Phyllosphere-Mediated Bacterial Degradation of Airborne Phenol Amarjyoti Sandhu Iowa State University

Iowa State University Capstones, Theses and Retrospective Theses and Dissertations Dissertations

2008 Phyllosphere-mediated bacterial degradation of airborne phenol Amarjyoti Sandhu Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd Part of the Microbiology Commons

Recommended Citation Sandhu, Amarjyoti, "Phyllosphere-mediated bacterial degradation of airborne phenol" (2008). Retrospective Theses and Dissertations. 15878. https://lib.dr.iastate.edu/rtd/15878

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. Phyllosphere-mediated bacterial degradation of airborne phenol

Amarjyoti Sandhu

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Major: Microbiology

Program of Study Committee: Gwyn A. Beattie, Major Professor Larry J. Halverson Joel R. Coats Martin H. Spalding Thomas B. Moorman

Iowa State University

Ames, Iowa

2008

UMI Microform 3307046 Copyright 2008 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, MI 48106-1346 ii

Table of Contents

Acknowledgements ...... iii Abstract...... iv General Introduction...... 1 Dissertation organization...... 2 Chapter 1. Literature Review ...... 3 Plant uptake of pollutants and pollutant availability to leaf-associated bacteria...... 3 Phenol - a model organic pollutant...... 5 Phenol degradation and genetic characterization of phenol-degrading bacteria ...... 6 Bacterial responses to phenol...... 7 References ...... 11 Chapter 2. Bacterial degradation of airborne phenol in the phyllosphere...... 16 Abstract...... 16 Introduction...... 17 Material and Methods...... 19 Results...... 23 Discussion...... 31 Acknowledgements...... 34 References ...... 35 Chapter 3. Identification and genetic characterization of phenol-degrading bacteria from leaf microbial communities...... 39 Introduction...... 39 Material and Methods...... 41 Results...... 46 Discussion...... 52 References ...... 58 Chapter 4. Effect of phenol on Pseudomonas syringae as determined by a transcriptome analysis of the response of P. syringae to phenol and other abiotic stresses...... 63 Introduction...... 63 Material and Methods...... 66 Results and Discussion ...... 69 References ...... 95 Chapter 5. Conclusion and future directions ...... 100 References ...... 104

iii

Acknowledgements

I am grateful to my major professor, Dr. Gwyn A. Beattie, for her guidance, patience, sound advice, and tremendous support throughout my Ph.D. Her scientific creativity provided me with an opportunity to improve my research skills. I would also like to thank my committee members, Dr. Larry J. Halverson, Dr. Martin Spalding, Dr. Thomas B. Moorman, and Dr. Joel Coats for their helpful suggestions and comments during meetings.

I thank all the people I got to know during my Ph.D for their immense moral and professional support. Thanks to all my past and present lab people, Jan Siebel, Dr. Chiliang Chen, Dr. Catherine Wright, Dr. Siva Sabaratnam, and Kelly Peterson, for their help and support whenever I needed. I thank Jan for being patient while she taught me to use Gas Chromatography, and other techniques I used in my research. Thanks to all the undergrads who worked in my lab for all their help and making my life so much easier. A special thanks to Dr. Larry Halverson and his lab for letting me use their equipment and for their useful comments and suggestions in joint lab meetings. Thanks to Dr. Halverson for being so easy to talk to and being available for one-on-one discussions whenever I was stuck on some problem. I would also like to extend my thanks to Microbiology department staff, Simi, Ann, and Kay for always being willing to help. I am thankful to Troy Bowman for driving me to the swine production facility site for sampling and to Iowa Select Farm people to allow sampling at their site. I would also like to thank Dr. Dan Nettleton in helping me in microarray experimental design and analysis.

Finally, I would like to thank my brothers, Davinder and Baldeep and my parents, especially my mom, for her support and encouragement throughout. I dedicate this thesis to my mother; without her, I would not have been able to reach where I am now. I also thank my husband, Harvinder, for encouragement and support, and for taking care of our little daughter, Harjot, while I was writing my thesis. Thanks to Harjot for being a great motivation for me to finish up fast. iv

Abstract

Bacterial degradation of organic pollutants in soil has been studied extensively but little is known about the availability and degradation of pollutants by bacterial resident on leaf surfaces. Here, we examined availability and degradation of phenol on leaves by introduced cells of a known phenol degrader and by natural microbial communities. Phenol-responsive gfp-expressing Pseudomonas fluorescens A506 bioreporter cells detected airborne phenol as well as phenol that had been adsorbed by leaf surfaces. The degradation of airborne 14C- phenol by leaf-associated bacteria showed that Pseudomonas sp. strain CF600 released eight 14 times more CO2 than a non-degrading mutant after introduction onto leaves. This study provided the first direct evidence of degradation of an organic air pollutant by leaf-associated bacteria. We evaluated phenol degradation by natural microbial communities on green ash leaves that were collected from a field site rich in airborne organic pollutants. We found that significantly more phenol was mineralized by these leaves when the communities were present than by these leaves following surface sterilization. Thus, phenol-degrading organisms were present in these natural communities and were metabolically capable of phenol degradation. We isolated and identified phenol-degrading bacterial isolates within these natural microbial communities. Ten phenol-degrading bacterial isolates were obtained that were most closely related to the genera Alcaligenes, Acinetobacter and Rhodococcus. The genetic characterization of these isolates revealed little diversity as phenol hydroxylase (PH) enzymes of all of the Gram-negative isolates were type II PHs, which have a moderate affinity for phenol, and these isolates harbored the ortho pathway of phenol degradation. Lastly, we attempted to understand the mechanism of phenol toxicity and tolerance in bacteria by comparing the transcriptional profiles of cells exposed to phenol and cells exposed to agents causing membrane and oxidative stress using the leaf-associated bacterial species, Pseudomonas syringae pv. tomato. Phenol resulted in the induction of heat shock genes that were not induced by the other stresses tested. The role of these genes in cellular and membrane protein damage in response to phenol is discussed.

General Introduction

Bacterial availability and degradation of organic pollutants has been studied extensively in the rhizosphere. Microbial population sizes and their pollutant degradation activity are generally much higher in the rhizosphere than in bulk soil due to the release of oxygen and nutrients by plant roots. Pollutants are brought into the rhizosphere from the surrounding soil by the plant roots via the transpiration stream. Pollutants can be translocated to shoots via evapotranspiration where they may undergo various fates, such as degradation by the plant enzymes to less toxic metabolites, compartmentalization or localization into specific plant tissues, and volatilization back into the air. Volatile organic compounds (VOCs), such as toluene and phenol, can also be taken up directly from air by leaves or adsorbed onto the leaf surfaces. Plant leaves make up a vast surface area of the terrestrial habitat and are known to be colonized by bacterial populations as high as 107-108 cells per g of leaf. Therefore, leaf surfaces of terrestrial plants could provide an excellent venue for promoting microbial access to VOCs. In the present study, we evaluated the ability of bacteria to access and degrade airborne organic pollutants on plant leaves by using phenol as a model compound. This was done not only with phenol degraders that were introduced onto leaf surfaces but also with naturally occurring bacterial communities on leaves taken from a site rich in airborne organic compounds. In addition, we attempted to isolate, identify and genetically characterize phenol-degrading bacterial isolates within the natural leaf-associated microbial communities. Identification and molecular characterization of the catabolic genes of pollutant-degrading bacterial communities in contaminated environments is important to determine the major players in degradation as well as the functional and genetic diversity of the degradation pathways involved. This information can be helpful in developing effective bioremediation strategies as well as in better understanding the microbial ecology of pollutant-rich sites. Because effective bioremediation of pollutants by bacteria requires the survival and maintenance of large populations in pollutant-rich sites, it is important to understand the mechanisms of toxicity in bacteria as well as the mechanisms that bacteria employ to tolerate stresses imposed by these toxic compounds. Previous studies have shown that phenol affects bacterial cell membranes, disrupting their integrity and function. Phenol has also been 2

suggested to cause oxidative stress in some bacterial species due to the action of enzymes catalyzing its degradation. To obtain a comprehensive understanding of the effects of phenol on bacterial cells, and to determine the toxicity and tolerance mechanisms of bacteria to phenol, we performed a transcriptome analysis of bacteria exposed to phenol, membrane stress, and oxidative stress. A model for the action of phenol was developed based on the expression data obtained.

Dissertation organization The dissertation is divided into five chapters. The first chapter provides a review of the literature on the availability and degradation of airborne phenol by leaf-associated bacteria, and the mechanism of phenol toxicity and tolerance in bacteria. The second chapter presents studies on the availability and degradation of airborne phenol by leaf-associated bacteria and phenol degradation by naturally occurring microbial communities from leaves. This chapter was published in Environmental Microbiology and was highlighted in the research highlights section of Nature Reviews in Microbiology. The third chapter describes the isolation, identification, and phylogenetic and genetic characterization of phenol-utilizing bacterial communities from leaves taken from a site rich in airborne volatile organic compounds. The fourth chapter presents the transcriptional profiling of the leaf-associated bacterial species Pseudomonas syringae in response to phenol to understand the mechanism of tolerance to, and toxicity of, phenol in these bacteria. This was done by comparing the transcriptome profiles of bacterial cells exposed to phenol and cells exposed to membrane stress (caused by sodium chloride and polyethylene glycol), and oxidative stress (caused by hydrogen peroxide). The fifth chapter contains my general conclusions and my possible future directions. 3

Chapter 1. Literature Review

Bacterial degradation of pollutants has been extensively studied in the rhizosphere. It is known that the rhizosphere harbors higher bacterial populations than the bulk soil due to the nutrients and oxygen released by plant roots. These microbial populations have been shown to play a major role in the degradation of pollutants brought into the rhizosphere from surrounding soil by the transpiration stream. The pollutants present in the soil can also be taken up by roots and translocated via the transpiration stream to the leaves, where these pollutants can undergo various fates, such as localization into specific plant parts, transformation into less toxic metabolites by plant enzymes, or volatilization back into the air. Leaves can also take up pollutants directly from the air. Because leaf surfaces are often colonized by bacteria, with populations as high as 107-108 cells per gram, these leaf- associated bacteria may be able to access and degrade pollutants. The degradation of pollutants by leaf-associated bacteria remains unexplored. In the present study, we hypothesized that phyllosphere bacteria may have access to and degrade the pollutants taken up by leaves, and tested this using phenol, an organic pollutant, as a model compound. We also attempted to identify and genetically characterize the phyllosphere bacteria responsible for degradation of phenol on leaves and to understand the mechanism of toxicity of phenol in a bacterial cell.

Plant uptake of pollutants and pollutant availability to leaf-associated bacteria. Bacterial degradation of pollutants has been extensively studied in the rhizosphere (Kuiper et al., 2004). Roots not only provide a nutrient-rich habitat for microbes, but also release secondary metabolites that can induce microbial genes involved in pollutant degradation or can act as co-metabolites to facilitate microbial degradation (Gilbert and Crowley, 1997; Rentz et al., 2005). Plants can also mobilize and accumulate considerable amounts of pollutants from the surrounding soil (Harvey et al., 2001); these pollutants can exert a positive selection pressure for pollutant-degrading bacteria in the rhizosphere (Miya, 2000) or in endophytic sites in the roots or throughout the plant (Siciliano et al., 2001).

