Gene transfer to plants by diverse species of bacteria: an open-source platform for plant biotechnology

Since the discovery in the 1970s that Agrobacterium tumefaciens is capable of transferring genes to plants, it has become the most important tool in plant biotechnology. It has also been widely considered that Agrobacterium is the only bacterial genus with this capacity. Here we show that several other genera of bacteria can be modified to mediate gene transfer to diverse plant species. Three bacterial species from three genera and two families, Rhizobium sp. NGR234, Sinorhizobium meliloti and Mesorhizobium loti, were made competent for gene transfer by acquiring both a modified, disarmed Ti plasmid and a binary vector. Stable transformation of three plant species—tobacco, rice and Arabidopsis—was achieved using these non-Agrobacterium species. Transformation was performed with minor modifications of published protocols, using the leaf disk method for tobacco, scutellum-derived callus for rice and floral dip for Arabidopsis. The resulting plants expressed hygromycin resistance and glucoronidase (GUS) activity, contained 1–3 copies of the T-DNA as indicated by southern blot analysis, and showed the normal range of T-DNA insertions into host genomes determined by the polymerase chain reaction (PCR)–mediated sequencing of integration sites. To ensure that the gene transfer did not result from contamination with Agrobacterium cells, controls including species-specific PCR, selective plating, and use of a tagged binary vector were implemented. Thus, diverse plant-associated bacteria, when harbouring a disarmed Ti plasmid and binary vector (or presumably a co-integrate or whole Ti plasmid), are readily able to transfer T-DNA to plants. The Ti plasmid is self-transmissible, perhaps indicating the existence of a ubiquitous natural mechanism effecting horizontal gene transfer from bacteria to plants. If so, then much of the plant genomic DNA sequence thus far determined, which seems bacterial in origin, may have derived from such transfer over recent evolutionary time scales.

This alternative to Agrobacterium-mediated technology may provide two distinct advantages. First, it may lead to more effective plant transformation for recalcitrant species or cell types, through use of natural bacterium–plant interactions, for instance with benign epiphytic bacteria as vectors for gene transfer. Second, the cumbersome patent thicket that surrounds Agrobacterium gene transfer technology has rendered this tool largely unusable for commercial purposes by most companies, institutions and individuals. However, the patents in the thicket explicitly refer to and claim Agrobacterium. Our observations will likely lead to a comprehensive ‘work-around’ of these patents, as the species now capable of gene transfer are decidedly distinct from Agrobacterium. This new gene transfer technology will be freely available through a novel open-source BIOS licence that will require sharing both improvements to the core technology and relevant biosafety data. The technology will be improved and further developed as a BioForge project—a new distributive, Internet-mediated, cooperative research approach (Nature 431: 494, 2004; www.bios.net). It is envisaged that this technology and the sharing paradigm it reflects will help break the dominance of plant transformation by a few multinational companies and will forge a highly cooperative and public-spirited process of technology development. This step towards independence from corporate control of enabling technology may foster diverse applications of biotechnology to emerge with a focus on public good and low-margin priorities typical of the needs of developing world agriculture, production environments and economies.