishes the principle of using DNA-mediated gene transfer to identify the genotype of diseases which can result from mutations at several different genetic loci. This type of differential genotypic diagnosis will be particularly important for establishing the applicability of somatic gene therapy in individual patients.

Cloning and expression of a mutant methylmalonyl coenzyme A mutase with altered cobalamin affinity that causes mut- methylmalonic aciduria.

Distinct genotypic and phenotypic forms of methylmalonyl CoA mutase (MCM) apoenzyme deficiency can be delineated by biochemical analysis of mutant fibroblasts. One form, designated mut-, expresses a phenotype in which residual enzyme activity is evident in cultured cells exposed to high concentrations of hydroxycobalamin. We describe cloning of an MCM cDNA from cells exhibiting a mut- phenotype and characterization of the mutant gene product overexpressed in primary muto human fibroblasts and Saccharomyces cerevisiae. Three novel base changes were observed. Recombinant clones containing one of these base changes (G717V) express four characteristics of the mut- phenotype: failure to constitute [14C]propionate incorporation activity in fibroblasts assayed under basal cell culture conditions, constitution of [14C]propionate incorporation activity in fibroblasts stimulated with 0.1-1.0 micrograms/ml hydroxycobalamin, interallelic complementation with alleles bearing an R93H mutation, and an apparent Km (adenosylcobalamin) 1,000-fold higher than normal. These results demonstrate that the G717V mutation produces the mut- phenotype and localizes determinants for adenosylcobalamin binding near the carboxyl terminus of MCM.

Genetic characterization of a MUT locus mutation discriminating heterogeneity in mut0 and mut- methylmalonic aciduria by interallelic complementation.

Genetic complementation of fibroblasts from patients with methylmalonic aciduria (MMA) defines a unique class of allelic mutations arising from mutations at the locus encoding the methylmalonyl coenzyme A (CoA) mutase apoenzyme. Various phenotypes of MMA have been delineated including complete absence of enzyme activity (mut0) and abnormal enzyme activity with an elevated Km for adenosylcobalamin (mut-). We describe genetic studies on a cell line (WG1130) from a patient with mut0 MMA which exhibited an unusual complementation phenotype, complementing with three of nine mut0 cell lines and four of five mut- cell lines. This suggests that interallelic complementation occurs between mutant alleles in WG1130 and subsets of alleles associated with both mut0 and mut- phenotypes. The methylmalonyl CoA mutase cDNA was cloned from WG1130 and found to contain a G354----A (Arg93----His) mutation. Gene transfer of this mutant clone into primary fibroblasts from patients with MMA confirms that this mutation expresses a mut0 phenotype when transferred into a mut0 cell line with low levels of mRNA but can contribute to apoenzyme function when transferred into mut cell lines which show correction with WG1130 by somatic cell complementation. These results point to further heterogeneity within both mut0 and mut- and may enable identification of mutations affecting discrete components of apoenzyme function.

Cloning and expression of mutations demonstrating intragenic complementation in mut0 methylmalonic aciduria.

The mut0 mutation resulting in methylmalonyl CoA mutase (MCM) apoenzyme deficiency and methylmalonic aciduria is characterized by undetectable enzyme activity in cell extracts and low incorporation of propionate into cultured cells which is not stimulated by hydroxycobalamin. A mut0 fibroblast cell line (WG1681) from an African-American male infant complemented another mut0 cell line (WG 1130). Cloning and sequencing of cDNA from WG 1681 demonstrated compound heterozygosity for two novel changes at highly conserved sites: G623R and G703R. In addition, two previously described homozygous polymorphisms, H532R and V671I, were found. Hybridization of allele-specific oligonucleotides to PCR amplified MCM exons from the proband and family members identified a clinically normal mother, half-sister, and half-brother as carriers of the G703R change in cis with both polymorphisms. Transfection of each change into a mut0 cell line with very low MCM mRNA (GM1673) demonstrated a lack of stimulation of propionate uptake in the absence and presence of hydroxycobalamin. Cotransfection of each mutation with the previously identified R93H mutation of WG 1130 stimulated propionate uptake, indicating that G623R and G703R are independently capable of complementing the R93H mutation.

Isodisomy of chromosome 6 in a newborn with methylmalonic acidemia and agenesis of pancreatic beta cells causing diabetes mellitus.