Plants can mediate the removal of organic pollutants from the environment not only by extracting pollutants from soil into the roots but also by translocating them to the shoots via the transpiration stream. For example, polyaromatic hydrocarbons such as pentachlorophenol have been shown to be taken up by a few plant species, including rice (Weiss et al., 1982), spinach and soybean (Casterline et al., 1985), and to be distributed in both roots and shoots after root uptake. A radiolabeled pesticide, bromacil, and two industrial waste compounds, phenol and nitrobenzene, were also shown to be taken up by soybean roots and translocated to the shoots when provided in hydroponic solution (McFarlane et al., 1987). Plants can take up organic compounds not only from soil through roots but also from air through leaves. The compounds can penetrate leaves in two ways: through stomata or through the cuticle of the epidermis. Stomatal uptake of gaseous pollutants is a major uptake pathway for small inorganic compounds such as sulfur dioxide, nitrogen oxides, and ozone (Slovik et al., 1996). Phaseolus vulgaris leaves exposed to gaseous ammonia in leaf chambers were also shown to take up ammonia via stomata whereas transport via the cuticle was negligible (Van Hove et al., 1987). Another study done to measure the uptake of a gaseous organic compound, 14C-toluene, by soybean leaves demonstrated that majority of the toluene provided was taken up by the stomatal pathway (Jen et al., 1995). The second route by which organic compounds penetrate leaves is the cuticle. The cuticle, being a lipophilic polymer membrane covering leaf surfaces, has been shown to take up or adsorb lipophilic compounds such as polychlorinated biphenyls (Barber et al., 2002) and polyaromatic hydrocarbons such as anthracene (Wild et al., 2004; Wild et al., 2005). Using two-photon excitation microscopy, Wild et al. (2004; Wild et al., 2005) could track the movement of anthracene, provided as an aqueous solution to intact maize leaves, through the epicuticular wax and plant cuticle and finally into the cell cytoplasm of maize leaves. These studies demonstrated that anthracene initially formed a thin band in the epicuticular wax, suggesting that cuticular uptake can result in the accumulation of organic compounds before diffusion into the epidermal cells. The other studies suggesting the sorption of organic compounds to the cuticle were performed with isolated plant cuticles. For example, Shafer and Schonherr (1985b) showed that organic compounds, such as phenol, 2-nitrophenol and 4-nitrophenol, sorbed onto the isolated cuticles of various plant species and suggested that this sorption was 5

due to hydrogen-bonding between the phenol and the free hydroxyl groups in the cutin polymer.

Leaves are colonized by a variety of microorganisms, including bacteria, yeasts and filamentous fungi, with bacteria being the most numerous (Hirano and Upper, 2000; Lindow and Brandl, 2003). Leaves often support bacterial populations as high as 107-108 cells per gram of leaf tissue (fresh weight). Like rhizosphere microbial communities, phyllosphere microbial communities are comprised of many distinct genera and species, with bacteria of the genus Pseudomonas widespread in both phyllosphere and rhizosphere communities (Legard et al., 1994; Benizri et al., 2005). Bacteria in this genus, along with those in related genera, exhibit a high level of metabolic diversity. Moreover, they are common among the bacteria known to degrade organic pollutants (Gulensoy and Alvarez, 1999; Whiteley et al., 2001), suggesting that pollutant degraders may be present on leaves as they are on roots. In addition, the majority of the leaf-associated bacteria are located on the leaf surface in depressions such as epidermal cell wall junctions and along the veins, although some, primarily pathogens, may establish large populations in the intercellular spaces. The bacteria present on leaf surfaces may, thus, have access to organic pollutants adsorbed by leaf surfaces. Some previous studies have suggested a role for leaf-associated bacteria in the removal of pollutants from air. For example, bacteria present on leaves of a common indoor plant have been shown to augment the removal of an airborne organic pollutant, toluene (De Kempeneer et al., 2004). Similarly, genetically modified endophytic Burkholderia cepacia reduced the amount of toluene volatilized from the aerial parts of plant by 50-70% after uptake from hydroponic solution containing toluene (Barac et al., 2004). These studies suggest that bacteria resident on or within the above ground regions of plants, i.e., the phyllosphere, may play an important role in removing organic pollutants from the environment.

Phenol - a model organic pollutant. In the present study, we evaluated the availability and degradation of airborne organic pollutants by leaf-associated bacteria using phenol as a model compound. Phenol is a priority 6

pollutant of the Environmental Protection Agency and is found in both indoor and outdoor air. It is a component of the emissions from wood, coal, and municipal solid waste combustion, as well as automobile exhaust and cigarette smoke (ATSDR, 1998). Moreover, it is an abundant pollutant generated by microbial communities during the anaerobic decomposition of livestock waste and is a major contributor to the odor intensity of animal wastes (Zahn, 2001). Moreover, several distinct phenol degradation pathways have been identified in Pseudomonas species (Feist and Hegeman, 1969; Herrmann et al., 1995; Whiteley et al., 2001), as well as in other bacterial species (Hughes et al., 1984; Arai et al., 1999; Duffner et al., 2000; Kim et al., 2002), supporting the possibility that phenol-degrading bacteria are widespread. Diverse phenol-degrading bacteria, such as Pseudomonas, Ralstonia, Burkholderia, Acinetobacter, Variovorax, Comamonas, and Klebsiella, have been isolated from the natural microbial communities of various ecosystems, including wastewater, rivers, soils and marine environments (van Schie and Young, 1998; Bastos et al., 2000; Abd-El-Haleem et al., 2002; Arutchelvan et al., 2005; Garcia et al., 2005; Geng et al., 2006).

Phenol degradation and genetic characterization of phenol-degrading bacteria. Aerobic bacterial phenol degradation occurs via two main pathways, ortho and meta. The first step that involves the hydroxylation of phenol to catechol is common for both the pathways and is catalyzed by phenol hydroxylase. Catechol is a common catabolic intermediate formed during the degradation of a variety of aromatic compounds, such as benzene, toluene, and naphthalene. Catechol is then metabolized via the ortho or meta pathway via the enzymes catechol 1,2-dioxygenase (C12O) and catechol 2,3-dioxygenase (C23O), respectively, resulting in intermediates used for energy generation via the tricarboxylic acid pathway, pyruvate metabolism and glycolysis. Bacteria can possess one or both of these pathways and in organisms with both, the expression of each pathway appears to be determined by the chemical nature of the aromatic substrate that the bacteria are grown with (Feist and Hegeman, 1969).

Genetic characterization of pollutant-degrading bacterial communities is important to identifying the structure and function of these communities in the environment. Phenol hydroxylase, the first enzyme in the phenol catabolic pathway, catalyzes the conversion of phenol into catechol, which is a rate-limiting step in phenol degradation. Phenol-degrading Gram-negative bacteria have been categorized into three groups based on the conserved amino acid residues of the large subunit of the phenol hydroxylase gene (LmPH). The three distinct kinetic groups of Gram-negative bacteria are groups I, II and III, which exhibit high, moderate and low affinity for phenol, respectively. Group-specific monitoring of phenol hydroxylase genes in contaminated soils has been shown to be important in predicting the catabolic potential of indigenous TCE-degrading bacterial populations (Futamata et al., 2001a; Watanabe et al., 2002) and thus could help in developing effective bioremediation strategies. No such grouping has been suggested for Gram-positive bacteria.

The type of phenol catabolic pathway present in bacteria can be determined by the presence of C12O, an enzyme for the ortho pathway, or C23O, an enzyme for the meta pathway. In previous studies, catechol dioxygenase genes have been used as targets to determine the composition and behavior of microbial communities present in a polluted environment (Sei et al., 2004). Specifically, Sei et al (2004) used these genes to monitor the behavior of aromatic compound-degrading bacterial populations and changes in microbial community structure in a sea water microcosm and found that bacterial populations possessing C12O genes were the primary degraders of phenol and benzoate and C23O- harboring bacteria dominated only under higher substrate concentrations. These studies demonstrate that genetically characterizing communities can identify the functionally relevant bacterial populations and indicate changes in microbial community structure in a contaminated environment.

Bacterial responses to phenol. Pseudomonas syringae was used as a model bacterial species for transcriptome studies for examining bacterial responses to phenol. P. syringae is both an epiphyte and a 8

pathogen colonizing a wide range of plant species. Enumeration of the P. syringae populations on healthy plant leaves has shown that 75 to 90% of the bacteria were epiphytic, i.e., residing on the leaf surfaces (Hirano and Upper, 2000). Epiphytic bacteria are exposed to various harsh and fluctuating environmental conditions such as variable temperatures, humidity, and UV radiation as well as toxic airborne pollutants. These conditions can adversely affect bacterial survival and their population sizes on leaf surfaces. Because effective removal of airborne pollutants from the environment by leaf-associated bacteria requires survival and maintenance of good population sizes, it is important to understand the mechanisms of toxicity and tolerance of epiphytic bacteria to the toxic pollutant compounds. Therefore, in the present study we used P. syringae as a model epiphytic bacterial species to study the effects of phenol on cellular physiology using transcriptome analysis.

Organic solvents, such as phenol, benzene, and toluene have been suggested to affect bacterial cells by partitioning into the cell membrane resulting in the disruption of membrane integrity and function. Most studies have suggested that bacterial cell membranes are modified in response to organic solvents. For example, the exposure of bacteria to phenol results in changes in cell membrane fatty acid composition by cis-to-trans isomerization of the fatty acids (Heipieper et al., 1992). It has been suggested that cis-to-trans conversion increases membrane ordering and consequently decreases the membrane fluidity (Sikkema et al., 1995). Most studies have examined the effect of phenol on cell membranes without considering its effects on cytoplasmic proteins or targets after entry into the cell. A few proteomic and transcriptome studies have shown that phenol induces heat shock proteins (Benndorf et al., 2001; Tam le et al., 2006). Heat shock proteins (HSPs) are the general stress proteins induced by a variety of stresses that denature proteins, including high temperature and toxic substances such as heavy metals, alcohols, and organic solvents. Denatured proteins in the form of incomplete peptides, unfolded and misfolded proteins, and protein aggregates, which are produced in response to stress exposure, induce the transcription of HSPs, suggesting that the HSPs play a crucial role in removing damaged proteins in the cell, thereby helping the cell adapt or survive under stress. Apart from the induction of heat shock proteins in bacteria in response to phenol, phenol has also been suggested to cause oxidative 9

stress. Proteomic analysis of Pseudomonas putida upon exposure to phenol revealed the upregulation of antioxidant genes, including catalases that convert H2O2 to H2O and minimize oxidative stress (Santos et al., 2004). Another study has suggested that the bacterial degradation of aromatic substrates, such as phenol and 4-chlorophenol by mono-and dioxygenase enzymes, also produces reactive oxygen species that cause oxidative stress (Tamburro et al., 2004). To generate a comprehensive picture of the effect of phenol on P. syringae, we compared the stress response of P. syringae cells to phenol, stressors that cause membrane damage (NaCl and PEG), and to an agent that causes oxidative stress (H2O2).

Oxidative stress response. Oxygen radicals resulting from aerobic growth, exposure to oxidative agents, such as

H2O2, and ionizing radiation cause oxidative stress in bacteria. Oxygen radicals, including superoxide, peroxide and hydroxyl radicals are known to damage cell macromolecules, such as DNA, lipids and amino acids by oxidation. This results in the induction of antioxidant genes, such as superoxide dismutases that convert superoxide into H2O2, and catalases that convert H2O2 to H2O and oxygen; these activities help prevent cell damage. Some antioxidant genes, including ahpC (alkyl hydroperoxide reductase, subunit C), sodB (superoxide dismutase), tpx (thiol peroxidase), dsbA (thiol:disulfide interchange protein) and tig (trigger factor presumably required for protein folding or accelerated protein export), have been shown to be induced by phenol in P. putida suggesting that phenol induces some oxidative stress in bacteria (Santos et al., 2004). The extent of oxidative stress caused by phenol, therefore, may be determined by comparing the genes induced by H2O2 and phenol in P. syringae.

The expression of many oxidative stress response systems is controlled by complex regulatory networks. One key regulator is OxyR, which senses the cytoplasmic redox state and is responsible for the activation of a number of genes involved in stress response networks in many bacteria. In E. coli, the OxyR transcription factor controls the expression of many H2O2-inducible genes, including hydroperoxidase I (catalase, katG), the 2 subunits of an alkyl hydroperoxide reductase (ahpCF), glutaredoxin 1(grxA), glutathione reductase 10

(gorA), the Fur repressor (fur), non-specific DNA binding protein (dps) and regulatory RNA

(oxyS). In addition to protecting against H2O2-induced stress, OxyR-regulated genes also confer resistance to hypochlorous acid, organic solvents, and reactive nitrogen species.

