UC Davis

Bacterial pigments are secondary metabolites that in addition to color have several useful properties including antimicrobial activities. Due to their promising applications, the discovery of new bacteria and new bacterial pigments is in the forefront of scientific research and economics.Food spoilage can be a fruitful area of new bacterial pigment discovery. In many cases spoilage characteristics such as color defects can be correlated to the presence of microorganisms. Color defects are not easily associated with enzymatic functions or metabolic pathways, however, combining genomics, phylogenomics, and metabolomics to spoilage analysis may help researchers understand food spoilage and possibly discover safe, new, bacterially produced colorants. Pigmented bacteria can also be isolated from natural environments, especially freshwater and marine habitats and should be studied in greater detail because out of all the secondary metabolites having antibiotic activity, pigments and associated bacteria are an understudied group. The violet pigment violacein is gaining attention from the scientific community because of its various antibiotic activities which can be used for human and animal health, and its brilliant color which can be used for industrial purposes. My dissertation consisted of three separate studies, each investigating aspects of bacterial taxonomy and pigments. I first investigated the pigmented bacterium Pseudomonas carnis, a contaminant of soymilk and tofu that can cause a brilliant blue color defect in refrigerated food substrates. The investigation of the blue coloration of soy foods involved taxonomic classification of P. carnis using whole genome comparative phylogenomics, and genomic analysis of pigment associated genes. The P. carnis strains investigated in my study fluoresced under ultraviolet light, and contain genes for homogentisate based pyomelanin synthesis, a brown pigment. The strains also contain at least two homologs of each of the trpABCDF genes. Tryptophan is an important amino acid in the biosynthesis of many secondary metabolites and pigments, and in this case, two copies of tryptophan biosynthesis genes trpABCDF are hypothesized to be needed to produce pigments in this and related strains. In my second study, nine new strains of bacteria were isolated from the environment, including a purple Janthinobacterium sp. strain which was observed growing on tofu. Of the nine total strains characterized, six showed purple growth and genomic analysis revealed that all of the five genes (vioABCDEF) needed to produce violacein were present in each of these genomes. The remaining three strains are closely related to violacein producers, but they do not contain all five violacein genes nor did they exhibit pigmented colonies. I used a multifaceted approach involving phylogenomic and genomic methods to assign strains to existing species and describe five new species of bacteria. Current classification methods include whole genome sequence analysis using average nucleotide identity and phylogenomic methods using established gene marker sets. I used all of the previously described methods with additional genomic population ecology approaches to maximize confidence in taxonomic placement of these novel strains and species which contributed to a better understanding of evolutionary relationships between pigmented bacteria and how pigment genes such as for violacein are transferred and retained between and within populations. My final study focused on creating a model for predicting violacein production in cultured or uncultured bacteria and used phylogenomic methods. I investigated how horizontal gene transfer of violacein genes has contributed to the pigment’s distribution throughout the bacterial tree of life taking into consideration that most violacein-producing bacteria are members of the Gammaproteobacteria phylum. Although most bacteria that produce violacein contain the entire operon, we found that some bacteria outside of the Gammaproteobacteria phylum contain only a subset of the five necessary genes in their genomes. These include members of the Cyanobacteria, Actinobacteria, and Deltaproteobacteria phyla. Myxococcus stipitatus, a Deltaproteobacteria, is missing the vioD gene from the operon and putatively produces deoxyviolacein. The others probably do not produce the pigment but may have co-opted the retained genes for other purposes. We predicted that a number of Gammaproteobacteria that are unknown to produce the pigment are capable of production because they contain all five genes. Using the hidden Markov model built for this investigation I discovered violacein production capability in unknown whole genomes. This model could be useful for future research where production of this pigment can be inferred from any bacterial whole genome without the need for culturing and extraction. My dissertation has helped lay a foundation for future studies on the evolution of pigment genes throughout bacterial populations and highlights the need for further studies focused on the potential functional importance of understudied secondary metabolites and the taxonomic classification of the bacteria that produce them.