Identification of functional lox sites in the plastid genome.

Our objective was to test whether or not cyclization recombination (CRE), the P1 phage site-specific recombinase, induces genome rearrangements in plastids. Testing was carried out in tobacco plants in which a DNA sequence, located between two inversely oriented locus of X-over of P1 (loxP) sites, underwent repeated cycles of inversions as a means of monitoring CRE activity. We report here that CRE mediates deletions between loxP sites and plastid DNA sequences in the 3'rps12 gene leader (lox-rps12) or in the psbA promoter core (lox-psbA). We also observed deletions between two directly oriented lox-psbA sites, but not between lox-rps12 sites. Deletion via duplicated rRNA operon promoter (Prrn) sequences was also frequent in CRE-active plants. However, CRE-mediated recombination is probably not directly involved, as no recombination junction between loxP and Prrn could be observed. Tobacco plants carrying deleted genomes as a minor fraction of the plastid genome population were fertile and phenotypically normal, suggesting that the absence of deleted genome segments was compensated by gene expression from wild-type copies. The deleted plastid genomes disappeared in the seed progeny lacking CRE. Observed plastid genome rearrangements are specific to engineered plastid genomes, which contain at least one loxP site or duplicated psbA promoter sequences. The wild-type plastid genome is expected to be stable, even if CRE is present in the plastid.

A novel approach to plastid transformation utilizes the phiC31 phage integrase.

Thus far plastid transformation in higher plants has been based on incorporation of foreign DNA in the plastid genome by the plastid's homologous recombination machinery. We report here an alternative approach that relies on integration of foreign DNA by the phiC31 phage site-specific integrase (INT) mediating recombination between bacterial and phage attachment sites (attB and attP, respectively). Plastid transformation by the new approach depends on the availability of a recipient line in which an attB site has been incorporated in the plastid genome by homologous recombination. Plastid transformation involves insertion of an attP vector into the attB site by INT and selection of transplastomic clones by selection for antibiotic resistance carried in the attP plastid vector. INT function was provided by either expression from a nuclear gene, which encoded a plastid-targeted INT, or expressing INT transiently from a non-integrating plasmid in plastids. Transformation was successful with both approaches using attP vectors with kanamycin resistance or spectinomycin resistance as the selective marker. Transformation efficiency in some of the stable nuclear INT lines was as high as 17 independently transformed lines per bombarded sample. As this system does not rely on the plastid's homologous recombination machinery, we expect that INT-based vectors will make plastid transformation a routine in species in which homologous recombination rarely yields transplastomic clones.