DEVELOPMENT AND IMPLEMENTATION OF BACTERIAL ISOLATES EXPRESSING BLUE AND GREEN FLUORESCING PROTEINS (BFP AND GFP) AS ALTERNATIVES TO CONVENTIONAL DNA- AND PROTEIN-SPECIFIC FLUORESCENCE STAINING FOR STUDYING BACTERIAL TRANSPORT BEHAVIOR IN GRANULAR MEDIA

Studies involving subsurface bacterial transport often employ populations stained with DNA-specific fluorescent dyes such as DAPI (4'6-diamidino-2-phenylindole). Although useful in tracking bacterial migration through subsurface media, the fluorochromes can affect surface charge and (or) impair microbial functions such as chemotaxis. Other methods such as ferromagnetic capture or PCR avoid these problems, but can be expensive and labor intensive. Therefore, insertion of bfp and gfp genes that code for production of blue (BFP) and green fluorescing proteins (GFP), respectively, were investigated as alternative labeling methods for future field-scale bacterial transport studies.

Three microbial isolates were employed. A motile Pseudomonas stutzeri isolated from a fractured-granite aquifer (Mirror Lake, NH) and a non-motile, mutant strain of the aforementioned isolate were transformed with plasmids containing gfp and bfp genes and ampicillin resistance. In addition, a chemotactic strain of Escherichia coli was transformed using a gfp-plasmid.

Results from survival experiments indicated that BFP-expressing P. stutzeri have decay rates about an order of magnitude higher than GFP-expressing E. coli. Flow-through column experiments (2.5 cm x 15 cm) involving sandy [0.59 mm, median grain size] aquifer sediments (3.5% Fe content) and a flow rate of 0.4 m/day were performed using DAPI, CFDA-SE (carboxyfluorescein diacetate succinimidyl ester, a protein-specific fluorescent stain) and BFP-labeled P. stutzeri. Fractional recoveries in the column effluent were 7.7%, 37.8%, and 44.1%, respectively. In another study employing similar grain size and flow rate, about 30% of introduced GFP-labeled E. coli were recovered.

These experiments suggest that bacterial isolates expressing fluorescent labels provide a useful alternative to use of DNA-specific fluorochrome dyes or to more laborious molecular-based labeling techniques. The fluorescent proteins appear to be stable for a number of weeks since transformed bacteria were still detectable after nearly a month in survival experiments. This suggests wide applicability for field-scale injection and recovery tests.