Viral vectors in gene therapy

The use of DNA as a drug is both appealing and simple in concept. Indeed in many instances the feasibility of such an approach has been established using model systems. In practical terms, however, the delivery of DNA to human tissues presents a wide variety of problems that differ with each potential therapeutic application. In this review, the design, production, and application of viral vectors for human gene therapy are considered. Although viral vectors are an obvious starting point because viruses have evolved efficient mechanisms to introduce and express their nucleic acid into recipient cells, by the same token the viral hosts have evolved sophisticated mechanisms to rid themselves of such pathogens. The challenge for the therapeutic use of viral vectors is to achieve efficient and often extended expression of the exogenous gene while evading the host defenses. Methodology used and progress towards that goal are reviewed.

Characterization of an adenovirus gene transfer vector containing an E4 deletion

We describe the construction and characterization of an adenovirus type 2 vector, Ad2E4ORF6, which has been modified in the E4 region to contain only open reading frame 6. When assayed in cultured cells, Ad2E4ORF6 virus replication is slightly delayed but viral DNA synthesis, host-cell protein synthesis shut-off, and virus yield are indistinguishable from wild type. Late protein synthesis is normal with the exception of fiber synthesis, which is reduced approximately 10-fold. Despite the reduced fiber synthesis, Ad2E4ORF6 viral particles appear to contain a full complement of fiber protein. Virus replication in cotton rats indicates that Ad2E4ORF6 is replication defective in vivo. This may have safety implications for the development adenovirus vectors in that virus arising by recombination in the E1 region of an Ad2E4ORF6-based vector would be defective for growth in vivo. The deletion of E4 open reading frames that are not required for virus growth in vitro increases the cloning capacity of adenovirus vectors by 1.9 kb and may be generally useful for the construction of adenovirus vectors containing large cDNA inserts and/or regulatory elements. We describe the inclusion of the A2E4ORF6 modification in a recombinant adenovirus vector, Ad2/CFTR-2, for gene transfer of the human cystic fibrosis transmembrane regulator (CFTR).

Delivery of purified, functional CFTR to epithelial cells in vitro using influenza hemagglutinin

To assess the feasibility of protein replacement as a potential therapy for cystic fibrosis, we have evaluated the ability of influenza hemagglutinin (HA) to mediate the delivery of purified cystic fibrosis transmembrane conductance regulator (CFTR) to recipient cells in vitro. CFTR was purified from both CHO cells and Sf9 cells and reconstituted into two different types of vesicular delivery vehicles. In one, CFTR and HA were co-reconstituted into the same lipid vesicle. After binding to the cell surface, delivery of CFTR to the recipient cell was achieved by a transient, low-pH activation of the fusion activity of HA. A second delivery strategy used HA virosomes together with purified CFTR that had been reconstituted into vesicles containing gangliosides, a receptor for HA. After binding of the HA virosomes and CFTR-containing vesicles to the recipient cells, delivery to the plasma membrane again was achieved by a transient pH drop. Delivery of functional CFTR was assessed using the SPQ fluorescence assay. Functional CFTR was detected in a fraction (> 20%) of the recipient cells using this assay. Quantitative binding and fusion assays using radiolabeled virosomes and lipid vesicles showed that on the order of 1,000 of the added CFTR-containing vesicles bound to each C127 cell under the conditions of our delivery protocols. However, only a fraction of these vesicles fused and delivered CFTR to the cell plasma membrane. The two delivery strategies were found to be approximately equivalent in their ability to deliver active CFTR, and there were no significant differences between deliveries using purified CFTR from either cell source. These feasibility studies suggest that purified CFTR can be delivered to a recipient cell in a functional form and therefore represent a significant step in establishing the concept of protein replacement as a therapy for cystic fibrosis.

Sequence homologies between nucleotide binding regions of CFTR and G-proteins suggest structural and functional similarities

Sequence homology between the alpha-subunits of G-proteins and other GTP-binding proteins and certain regions within the nucleotide binding domains (NBDs) of cystic fibrosis transmembrane conductance regulator (CFTR) indicates that these protein structures may be similar. A sequence alignment of the NBDs of CFTR and NBDs from other membrane transporters, forms the basis of a structural model. This model predicts that one of the conserved sequences GGQR, within which a number of CF mutations occur, forms part of the nucleotide binding pocket and serves as an ON/OFF conformational switch as observed in GTP binding proteins. Furthermore, based on subtle sequence differences between the first and second NBDs of CFTR and from structure-activity data, we suggest that the nucleotide binding site environments of the two NBDs are different.