To obtain a comprehensive picture of the effects of phenol in bacteria and the mechanism of bacterial tolerance to phenol, transcriptome analysis was performed by exposing P. syringae pv. tomato to phenol and agents that cause membrane stress, including

NaCl and PEG, and oxidative stress, such as H2O2. There are numerous previous studies where transcriptome analysis has been used to evaluate bacterial responses to stress agents, including H2O2, phenol, catechol, and antibiotics (Zheng et al., 2001; Phadtare et al., 2002; Palma et al., 2004; Weber et al., 2005; Zeller et al., 2005; Tam le et al., 2006), and some genes that help bacteria survive under the stressful conditions have been identified. These studies have been effective at identifying the members of specific stress regulons, thus improving our understanding of both global and stress-specific regulatory networks that control the expression of these genes. Collectively these studies have provided important insights into bacterial stress physiology. In the present study, we have examined the mechanism of phenol toxicity in bacteria by evaluating the effects of phenol on genes influencing cell membranes and the cell cytosol, and the mechanism of phenol tolerance by evaluating the physiological role of the stress-responsive genes that were induced the most by phenol and were not induced by other stresses.

Water stress response. In natural environments, such as soil, bacteria can experience stress due to low water availability. Water stress can be either due to limited water activity because of the high solute concentrations in the environment in which a bacterial cell is present or the physical unavailability of water because of the interactions of water with a matrix, such as soil (Papendick and Campbell, 1980). The former type of water stress is referred to as osmotic stress and the latter as matric stress. Because the cytoplasmic membrane of bacteria is permeable to water but not to most other metabolites, hyper- or hypoosmotic shock causes an instantaneous efflux or influx of water, respectively, which is accompanied by a concomitant 11

decrease or increase in the cytoplasmic volume (Csonka, 1989). Proteins and other biological macromolecules have evolved to function only within certain normal ranges of water activities, outside of which some essential cellular function(s) becomes impaired. Sudden exposure to a hyperosmotic environment causes plasmolysis, i.e., cell shrinkage due to loss of water from the cell cytosol, resulting in the inhibition of a variety of physiological processes ranging from nutrient uptake (Roth et al., 1985a; Roth et al., 1985b; Rudd and Menzel, 1987) to DNA replication (Meury, 1988). Importantly, water stress has been shown to cause changes in the membrane by altering the amounts of trans and cis isomers of monounsaturated fatty acids (Halverson and Firestone, 2000) or by differential synthesis of outer membrane proteins (Lugtenberg et al., 1976; Epstein, 1983). Therefore, a comparison of the genes induced by NaCl and PEG to those induced by phenol may help determine the extent and nature of the membrane stress caused by phenol in bacteria.

References

Abd-El-Haleem, D., Moawad, H., Zaki, E.A., and Zaki, S. (2002) Molecular characterization of phenol-degrading bacteria isolated from different Egyptian ecosystems. Microb Ecol 43: 217-224. Arai, H., Akahira, S., Ohishi, T., and Kudo, T. (1999) Adaptation of Comamonas testosteroni TA441 to utilization of phenol by spontaneous mutation of the gene for a trans-acting factor. Mol Microbiol 33: 1132-1140. Arutchelvan, V., Kanakasabai, V., Nagarajan, S., and Muralikrishnan, V. (2005) Isolation and identification of novel high strength phenol degrading bacterial strains from phenol-formaldehyde resin manufacturing industrial wastewater. J Hazard Mater 127: 238-243. Agency for Toxic Substances and Disease Registry (ATSDR). (1998) Toxicological profile for phenol. Atlanta, GA: U. S. Department of Health and Human Services, Public Health Service. Barac, T., Taghavi, S., Borremans, B., Provoost, A., Oeyen, L., Colpaert, J. V. Vangronsveld, J., and van der Lelie, D. (2004) Engineered endophytic bacteria improve phytoremediation of water-soluble, volatile, organic pollutants. Nat Biotechnol 22: 583-588. Barber, J.L., Kurt, P.B., Thomas, G.O., Kerstiens, G., and Jones, K.C. (2002) Investigation into the importance of the stomatal pathway in the exchange of PCBs between air and plants. Environ Sci Technol 36: 4282-4287. Bastos, A.E., Moon, D.H., Rossi, A., Trevors, J.T., and Tsai, S.M. (2000) Salt-tolerant phenol-degrading microorganisms isolated from Amazonian soil samples. Arch Microbiol 174: 346-352. 12

Benizri, E., Piutti, S., Verger, S., Pages, L., Vercambre, G., Poessel, J.L., and Michelot, P. (2005) Replant diseases: Bacterial community structure and diversity in peach rhizosphere as determined by metabolic and genetic fingerprinting. Soil Biol Biochem 37: 1738-1746. Benndorf, D., Loffhagen, N., and Babel, W. (2001) Protein synthesis patterns in Acinetobacter calcoaceticus induced by phenol and catechol show specificities of responses to chemostress. FEMS Microbiol Lett 200: 247-252. Casterline, J.L., Jr., Barnett, N.M., and Ku, Y. (1985) Uptake, translocation, and transformation of pentachlorophenol in soybean and spinach plants. Environ Res 37: 101-118. Csonka, L.N. (1989) Physiological and genetic responses of bacteria to osmotic stress. Microbiol Rev 53: 121-147. De Kempeneer, L., Sercu, B., Vanbrabant, W., Van Langenhove, H., and Verstraete, W. (2004) Bioaugmentation of the phyllosphere for the removal of toluene from indoor air. Appl Microbiol Biotechnol 64: 284-288. Duffner, F.M., Kirchner, U., Bauer, M.P., and Muller, R. (2000) Phenol/cresol degradation by the thermophilic Bacillus thermoglucosidasius A7: cloning and sequence analysis of five genes involved in the pathway. Gene 256: 215-221. Epstein, W. (1983) Membrane-mediated regulation of gene expression in bacteria. In Gene function in prokaryotes. J. Beckwith, J. Davies, and Gallant, J.A. (eds). Cold Spring Harbor, N.Y: Cold Spring Harbor Laboratory, pp. 281-292. Feist, C.F., and Hegeman, G.D. (1969) Phenol and benzoate metabolism by Pseudomonas putida: regulation of tangential pathways. J Bacteriol 100: 869-877. Futamata, H., Harayama, S., and Watanabe, K. (2001) Group-specific monitoring of phenol hydroxylase genes for a functional assessment of phenol-stimulated trichloroethylene bioremediation. Appl Environ Microbiol 67: 4671-4677. Garcia, M.T., Ventosa, A., and Mellado, E. (2005) Catabolic versatility of aromatic compound-degrading halophilic bacteria. FEMS Microbiol Ecol 54: 97-109. Geng, A., Soh, A.E., Lim, C.J., and Loke, L.C. (2006) Isolation and characterization of a phenol-degrading bacterium from an industrial activated sludge. Appl Microbiol Biotechnol 71: 728-735. Gilbert, E.S., and Crowley, D.E. (1997) Plant compounds that induce polychlorinated biphenyl biodegradation by Arthrobacter sp. strain B1B. Appl Environ Microbiol 63: 1933-1938. Gulensoy, N., and Alvarez, P.J. (1999) Diversity and correlation of specific aromatic hydrocarbon biodegradation capabilities. Biodegradation 10: 331-340. Halverson, L.J., and Firestone, M.K. (2000) Differential effects of permeating and nonpermeating solutes on the fatty acid composition of Pseudomonas putida. Appl Environ Microbiol 66: 2414-2421. Harvey, P.J., Campanella, B.F., Castro, P.M.L., Harms, H., Lichtfouse, E., Schaeffner, A.R. et al. (2001) Phytoremediation of polyaromatic hydrocarbons, anilines and phenols. Environ Sci & Pollut Res. 9: 29-47. Heipieper, H.J., Diefenbach, R., and Keweloh, H. (1992) Conversion of cis unsaturated fatty acids to trans, a possible mechanism for the protection of phenol-degrading 13

Pseudomonas putida P8 from substrate toxicity. Appl Environ Microbiol 58: 1847- 1852. Herrmann, H., Muller, C., Schmidt, I., Mahnke, J., Petruschka, L., and Hahnke, K. (1995) Localization and organization of phenol degradation genes of Pseudomonas putida strain H. Mol Gen Genet 247: 240-246. Hirano, S.S., and Upper, C.D. (2000) Bacteria in the leaf ecosystem with emphasis on Pseudomonas syringae-a pathogen, ice nucleus, and epiphyte. Microbiol Mol Biol Rev 64: 624-653. Hughes, E.J., Bayly, R.C., and Skurray, R.A. (1984) Evidence for isofunctional enzymes in the degradation of phenol, m- and p-toluate, and p-cresol via catechol meta-cleavage pathways in Alcaligenes eutrophus. J Bacteriol 158: 79-83. Jen, M.S., Hoylman, A.M., Edwards, N.T., and Walton, B.T. (1995) Experimental method to measure gaseous uptake of 14C-toluene by foliage. Environ and Exp Bot 35: 389-398. Kim, D., Park, M.J., Koh, J.S., So, J.S., and Kim, E. (2002) Three separate pathways for the initial oxidation of limonene, biphenyl, and phenol by Rhodococcus sp. strain T104. J. Microbiol. 40: 86-89. Kuiper, I., Lagendijk, E.L., Bloemberg, G.V., and Lugtenberg, B.J. (2004) Rhizoremediation: a beneficial plant-microbe interaction. Mol Plant Microbe Interact 17: 6-15. Legard, D.E., McQuilken, M.P., Whipps, J.W., Felon, J.S., Fermor, T.R., Thompson, I.P., Bailey, M.J., and Lynch, J.M. (1994) Studies of seasonal changes in the microbial populations on the phyllosphere of spring wheat as a prelude to the release of a genetically modified microorganism. Agri Ecosys Environ 50: 87-101. Lindow, S.E., and Brandl, M.T. (2003) Microbiology of the phyllosphere. Appl Environ Microbiol 69: 1875-1883. Lugtenberg, B., Peters, R., Bernheimer, H., and Berendsen, W. (1976) Influence of cultural conditions and mutations on the composition of the outer membrane proteins of Escherichia coli. Mol Gen Genet 147: 251-262. McFarlane, J.C., Pfleeger, T., and Fletcher, J. (1987) Transpiration effect on the uptake and distribution of bromacil, nitrobenzene, and phenol in soybean plants. J Environ Qual 16: 372-376. Meury, J. (1988) Glycine betaine reverses the effects of osmotic stress on DNA replication and cellular division in Escherichia coli. Arch Microbiol 149: 232-239. Miya, R.K., and Firestone, M.K. (2000) Phenanthrene-degrader community dynamics in rhizosphere soil from a common annual grass. J Environ Qual 29: 584-592. Palma, M., DeLuca, D., Worgall, S., and Quadri, L.E. (2004) Transcriptome analysis of the response of Pseudomonas aeruginosa to hydrogen peroxide. J Bacteriol 186: 248- 252. Papendick, R.I., and Campbell, G.S. (1980) Water potential relations in soil microbiology. Madison, WI: Soil Society of America. Phadtare, S., Kato, I., and Inouye, M. (2002) DNA microarray analysis of the expression profile of Escherichia coli in response to treatment with 4,5-dihydroxy-2- cyclopenten-1-one. J Bacteriol 184: 6725-6729. Rentz, J.A., Alvarez, P.J., and Schnoor, J.L. (2005) Benzo[a]pyrene co-metabolism in the presence of plant root extracts and exudates: Implications for phytoremediation. Environ Pollut 136: 477-484. 14