Isodisomy (ID) is a genetic anomaly defined as the inheritance of two copies of the same genetic material from one parent. ID in an offspring is a rare cause of recessive genetic diseases via inheritance of two copies of a mutated gene from one carrier parent. We studied a newborn female with a mut(o) of methylmalonic acidemia and complete absence of insulin-producing beta cells in otherwise normal-appearing pancreatic islets, causing insulin-dependent diabetes mellitus. The patient died 2 wk after birth. Serotyping of the HLA antigens, DNA typing of HLA-B and HLA class II loci, study of polymorphic DNA markers of chromosome 6, and cytogenetic analysis demonstrated paternal ID, involving at least a 25-centiMorgan portion of the chromosome pair that encompasses the MHC. ID probably caused methylmalonic acidemia by duplication of a mutated allele of the corresponding gene on the chromosome 6 inherited from the father. It is also very likely that ID was etiologically related to the agenesis of beta cells and consequent insulin-dependent diabetes mellitus in our patient. We thus speculate on the existence of a gene on chromosome 6 involved in beta cell differentiation.

Identification of a point mutation in the topoisomerase II gene from a human leukemia cell line containing an amsacrine-resistant form of topoisomerase II.

HL-60/AMSA is a human leukemia cell line that is 50- to 100-fold more resistant to the cytotoxic actions of the topoisomerase II-reactive intercalator amsacrine than is its drug-sensitive HL-60 parent line. Previously, we have shown that the topoisomerase II from HL-60/AMSA is also resistant to inhibition by amsacrine and other intercalating agents. We therefore sought the molecular basis for the resistance of the topoisomerase II of HL-60/AMSA and, by inference, of the HL-60/AMSA line itself. We report the cloning and sequencing of the topoisomerase II genes from both the sensitive and resistant leukemia cell lines using polymerase chain reaction technology. We have identified a single base change associated with the drug-resistant form of topoisomerase II. This mutation is present in both cloned HL-60/AMSA complementary DNA and extracted HL-60/AMSA genomic DNA. A rapid assay for this mutation in clinical samples has been developed and applied to the DNA of cells from both normal volunteers and leukemia patients. Thus far, the HL-60/AMSA genotype has not been identified in the cells from any individual, suggesting that this genotype is indeed a mutation and not an allelic form of topoisomerase II. The novel assay developed will allow a rapid search for the prevalence of this mutation in clinical samples from patients with leukemia who have relapsed following intercalator therapy.

Non-viral gene therapy.

Non-viral gene therapies are currently under development that employ drug-delivery methods for targeting genes to selected cells in the body, where they express therapeutic gene products. Various methods have been described for non-viral gene therapy, ranging from the direct intramuscular injection of purified DNA to the systemic administration of formulations comprising DNA and lipids, proteins, peptides, or polymers. Products for non-viral gene therapies are designed both for direct administration to patients by conventional routes and for expression of a therapeutic product over a finite period of time in a manner similar to conventional medicines. Initial preclinical and clinical studies indicate that non-viral gene delivery methods exhibit safety profiles similar to conventional pharmaceutical or biological products. Clinical trials have been proposed, or are currently under way, to assess the applicability of non-viral gene therapy for a variety of disorders, including cystic fibrosis, cancer, and peripheral vascular disease. Non-viral techniques may soon allow gene therapy to be applied in clinical practice alongside conventional medicines for the treatment of common diseases.

Clinical considerations in the design of protocols for somatic gene therapy.

Despite two decades of investigating the potential for somatic gene therapy in curing human disease, few clinical trials are being proposed. This delay is due in part to limitations of existing methods for gene transfer and to the recognized need to proceed judiciously into this controversial arena. Delay is also caused by a disjunction between the traditional precedents and principles of clinical investigation and the procedures instituted to regulate somatic gene therapy. The premise of this report is twofold: first, that more extensive clinical investigation of gene transfer technologies would be beneficial to patients, medicine, and basic science and second, that clinical investigations could be expedited by appealing to the established experience in clinical investigation without compromising the scientific excellence and discipline essential for this highly public process. This report develops a clinical perspective on potential applications of existing gene transfer technologies and the issues involved in developing experimental protocols. The initiation of clinical trials should be a primary goal of gene therapy research programs.

Nonviral gene therapy: the promise of genes as pharmaceutical products.