Adenoviral-mediated gene transfer to fetal pulmonary epithelia in vitro and in vivo

Vector-mediated gene transfer offers a direct method of correcting genetic pulmonary diseases and might also be used to correct temporary abnormalities associated with acquired, nongenetic disorders. Because the fetus or newborn may be a more immune tolerant host for gene transfer using viral vectors, we used replication defective recombinant adenoviral vectors to test the feasibility of gene transfer to the fetal pulmonary epithelium in vitro and in vivo. Both proximal and distal epithelial cells in cultured fetal lung tissues from rodents and humans diffusely expressed the lacZ transgene 3 d after viral infection. In vivo gene delivery experiments were performed in fetal mice and lambs. Delivery of Ad2/CMV-beta Gal to the amniotic fluid in mice produced intense transgene expression in the fetal epidermis and amniotic membranes, some gastrointestinal expression, but no significant airway epithelial expression. When we introduced the adenoviral vector directly into the trachea of fetal lambs, the lacZ gene was expressed in the tracheal, bronchial, and distal pulmonary epithelial cells 3 d after viral infection. Unexpectedly, reactive hyperplasia and squamous metaplasia were noted in epithelia expressing lacZ in the trachea, but not in the distal lung of fetal lambs. 1 wk after infection, adenovirus-treated fetuses developed inflammatory cell infiltrates in the lung tissue with CD4, CD8, IgM, and granulocyte/macrophage positive immune effector cells. Transgene expression faded coincident with inflammation and serologic evidence of antiadenoviral antibody production. While these studies document the feasibility of viral-mediated gene transfer in the prenatal lung, they indicate that immunologic responses to E1-deleted recombinant adenoviruses limit the duration of transgene expression.

Specific competitive inhibitor of secreted phospholipase A2 from berries of Schinus terebinthifolius

Two structurally related triterpenoids 1 and 2 from pink peppercorn (berries of Schinus terebinthifolius) are identified and characterized as active site-directed specific competitive inhibitors of the three classes of secreted 14 kDa phospholipase A2. The inhibitors not only protect the active site histidine from alkylation but also inhibit the action of secreted phospholipase A2 from pig pancreas, human synovial fluid, and bee venom. Detailed X-ray crystallographic results on the structures of the inhibitors are provided. By physical methods and X-ray crystallography the triterpenoids were identified as masticadienoic acid and masticadienolic acid (schinol). Several other triterpenoids were ineffective as inhibitors of phospholipase A2; however certain ganoderic acid derivatives showed noticeable inhibition. Results show that the side chain of these acidic tetracyclic terpenoids can access the catalytic-site region of phospholipase A2, whereas the acyclic nucleus is at the interfacial recognition region. The selectivity of the assay protocol used here is demonstrated by the fact that the original screen of ethyl acetate extracts of 60 commercially available spices and herbs was carried out with phospholipase A2 from pig pancreas, and only one extract showed inhibitory action on the hydrolytic activity in the scooting mode. Under such assay conditions the enzyme remains tightly bound to the surface of the substrate vesicles. In this way, nonspecific effects of additives that promote desorption of the enzyme from the substrate vesicle surface, under conditions in which the binding of the enzyme to the vesicle is weak, are avoided. The assay protocol is useful for the kinetic characterization of the inhibitors of phospholipase A2, and it does not give false positive results with amphiphilic and hydrophobic compounds, as is the case with virtually all assay systems in use.