Roth, W.G., Leckie, M.P., and Dietzler, D.N. (1985a) Osmotic stress drastically inhibits active transport of carbohydrates by Escherichia coli. Biochem Biophys Res Commun 126: 434-441. Roth, W.G., Porter, S.E., Leckie, M.P., Porter, B.E., and Dietzler, D.N. (1985b) Restoration of cell volume and the reversal of carbohydrate transport and growth inhibition of osmotically upshocked Escherichia coli. Biochem Biophys Res Commun 126: 442- 449. Rudd, K.E., and Menzel, R. (1987) his operons of Escherichia coli and Salmonella typhimurium are regulated by DNA supercoiling. Proc Natl Acad Sci U S A 84: 517- 521. Santos, P.M., Benndorf, D., and Sa-Correia, I. (2004) Insights into Pseudomonas putida KT2440 response to phenol-induced stress by quantitative proteomics. Proteomics 4: 2640-2652. Sei, K., Inoue, D., Wada, K., Mori, K., Ike, M., Kohno, T., and Fujita, M. (2004) Monitoring behaviour of catabolic genes and change of microbial community structures in seawater microcosms during aromatic compound degradation. Water Res 38: 4405- 4414. Shafer, W.E., and Schonherr, J. (1985) Accumulation and transport of phenol, 2-nitrophenol, and 4-nitrophenol in plant cuticles. Ecotoxicol Environ Saf 10: 239-252. Siciliano, S.D., Fortin, N., Mihoc, A., Wisse, G., Labelle, S., Beaumier, D., Ouellette, D., Roy, R., Whyte, L.G., Banks, M.K., Schwab, P., Lee, K., and Greer, C.W. (2001) Selection of specific endophytic bacterial genotypes by plants in response to soil contamination. Appl Environ Microbiol 67: 2469-2475. Sikkema, J., de Bont, J.A., and Poolman, B. (1995) Mechanisms of membrane toxicity of hydrocarbons. Microbiol Rev 59: 201-222. Slovik, S., Siegmund, A., Fuhrer, H.W., and Heber, U. (1996) Stomatal uptake of SO2, NOx and O3 by spruce crowns (Picea abies) and canopy damage in Central Europe. New Phytol 132: 661-676. Tam le, T., Eymann, C., Albrecht, D., Sietmann, R., Schauer, F., Hecker, M., and Antelmann, H. (2006) Differential gene expression in response to phenol and catechol reveals different metabolic activities for the degradation of aromatic compounds in Bacillus subtilis. Environ Microbiol 8: 1408-1427. Tamburro, A., Robuffo, I., Heipieper, H.J., Allocati, N., Rotilio, D., DI Ilio, C., and Favaloro, B. (2004) Expression of glutathione S-transferase and peptide methionine sulphoxide reductase in Ochrobactrum anthropi is correlated to the production of reactive oxygen species caused by aromatic substrates. FEMS Microbiol Lett 241: 151-156. Van Hove, L.W.A., Koops, A.J., Adema, E.H., Vredenberg, W.J., and Pieters, G.A. (1987) Analysis of the uptake of atmospheric ammonia by leaves of Phaseolus vulgaris L. Atmos Environ 21: 1763-1763. van Schie, P.M., and Young, L.Y. (1998) Isolation and characterization of phenol-degrading denitrifying bacteria. Appl Environ Microbiol 64: 2432-2438. Watanabe, K., Futamata, H., and Harayama, S. (2002) Understanding the diversity in catabolic potential of microorganisms for the development of bioremediation strategies. Antonie van Leeuwenhoek 81: 655-663. 15

Weber, H., Polen, T., Heuveling, J., Wendisch, V.F., and Hengge, R. (2005) Genome-wide analysis of the general stress response network in Escherichia coli: sigmaS-dependent genes, promoters, and sigma factor selectivity. J Bacteriol 187: 1591-1603. Weiss, U.M., Moza, P., Schuenert, I., Haque, A., and Korte, F. (1982) Fate of 14C-labeled pentachlorophenol in rice plants under controlled conditions. J Agric Food Chem 30: 1186-1190. Whiteley, A.S., Wiles, S., Lilley, A.K., Philp, J., and Bailey, M.J. (2001) Ecological and physiological analyses of Pseudomonad species within a phenol remediation system. J Microbiol Methods 44: 79-88. Wild, E., Dent, J., Thomas, G.O., and Jones, K.C. (2005) Real-time visualization and quantification of PAH photodegradation on and within plant leaves. Environ Sci Technol 39: 268-273. Wild, E., Dent, J., Barber, J.L., Thomas, G.O., and Jones, K.C. (2004) A novel analytical approach for visualizing and tracking organic chemicals in plants. Environ Sci Technol 38: 4195-4199. Zahn, J.A., DiSpirito, A.A.,Do, Y.S., Brooks, B.E., Cooper, E.E., and Hatfield, J.L. (2001) Correlation of human olfactory responses to airborne concentrations of malodorous volatile organic compounds emitted from swine effluent. J Environ Qual 30: 624- 634. Zeller, T., Moskvin, O.V., Li, K., Klug, G., and Gomelsky, M. (2005) Transcriptome and physiological responses to hydrogen peroxide of the facultatively phototrophic bacterium Rhodobacter sphaeroides. J Bacteriol 187: 7232-7242. Zheng, M., Wang, X., Templeton, L.J., Smulski, D.R., LaRossa, R.A., and Storz, G. (2001) DNA microarray-mediated transcriptional profiling of the Escherichia coli response to hydrogen peroxide. J Bacteriol 183: 4562-4570.

Chapter 2. Bacterial degradation of airborne phenol in the phyllosphere1

Amarjyoti Sandhu, Larry J. Halverson, and Gwyn A. Beattie Department of Plant Pathology and Interdepartmental Microbiology Program, Iowa State University, Ames, IA 50011-3211, USA.

Abstract Despite the vast surface area of terrestrial plant leaves and the large microbial communities they support, little is known of the ability of leaf-associated microorganisms to access and degrade airborne pollutants. Here, we examined bacterial acquisition and degradation of phenol on leaves by an introduced phenol degrader and by natural phyllosphere communities. Whole-cell gfp-based Pseudomonas fluorescens bioreporter cells detected phenol on leaves that had previously been transiently exposed to gaseous phenol, indicating that leaves accumulated phenol; moreover, they accumulated it in sites that were accessible to epiphytic bacteria and to concentrations that were at least 10-fold higher than those in the air. After inoculated leaves were exposed to gaseous 14C-phenol, leaves 14 harboring the phenol-degrading Pseudomonas sp. strain CF600 released 8 times more CO2 than did leaves harboring a non-degrading mutant, demonstrating that CF600 actively mineralized phenol on leaves. We evaluated phenol degradation by natural microbial communities on green ash leaves that were collected from a field site rich in airborne organic pollutants. We found that significantly more phenol was mineralized by these leaves when the communities were present than by these leaves following surface sterilization. Thus, phenol-degrading organisms were present in these natural communities and were metabolically capable of phenol degradation. Collectively, these results provide the first direct evidence that bacteria on leaves can degrade an organic pollutant from the air, and indicate that bacteria on leaves could potentially contribute to the natural attenuation of organic air pollutants.

1 Environ Microbiol (2007) 9(2): 383-392 17

Introduction Plants can mediate the removal of organic pollutants from the environment. They can do this directly, such as by extracting pollutants from the soil, and indirectly by fostering populations of pollutant-degrading bacteria. The transpiration stream of plants promotes the movement of soil pollutants toward roots and their translocation to the shoots, thus allowing plants to act as wicks that remove pollutants from soil. Similarly, plants may remove pollutants from the air via stomatal uptake and sorption. Stomatal uptake of gaseous pollutants is a major uptake pathway for small inorganic compounds such as sulfur dioxide, nitrogen oxides, and ozone (Slovik et al., 1996), whereas sorption to the plant cuticle, a waxy layer on the leaf surface, can contribute to the uptake of airborne organic compounds (Schmitz et al., 2000). Following uptake, pollutants may be re-released into the environment, accumulated in the plant tissue, and transformed or degraded by plant metabolic processes or by photodegradation (Pilon-Smits, 2005; Wild et al., 2005). Soil contaminants may also be degraded by microbes in the rhizosphere. In this study, we explore the potential for phyllosphere, i.e., leaf-associated, microbial communities to access and degrade organic pollutants from the air.

The availability of organic air pollutants to phyllosphere microorganisms is likely influenced by their uptake pathway. Stomatal uptake should lead to movement of pollutants to the leaf intercellular spaces, a region typically devoid of significant microbial populations. In contrast, cuticular uptake may lead to accumulation on the leaf surface, a region often rich in microorganisms. Several studies support a cuticular uptake pathway for lipophilic gaseous organic compounds, including phenol (G.A. Beattie and J.S. Seibel, unpublished data; McCrady, 1994; Dienum et al., 1995; Keymeulen et al., 1995). Cuticular uptake can involve sorption to the outermost surface of the cuticle, diffusion into the underlying cuticle, and eventual movement into the epidermal cell wall and cytoplasm, as shown for anthracene and other compounds when applied as aqueous solutions to leaf surfaces (Schönherr and Riederer, 1989; Wild et al., 2004). We predict that a consequence of this sorption to the cuticle is enhanced availability to the microorganisms on the leaf surface.

Leaves often support diverse microbial communities with individual populations as large as 107-108 cells per gram of leaf. The majority of these bacteria are present on the leaf surface, although some, primarily pathogens, may establish large populations in the intercellular spaces. Bacteria on leaf surfaces are not uniformly distributed, but rather may occur as individual cells or aggregates (Monier and Lindow, 2004). Roots also support diverse microbial communities, and these communities have been shown to promote pollutant degradation in the rhizosphere (Kuiper et al., 2004). Roots not only provide a nutrient-rich habitat for microbes, but also release secondary metabolites that can induce microbial genes involved in pollutant degradation (Gilbert and Crowley, 1997; Casavant et al., 2003) or can act as co-metabolites that facilitate degradation (Rentz et al., 2005). Roots can also mobilize and accumulate pollutants (Harvey et al., 2002), thus enriching pollutant- degrading bacteria near roots (Miya and Firestone, 2000) and in endophytic sites (Siciliano et al., 2001). Leaves may exhibit analogous properties, including releasing secondary metabolites that promote degradation and accumulating gaseous pollutants that enrich for pollutant-degrading bacteria. The large surface area of leaves in terrestrial ecosystems suggests that such processes could be significant to air quality and the ecology of leaf- associated microbes.

Here, we investigated the potential for pollutant degradation in the phyllosphere using phenol as a model compound and Pseudomonas species on bean (Phaseolus vulgaris) and maize (Zea mays) leaves as model phyllosphere systems. Phenol is a priority pollutant of the U.S. Environmental Protection Agency; it is a component of the emissions from wood, coal, and municipal solid waste combustion, as well as automobile exhaust and cigarette smoke (Agency for Toxic Substances and Disease Registry, 1998). Moreover, it is an abundant pollutant generated by microbial communities during the anaerobic decomposition of livestock waste (Zahn et al., 1997) and is a major contributor to the odor intensity of animal wastes (Zahn et al., 2001). In preliminary studies, we found that bean and maize leaves, which are known to support significant bacterial communities, removed detectable levels of gaseous phenol from the air in a closed system (A. Sandhu, J.S. Seibel and G.A. Beattie, unpublished data). The objectives of the present study were to evaluate 1) the availability of 19

phenol to leaf-associated bacteria following its uptake into bean and maize leaves, 2) the potential for phenol degradation by phenol-degrading bacteria on leaves, and 3) the phenol degradation ability of naturally-occurring epiphytic microbial communities from areas rich in airborne organic pollutants.

Material and Methods

Bacterial strains, plasmids, and growth conditions Pantoea agglomerans strain BRT98 (Sabaratnam and Beattie, 2003), Pseudomonas syringae pv. syringae strain B728a (Loper and Lindow, 1987), Ps. syringae pv. tomato strain DC3000 (Whalen et al., 1991), Ps. fluorescens strain A506 (Wilson and Lindow, 1994), and Escherichia coli strain AW101 (Wise and Kuske, 2000), which contains a chromosomal fusion of the dmp promoter from Pseudomonas sp. Strain CF600 and lacZ, were grown in Luria-Bertani (LB) medium (Miller, 1972), or in yeast extract succinate (YES) medium (Stiner and Halverson, 2002) supplemented with Hutner’s mineral solution. Pseudomonas sp. strain CF600 (Shingler et al., 1989) harbors the phenol catabolic operon dmp on plasmid pVI150 and is able to degrade phenol. The CF600 derivative CF427 (Pavel et al., 1994) has a Kmr cassette in place of an internal dmpR fragment and does not degrade phenol. Strains CF600 and CF427 were grown in M9 minimal medium (Miller, 1972) containing 2.5 mM phenol or 2.5 mM glucose, respectively, and supplemented with Hutner’s mineral solution (Smibert and Krieg, 1994). Antibiotics were used at the following concentrations (g/ml): rifampicin, 50; kanamycin, 50; and tetracycline, 20. Mercuric chloride was used at a concentration of 25 g/ml.