Although most research on gene therapy has focused on the use of recombinant viruses to deliver genes to cells in vivo, progress also has been made toward developing nonviral, pharmaceutical formulations of genes for in vivo human therapy. Various methods for nonviral gene therapy have been proposed. Some approaches are aimed at developing "artificial viruses" that attempt to mimic the process of viral infection using synthetic materials. Others apply the theory and methods of advanced, particulate drug delivery to deliver DNA to select somatic targets. These approaches employ DNA complexes containing lipid, protein, peptide, or polymeric carriers as well as ligands capable of targeting the DNA complex to cell-surface receptors on the target cell and ligands for directing the intracellular trafficking of DNA to the nucleus. Nonviral systems have been used to deliver genes to the lung, liver, endothelium, epithelium, and tumor cells and have been shown to be generally safe. More than a dozen clinical trials are currently underway using nonviral systems for disease indications including cystic fibrosis and cancer. Future advances in nonviral systems will be based on an emerging appreciation of the biological constraints on the fate and function of DNA within the body and within the cell.

Multicompartment, numerical model of cellular events in the pharmacokinetics of gene therapies.

DNA expression vectors may be administered to patients like conventional medicines to have a finite and controlled duration of action. The clinical application of these medicines will require a precise understanding of the kinetics of the administered gene, the mRNA transcript, and the gene product. The apparent kinetic properties of the therapeutic gene product, including the level and duration of action, will be determined by various intrinsic kinetic processes including: (i) distribution and biological fate of the DNA expression vector; (ii) rates of DNA uptake into cells and dynamics of intracellular trafficking; (iii) half-life of the DNA vector in the cell; (iv) transcription rate; (v) half-life of mRNA; (vi) translation rate; and (vii) post-translational processing, distribution, and fate of the gene product. To consider in a theoretical manner how the intrinsic kinetics of cellular processes may affect the apparent level of a therapeutic gene product over time, we have constructed a multicompartment, numerical model. The model has six compartments, designated MILIEU, ENDOSOME, CELL, RNA, PROTEIN, and PRODUCT. The apparent level and kinetics of the gene product over time are calculated with different values for the intrinsic t1/2 of DNA in the MILIEU, ENDOSOME, and CELL; the intrinsic t1/2 of mRNA; the intrinsic t1/2 of the gene product; endosomal stability; and transcription rate. The model demonstrates how first-order kinetics can result from the summation of complex kinetic processes and provides a theoretical basis for future pharmacokinetic studies. This theoretical model illustrates how the half-lives of DNA, RNA, and gene product each affect the level of the product and highlights strategies for enhancing the therapeutic profile of gene therapies.

The challenge of follow-up for clinical trials of somatic gene therapy.

Although extensive efforts have been made to optimize the safety of vectors for somatic gene transfer and somatic gene therapy, the safety of these agents must ultimately be assessed in clinical trials. A statistically significant assessment of safety will be complicated by the relatively small number of patients who will be enrolled in initial clinical trials, the need for long-term longitudinal follow-up of patients and perhaps their progeny, and the traditionally poor participation in long-term follow-up by many patients. This article reviews the potential risks of retroviral-mediated gene transfer, the statistical power required to assess the true incidence of potential complications, the number of patients who may participate in clinical trials involving retroviral vectors, and the factors that make thorough follow-up and uniform data ascertainment difficult. The role of the FDA in assessing the safety of retroviral vectors and the potential role of registries for patient tracking and data ascertainment are discussed.

Gene transfer to synovial cells by intra-articular administration of plasmid DNA.

We studied reporter gene expression in synovial tissue after intra-articular administration of an expression plasmid into the knees of rabbits and rats. In both species, administration of a plasmid encoding beta-galactosidase led to gene expression in the synovial cells lining the joint. Expression correlated with the presence of plasmid DNA in synovial tissue extracts. Studies with a plasmid encoding chloramphenicol acetyltransferase demonstrated that gene expression persists for 2-5 days after administration. Southern blotting demonstrated that the administered plasmid was taken up rapidly by synovial tissue and degraded. By 24 hr after administration, no intact plasmid could be detected by Southern blotting, although small amounts of plasmid could be amplified by PCR up to 7 days. Administration of a plasmid encoding human growth hormone demonstrated that this product could be expressed from synovial cells and secreted into the synovial fluid. The histological distribution of gene expression in synovium resembles the known distribution of particulate materials injected into the joint and suggests that plasmid DNA is taken up by nonspecific endocytosis like other particulate materials during the remodeling of synovial fluid.

In vivo gene transfer into rabbit thyroid follicular cells by direct DNA injection.

