In a previous study, we demonstrated that transgenic Lotus plants producing opines (which are small amino acid and sugar conjugates) specifically favor growth of opine-degrading rhizobacteria. The opine-induced bias was repeated and demonstrated with another soil type and another plant species (Solanum nigrum). This phenomenon is therefore independent of both soil type and plant species.The use of microorganisms as biopesticides or plant growth enhancers is an attractive alternative to the use of chemical pesticides and fertilizers (3,10,12,35,37). However, introduction of plant-growth-promoting bacteria in open fields often fails. This is attributed to limited survival of the inoculant strain in the rhizosphere, where it faces competition from resident microorganisms, a diverse community well adapted to the biological and physicochemical properties of the plant-soil interface (37). It is therefore crucial to develop methods to extend the fitness and persistence of the inoculant microorganisms, possibly by introducing a bias in the competition that benefits the isolate inoculated (20). This bias may be generated by addition to the soil, or release by the plant, of one or more substrates utilizable only by the introduced strain. This approach has been successfully used to sustain growth of various microbes in soil (1, 2, 5). Similarly, plants engineered to produce bacterial growth substrates have been shown to specifically select populations of microbes utilizing these substrates in the rhizospheres of Lotus (8, 21) and tobacco plants (31). Most often, these growth substrates have been opines (4), a family of compounds derived from amino acids and/or sugars and specifically detected in the crown gall tumors and hairy root formations induced by members of the genus Agrobacterium (4).Bacterial populations are highly dependent upon soil type (13,14,24,23,32) and plant exudates (7,15,17,38). Therefore, there is a risk that a selective microbial substrate strategy might be successful for a single soil type or a single plant species or cultivar. The work described here was aimed at determining whether the impact of opine production on soil bacteria is independent of the type of opines produced by the plant, the origin of the soil, and the plant species producing the opines. Such investigations are crucial to evaluate whether opine-producing plants and biased rhizosphere strategies could be used to engineer plant-microbe interactions under various conditions.To address the questions above, plants of the legume Lotus corniculatus cv. Rodéo and Solanum nigrum plants were engineered via Agrobacterium rhizogenes transformation to produce opines, as described by Petit et al. (26). The transformed plants produced the opines mannopine, mannopine and nopaline, or mannopine and octopine; the latter opine has not been tested previously (Fig. 1) (for a review, see reference 4). Transformed control plants harboring the pRi oncogenes but producing no opines (ONC plants) were also generated by using the same procedure (26). Plants we...