Bioreporter plasmid construction

The 574-bp dmp promoter (Pdmp) was amplified by PCR from E. coli strain AW101, ligated into pGEM-T easy vector (Promega, Madison, WI), and cloned as an EcoRI-BamHI fragment into pPROBE-KT’ (Miller et al., 2000) upstream of a promoterless gfp gene (Miller and Lindow, 1997). The plasmid pAW50-B23 (Wise and Kuske, 2000) was linearized with

ClaI and inserted into a unique ClaI site in the Pdmp-gfp-fusion-containing pPROBE-KT’ 20

plasmid. The resulting plasmid, pPhenol, was transferred from E. coli strain DH5α to the P. agglomerans, Ps. fluorescens, and Ps. syringae strains by conjugation in triparental matings with a helper strain containing the plasmid pRK2073 (Better and Helinski, 1983). Similarly, the plasmid pAW50-B24 (Wise and Kuske, 2000) was cloned into the ClaI site of pPROBE- KT’ plasmid, forming pPhenB24, which was then transferred by conjugation into the P. agglomerans, Ps. fluorescens, and Ps. syringae strains.

Phenol induction of Pdmp-gfp in culture medium

YES-grown cells were washed twice and resuspended in YES (106 cells/ml) in airtight glass tubes. Phenol was added to final concentrations of 0 to 100 g/ml and after 4 h of incubation on a shaker, and GFP fluorescence by cells was measured using a Beckman-Coulter EPICS- XL-MCL flow cytometer, as described previously (Axtell and Beattie, 2002). Bacteria were identified based on their fluorescence. Samples were subjected to flow cytometry until either 25,000 fluorescent particles had been examined or the sample had been run for 30 sec. The geometric mean fluorescence was reported as the mean cellular fluorescence per sample. Phenol induction of strains containing pPhenB24 was characterized and was similar to that of strains with pPhenol.

Availability of phenol to bacteria on leaves

Bean (Phaseolus. vulgaris L. cv. Bush Blue Lake 274) and maize (Zea mays L. inbred B73) seeds were grown in a sterile peat-perlite-soil (1-2-1) potting mixture at 45% relative humidity and 28°C with a 12-h photoperiod. Plants were used at the time of emergence of the first trifoliate (bean) or third leaf (maize), at which time the leaves that were used in the experiments were approximately 1 g (bean) or 0.3 g (maize). Inocula were prepared by suspending A506(pPhenol) cells grown on solid LB media in one-tenth strength YES (1/10 YES) (108 cells/ml). Airtight exposure chambers were constructed using glass jars (3.75 L) with custom-machined polypropylene lids equipped with an injection port. This port was a 24-mm screw cap mininert valve that allowed liquid phenol to be introduced onto a glass 21

dish on a stainless steel holder attached to the lid; the phenol subsequently volatilized into the chamber.

To evaluate the availability of airborne phenol to bacteria on bean leaves, primary leaves of intact, potted plants were inoculated by submersion in the A506(pPhenol) inoculum or 1/10 YES. Plants were air dried until liquid was no longer visible on the leaves. Plants were severed approximately 3 cm below the point of attachment of the primary leaves, and the stem was placed in 2 ml of water in a microfuge tube, which was in a glass serum vial for support. Two plant leaves were introduced into each of 20 exposure chambers. Ten l of a phenol solution (30 g/ml in water) were injected into 12 of the chambers, 8 of which had plants inoculated with bacteria and 4 of which had plants inoculated with 1/10 YES. After incubation of the chambers at 25°C for 4 h, the leaves were removed from the chambers and the bacterial cells were recovered from each leaf by submerging in 20 ml of 1/10 YES medium and sonicating for 7 min (O'Brien and Lindow, 1989). Cells were harvested by centrifugation, resuspended in 1/10 YES, and examined using flow cytometry. The fluorescence of leaf-derived cells was compared to that of a suspension of cells (106 cells/ml) that were exposed to 1 g/ml of phenol in 1/10 YES in airtight glass tubes for 4 h. Flow cytometry samples were run until either 25,000 fluorescent particles had been examined or the sample had been run for 50 sec.

To evaluate the availability of leafborne phenol to bacteria on bean and maize leaves, plants were exposed to airborne phenol prior to plant inoculation with A506(pPhenol) cells. Uninoculated plants were severed 3 cm below the primary leaves (beans) or 3 cm below the emergence point of the first true leaf (maize) and placed in water, as above, and in exposure chambers. Phenol was introduced into half of the chambers for each plant species. After a 4- h incubation at 25°C, the plants were removed from the chambers, inoculated with A506(pPhenol) cells by submersion, placed in new chambers that did not contain airborne phenol, and incubated for an additional 4 h before bacterial recovery and analysis.

Phenol degradation by bacteria on bean leaves

Gas-tight chambers (1L) were constructed by grinding the open ends of 85-mm diameter Pyrex glass beakers to conform to a flat surface and securing them to a glass plate with rubber bands that bound glass pegs fused to the beakers with metal pegs on the glass. Stopcock grease was used to make a gas-tight seal. Liquid phenol was introduced into a Pyrex dish via a mininert valve injection port that was fused to the chamber, and the phenol rapidly volatilized after its introduction.

Bacterial degradation of phenol on leaves was evaluated using Pseudomonas sp. strains CF600 and CF427. Bacterial inocula were prepared by growing cells in M9 minimal medium with 2.5 mM phenol or 2.5 mM glucose, respectively, at 28 °C for 24 h, washing cells twice, and resuspending cells in 10 mM phosphate buffer (PB) (108 cells/ml). Primary bean leaves of intact plants were submerged in bacterial inocula or in PB and were air dried for 1 h. For each treatment, four replicate leaves were used, and each was excised from the stem, placed in water in a microfuge tube, as described above, and then into a gastight chamber. Bacteria were recovered from the second primary leaf of each inoculated plant and were enumerated by plating. Ten l of 0.054 mol of 14C-phenol (Phenol-UL (uniformly labeled)-14C, Sigma-Aldrich, St. Louis, MO) with a specific activity of 5.45 Ci/mol were 14 injected into each chamber for an estimated final concentration of 17 g/L of phenol. CO2 was trapped in 1 ml of 5M KOH in a glass vial (12 x 31 mm) in each chamber. After 24 h, bacteria were recovered from the leaves and enumerated, and the 14C in 1 ml of each the leaf wash, the KOH, and the water from the microfuge tubes supporting the leaves was measured after adding 15 ml of Scintisafe Plus 50% cocktail (Fisher Scientific, Fair Lawn, NJ).

Phenol degradation by naturally occurring microbial communities on leaves

Green ash (Fraxinus pennsylvanica) leaves were collected from trees growing within 5 to 6 m of a swine production facility in Hamilton County, Iowa, in August and October, 2005. A subset of the leaves were surface-sterilized by immersion in 0.02% Tween 20 for 20 s, 0.2% sodium hypochlorite for 3 min, and sterile water 3 times for at least 1 min; leaves were dried 23

in a laminar flow hood until surface water was no longer visible (approximately 30 min). Bacteria were recovered by sonication from unsterilized leaves, as well as from sterilized leaves to verify sterilization, and were enumerated on one-tenth strength tryptic soy agar (Microtech Scientific, Orange, CA) amended with the fungicide cycloheximide (100 g/ml). At least three replicate leaves, each having 5 leaflets, were examined for each treatment. These leaves were placed in gastight chambers, with one leaf per chamber, and were exposed to radiolabeled phenol for 24 h. The radioactivity in the KOH traps, wash fluid from the leaves, and water from the microfuge tubes was measured, as described above.

Statistical analysis

Unless otherwise stated, P-values were determined based on Student’s t-test. Proportional data was subjected to arcsine squareroot transformation before analysis.

Results

Phenol induction of Pdmp-gfp in Pseudomonas fluorescens A506 and other bacterial strains The plasmid pPhenol (Fig. 1) was constructed as a reporter of phenol exposure by bacteria. pPhenol carries a transcriptional fusion between the promoter of the dmp phenol catabolic operon and gfp, as well as dmpR-B23, which encodes a derivative of the transcriptional activator DmpR that mediates an enhanced response to phenol (Wise and Kuske, 2000).

Pdmp-gfp and dmpR-B23 are divergently transcribed in pPhenol. This plasmid was introduced into Pantoea agglomerans strain BRT98, Pseudomonas syringae pv. syringae strain B728a, Ps. fluorescens strain A506 and Ps. syringae pv. tomato strain DC3000. Nonpathogenic bacteria such as P. agglomerans BRT98 and Ps. fluorescens A506 have previously been shown to colonize primarily or exclusively external leaf surface sites (Wilson et al., 1999; Sabaratnam and Beattie, 2003), whereas phytopathogenic bacteria, including Ps. syringae pv. syringae B728a and Ps. syringae pv. tomato DC3000, can colonize internal leaf sites, at least on host plants (Boureau et al., 2002; Sabaratnam and Beattie, 2003). The Pdmp-gfp fusion on pPhenol exhibited relatively low levels of induction by phenol in BRT98(pPhenol) and 24

B728a(pPhenol) (Fig. 2A), in contrast to its responsiveness to phenol in A506(pPhenol) (Fig.

2B) and DC3000(pPhenol) (Fig. 2C). Thus, the responsiveness of Pdmp to phenol in these strains was not correlated with their taxonomy, nor was it correlated with their sensitivity to phenol based on their growth in the presence of up to 100 g/ml of phenol (data not shown). Regression analyses of the A506(pPhenol) and DC3000(pPhenol) data resulted in sigmoidal curves with equations y = 34.85/(1+(x/1.453)-1.405 (R2 = 0.95, P < 0.0001) and y = 51.17/(1+(x/6.529)-0.966 (R2 = 0.95, P < 0.0001), respectively. Moreover, the induction response of A506(pPhenol) and DC3000(pPhenol) by phenol was linear from 0.3 g/ml to 25 g/ml phenol (Figs. 2B and 2C insets), indicating that this range provided the most dynamic response and thus optimal efficacy for these strains as bioreporters of phenol. In leaf colonization studies over 24 h, A506(pPhenol) exhibited 30-fold better survival than DC3000(pPhenol) during the population decline that frequently occurs on leaves following bacterial inoculation (data not shown). Because of its superior leaf surface survival, A506(pPhenol) was selected as the bioreporter for the following studies. EcoRI ClaI T1(4) BamHI

gfp T1 Pdmp dmpR-B23

Tet pVS1 Km replicon ClaI

Fig. 1. Map of pPhenol. The 21.5-kb plasmid has a fusion between the dmp promoter and gfp and carries the regulatory gene dmpR-B23 and the pVS1 origin of replication (Heeb et al., 2000). T1 represents the E. coli rrnB transcriptional terminator. The arrows indicate the direction of transcription.

20 40 40 50 60 20 25 15 30 RFU 50 RFU 0 40 0 -0.5 0.5 1.5 -0.5 0.5 1.5 10 20 Log (phenol conc) 30 Log (phenol conc) RFU 20 5 10 10 0 AC0 B 0 0.1 1 10 100 0.0010.01 0.1 1 10 100 0.0010.01 0.1 1 10 100 Phenol conc (ug/ml)

Fig. 2. Induction of the Pdmp-gfp fusion in bioreporter cells in YES medium after phenol exposure for 4 h. A) P. agglomerans BRT98(pPhenol) (●) and Ps. syringae pv. syringae B728a(pPhenol) (○), B) Ps. fluorescens A506(pPhenol), and C) Ps. syringae pv. tomato DC3000(pPhenol). In B and C, distinct symbols indicate data generated in separate, independent experiments. Each data point is the mean of 3 independent measurements. Relative fluorescence units (RFU) at 0 g/ml of phenol were 3.0, 1.7, 1.5 and 0.5 for BRT98(pPhenol), B728a(pPhenol), A506(pPhenol) and DC3000(pPhenol), respectively. Linear regression of the RFU values at 0.3 to 25 g/ml phenol (insets) yielded the equations B) y = 18.62x + 13.42 (R2 = 0.90, P < 0.0005) and C) y = 23.27x + 7.11 (R2 = 0.83, P = 0.01).

