Construction of Non-Viral Vectors
Identification of cellular and molecular processes associated with gene expression and concurrent advancement of recombinant DNA technology has led to production of highly sophisticated recombinant DNA molecules. Several issues that must be consid-ered when designing a plasmid-based vector are briefly outlined below.
Origin of Replication (ori). Identification of the rop gene as the primary gatekeeper that limits the plasmid number to 20 to 40 copies/cell in bacteria led to the development of the pUC series of plasmids. Elimination of rop increased the acceptable copy number to more than 500 copies/cell for plasmids that encode proteins of approximately 1,500 base pairs. Most of the plasmids used in the clinic in the last 2 to 3 years have been constructed using pUC plasmids (Manthorpe, 2005).
Selection Marker. Selection markers have commonlybeen those that conferred resistance to penicillin-based drugs since growth of bacteria in antibiotic supplemented media is convenient, inexpensive and efficient. However, the U.S. Food and Drug Administration has stated that (a) selection markers for reagents suitable for human use must have the least impact on human health and should not confer antibiotic resistance to symbiotic bacteria, (b) selection by using drugs within the penicillin family could induce adverse reactions in those allergic to these drugs in the event that drug residue remained in purified plasmid preparations, and (c) the TN903 gene, conferring resistance to kanamycin, an antibiotic sparingly used in humans, is acceptable for use in plasmids for clinical gene transfer protocols (U.S. Dept. of Health and Human Services, 1998). Despite these recommendations, the risk of introduction of antibiotic resistance is still significant in protocols that require high doses of plasmid.
Advances in recombinant DNA technology suggest that properties associated with viral vectors can be mimicked by elements that facilitate nuclear main-tenance and replication of plasmid constructs.
Site-Specific Integration. Viral integration in the hu-man chromosome is driven by site-specific recombi-nases (SSR) that recognize unique sites within pathogen and host genomes. SSRs derived from bacteriophage that do not require bacterial cofactors for stable integration have been used for gene transfer (Glover, 2006). The most efficient integrase, fC31, integrates at a number of sites in the human genome and has significant potential for treatment of recessive genetic disorders (Thyagarajan, 2001). Proximity of integration sites to tumor-suppressor genes and genes that regulate cell proliferation is currently unknown and must be investigated prior to clinical testing.
Extra-Chromosomal Replication and Stability. SV40 andEpstein–Barr virus (EBV) viruses replicate episo-mally in mammalian cells via Tag and EBV nuclear antigen 1 (EBNA1) proteins respectively. Inclusion of viral sequences with the corresponding viral protein that promotes segregation of viral episomes during cell division can markedly sustain the expression of a therapeutic transgene. Unfortunately, viral proteins like Tag and EBNA1 can disable retinoblastoma and p53 tumor suppres-sor pathways, making them unsuitable for human gene transfer. To address this problem, a plasmid in which the scaffold/matrix attachment region (S/ MAR) from the human-b-interferon gene cluster was generated (Glover, 2006). Mitotic stability of this plasmid is conferred by interaction of the S/MAR element with components of the nuclear matrix. This plasmid has been maintained for over 100 divisions in CHO and HeLa cells (Piechaczek, 1999). The safety and stability of S/MAR-based episomal-based plasmids in vivo has not been established. Extensive testing must be completed before they can be considered for human use.
Human Artificial Chromosomes. Chromosomes repli-cate and maintain a low, defined copy number within host cells and accommodate many large genes and regulatory elements. Human artificial chromosomes (HACs) are susceptible to recombination and often integrate into and disrupt the host genome in an unpredictable manner. However, use of HACs to provide long-term expression of HPRT and CFTR transgenes has been described (Grimes, 2001; Auriche, 2002). The size of HACs prevents them from being packaged in viral vectors. Non-viral delivery methods for HACs have not been successful in vivo. Thus, even though elegant recombinant DNA molecules for gene transfer have been constructed, therapeutic efficacy relies on efficient methods of gene delivery.