Viral vs. Nonviral in Gene Therapy: Which Vector Will Prevail?

Pointing to a few empty chairs in a conference room, moderator David T. Curiel joked that a symposium on emerging nonviral vectors was a "failure" compared to other packed talks at the American Society of Gene Therapy's inaugural meeting last month. The director of the University of Alabama at Birmingham's gene therapy program also noted that viral vectors have generated more research activity than nonviral vectors. However, nonviral backers think their delivery vehicles for therapeutic genetic material promise greater impact in the future. And even some pioneers of viral vectors attending the Seattle meeting agreed. EASILY ADAPTABLE: The chemical flexibility of nonviral systems means they are easier to modify, explains the University of Pittsburgh's Leaf Huang. For now, the gene therapy arena remains divided. "The field will remain bifurcated for many years," W. French Anderson, professor and director of the Gene Therapy Laboratories, University of Southern California School of Medicine in Los Angeles, predicted during a press briefing at the meeting. The division exists as a result of each system's inherent strengths and weaknesses. Some viral vectors have demonstrated higher levels of expression and genomic incorporation during clinical trials than nonviral vectors have thus far. But nonviral vectors, while not as efficient as their viral counterparts, avoid triggering an immune response from the body--perhaps gene therapy's biggest obstacle to success (W.F. Anderson, Nature, 392(suppl):25-30, 1998). However, Philip L. Felgner, who developed the first nonviral vector, a cationic lipid he named Lipofectin, thinks that the efficiency issue can be overcome fairly easily. "In a number of systems, we can produce the same efficacy as we can with a viral system, but we have to do it by dumping huge quantities of plasmid onto the system," the chief scientific officer of Gene Therapy Systems Inc., a San Diego biotech firm, asserted in an interview with The Scientist following his talk at the meeting. "Efficacy is not the same as efficiency." Felgner further notes that higher doses of the lipid, unlike higher doses of a virus, does not initiate immunogenesis. The more transient expression of the nonviral vectors also poses little problem, he explains, because the molecules could, unlike immunogenic viral vectors, be injected repeatedly. NONVIRAL BOOSTER: Gene Therapy Systems' Philip Felgner expects nonviral vectors to supplant viral ones, once gene therapy becomes a clinically routine procedure. Still, nonviral vectors have some obstacles to overcome before they can be judged truly successful, admits Felgner, who adds that the vectors still need to become more efficient. They aren't as fully understood as first-generation viral vectors, because they haven't been in as many clinical trials. On the way to the clinic, researchers can pursue five basic goals to improve the vectors, Felgner says: American Society of Gene Therapy - http://www.asgt.org Condense the DNA into a small package Target specific tissue via cell surface receptors Avoid nonspecific uptake Escape the endosome Cross the nuclear membrane In an interview with The Scientist prior to the conference, James M. Wilson , director of the Institute for Human Gene Therapy at the University of Pennsylvania Medical Center in Philadelphia, notes that many viral vectors can already do many of these things naturally. For example, unlike any other vector thus far, lentiviral vectors--stripped-down versions of the HIV retrovirus--can enter the nucleus. Building a synthetic molecule mimicking those qualities could prove difficult. "You are asking a lot of a molecular particle." Wilson admits both viral and nonviral vector builders are approaching the same problems from different sides. Viral vector builders are paring away pieces of genes so the vectors become more "stealthlike" and thus avoid detection from the body's immune system. Meanwhile nonviral vector architects are adding pieces to their vehicles--such as ligands to help the vectors bind to specific tissue--to make the molecules imitate some desirable properties of viruses. Other scientists are working to find the best uses for all vectors--both nonviral and viral. Inder M. Verma, gene therapy researcher at Children's Hospital of Los Angeles, suspects that many will have applications. "There is no perfect vector for every disease," Verma notes. "It really depends on what you want to do." Indeed, some researchers have already turned perceived shortcomings of viral vectors into advantages by directing them toward different diseases. For example, one team has begun using adenoviruses to initiate angiogenesis. Members of that same team grew frustrated with that virus as a vector for cystic fibrosis, because its high initial transfection was thwarted by short-term expression. However, those conditions fit the needs of angiogenesis well (P. Smaglik, The Scientist, 12(10):4-5, 1998). Just as the viral vector field continues to expand, the newer field of nonviral vectors continues to develop. The following are among many new approaches presented at the inaugural meeting of the American Society of Gene Therapy, May 28-31, in Seattle: Photo courtesy of Leaf Huang Human Artificial Chromosomes Building a chromosome from the ground up, using a set of telomeres, a centromere and therapeutic material could mimic one advantage of herpes-based vectors--without the toxicity, stated Huntington Willard, professor of human genetics at Case Western Reserve University School of Medicine in Cleveland, Ohio. "They can carry many large insertions and express multiple genes," Willard stated. "These chromosomes are enormous." He notes that the artificial chromosomes can be built as either circular or linear systems. Some disadvantages exist, too. Some centromere functions remain unknown, Willard reported. Researchers inserting human artificial chromosomes into mouse cells produced the artificial chromosomes at high efficiency, but the human centromere failed to function. "We are in the process of developing mouse artificial chromosomes, to get around this possible species effect," Willard told The Scientist following the talk. Cationic Lipids The first nonviral vector's single biggest failing is its low efficiency, Philip Felgner, Gene Therapy Systems Inc. chief science officer, admitted. "We definitely have plenty of room to improve these kinds of systems." However, Felgner thinks the problem could be addressed by adapting a "clamp" made from peptide nucleic acid (PNA) that the company has already developed as a fluorescent probe to track the vector's progress in cells. Felgner posits that adding ligands to the PNA could improve both efficiency and targeting of this vector. Infectious Mammalian Chromosomes This approach represents a synthesis of viral and nonviral