Availability of phenol to A506(pPhenol) on bean leaves The availability of phenol to bacteria on leaves was evaluated based on the fluorescence of A506(pPhenol) cells recovered from leaves that were or were not exposed to phenol in closed chambers. When uninoculated leaves were exposed to phenol and the leaf washings were examined by flow cytometry (Fig. 3, trt 1), the number of fluorescent particles and their fluorescence was low, indicating a low background. Most or all of these particles were probably leaf debris. When leaves inoculated with A506(pPhenol) cells were incubated in chambers without phenol (Fig. 3, trt 2), the number of fluorescent particles and their fluorescence was also low, indicating that leaves did not produce phenolic molecules that induced fluorescence in A506(pPhenol). When leaves inoculated with A506(pPhenol) cells were incubated with phenol for 4 h (Fig. 3, trt 3), the A506(pPhenol) cells were significantly more fluorescent than the cells from leaves that were not exposed to phenol (Fig. 3, trt 2) 26

(P<0.0005), indicating that epiphytic A506(pPhenol) cells detected phenol that had been introduced into the air. The fluorescence of the A506(pPhenol) inoculum cells exposed to 1 g/ml of phenol in broth culture (Fig. 3, trt 4) was similar to that predicted by the dose- response curve (Fig. 2B). Based on this dose-response curve, the A506(pPhenol) cells on leaves sensed 2 to 5 g/ml, with an average of 3 g/ml, of phenol. This is substantially greater than the amount of phenol introduced into the air (0.08 g/ml), suggesting that the phenol was concentrated at the leaf surface, presumably because of sorption to the leaf.

Mean Mean fluorescence # fluorescent Treatment per treatment particles (1) Background particles from phenol- exposed leaves 9.0 ± 0.0 1,809 ± 416

(2) Bacteria from unexposed leaves 9.3 ± 0.1 2,276 ± 673

(3) Bacteria from phenol-exposed leaves 32.9 ± 1.1 22,028 ± 659

17.7 ± 2.7 17,680 ± 613 (4) Bacteria exposed to phenol in culture 0 10 20 30 40 50 Mean fluorescence of particles/ sample

Fig. 3. Availability of phenol provided in gaseous form to A506(pPhenol) on bean leaves. In trt 1, uninoculated leaves were incubated for 4 h following the introduction of 300 g of phenol into the chambers. In trt 2, A506(pPhenol) cells were recovered from inoculated leaves after incubating for 4 h without phenol exposure. In trt 3, A506(pPhenol) cells were inoculated onto leaves and the leaves were incubated for 4 h following the introduction of 300 g of phenol into the chambers. In trt 4, A506(pPhenol) cells were exposed to 1 g/ml of phenol in broth culture for 4 h. The fluorescence of the bacterial cells was measured by flow cytometry. Symbols represent the mean fluorescence of the cells recovered from individual leaves, with the values on the right reflecting the mean fluorescence per treatment – standard error of the mean (SEM) and mean number of fluorescent particles – SEM (n = 4, 16 and 16 leaves for trts 1-3, respectively, and n = 3 cultures for trt 4).

Availability of leafborne phenol to A506(pPhenol) on bean and maize leaves The availability of phenol to epiphytic bacteria following transient exposure of plants to phenol was also evaluated. Bean and maize leaves were exposed to phenol for 4 h and then were removed from the chambers, inoculated with A506(pPhenol) cells, and placed in chambers with no airborne phenol. A506(pPhenol) cells recovered from leaves that had not been exposed to phenol showed a relatively low level of fluorescence induction (Fig. 4, trts 1 and 3). The wide variability in the mean fluorescence of the samples from maize leaves (Fig.

4, trt 3) suggests that either maize leaves produce compounds that induce Pdmp expression or leaf-derived fluorescent particles, such as leaf debris, are recovered at a higher rate from maize leaves than from bean leaves. A506(pPhenol) cells recovered from leaves that were exposed to phenol exhibited significantly more fluorescence than those recovered from leaves of the same species that had not been exposed to phenol (Fig. 4) (P<0.05). The amount of phenol accessible to bacteria on bean leaves under these conditions was estimated to be between 1.5 and 1.8 g/ml, with an average of 1.7 g/ml based on the dose-response curve shown in Fig. 2B. In contrast, A506(pPhenol) cells from maize leaves exhibited a much broader range in their mean fluorescence, from 11.2 to 36.7, with an average mean fluorescence of 27.0 ± 1.8, which was significantly higher than that of the cells from bean leaves (P<0.05). Based on the dose response curve, this reflects exposure to 1.0 to 3.5 g/ml of phenol, with an average of 2 g/ml. These estimates of phenol exposure only approximate the true exposure levels because they do not account for possible differences in cellular physiology between the cells on the leaves and the cells in the cultures used to generate the dose-response curves. The broader range of fluorescence among the cells recovered from maize than from bean leaves suggests that either the two plant species differed in the distribution of phenol adsorbed onto the leaf surfaces or that the bacterial cells were less uniformly distributed on maize leaves than on bean leaves. This greater fluorescence was observed for cells recovered from the first leaf of the maize seedlings as well as those separately recovered from the second leaf (data not shown). The fluorescence of bacteria that were present on leaves during phenol exposure (Fig. 3, trt 3) was significantly higher than the fluorescence of bacteria that were on leaves previously exposed to phenol (Fig. 4, trt 2) 28

(P<0.05). This result suggests that bacteria on leaves can access both airborne and leafborne phenol. Mean Mean fluorescence # fluorescent Treatment per treatment particles (1) Bacteria from unexposed bean leaves 5.7 ± 0.2 137 ± 22 (2) Bacteria from phenol-exposed bean leaves 23.0 ± 0.2 25,000 ± 0 (3) Bacteria from unexposed maize leaves 8.7 ± 1.1 23 ± 5 (4) Bacteria from phenol-exposed maize 27.0 ± 1.8 14,161 ± 2502 leaves (5) Bacteria exposed to phenol in culture 18.6 ± 2.4 18,093 ± 4161 0 10 20 30 40 50 Mean fluorescence of particles/sample

Fig. 4. Bioavailability to A506(pPhenol) of phenol that had been taken up and accumulated in bean (trts 1 and 2) and maize (trts 3 and 4) leaves. Excised leaves were incubated in the chambers without phenol exposure (trts 1 and 3) or with 300 g of phenol (trts 2 and 4) for 4 h after which the leaves were removed from the chambers, inoculated with A506(pPhenol) cells and incubated for another 4 h. In trt 5, A506(pPhenol) cells were exposed to 1 g/ml phenol in 1/10 YES. The bacteria were recovered from each leaf and their fluorescence was measured by flow cytometry. Symbols and values are as described in Fig. 3, with n=15, 16, 15, 16, and 3 for trts 1-5, respectively.

Phenol degradation by bacteria on bean leaves The ability of bacteria to degrade phenol while resident on leaves was evaluated based on detectable mineralization or cellular incorporation of radiolabeled phenol. Bacteria were suspended on a filter disk, which was placed on water agar and exposed to 14C-phenol for 24 14 h in the presence of a KOH trap for C02. The amount of radioactivity in KOH incubated with Pseudomonas sp. strain CF600 was 79% of the radioactivity introduced into the chamber as compared to only 12% for CF427, a mutant strain defective in the dmpR regulatory gene, and 6% for a filter disk without bacteria. Thus, mineralization ability was confirmed for Pseudomonas sp. strain CF600 and was confirmed to be greatly reduced for strain CF427. To evaluate if CF600 can degrade phenol when present on leaves, primary bean leaves were submerged in a CF600 cell suspension, a CF427 cell suspension, or phosphate buffer; these leaves were then exposed to uniformly labeled 14C-phenol (specific 29

activity 5.45 Ci/mol) for 24 h in chambers that each contained a KOH trap. The CF600 population on bean leaves did not change under these conditions for at least 24 h, remaining at approximately 108 CFU/g leaf After incubation with 14C-phenol, the radioactivity in the KOH was significantly higher (approximately 8-fold) in the chambers with the CF600- inoculated leaves than in the chambers with the CF427-inoculated leaves or in those with the uninoculated leaves (single factor ANOVA, P<0.005) (Fig. 5), indicating that CF600 cells on leaves could access and degrade phenol introduced into the air. The low level of radioactivity in the KOH from chambers with the uninoculated leaves provides strong evidence against plant-mediated phenol mineralization, at least under the conditions examined. Although more radioactivity was assimilated into the CF600 cells than into the CF427 cells, this difference was not significant (Fig. 5). The negligible amount of radioactivity in the water in the microfuge tubes supporting the leaves (Fig. 5) indicates that the gaseous radiolabeled phenol did not partition appreciably into the water; this is attributable, in part, to the small surface area of the water. The CF600 and CF427 population sizes on leaves after 24 h of exposure to phenol were not different than they were at 0 h (data not shown), indicating that the degradation of phenol by CF600 was not due to a larger population size on leaves and that degradation activity did not detectably influence epiphytic population sizes.

25 CF600-inoculated leaves CF427-inoculated leaves

20 Uninoculated leaves

radioactivity%Mean 5

0 KOH Leaf wash Water 14 14 14 ( CO2) ( C-labeled (Dissolved CO2 cells) or 14C-phenol) Fig. 5. 14C-phenol degradation by Pseudomonas sp. strains CF600 and CF427 on bean leaves in closed chambers. The radioactivity was measured in the KOH from reservoirs placed in the chambers, in the bacterial suspension recovered from washing the leaves, and in the water from the microfuge tubes supporting the leaves. Values represent the mean percentage – SEM of the total radioactivity introduced into the chambers at 0 h (n = 4).

Phenol degradation by naturally occurring microbial communities on leaves Leaves were collected from a stand of green ash trees adjacent to a swine production facility, and thus from an environment that was likely rich in airborne volatile organic compounds, including phenol (Zahn et al., 1997). Leaves from both sampling times harbored 105 to 107 CFU/g bacteria based on culturable counts. Subsets of the leaves were surface-sterilized with 0.2% sodium hypochlorite; the effectiveness of this sterilization was confirmed based on the subsequent inability to recover culturable cells from the treated leaves. After exposure of the unsterilized and surface-sterilized leaves to radiolabeled phenol in closed chambers for 24 h, the amount of radioactivity that accumulated in the KOH in chambers with the unsterilized leaves was significantly greater than that in chambers with the sterilized leaves (P < 0.005); this was observed with leaves sampled in August, in which the difference was 3.1-fold (data not shown), and in October, in which the difference was 2.5 fold (Fig 6). These results indicate that organisms that were susceptible to the surface-sterilization treatment were 31

capable of mineralizing airborne phenol. The relatively low level of radioactivity in the KOH from chambers with the sterilized leaves provides good evidence against plant- mediated phenol mineralization by the green ash leaves, as was observed with the uninoculated bean leaves (Fig. 5). The amount of radioactivity in the microbial cells washed from the unsterilized leaves was significantly higher (approximately 10-fold) than in the wash fluid recovered from the sterilized leaves (P<0.0005), suggesting that phenol was assimilated into at least some of the microbial cells recovered from the green ash leaves.

10 sterilized leaf unsterilized leaf 8

Mean % radioactivity % radioactivity Mean 2

0 KOHKOH Leaf Washate wash WaterWater 14 14 14 ( CO2) ( C-labeled (Dissolved CO2 cells) or 14C-phenol)

Fig. 6. 14C-phenol degradation by the naturally occurring microbial community on green ash leaves sampled in October from an area near a swine production facility. Data are as described in Fig. 5, with n = 4 and 3 for the sterilized and unsterilized treatments, respectively.

