Abstract
Genetically altered or tagged Vibrio fischeri strains can be observed in association with their mutualistic host Euprymna scolopes, providing powerful experimental approaches for studying this symbiosis. Two limitations to such in situ analyses are the lack of suitably stable plasmids and the need for a fluorescent tag that can be used in tandem with green fluorescent protein (GFP). Vectors previously used in V. fischeri contain the p15A replication origin; however, we found that this replicon is not stable during growth in the host and is retained by fewer than 20% of symbionts within a day after infection. In contrast, derivatives of V. fischeri plasmid pES213 were retained by approximately 99% of symbionts even 3 days after infection. We therefore constructed pES213-derived shuttle vectors with a variety of selectable and visual markers. To include a visual tag that can be used in conjunction with GFP, we compared seven variants of the DsRed2 red fluorescent protein (RFP): mRFP1, tdimer2(12), DsRed.T3, DsRed.T4, DsRed.M1, DsRed.T3_S4T, and DsRed.T3(DNT). The last variant was brightest, displaying >20-fold more fluorescence than DsRed2 in V. fischeri. RFP expression did not detectably affect the fitness of V. fischeri, and cells were readily visualized in combination with GFP-expressing cells in mixed infections. Interestingly, even when inocula were dense enough that most E. scolopes hatchlings were infected by two strains, there was little mixing of the strains in the light organ crypts. We also used constitutive RFP in combination with the luxICDABEG promoter driving expression of GFP to visualize the spatial and temporal induction of this bioluminescence operon during symbiotic infection. Our results demonstrate the utility of pES213-based vectors and RFP for in situ experimental approaches in studies of the V. fischeri-E. scolopes symbiosis.