UC Berkeley

Perchlorate (ClO4-) is a soluble anion that occurs naturally in small concentrations, however, as a widely used oxidant in solid munitions, it has become a significant contaminant in ground water throughout the United States due to unregulated disposal of this compound prior to 1997. As a competitive inhibitor of iodine uptake in the thyroid gland, perchlorate ingestion can lead to lower thyroid hormone production, which is of particular concern for proper pre- and neonatal development. Recent reports have documented perchlorate in dairy and human breast milk, indicative of its movement to the top of the food chain. Current remediation of this compound usually involves ion exchange technologies, which although effective, simply concentrate the perchlorate out of the treated water into brine solutions. In contrast, many microorganisms are capable of respiring perchlorate, transforming it into harmless chloride. As a result, bioremediation has been identified as the most effective means of contaminant removal and degradation, and many strategies have been developed to take advantage of these dissimilatory perchlorate reducing bacteria (DPRB).

Traditional bioremediation strategies have been based on stimulating DPRB with cheap and easily available organic electron donors such as ethanol and acetate. While effective at stimulating perchlorate reduction, these compounds also stimulate considerable microbial growth, both of DPRB and non-target organisms. The excessive growth of organisms leads to biofouling, which can cause treatment failure and the stimulation of unwanted metabolisms such as iron and sulfate reduction resulting in the production of toxic and malodrous compounds. Further, addition of labile organics gives poor feedback control over bioremediation schemes and in the case of drinking water treatment, can contribute to downstream disinfection byproducts (DBPs).

To address these issues, an electrochemical system was investigated for stimulation of DPRB. A variety of electrochemical systems have been developed to stimulate microbial metabolism (Chapter 1), but none had been applied to perchlorate reduction. The system was attractive due to the ability to supply reducing equivalents for microorganisms to utilize in reducing perchlorate without adding carbon that would stimulate growth. In addition, the ability to alter both the available potential and current offered the possibility of tighter feedback control and thermodynamic targeting of perchlorate but not more electronegative electron acceptors.

Experiments to make use of cathodic electrodes as electron donors for perchlorate reduction were investigated (Chapter 2). Pure cultures of previously isolated DPRB were tested in the cathodic chamber of the bioelectrical reactor (BER) in cell suspensions utilizing anthraquinone-2,6-disulfonate (AQDS) as an electron shuttle. These experiments served as proof of concept, and demonstrated that organisms could successfully reduce perchorate in this manner. However, since these pure cultures could not survive in the BER under growth conditions, an enrichment was performed in the cathodic chamber to isolate organisms that would function for extended periods. Two novel DPRB were isolated from this enrichment, and one, strain VDY, was tested further. Strain VDY was capable of reducing perchlorate in the cathodic chamber both with and without AQDS as a shuttle. As a result, the organism was used to inoculate up-flow BERs that were designed for continous treatment of perchlorate. These reactors functioned effectively under a variety of perchlorate concentrations, both with and without the shuttle AQDS, and did not suffer from biofouling.

The other organism, strain MP, isolated from the BER enrichment was of unique phylogenetic affiliation for DPRB (Chapter 3). Most DPRB have been isolated from two genera within the Betaproteobacteria- the Dechloromonas and the Azospira. Strain MP, in contrast, was most closely related to another undescribed DPRB, strain CR, and during phylogenetic characterization of these and one other strain, LT-1, all were identified as members of novel clades with no previously known DPRB. Strains MP and CR were members of the genus Propionivibrio, which previously consisted of obligate fermentative organisms. Strain LT-1 constituted a new genus in the Rhodocyclaceae, named Dechlorobacter, which was most closely related to the Azonexus, a genus also with no known DPRB.

Strain VDY was physiologically unique compared to other DPRB because it could survive and function in the cathodic chamber of the BER under growth conditions. As a result, the organism was fully characterized physiologically and phylogenetically (Chapter 4) in an attempt to better understand the basis for this success. VDY was most closely related to another DPRB in the Alphaproteobacteria, strain WD. Together these organisms make up a perchlorate-reducing, non-magnetosome forming clade within the genus MagnetospirillummamI or mamL genes, necessary for magnetosome formation, and was unable to form magnetosomes under the physiological conditions tested. The ability of strain VDY to exist in the BER without an added carbon source was most likely due to autotrophic carbon fixation. This was supported by the demonstrated presence of the RuBisCO cbbM gene, which was expressed under autotrophic growth on hydrogen, but not during heterotrophic growth on acetate.

Since VDY was capable of reducing perchlorate in the BER without an added mediator, it was hypothesized that the organism might be able to utilize other inorganic electron donors, including reduced iron coupled to perchlorate reduction. VDY was tested for its ability to oxidize iron(II) in the form of FeCl2 (Chapter 5). Washed cell suspensions were capable of oxidizing iron coupled to both perchorate and nitrate. However, Fe(II) oxidation was not stoichiometrically balanced when the cells were reducing perchlorate, and indicated the presence of stored reducing equivalents, even when the culture was pre-grown in electron donor-limited conditions. The organism was capable of oxidizing iron(II) under growth conditions, but could not couple this metabolism to growth. Cells without any added electron donor reduced an equivalent amount of perchlorate as those cultured with iron(II), further demonstrating that the observed Fe(II) oxidation was not coupled to perchlorate reduction. Iron toxicity tests showed that in the presence of either ferrous or ferric iron, growth of VDY on acetate and perchlorate was inhibited, although the effect of ferrous iron was more deleterious. As a result, it was determined that while strain VDY is capable of concomitant iron(II) oxidation during perchlorate reduction, this process is not metabolically coupled and growth is in fact prevented by the presence of iron(II).

The work within this dissertation describes the successful development of a new reactor system for the treatment of perchlorate-contaminated influents and the complete characterization of two novel organisms isolated from original enrichments with that system. Further investigation of perchlorate-dependent iron(II) oxidation was also carried out in strain VDY. As a result of these studies, a new bioelectrical option now exists for perchlorate remediation strategies, two new DPRB have been fully characterized and described, and the known phylogenetic diversity of DPRB within the Betaproteobacteria has been significantly expanded.