Discussion

Leaf surfaces of terrestrial plants provide an excellent venue for promoting microbial access to volatile organic compounds (VOCs). Here, we found that in addition to their known ability to foster significant microbial communities, plant leaves can accumulate the VOC phenol and make it available for subsequent degradation by phyllosphere bacteria. 32

Furthermore, we found that introduced bacteria can degrade phenol during leaf colonization and that such degradation may be a common feature of some terrestrial ecosystems based on the degradation activity detected in natural phyllosphere communities. The relatively rapid rate of degradation by these natural communities, as evidenced by detectable degradation within 24 h, suggests that phyllosphere communities near sites rich in airborne pollutants are naturally primed for such degradation. Collectively, these studies indicate that phylloremediation, so named by analogy to rhizoremediation in roots, may serve as a natural attenuation mechanism for air pollutants as well as could provide an alternative source of energy and nutrients to bacteria on leaves.

Phylloremediation of VOCs has been suggested by several previous studies. In particular, De Kempeneer et al. (2004) found that the rate of removal of airborne toluene by azalea leaves in a closed system was greater when leaves were inoculated with a toluene- degrading enrichment culture of Pseudomonas putida than when they were mock-inoculated with water. Although this study did not directly evaluate bacterial degradation activities, and thus did not rule out enhanced plant uptake in the presence of Ps. putida, it indicated that phyllosphere bacteria can contribute to reductions in airborne toluene. Another study suggests that phylloremediation can attenuate pollutants that become airborne via evapotranspiration. Specifically, the leaves of yellow lupine plants grown in a hydroponic system amended with toluene released less toluene by evapotranspiration when the plants were colonized with an endophytic toluene-degrading Burkholderia cepacia than when they were colonized with non-degrading B. cepacia (Barac et al., 2004). And lastly, a Methylobacterium sp. that was isolated from leaves and other hybrid poplar tissues was able to degrade xenobiotic compounds, including several volatile organic explosives (Van Aken et al., 2004), demonstrating that bacteria capable of VOC degradation are present in at least some phyllosphere communities.

We have provided direct evidence of active VOC degradation by natural phyllosphere communities. These communities were on leaves grown near a swine production facility, a site where VOCs are often released due to anaerobic degradation of organic compounds by 33

bacteria. Up to 168 airborne chemicals have been identified near swine production facilities (Curtis, 1983), with volatile fatty acids, phenols, alcohols, ammonia, and sulfur-containing compounds being the most abundant. Among 29 swine production facilities examined, phenol was present at an average air concentration of 0.025 µg/ml, which is similar to the levels used in this study (0.02 to 0.08 µg/ml). The natural bacterial communities on the sampled leaves contained culturable bacterial populations of 105 to 107 CFU/leaf and included a broad diversity of colony types. Based on the diversity of organisms that can degrade phenol and are common to phyllosphere communities, including Acinetobacter, Arthrobacter, Bacillus, Pseudomonas, Ralstonia, and Rhodococcus (Mishra et al., 2001; Kirchner et al., 2003; Shingler, 2004; Lambais et al., 2006), multiple candidate organisms could have been responsible for phenol degradation on these leaves.

We found that bean and maize leaves were highly effective bioaccumulators of phenol. Phenol was introduced into the closed chambers at a concentration of 0.08 µg/ml, assuming a uniform distribution, whereas the bioreporter cells detected concentrations of 1 to 3.5 µg/ml on leaves, demonstrating an increase in concentration of over 10-fold. The leaf surface properties of bean and maize are highly distinct, suggesting that phenol accumulation occurs in many plant species. VOCs may accumulate in the cuticle, as indicated by the visible localization of anthracene in the outermost layer of the cuticle of maize leaves, with subsequent diffusion into the epidermal cell cytoplasm over a 96-h period (Wild et al., 2004). We predict that phenol sorbed to the cuticle in the 4-h periods used in our studies based on its detection by the Ps. fluorescens bioreporter cells, which as nonpathogenic bacteria are likely restricted to leaf surface sites (Wilson et al., 1999; Sabaratnam and Beattie, 2003).

The dmp catabolic pathway in Pseudomonas sp. CF600 mediated complete mineralization of phenol on leaves. Such mineralization accounted for approximately 20% of the introduced radiolabel, whereas 8% was associated with the CF600 cells, presumably due to incorporation into macromolecules. These results indicate that phenol was used primarily as a source of energy and secondarily as a source of carbon under these conditions. This is consistent with the lack of significant population growth of CF600, based on 34

culturable counts, during the 24-h incubation period with 14C-phenol (data not shown). Although we did not detect such growth, detectable population increases due to the presence of a VOC as an additional nutrient source would likely require exposure periods longer than 24 h.

Collectively, terrestrial leaves provide a vast surface area, as much as 400,000,000 km2 globally (Myneni et al., 1997; Morris and Kinkel, 2002), for the sorption and uptake of air pollutants and as a habitat for microorganisms. Exposure of phyllosphere communities to airborne organic pollutants could affect their ecology by augmenting phyllosphere nutrient pools, enriching for pollutant-degraders, and inhibiting pollutant-sensitive organisms. This exposure appears to occur on the leaf surface, based on the probable surface location of our Ps. fluorescens bioreporter cells (Wilson et al., 1999; Sabaratnam and Beattie, 2003), but could also occur in endophytic sites, where degradation would affect pollutant phytotoxicity as well as airborne pollutant levels. We found that a bioreporter derivative of Ps. syringae pv. tomato (Pst) was similar to the Ps. fluorescens A506 bioreporter in its responsiveness to phenol, but as a pathogen, Pst can establish large endophytic populations in host plants with a delay of at least 48 h before disease symptoms occur in some plants (Boureau et al., 2002). This strain would thus be appropriate for future investigations into the bioavailability of phenol in endophytic sites. Similarly, future studies aimed at quantifying the impact of phylloremediation on airborne VOC concentrations are needed to understand the role of phylloremediation in natural pollutant attenuation and to design strategies for minimizing indoor and outdoor airborne organic pollutants.

Acknowledgements

We are thankful for the contributions of Janice Seibel in designing the closed chamber systems used in these studies and to Dr. Janice Thompson, Troy Bowman, and Iowa Select farms for their help in obtaining leaf samples from the field. The work was supported by funds from the Department of Agronomy Endowment and the College of Agriculture at Iowa State University.

References

Agency for Toxic Substances and Disease Registry (ATSDR). (1998) Toxicological profile for phenol. Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service. Axtell, C.A., and Beattie, G.A. (2002) Construction and characterization of a proU-gfp transcriptional fusion that measures water availability in a microbial habitat. Appl Environ Microbiol 68: 4604-4612. Barac, T., Taghavi, S., Borremans, B., Provoost, A., Oeyen, L., Colpaert, J.V. Vangronsveld, J., and van der Lelie, D. (2004) Engineered endophytic bacteria improve phytoremediation of water-soluble, volatile, organic pollutants. Nat Biotechnol 22: 583- 588. Better, M., and Helinski, D.R. (1983) Isolation and characterization of the recA gene of Rhizobium meliloti. J Bacteriol 155: 311-316. Boureau, T., Routtu, J., Roine, E., Taira, S., and Romantschuk, M. (2002) Localization of hrpA-induced Pseudomonas syringae pv. tomato DC3000 in infected tomato leaves. Mol. Plant Pathol. 3: 451-460. Casavant, N.C., Thompson, D., Beattie, G.A., Phillips, G.J., and Halverson, L.J. (2003) Use of a site-specific recombination-based biosensor for detecting bioavailable toluene and related compounds on roots. Environ. Microbiol. 5: 238-249. Curtis, S.E. (1983) Environmental management in animal agriculture. Ames: Iowa State University Press. De Kempeneer, L., Sercu, B., Vanbrabant, W., Van Langenhove, H., and Verstraete, W. (2004) Bioaugmentation of the phyllosphere for the removal of toluene from indoor air. Appl. Microbiol. Biotechnol. 64: 284-288. Dienum, G., Baart, A., Bakker, D.J., Duyzer, J.H., and Van Den Hout, D. (1995) The influence of uptake by leaves on atmospheric deposition of vapor-phase organics. Atmos. Environ. 29: 997-1005. Gilbert, E.S., and Crowley, D.E. (1997) Plant compounds that induce polychlorinated biphenyl biodegradation by Arthrobacter sp. strain B1B. Appl. Environ. Microbiol. 63: 1933-1938. Harvey, P.J., Campanella, B.F., Castro, P.M.L., Harms, H., Lichtfouse, E., Schaeffner, A.R. et al. (2002) Phytoremediation of polyaromatic hydrocarbons, anilines and phenols. Environ. Sci. Poll. Res. 9: 29-47. Heeb, S., Itoh, Y., Nishijyo, T., Schnider, U., Keel, C., Wade, J. et al. (2000) Small, stable shuttle vectors based on the minimal pVS1 replicon for use in gram-negative, plant- associated bacteria. Molecular Plant-Microbe Interactions 13: 232-237. Keymeulen, R., Schamp, N., and Vanlangenhove, H. (1995) Uptake of gaseous toluene in plant-leaves - a two compartment model. Chemosphere 31: 3961-3975. Kirchner, U., Westphal, A.H., Muller, R., and van Berkel, W.J.H. (2003) Phenol hydroxylase from Bacillus thermoglucosidasius A7, a two-protein component monooxygenase with a dual role for FAD. J. Biol. Chem. 278: 47545-47553. Kuiper, I., Lagendijk, E.L., Bloemberg, G.V., and Lugtenberg, B.J.J. (2004) Rhizoremediation: A beneficial plant-microbe interaction. Mol. Plant-Microbe Interact. 17: 6-15. 36

Lambais, M.R., Crowley, D.E., Cury, J.C., Büll, R.C., and Rodrigues, R.R. (2006) Bacterial diversity in tree canopies of the Atlantic Forest. Science 312: 1917. Loper, J.E., and Lindow, S.E. (1987) Lack of evidence for in situ fluorescent pigment production by Pseudomonas syringae pv. syringae on bean leaf surfaces. Phytopathology 77: 1449-1454. McCrady, J.K. (1994) Vapor-phase 2,3,7,8-TCDD sorption to plant foliage - a species comparison. Chemosphere 28: 207-216. Miller, J.H. (1972) Experiments in molecular genetics: Cold spring harbor laboratory, NY. Miller, W.G., and Lindow, S.E. (1997) An improved GFP cloning cassette designed for prokaryotic transcriptional fusions. Gene 191: 149-153. Miller, W.G., Leveau, J.H., and Lindow, S.E. (2000) Improved gfp and inaZ broad-host- range promoter-probe vectors. Mol. Plant-Microbe Interact. 13: 1243-1250. Mishra, V., Lal, R., and Srinivasan (2001) Enzymes and operons mediating xenobiotic degradation in bacteria. Crit. Rev. Microbiol. 27: 133-166. Miya, R.K., and Firestone, M.K. (2000) Phenanthrene-degrader community dynamics in rhizosphere soil from a common annual grass. J. Environ. Qual. 29: 584-592. Monier, J.M., and Lindow, S.E. (2004) Frequency, size, and localization of bacterial aggregates on bean leaf surfaces. Appl. Environ. Microbiol. 70: 346-355. Morris, C.E., and Kinkel, L.L. (2002) Fifty years of phyllosphere microbiology: significant contributions to research in related fields. In Phyllosphere Microbiology. Lindow, S.E., Hecht-Poinar, E.J., and Elliot, V.J. (eds). St. Paul: The American Phytopathological Society, pp. 365-375. Myneni, R.G., Nemani, R.R., and Running, S.W. (1997) Estimation of global leaf area index and absorbed par using radiative transfer models. IEEE Trans. Geosc. Remote Sens. 35: 1380-1393. O'Brien, R.D., and Lindow, S.E. (1989) Effect of plant species and environmental conditions on epiphytic population sizes of Pseudomonas syringae and other bacteria. Phytopathology 79: 619-627. Pavel, H., Forsman, M., and Shingler, V. (1994) An aromatic effector specificity mutant of the transcriptional regulator DmpR overcomes the growth constraints of Pseudomonas sp. strain CF600 on para-substituted methylphenols. J. Bacteriol. 176: 7550-7557. Pilon-Smits, E. (2005) Phytoremediation. Annu. Rev. Plant Biol. 56: 15-39. Rentz, J.A., Alvarez, P.J.J., and Schnoor, J.L. (2005) Benzo[a]pyrene co-metabolism in the presence of plant root extracts and exudates: Implications for phytoremediation. Environ. Poll. 136: 477-484. Sabaratnam, S., and Beattie, G.A. (2003) Differences between Pseudomonas syringae pv. syringae B728a and Pantoea agglomerans BRT98 in epiphytic and endophytic colonization of leaves. Appl. Environ. Microbiol. 69: 1220-1228. Schmitz, H., Hilgers, U., and Weidner, M. (2000) Assimilation and metabolism of formaldehyde by leaves appear unlikely to be of value for indoor air purification. New Phytol. 147: 307-315. Schönherr, J., and Riederer, M. (1989) Foliar penetration and accumulation of organic chemicals in plant cuticles. Rev. Environ. Contam. Toxicol. 108: 1-70. 37