Direct injection of DNA expression vectors into muscle leads to expression of encoded recombinant gene products in mature muscle cells. This phenomenon is not shared by most other organs. We have surveyed various organs in the rabbit to identify other cell types that would express DNA vectors after direct injection. We observed that thyroid follicular cells were capable of acquiring plasmid DNA and expressing recombinant gene products after direct interstitial injection of plasmid vectors into the thyroid gland. The level of expression of a chloramphenicol acetyltransferase (CAT) reporter gene in thyroid tissue was similar to that seen in muscle tissue three days after injection in controlled experiments. Using a beta-galactosidase reporter gene, expression was localized to thyroid follicular cells. CAT activity decreased with first-order kinetics and a half-life t1/2 of 40 hr. DNA was identified in thyroid tissue by polymerase chain reaction (PCR) analysis and displayed first-order elimination kinetics with a half-life t1/2 of 10 hr. The persistence of the gene and gene product in the thyroid was significantly different from that observed after injection of DNA vectors into muscle or delivery of DNA vectors to the liver using asialoglycoprotein/polylysine/DNA complexes, suggesting that there are significant differences in the process of DNA uptake or compartmentalization in these experimental systems. These results introduce the possibility of developing the thyroid as a novel target for treating certain thyroid or systemic diseases using DNA vectors.

Overexpression of human methylmalonyl CoA mutase in mice after in vivo gene transfer with asialoglycoprotein/polylysine/DNA complexes.

Methylmalonic acidemia resulting from genetic deficiency of methylmalonyl CoA mutase (MCM) is an often fatal metabolic disease. Somatic gene therapy for this disorder may require gene replacement in the liver. We describe overexpression of MCM in the liver of mice after in vivo gene delivery using asialoglycoprotein/polylysine/DNA (ASO/PL/DNA) targeted delivery to the liver of plasmids expressing recombinant MCM. After intravenous administration of the ASO/PL/DNA complex, the vector sequences are cleared from the blood with t1/2 = 2.5 min and > 95% of the vector is taken up by the liver. Vector sequences are cleared from the liver with t1/2 = 1.0-1.3 hr. MCM enzyme activity in the liver increases to levels 30-40% over baseline 6-24 hr after injection. No acute or chronic toxicity was observed. This net level of expression is likely to be therapeutic for MCM if the complex could be administered repetitively to treat acute episodes of life-threatening acidosis or establish a steady-state level of MCM activity. Repetitive administration of the ASO/PL/DNA complexes in mice was associated with formation of antibodies against asialo-orosomucoid and the asialo-orosomucoid complex but not against DNA.

Retrovirus-mediated gene transfer into canine thyroid using an ex vivo strategy.

We describe studies in a canine model aimed at establishing methods for ex vivo gene delivery to thyroid follicular cells. Canine follicular cells were harvested from tissue obtained by unilateral lobectomy, grown in thyrotropin-containing media, and transduced with amphotropic retroviral vectors carrying Escherichia coli beta-galactosidase or Tn7 neomycin-resistance genes. Up to 30% of cells were transduced with retroviral vectors containing the neomycin resistance gene, and transduced cells could be selected with G418. Significantly, transduced and selected cells exhibited the morphology of thyroid follicular cells and continued to express thyroglobulin. To assess the viability of cultivated and transduced cells for transplantation, cells were stained with the vital fluorescent dye DiI, recovered by trypsinization, and transplanted into the contralateral thyroid lobe of autologous animals. Engraftment was demonstrated by fluorescence microscopy and identification of proviral sequences 7-10 days after transplantation. Proviral transcripts were evident using coupled reverse transcription and the polymerase chain reaction using total RNA from transplanted glands. Thyroid follicular cells may represent an attractive target for gene therapy due to their proliferative potential, their large protein synthetic and secretory capacity, and their susceptibility to regulation. The thyroid might be a target for therapy of congenital or acquired thyroid diseases as well as disorders requiring regulated expression of proteins in the circulation. This work demonstrates the feasibility of ex vivo gene delivery to thyroid follicular cells that may be used in future investigations.

Retroviral gene transfer into the intestinal epithelium.

