MONITORING OF GENETICALLY MODIFIED MICROBES IN SOIL OR ON PLANT SURFACES.

Janet Jansson, Department of Biochemistry, Arrhenius Laboratories, Stockholm University,

S-10691 Stockholm, Sweden.

Microorganisms are used for an increasing number of environmental applications, including bioremediation and plant growth enhancement.To assess the efficacy of these biotechnology products in nature, as well as to determine their behaviour, pattern of distribution and mode of action, it is necessary to have specific and sensitive monitoring methods.Traditional monitoring techniques generally lack the specificity required for monitoring of a specific bacterial population in environmental samples, such as soil or plant surfaces. Therefore, novel molecular-based techniques have recently been developed to identify and quantitate populations of specific microorganisms in nature. One of the most promising methods relies on tagging of the target microbes with a biomarker that endows the cells with a unique phenotype that can easily be detected, such as the gfp gene, encoding the green fluorescent protein (GFP). An advantage of GFP is that, unlike other biomarkers, it does not require any substrate or additional cofactors in order to fluoresce. Also, single cells with a chromosomal integration of gfp can easily be identified by epifluorescence microscopy or confocal microscopy. In addition, GFP-tagged cells can easily be quantitated by flow cytometry.

We developed a gfp cassette that was optimized for expression of gfp in a variety of bacteria, including gram negative and gram positive strains. This cassette was chromosomally integrated in the cells based on a mini-Tn5 vector system. In some cells, two copies of gfp were integrated into the chromosome for enhanced fluorescence intensity. The tagged cells could be monitored under growth or starvation conditions, demonstrating the utility of GFP as a biomarker. Although GFP was ideal for enumeration of total cell numbers, it was not possible to ascertain whether the GFP-tagged cells were metabolically active. Therefore, we constructed some vectors that also contained the luxAB genes (encoding bacterial luciferase) or the luc gene (encoding eukaryotic luciferase), since bioluminescence in known to be dependent on cellular energy status and only metabolically active cells emit light. A dual marker system with both gfp and luxAB under control of the same consitutive promoter allowed simultaneous quantitation of the total cell number and metabolic activity, respectively, of specific marked populations.

The monitoring strategies we developed were used to quantitate a 4-chlorophenol-degrading strain, Arthrobacter sp. A-6, in 4-chlorophenol contaminated soil. Arthrobacter sp. A-6 tolerates high substrate loads of 4-chlorophenol (>350 ppm) and appears to degrade the compound by an unusual catabolic route. Arthrobacter sp. A-6 cells tagged with the luc gene were quantitated in soil by luminometry, whereas gfp-tagged cells were quantitated by flow cytometry. The cells were extracted from soil by density gradient centrifugation before analysis. This combination of approaches enabled us to relate the population density and metabolic activity of Arthrobacter sp. A-6 to the disappearance of 4-chlorophenol in the contaminated soil.

In addition we monitored the pattern of plant colonization of specific bacteria used for biocontrol of plant diseases. The bacteria could easily be distinguished on plant roots and seeds as single GFP-fluorescing cells by epifluorescence microscopy. Some internal colonization of plant seeds was detected by confocal microscopy. Strains tagged with the dual gfp/luxAB marker system were monitored by confocal microscopy (GFP fluorescence) and by CCD imaging (luminescence) directly on plant surfaces to determine factors influencing activity of the cells on the plant. In addition, the dual tagged strains were monitored in soil by flow cytometry and luminometry. The GFP fluorescent cell counts remained at a high level even during long term incubation in soil. By contrast, the luciferase activity decreased during incubation in soil, indicating a decrease in metabolic activity of the population under these nutrient limited conditions.

In conclusion, we have demonstrated that it is possible to monitor specific bacteria, in situ, in environmental samples in order to gain knowledge about their fate and behaviour in nature.