Since the death last September of an Arizona teenager, the first person to die of gene therapy, many gene therapists are looking to gutless adenovirus vectors to mend the reputation of adenovirus-based gene therapy. Gutless vectors are named for their lack of adenovirus genes. If, as expected, they prove safer and provide longer lasting therapeutic gene expression, adenovirus vectors like the one that killed Jesse Gelsinger will likely be phased out for most diseases.
Why Gelsinger died remains unknown. But multiple investigations of both his clinical trial and gene therapy regulation have been launched. The Food and Drug Administration, after finding numerous errors in the Gelsinger trial, in January suspended all seven trials in progress where he was treated, the Institute for Human Gene Therapy in Philadelphia. On the heels of that event is publication in February of long-anticipated work on gutless vectors1 by Volker Sandig, Thomas Caskey, and their colleagues at Merck & Co. They describe an optimized gutless vector system that may overcome major problems earlier vectors have never solved--inflammatory immune responses to adenoviral proteins and inadequate transgene expression, usually lasting only days instead of months. Therapeutic gene expression stops when the cell-mediated immune response, reacting against adenoviral proteins, kills vector-transduced cells. A variety of adenoviral vectors, Gelsinger's among them, dealt with those problems by deleting various regulatory genes controlling expression of genes for antigenic viral proteins; it was believed the deletions would prevent viral protein production. Unfortunately, it didn't work out that way--viral antigen expression was lowered rather than stopped, so immune responses still shut down transgene expression.
Gutless vectors were recognized years ago as the logical next step in vector design. They contain no adenoviral sequences except those for DNA replication and packaging DNA into the viral capsid. Producing them, however, eluded nearly everyone who tried.
Making gutless viruses requires supplying of adenoviral proteins in trans, which is done by coinfecting vector- producing cells with helper viruses. Helper viruses, however, introduce two new problems: how to separate helper and gutless viruses produced by cells, and how to prevent their DNA from recombining to create unwanted mutant viruses and diminish vector production.
A few years ago, a solution to the separation problem seemed at hand. Frank Graham and Robin Parker at McMaster University in Canada introduced an inducible means of shutting down production of helper virus, so only gutless viruses emerged from production cells. They used the inducible Cre-lox DNA recombination system to make helper virus delete packaging signals by intramolecular DNA recombination; without DNA packaging, no helper viruses form as contaminates. But in practice, many labs found the technique failed to make gutless vectors more often than it succeeded. At Merck, which licensed Graham's system, Sandig, Caskey, and their coworkers decided to find out why what seemed ingenious often failed to perform. The main problem was DNA recombination between the packaging signals of the helper and gutless viruses, which allowed helper viruses to acquire packaging signals that could not be excised by Cre-lox recombination. As a result, recombinant viruses contaminated and nearly always overgrew gutless viruses. The team solved the problem by giving the helper virus a packaging signal designed not to recombine with the vector signal.
They also made two other changes. They added a 400 bp fragment located outside the packaging signal to the gutless vector to improve vector packaging efficiency. And they redesigned the vector's stuffer sequence to increase resistance to DNA rearrangement and boost production. Stuffers are needed because viable adenoviruses require a minimum amount of DNA, and therapeutic genes are usually far smaller than the 28-36 kb of DNA in a vector; stuffers take up the space unoccupied by transgenes. Merck's stuffer uses human DNA free of known genes, repeat sequences, and retroviral elements, and is constructed to inhibit homologous recombination with chromosomal DNA.
The modifications have allowed Merck to make gutless vectors for a number of transgenes, with far less contamination than previously possible. "We have not designed the system," Sandig comments. "Frank Graham's group did that. But we have changed the small pieces that make it work." Animal tests show the benefits of making less-inflammatory vectors. In rodents, the researchers find transgenes delivered by gutless vectors can be expressed at high levels for more than 200 days. Similar results have been found in baboons by Arthur Beaudet's lab at the Baylor College of Medicine.2
Better therapeutic gene expression may allow vector doses far lower than those used today, which often are set high to compensate for poor transgene expression. Although dosing will depend on target tissues, transgene promoters, and other considerations, Sandig thinks it is conceivable that gutless vector doses will be orders of magnitude smaller than those of early-generation vectors.
Smaller doses, in turn, may ease the problems of scaling up vector production. Indeed, production concerns provoke interest in Graham's system: "The need to produce at Phase III levels," Sandig declares, "is the reason people want biological systems that eliminate the helper without physical separation." Sandig points out that at least one other major hurdle lies ahead for adenovirus gene therapy. He calls it "neutralizing immunity"--millions of people produce antibodies against the adenovirus strains now used for vector development. Those people, immunized after catching colds caused by adenoviruses, might not be helped by adenovirus gene therapy. Vectors are needed from less-common adenovirus strains. Another problem looms: regulatory liability. Merck has no plans to produce gutless vectors for gene therapy trials at the moment; it is still in animal testing. But assuming they eventually prove safe in humans, Merck's regulatory concerns may delay the new vectors from coming into widespread use. The FDA could hold Merck ultimately responsible for their use in humans, even if investigators outside Merck offer to sign liability waivers in order to test them. Merck cited those concerns when it recently refused a request by Beaudet to use part of Merck's system.
Merck's view of its responsibilities may leave others no choice but to find other means of producing gutless vectors. One company that has done that is Genstar Therapeutics, a San Diego gene therapy company planning a Phase I clinical trial this year to test a gutless vector treatment for hemophilia A. Genstar doesn't use Cre-lox recombination to eliminate helper viruses. Instead, Genstar separates helper and gutless viruses by CsCl density gradient centrifugation. If either vector proves safer and more effective than the ones in use today, adenovirus-based gene therapy should have a much brighter future.
Tom Hollon (firstname.lastname@example.org) is a freelance writer in Rockville, Md.
1. V. Sandig et al., "Optimization of the helper-dependent adenovirus system for production and potency in vivo," Proceedings of the National Academy of Sciences, 97:1002-7, Feb. 1, 2000.
2. N. Morral et al., "Administration of helper-dependent adenoviral vectors and sequential delivery of different vector serotype for long-term liver-directed gene transfer in baboons," Proceedings of the National Academy of Sciences, 96:12816-21, Oct. 26, 1999.