The ability of predicted internal dose measures to reconcile tumor bioassay data for chloroform

PBPK models are developed in the hope that they will improve our ability to extrapolate from one species to another and from one exposure regime to another. Evidence that a dose measure was successful at reconciling the available animal bioassay data would be encouraging. It would give us some confidence that the dose measure (as evaluated by a PBPK model) might yield reasonable predictions for yet other species (e.g., humans) and other dose routes. We have investigated the ability of a modified version of the Corley et al. (R. A. Corley, A. L. Mendrala, F. A. Smith, D. A. Staats, M. L. Gargas, R. B. Conolly, M. E. Andersen, and R. H. Reitz, 1990, Toxicol. Appl. Pharmacol. 103, 512-527) PBPK model for chloroform to reconcile the available bioassay data. Two rate-dependent dose measures, maximal rate of metabolism in the liver, and percentage of hepatocytes killed per day performed well at reconciling the rodent liver bioassay data, while all rate-independent dose measures performed less well. In contrast, none of the PBPK dose measures were capable of reconciling the rat and mouse kidney tumor response data. Here, administered dose scaled to body surface area performed the best.

Primary and secondary prevention services in clinical practice. Twenty years' experience in development, implementation, and evaluation

This article reviews lessons from 20 years of experience in development and provision of clinical preventive services at Group Health Cooperative of Puget Sound, a large health maintenance organization. Critical factors for enhancing service include the use of a population-based epidemiologic viewpoint coupled with specific evidence-based criteria to examine issues; involvement of practitioners in the process; a systems approach to implementation focused on predisposing factors of the practitioners and enabling factors in the practice, organizational, and community environments; feedback of program outcomes; and the use of automated clinical information systems. Outcome results from our clinical prevention efforts include a 32% decrease in late-stage breast cancer (1989 to 1990); 89% of 2-year-old children with complete immunizations (1994); decrease in adult smokers from 25% to 17% (1985 to 1994); and an increase in bicycle safety helmet use among children from 4% to 48% along with a 67% decrease in bicycle-related head injuries (1987 to 1992). Systematic population-based approaches to the development and provision of clinical preventive services targeting the one-to-one level of primary care and multiple infrastructure levels of care are forging a synthesis of clinical medicine and public health approaches. This approach will become pervasive as clinical information systems improve, risk information is captured routinely, and practitioners gain skills in the art of patient risk behavior change and population-based care.

Uncertainty in fitted estimates of apparent in vivo metabolic constants for chloroform

It has become common practice to rely on fitted estimates of apparent in vivo metabolic constants (e.g., Vmax and KM) in the parameterization of PBPK models. Yet, quantitative estimates of precision in these fitted parameters are not routinely reported. Such information is needed to assess the reliability of model predictions. The purpose of this study was to assess the precision in estimates of Vmax and KM for chloroform, accounting for both the statistical uncertainties in parameter estimates from individual data sets and any additional uncertainty due to differences in the parameter estimates derived from various experiments. Joint confidence regions for Vmax and KM from each experiment, generated using maximum likelihood techniques, were used to evaluate these questions. Three previously published data sets were considered. Estimates of Vmax and KM obtained from these data sets differed more than could be explained as a consequence of a limited number of observations, measurement error, or stochastic error. Issues associated with the use of maximum likelihood techniques to estimate joint confidence regions, the estimation of metabolic constants from individual experiments within a gas uptake study versus the full data set, the degree of overlap in the joint confidence regions for metabolic constants obtained from separate data sets, and the implications for risk assessment are discussed.

Functional activation of the cystic fibrosis trafficking mutant delta F508-CFTR by overexpression

The most common mutation in the gene associated with cystic fibrosis (CF) causes deletion of phenylalanine at residue 508 (delta F508) of the gene product called CFTR. This mutation results in the synthesis of a variant CFTR protein that is defective in its ability to traffic to the plasma membrane. Because earlier studies showed delta F508-CFTR retains significant phosphorylation-regulated chloride (Cl-) channel activity, processes capable of restoring the mislocalized delta F508-CFTR to the correct cellular destination may have therapeutic benefit. Here we report one such process that involves overexpression of the mutant protein and appears to result in the escape of a small amount of delta F508-CFTR to the plasma membrane. In recombinant cells where expression of delta F508-CFTR is controlled by the metallothionein promoter, this effect can be brought about by treatment with sodium butyrate. Although cAMP-activated Cl- channel activity could also be detected in immortalized human airway epithelial cells homozygous for the delta F508 mutation at the single cell level, treatment with butyrate did not generate a measurable cAMP-stimulated Cl- current in polarized monolayers of primary CF airway epithelia. However, the observation that overexpression can effect the presence of recombinant delta F508-CFTR at the plasma membrane suggests that perhaps other butyrate-like compounds that are more potent and more specific for the promoter of the CF gene may be efficacious in alleviating the Cl- channel defect associated with CF.

