endogenous provirus in all cells of a transgenic animal provides the potential for generating pathogenic recombinant viruses by straightforward and well-understood mechanisms. Such concerns are particularly acute in chickens and pigs, where infectious proviruses very similar in sequence to those used for vectors are known to be present (Boeke and Stoye, 1997). In mice, there is a well-studied model in which recombination between benign endogenous proviruses or endogenous proviruses, and infecting viruses early in the life of the animal, can cause a high incidence of lymphoma (nearly 100 percent in some mouse strains) 6 months later (Stoye et al., 1991; Rosenberg and Jolicoer, 1997). Given this example, it is reasonable to expect that viruses of much greater pathogenicity are likely to arise in an animal when there is a possibility of recombination between vector and endogenous viral sequences.

Similar concerns arise with the use of vectors based on lentiviruses for the introduction of genes (see Chapter 2). Recombination of lentiviruses in circulation in domestic animal populations, such as Feline Immunodeficiency Virus (FIV) in cats and Bovine Immunodeficiency Virus (BIV) in cattle, with vectors based on Human Immunodeficiency Virus (HIV) is improbable due to the large genetic distance between them. However, vectors based on FIV and BIV are being developed (Curran et al., 2000; Berkowitz et al., 2001), and their use to introduce transgenes into the corresponding species would significantly increase the probability of generating more pathogenic recombinants.