The epithelial cells of the gastrointestinal tract may be attractive targets for somatic gene therapy. In these studies, we have used rats and mice to explore the feasibility of gene transfer into the small intestinal epithelium using retroviral vectors. The first series of experiments was conducted in mature Sprague-Dawley rats using an ecotropic retroviral vector that has bacterial beta-galactosidase (beta-Gal) as the reporter gene. The vector was introduced into the lumen of ligated segments of terminal ileum. After a 4-hr exposure period, the ligatures were removed. Sham-operated animals were subjected to the same ligation procedure but received only tissue culture medium in the ligated segment. All animals were sacrificed 6 days later, and tissue from both the experimental segment and an upstream control segment was assessed for cytoplasmic beta-Gal activity using X-Gal histochemistry. Expression of the reporter gene was observed in the crypt epithelium of tissue exposed to the vector. In the villus epithelium, high background staining precluded accurate assessment of reporter gene expression. To obviate the latter problem, we sought an alternative reporter gene for which there would be no background staining in control animals. We repeated the experiments with beta-glucuronidase as the reporter gene in MPS VII mutant mice, which are devoid of this enzyme. In these studies, ileal segments exposed to the vector demonstrated expression of the reporter gene in both the crypt and villus epithelium 4 days after exposure. These results indicate that genes can be transferred into the intestinal epithelium using retroviral vectors introduced luminally.(ABSTRACT TRUNCATED AT 250 WORDS)

Hepatocellular transplantation in acute hepatic failure and targeting genetic markers to hepatic cells.

Orthotopic liver transplantation (OLT) represents the only therapeutic option for many patients with end-stage liver disease as well as many inborn genetic errors of hepatic metabolism. Despite dramatic progress in methods for OLT, the utilization of this procedure is limited by its considerable morbidity and mortality, by a chronic shortage of organs for transplant, and by difficulty arranging funding for many patients. Many children with fulminant hepatic failure do not receive OLT because this technology is unavailable or unaffordable. Hepatocellular transplantation (HCT), in which isolated, heterologous hepatocytes from a donor liver would be infused into the diseased organ in order to provide essential hepatic functions, could provide a much needed therapeutic alternative to OLT in the treatment of some causes of hepatic insufficiency. Experiments in animals have demonstrated that several genetic deficiencies of hepatic metabolism as well as experimental induced hepatic failure in animals can be reversed by HCT. Despite this experience, HCT has never been attempted in human subjects. This protocol represents the first proposed clinical trial of HCT. We are proposing a clinical trial in which HCT would be attempted as a therapeutic intervention in children with acute hepatic failure who have no other medical or surgical options. This proposal is intended to establish surgical methods for HCT and to evaluate the feasibility of this procedure for treating hepatic disease in humans. It is our expectation that HCT may provide short-term support for patients awaiting organ availability, a "bridge to recovery" allowing patients with fulminant hepatic failure to recover, or a long-term repopulation of the patient's liver with healthy donor cells. One of the major limitations of many animal studies in HCT is that, since the donor hepatocytes are often indistinguishable from those of the host, it has often been difficult to demonstrate a clear correlation between engraftment and the therapeutic effect. In order to verify engraftment independent of the therapeutic response, we propose to "mark" the donor hepatocytes by transducing these cells with a recombinant retroviral vector (LNL6) carrying a marker gene (NEO-R, neomycin phosphoribosyl transferase). The presence of this marker will enhance the ability to identify transplanted cells in the host using assays for the NEO-R gene or transcribed NEO-R mRNA. The LNL6 vector has been approved for human use and has been used as a marker gene for transplanted cells in human subjects without any reported adverse effects. We would like to emphasize that this is a proposal with therapeutic intent.(ABSTRACT TRUNCATED AT 400 WORDS)

Therapeutic promise of molecular genetics.

The application of molecular genetics to medicine is based on the observation that even common diseases have both genetic and environmental components. Conventional medicines are often effective in managing environmental components of disease but are generally ineffective in managing genetic diseases or manipulating the genetic component of multifactorial diseases. The development of therapies aimed at the genetic component of disease will require non-conventional medicinal applications of molecular genetics. Various approaches have been proposed such as diagnosing the propensity for disease to facilitate early intervention with conventional therapies, selectively eliminating mutant genes from human populations, correcting mutations in human chromosomes, and using genes as medicines to modify the genetic components of disease. Of these, it is the development of gene medicines that has the greatest practical potential. The combination of conventional medicines, focused on the environmental components of disease, and gene medicines, focused on the genetic components, will provide the clinician with broad options for managing health and disease. The challenge to molecular biology is to develop gene medicines that are effective, safe, and socially acceptable, and therapies that map well to established clinical practice and may be employed efficaciously alongside conventional medicines.