Treatment of cystic fibrosis based on understanding CFTR

In the five years since the gene associated with cystic fibrosis was isolated, rapid progress has been made in understanding the structure and function of CFTR, the gene product. Based on the knowledge that CFTR is a cAMP-regulated chloride channel in the apical surface of epithelia in many tissues, new approaches to treating CF disease have emerged. These include development of agents to open alternative chloride channels, to relocate mutant CFTR, which commonly is mislocalized because of a trafficking defect, and to deliver CFTR using protein or gene therapy. Although the underlying basis of each of these approaches is now well understood, major practical problems face each of them. Nevertheless, it is difficult not to be optimistic that new therapies will emerge.

Use of probabilistic expert judgment in uncertainty analysis of carcinogenic potency

A new approach to characterizing the state of knowledge about carcinogenic potency is described. In this approach, the carcinogenic risk posed by a specific dose is characterized by a probability distribution, indicating the relative likelihood of different risk estimates. The approach utilizes expert judgment and a probability tree and is illustrated in a case study of chloroform exposure. Experts in cancer biology/toxicology, pharmacokinetics, and dose-response modeling were identified by a panel of science-policy specialists. In a workshop, experts reviewed the chloroform data, received training in probability elicitation, and constructed a consensual probability tree based on biological theories of cancer causation. Distributions of carcinogenic risk were developed based on the probability tree, chloroform data, judgmental probabilities provided by the experts, and classical statistical techniques. Risk distributions varied considerably between experts, with some predicting essentially no risk from 100 ppb chloroform in drinking water while other have at least some probability on risks generally considered of regulatory significance. Estimated human risk was much lower when extrapolating from liver tumors in animals than from kidney tumors. Issues of scientific disagreement leading to different risk distributions between experts are discussed. The resulting risk distributions are compared to standard EPA risk calculations for the same exposure scenario as well as to the expert judgement of epidemiologists about cancer risks of chlorinated drinking water. Issues in combining expert judgments are discussed, and several alternative methods are presented. Strengths and weaknesses of the distributional approach are discussed.

Comparison of survival. Chemotherapeutically treated and untreated older patient with acute myeloid leukemia

Survival of older patients with acute nonlymphocytic leukemia treated with chemotherapy was compared with those given only supportive care. The treated group was younger and had better survival. The benefits which measure in months must be balanced against socioeconomic cost, toxicity of therapy, patient wishes, and evolving information about the standards of management in the older leukemia patient.

Adenovirus-mediated gene transfer to murine retinal cells in vitro and in vivo

Adenovirus-mediated gene transfer to retinal cells was evaluated using the replication-defective recombinant adenovirus vector Ad2/CMVlacZ-1 (coding for beta-galactosidase) both in an in vitro murine culture model and in vivo in adult mice. In vitro, no difference in infectability of neuronal and glial cells was observed, and 50% of neurons expressed the exogenous gene at low viral concentration (10 pfu/cell). In vivo, intraocular injection of 3 x 10(6) pfu Ad2/CMVlacZ-1 resulted in expression of the transferred beta-galactosidase gene in retinal pigment epithelium and ganglion cells. These results demonstrate that Ad2/CMVlacZ-1 is an effective vector for gene transfer into retinal cells.

Cellular transformation by a transmembrane peptide: structural requirements for the bovine papillomavirus E5 oncoprotein

The E5 oncoprotein of bovine papillomavirus, only 44 amino acids long, occurs as a disulfide-bonded transmembrane dimer. This remarkable oncoprotein stimulates signal transduction through activation of the platelet-derived growth factor (PDGF) receptor, and E5 exhibits limited amino acid sequence similarity with PDGF. Results presented here suggest that a key feature of the hydrophobic transmembrane domain is an amino acid side chain that participates in interhelical hydrogen bond formation. These data are reminiscent of the activated neu oncogene, in which a point mutation in the transmembrane domain leads to ligand-independent dimerization and activation of a receptor tyrosine kinase. Significantly, the transmembrane domain of E5 can be largely replaced by the transmembrane domain from the activated neu receptor tyrosine kinase. Extensive mutagenesis defines the minimal structural features required for transformation by the E5 oncoprotein as, first, the ability to dimerize and, second, presentation of a negatively charged residue at the extracellular side of the membrane. The biological activity of E5 mutants that lack most amino acid residues similar to PDGF suggests that E5 and PDGF activate the PDGF receptor by distinct mechanisms.

