A strain of Pseudomonas putida that exhibited seedling disease control on cotton was among 5000 strains examined for unique properties that could be used to selectively recover genetically engineered pseudomonads from environmental samples. One isolate (M-17) was found to produce a red halo around single colonies grown on media containing 10 mg/l crystal violet. This decolorization reaction was constitutively produced when the growth medium contained glucose and asparagine, but was inhibited by the substitution of ammonia, nitrate or urea for the amino acid nitrogen source. However, a different medium containing succinic acid and using ammonia as the sole nitrogen source was found to induce the reaction. Factors that did not affect the decolorization reaction included temperature (14-30° C) and pH (3-8). Cultures of M-17 grown in broth containing crystal violet were able to decolorize still broth solutions, but not when incubated on a shaker (150 rpm). Stationary cultures formed a red precipitate. Efforts to characterize the precipitate revealed that it was soluble in polar organic solvents and insoluble in non-polar organic solvents. Thin layer chromatography revealed the presence of six bands that possibly represented various demethylated forms of crystal violet. Chemically derived mutants (cry-) did not produce a red halo but a clear, colorless region surrounding bacterial growth was observed. A greenhouse study demonstrated that strain M-17 provided protection against fungal disease as shown by plant stands on cotton (67% stand) equivalent to the commercial fungicides (70% stand) and significantly improved over the unamended control (38% stand). / Master of Science

A major factor in the ability of Pseudomonas fluorescens Pf-5 to act as a biological control agent is its production of antibiotics, including pyoluteorin (PLT), 2,4-diacetylphloroglucinol (2,4-DAPG) and pyrrolnitrin (PRN). The data provided in this thesis demonstrate that the presence of any of these antibiotics in the extracellular milieu affects production of that same antibiotic, as well as others, by Pf-5. Amending the growth medium with antibiotics had multiple effects on secondary metabolism in Pf-5. i) PLT positively regulated its own production, ii) 2,4-DAPG positively regulated its own production. iii) PLT suppressed 2,4-DAPG production. iv) 2,4- DAPG inhibited PLT production. v) PLT suppressed transcription of a heterologous ferric-pyoverdine uptake gene. vi) PRN exerted a slight inhibitory effect on PLT gene transcription and production. PLT autoinduction by Pf-5 was extensively characterized, and was shown to require concentrations of exogenous PLT in the nanomolar range. These low concentrations are comparable to those of many molecules proposed to function in signaling roles. PLT served as a signal between distinct populations of cells within the rhizosphere, where it prompted autoinduction by those cells. Aside from effects of Pf- 5 antibiotics on one another, I also described the positive effect of exogenous PLT on expression of a set of transport genes flanking the PLT biosynthetic gene cluster. Sequence data and experimental evidence suggests that these genes encode a transport apparatus for PLT. The deduced amino acid sequences for four adjacent open reading frames together resemble Type I secretion apparatuses, which typically function in transport of proteins rather than secondary metabolites. The intact transporter genes are necessary for optimal PLT production. Taken together, the data from the studies described herein demonstrate that i) the production of PLT by Pf-5 can affect the production of PLT by neighboring cells, and ii) PLT and other exogenous secondary metabolites have both autoregulatory and cross-regulatory effects in culture. Because Pf-5 derivatives engaged in PLT crossfeeding in the rhizosphere, it is likely that cross-feeding occurs for other secondary metabolites as well. Thus, production of an antibiotic by one cell can profoundly affect secondary metabolism in neighboring cells occupying natural habitats. / Graduation date: 2004

Pseudomonas is a diverse genus of Gram-negative bacteria that includes pathogens of plants, insects, and humans as well as environmental strains with no known pathogenicity. Pseudomonas fluorescens itself encompasses a heterologous group of bacteria that are prevalent in soil and on foliar and root surfaces of plants. Some strains of P. fluorescens suppress plant diseases and the genomic sequences of many biological control strains are now available. I used a combination of bioinformatic and phylogenetic analyses along with mutagenesis and biological assays to identify and compare the TonB-dependent outer-membrane proteins (TBDPs) of ten plant-associated strains of P. fluorescens and related species. TBDPs are common in Gram-negative bacteria, functioning in the uptake of ferric-siderophore complexes and other substrates into the cell. I identified 14 to 45 TBDRs in each strain of P. fluorescens or P. chlororaphis. Collectively, the ten strains have 317 TBDPs, which were grouped into 84 types based upon sequence similarity and phylogeny. As many as 13 TBDPs are unique to a single strain and some show evidence of horizontal gene transfer. Putative functions in the uptake of diverse groups of microbial siderophores, sulfur-esters, and other substrates were assigned to 28 of these TBDP types based on similarity to characterized orthologs from other Pseudomonas species. Redundancy of TBDP function was evident in certain strains of P. fluorescens, especially Pf-5, which has three TBDPs for ferrichrome/ferrioxamine uptake, two for ferric-citrate uptake and three for heme uptake. Five TBDP types are present in all ten strains, and putative functions in heme, ferrichrome, cobalamin, and copper/zinc uptake were assigned to four of the conserved TBDPs. The fluorescent pseudomonads are characterized by the production of pyoverdine siderophores, which are responsible for the diffusible UV fluorescence of these bacteria. Each of the ten plant-associated strains of P. fluorescens or P. chlororaphis has three to six TBDPs with putative roles in ferric-pyoverdine uptake (Fpv). To confirm the roles of the six Fpv outer membrane proteins in P. fluorescens Pf-5, I introduced deletions into each of the six fpv genes in this strain and evaluated the mutants and the parental strain for heterologous pyoverdine uptake. I identified at least one ferric-pyoverdine that was taken up by each of the six Fpv outer-membrane proteins of Pf-5. By comparing the ferric-pyoverdine uptake assay results to a phylogenetic analysis of the Fpv outer-membrane proteins, I observed that phylogenetically-related Fpv outer-membrane proteins take up structurally-related pyoverdines. I then expanded the phylogenetic analysis to include nine other strains within the P. fluorescens group, and identified five additional types of Fpv outer-membrane proteins. Using the characterized Fpv outer-membrane proteins of Pf-5 as a reference, pyoverdine substrates were predicted for many of the Fpv outer-membrane proteins in the nine other strains. Redundancy of Fpv function was evident in Pf-5, as some pyoverdines were recognized by more than one Fpv. It is apparent that heterologous pyoverdine recognition is a conserved feature, giving these ten strains flexibility in acquiring iron from the environment. Overall, the TBDPs of the P. fluorescens group are a functionally diverse set of structurally-related proteins present in high numbers in many strains. While putative functions have been assigned to a subset of the proteins, the functions of most TBDPs remain unknown, providing targets for further investigations into nutrient uptake by P. fluorescens spp.. The work presented here provides a template for future studies using a combination of bioinformatic, phylogenetic, and molecular genetic approaches to predict and analyze the function of these TBDPs. / Graduation date: 2012