approaches. "We're using some components of a virus," reported Jean-Michel Vos, associate professor of biochemistry at the University of North Carolina at Chapel Hill. Researchers formed components of the Epstein-Barr virus into a large circular replicating molecule that, in culture, shows stable expression for over a year. "It's a very large virus," Vos said. "It belongs to the Herpes family." In early transgenic mouse models, it indicated an efficiency of over 25 percent--much higher than other nonviral vectors. The challenge will be to see if researchers can match that rate in human in vivo trials and also direct its replication to affected tissues. Photo courtesy of Szoka & Tang Starburst Dendrites and New Polymers These polymers, shaped, as their names suggest, like stars built off a solid core, provide something that some other nonviral vectors lack--compactness. "All the polymers, regardless of their structures, gives you a collapsed DNA," explains Frank Szoka a professor of pharmaceutical chemistry at the University of California at San Francisco. However, partial "fractured" versions also have the ability to release DNA from the endosomes--another desirable quality lacking in some nonviral approaches. Once in the endosome, the compact structure expands in a process resembling the opening of an umbrella. In early tests, the molecules were applied to a rabbit's carotid artery and effectively transfected genes in the targeted area. They have yet to be tested on humans. Endothelial Cells Modifying endothelial cells into vectors could provide a more specific delivery vehicle, reported James A. Zwiebel, senior investigator at the National Cancer Institute. "Specific cell types can be isolated and genetically modified." Zwiebel and colleagues tested the modified cells as cancer fighters in mice. He reported that the cells--modified to carry therapeutic genes--shrank the tumors and prolonged the animals' lives. While no clinical human tests have yet been approved, Zwiebel notes that isolating the cells, then modifying them, could be performed with existing technology. "The cells can be readily harvested from subcutaneous fat, or even from the bloodstream." -P.S. Felgner predicts that the pharmaceutical industry will be more willing to back a synthetic vector than a viral one. "When you try to engage the [pharmaceutical] people with your gene therapy program and you're advocating the use of a viral vector, you come in with a tremendous liability, because those people aren't just thinking about having a successful animal model demonstration of gene therapy or even of having a successful clinical experiment. They want an actual product." Anderson agrees, even though he is the senior author of a paper documenting the first human gene therapy clinical trial, which employed a retroviral vector to correct the severe combined immunodeficiency disorder caused by a shortage of adenosine deaminase (R.M. Blaese, et al. Science, 270:475-80, 1995). "A nonviral system is easier to manufacture," French asserts. He adds that considerations such as cost and ease of manufacturing are key to the pharmaceutical industry, which will likely dictate the direction of delivery. "How does a product get out to the public? It goes out through the pharmaceutical industry." Felgner notes that if viral and nonviral vectors performed the same, the pharmaceutical industry would opt for the nonviral methods because of ease of manufacturing and lower levels of toxicity. Alan McClelland, vice president of research at Genetics Therapy Inc., a Novartis company based in Gaithersburg, Md., agrees--partially. "All things being equal, a synthetic system would be preferable," McClelland explains, adding that the company is developing both nonviral and viral systems. Some systems will be more effective against some kinds of diseases than others, he thinks. Both systems will likely find their way into use if they navigate successfully through clinical trials. "We're not going to end up with viral systems completely supplanted by nonviral systems." NATURAL ADVANTAGE: Penn's James Wilson notes that many viral vectors can already do many things that existing nonviral vectors cannot. The issue of manufacturing ease also remains somewhat less clear-cut than Felgner claims, McClelland explains. "I don't think that's enough of a barrier." For example, GTI routinely creates retroviral and adenoviral vectors; adeno-associated viruses, which even research labs have had difficulty producing in quantities large enough for experiments, present the only real production problem. The company provided the vectors for French and colleagues' Phase I. And issues of manufacturing control for nonviral vectors haven't been completely worked out yet, either, he says. Felgner admits that the cationic lipid production process isn't completely controlled yet--a properly mixed lipid yields a consistent solution, while too much or too little DNA in it yields a gloppy substance that sticks to the sides of the vial. "One of the challenges of the field is to bring these things under good control." Still, Felgner thinks that will be easier for nonvirals than for virals--especially in large quantities. A philosophical difference about the nature of gene therapy may lie at the heart of the debate. "A lot of the groups here that work on viruses come from a different perspective. They're thinking that it's not a pharmaceutical that they are developing; [That] what they're developing is a new paradigm for the treatment of disease where one-time treatment leads to a permanent cure." Felgner thinks that, ultimately, gene therapy will be considered a pharmaceutical product and should be treated as such during the design stage, as well as during manufacturing. "These agents are in a vial. They are druglike. They're handled by drug companies. And they are running those clinical trials like anyone would run a clinical trial for drugs," Felgner notes of cationic lipids. Photo: Mark Lieberman/The Salk Inst. VECTOR CORNUCOPIA: The myriad of both nonviral and viral vectors under development may have qualities that make them more suitable to treating specific conditions, concludes Inder M. Verma, of Children's Hospital of Los Angeles. Leaf Huang , professor of pharmacology at the University of Pittsburgh, agrees, adding that the synthetic nature of nonviral vectors makes them more easily adaptable. "They are chemically very flexible, Huang explains. "One can chemically modify these vectors." He notes that as viral vector researchers advance the basic science of how viruses get into cells, nonviral researchers will capitalize by aping these functions in their artificial systems. "Somewhere down the road, the field will merge," Huang predicts. "The best vector could be a hybrid of both."

Viral vs. Nonviral in Gene Therapy: Which Vector Will Prevail?

Interested in reading more?

Become a Member of