Correction of cAMP-stimulated fluid secretion in cystic fibrosis airway epithelia: efficiency of adenovirus-mediated gene transfer in vitro

Adenovirus vectors are a promising vehicle to deliver cystic fibrosis transmembrane conductance regulator (CFTR) cDNA to airway epithelia. However, the value of adenovirus vectors will depend on the efficiency with which the vector can correct the defective fluid transport that is though to underlie the pathogenesis of the disease. To address the efficiency of gene transfer, we applied adenovirus vectors expressing CFTR (Ad2/CFTR-1) or beta-galactosidase to the mucosal surface of primary cultures of airway epithelial cells grown as polarized epithelial monolayers on permeable filter supports. These conditions provide a model that reproduces the physiology of the airways in vivo. We found that after adding 1 moi Ad2/CFTR-1 to the mucosal surface, cAMP agonists stimulated fluid secretion that was within the range observed in epithelia from normal subjects. When we measured electrolyte transport, we found that as little as 0.1 moi partially restored cAMP-stimulated Cl- secretion, and at 10 moi Cl- secretion was in the normal range. A related vector encoding beta-galactosidase generated activity in approximately 20% of cells at an moi of 1 and 90% of cells at an moi of 10. These data suggest that Ad2/CFTR-1 is very efficient at restoring normal fluid and electrolyte transport to CF airway epithelia. Thus, they suggest that relatively low input doses could be used for gene transfer to CF airway epithelia.

Corpus luteum growth and function in lactating Holstein cows during spring and summer

Corpus luteum growth and function were monitored daily for a complete estrous cycle by ultrasonography and peripheral serum progesterone in lactating, cyclic Holstein cows during spring (n = 8) and summer (n = 8). For spring and summer, respectively, daily ambient temperatures (mean +/- SEM) were 21.2 +/- .9 and 31.1 +/- .3 degrees C. In summer compared with spring, early morning rectal temperatures were higher on d 1, 2, and 3, when the corpus luteum was forming, and on d 15 and 19. The length of the luteal phase and the corpus luteum cross-sectional area were similar for the two seasonal groups. Central luteal cavities were observed during spring and summer. Serum progesterone secreted between d 6 and 18 was lower during summer. In addition, progesterone concentrations during the entire luteal phase and the peak magnitude of progesterone tended to be lower during summer. Suppressed luteal function may contribute to low fertility when cows are inseminated during summer.

Stoichiometry of recombinant cystic fibrosis transmembrane conductance regulator in epithelial cells and its functional reconstitution into cells in vitro

We have generated several clones of Chinese hamster ovary, mouse epitheloid C127, and pig kidney epithelial LLCPK1 cells producing high levels of functional recombinant human cystic fibrosis transmembrane conductance regulator (CFTR). Processing of CFTR to the mature and fully glycosylated form in these cells is inefficient with only approximately 40% of all newly synthesized CFTR being converted to the mature form. Furthermore, expression of the most frequent mutant allele of the cystic fibrosis (CF) gene, the delta F508 mutant in these epithelial cells, indicated that it is biosynthetically arrested at the endoplasmic reticulum and fails to traffic to the plasma membrane. Using a combination of CFTR mutants and monoclonal antibodies, all the detectable recombinant CFTR in these cells was determined at least under the conditions used, to be present as a monomer. To demonstrate the feasibility of protein replacement therapy, we were able to effect the physical transfer of functional recombinant CFTR produced in Chinese hamster ovary cells to the plasma membranes of Ha3b fibroblasts, a cell line devoid of cAMP-stimulated chloride channels. Transfer of CFTR was mediated by the hemagglutinin viral fusion protein of influenza virus present on the Ha3b cells. Efficiency of transfer was up to 25% of the target cells, and CFTR chloride channel activity was detectable for up to 12 h post-fusion. Therefore, with the development of an appropriate formulation of fusogenic proteoliposome or virosome containing reconstituted purified CFTR, it should be feasible to introduce functional CFTR into CF-affected cells.