Shingler, V. (2004) Transcriptional regulation and catabolic strategies of phenol degradative pathways. In Pseudomonas. Ramos, J.L. (ed). New York: Kluwer Academic / Plenum Publishers, pp. 451-477. Shingler, V., Franklin, F.C.H., Tsuda, M., Holroyd, D., and Bagdasarian, M. (1989) Molecular analysis of a plasmid-encoded phenol hydroxylase from Pseudomonas CF600. J. Gen. Microbiol. 135: 1083-1092. Siciliano, S.D., Fortin, N., Mihoc, A., Wisse, G., Labelle, S., Beaumier, D. et al. (2001) Selection of specific endophytic bacterial genotypes by plants in response to soil contamination. Appl. Environ. Microbiol. 67: 2469-2475. Slovik, S., Siegmund, A., Führer, H.W., and Heber, U. (1996) Stomatal uptake of SO2, NOx and O-3 by spruce crowns (Picea abies) and canopy damage in central Europe. New Phytol. 132: 661-676. Smibert, R.M., and Krieg, N.R. (1994) Phenotypic characterization. In Methods for general and molecular bacteriology. Gerhardt, P., Murray, R.G.E., Wood, W.A., and Kreig, N.R. (eds). Washington, D.C.: American Society for Microbiology, pp. 607-654. Stiner, L., and Halverson, L.J. (2002) Development and characterization of a green fluorescent protein-based bacterial biosensor for bioavailable toluene and related compounds. Appl. Environ. Microbiol. 68: 1962-1919. Van Aken, B., Yoon, J.M., and Schnoor, J.L. (2004) Biodegradation of nitro-substituted explosives 2,4,6-trinitrotoluene, hexahydro-1,3,5-trinitro-1,3,5-triazine, an octahydro- 1,3,5,7-tetranitro-1,3,5-tetrazocine by a phytosymbiotic Methylobacterium sp associated with poplar tissues (Populus deltoides x nigra DN34). Appl. Environ. Microbiol. 70: 508- 517. Whalen, M.C., Innes, R.W., Bent, A.F., and Staskawicz, B.J. (1991) Identification of Pseudomonas syringae pathogens of Arabidopsis and a bacterial locus determining avirulence on both Arabidopsis and soybean. Plant Cell 3: 49-59. Wild, E., Dent, J., Thomas, G.O., and Jones, K.C. (2005) Real-time visualization and quantification of PAH photodegradation on and within plant leaves. Environ. Sci. Technol. 39: 268-273. Wild, E., Dent, J., Barber, J.L., Thomas, G.O., and Jones, K.C. (2004) A novel analytical approach for visualizing and tracking organic chemicals in plants. Environ. Sci. Technol. 38: 4195-4199. Wilson, M., and Lindow, S.E. (1994) Ecological similarity and coexistence of epiphytic ice- nucleating (Ice+) Pseudomonas syringae strains and a non-ice-nucleating (Ice-) biological control agent. Appl. Environ. Microbiol. 60: 3128-3137. Wilson, M., Hirano, S.S., and Lindow, S.E. (1999) Location and survival of leaf-associated bacteria in relation to pathogenicity and potential for growth within the leaf. Appl. Environ. Microbiol. 65: 1435-1443. Wise, A.A., and Kuske, C.R. (2000) Generation of novel bacterial regulatory proteins that detect priority pollutant phenols. Appl. Environ. Microbiol. 66: 163-169. Zahn, J.A., Hatfield, J.L., Do, Y.S., DiSpirito, A.A., Laird, D.A., and Pfeiffer, R.L. (1997) Characterization of volatile organic emissions and wastes from a swine production facility. J. Environ. Qual. 26: 1687-1696. 38

Zahn, J.A., DiSpirito, A.A., Do, Y.S., Brooks, B.E., Cooper, E.E., and Hatfield, J.L. (2001) Correlation of human olfactory responses to airborne concentrations of malodorous volatile organic compounds emitted from swine effluent. J. Environ. Qual. 30: 624-634.

Chapter 3. Identification and genetic characterization of phenol-degrading bacteria from leaf microbial communities

Introduction A wide diversity of phenol-degrading bacteria, including Pseudomonas, Ralstonia, Acinetobacter, Burkholderia, Comamonas and Variovorax, have been isolated from natural microbial populations in various ecosystems including wastewaters, rivers, seas, and soils (van Schie and Young, 1998; Bastos et al., 2000; Abd-El-Haleem et al., 2002; Arutchelvan et al., 2005; Garcia et al., 2005; Geng et al., 2006). The presence of pollutants in these sites can select for pollutant-degrading bacteria (Ghiorse et al., 1995; Van der Meer, 2006). Whereas the structure and function of microbial communities in contaminated wastewaters and polluted soils have been well characterized (Shafer and Schonherr, 1985a; Watanabe et al., 2002; Sei et al., 2003; Brakstad and Lodeng, 2005; DeRito et al., 2005) the nature of those on plant leaves, especially near sites abundant in airborne organic compounds, has not. In our previous study (Sandhu et al., 2007), we provided the first direct evidence of phenol degradation by bacteria on plant leaves; we observed this with introduced bacteria and natural microbial communities present on green ash (Fraxinus pennsylvanica) leaves that were taken from a site rich in airborne phenol. We found that significantly more phenol was mineralized when microbial communities were present on the leaves than when leaves were surface-sterilized. A few previous studies have shown that bacteria capable of pollutant degradation are present in at least some phyllosphere communities. For example, phyllosphere bacteria related to the genera Acinetobacer, Alcaligenes, Bacillus, and Micrococcus were isolated from the leaves of various tropical plants and shown to degrade the hydrocarbons diesel, kerosene and lubricating oil in culture (Ilori et al., 2006). In another study, a Methylobacterium sp. that was isolated from leaves and other hybrid poplar tissues was able to degrade xenobiotic compounds, including several volatile organic explosives (Van Aken et al., 2004). Although these studies clearly show that pollutant-degrading 40

bacteria can be present in the phyllosphere, little has been done to identify and characterize these pollutant-degrading phyllosphere bacterial communities. In the present study, we identified members of the microbial communities that are capable of degrading airborne pollutants on green ash leaves, thus improving our understanding of phyllosphere microbial ecology, at a site rich in airborne organic pollutants.

Characterization of the genetic and functional diversity of pollutant-degrading bacterial communities can provide insight into the catabolic potential of those communities (Watanabe et al., 2002). Identification of the bacterial species that are capable of degrading most of the pollutants in contaminated sites should help in the development of effective bioremediation strategies (Watanabe et al., 1998; Watanabe et al., 2002). Molecular characterization of pollutant-degrading isolates is important for developing genetic markers for monitoring the presence and expression of catabolic genes in situ. For example, genes encoding the catabolic enzymes phenol hydroxylase (Watanabe et al., 1998; Futamata et al., 2001a) and catechol dioxygenase (Okuta et al., 1998) have been commonly used to assess the genetic diversity of phenol-degrading bacterial populations in soil and marine environments (Sei et al., 2004). The sequence diversity of catabolic genes in environmental samples may reflect the diversity in functional potential within the microbial communities in those samples.

Phenol is a priority pollutant of the US Environmental Protection Agency (ATSDR, 1998); it is a component of the emissions from wood, coal and municipal solid waste combustion, as well as automobile exhaust and cigarette smoke. It is also one of the compounds responsible for malodors associated with livestock production facilities resulting from the anaerobic decomposition of animal wastes (Zahn et al., 1997; Zahn, 2001). The objectives of the present study were to (i) isolate and identify phenol-degrading bacteria on green ash leaves, and (ii) characterize the genetic diversity of the phenol catabolic genes present in the phenol-degrading isolates and total leaf communities. This information will help in identifying bacterial contributors to phenol degradation and in assessing the diversity of the phenol degradation genes or pathways employed by these bacteria in the phyllosphere. 41

Material and Methods

Confirmation of phenol degradation by naturally occurring microbial communities on leaves. Green ash (Fraxinus pennsylvanica) leaves were collected from trees growing within 5-6 m of a swine production facility in Hamilton County, Iowa, in September, 2006 and were transported in plastic bags in a cooler and stored at 4°C until used. A subset of leaves were surface-sterilized by immersion in 0.02% Tween 20 for 20 s, 0.2% sodium hypochlorite for 3 min, and sterile water three times for at least 1 min; leaves were dried in a laminar flow hood until surface water was no longer visible (approximately 30 min). Two replicate leaves, each having five leaflets, were examined for each treatment. These leaves were placed in gastight chambers, with one leaf per chamber, and were exposed to 10 l of 0.054 mol of 14C-phenol (Phenol-UL (uniformly labeled)-14C, Sigma-Aldrich, St. Louis, MO) with a specific activity of 5.45 Ci/mol for 24 h. For each chamber, the radioactivity was measured after adding 15 ml of Scintisafe Plus 50% cocktail (Fisher Scientific, Fair Lawn, NJ) to each 1 ml of the 14 KOH used for trapping CO2 released due to microbial phenol degradation, 1 ml of the wash fluid from the leaves, and 1 ml of water from the microfuge tube supporting the leaves. This experiment was performed to evaluate the phenol degradation activity of the phyllosphere community; such activity was observed with green ash leaves collected in August and October, 2005.

Recovery of microbial communities from green ash leaves. Microbial communities were recovered from each leaf by suspending the leaf, with its 5-6 leaflets, in 25 ml of 10 mM phosphate buffer (PB), sonicating for 7 min and vortexing. The leaf was removed from the tube and the suspension was centrifuged at 5000x g for 5 min to remove soil particles. The supernatant was centrifuged at 8000x g for 10 min and the final pellet was resuspended in 1 ml of M9 minimal medium with no carbon source.

Isolation of phenol-utilizing bacteria Enrichments. Three leaves were used to enrich for phenol-utilizing isolates. Serial enrichments were performed on the communities recovered from each of the three leaves by increasing the phenol concentration during each subculturing. A 1 ml bacterial suspension was transferred into 4 ml of M9 minimal medium with 1 mM phenol as a carbon source and incubated for 4 days at 28°C. For the second enrichment, 500 µl of cells from the first enrichment were transferred to 4.5 ml of M9 medium with 2 mM phenol and grown for 2 days; the remainder was stored at -20°C for microbial community DNA isolation. The third enrichment was performed in the same manner as the second, also with 2 mM phenol and grown for 2 days. Dilutions were plated on M9 plates with 2 mM phenol and streaked for purity. A total of 17 pure isolates were obtained by this method. Colony autoradiography. In addition to subjecting the communities recovered from each of the three leaves to enrichments, the communities recovered from a different set of three leaves were directly plated onto 1/10 strength Tryptic Soy Agar (1/10 TSA) plates for autoradiography. From these plates, which supported densely packed colonies, two hundred colonies, each likely impure to the high density of the source plate, were transferred as short streaks onto 1/10 TSA plates and the plates were incubated overnight at 28°C. The plates were exposed to 0.2 µCi of 14C-phenol in an airtight glass container for 2 days to allow cells that degraded phenol to accumulate 14C. The colonies were replica printed on filter paper (Whatman no. 3), the filter was exposed to a phosphorscreen (Amersham Biosciences) for 2 days, and the image was analyzed using a Typhoon 9410 phosphorimager (Amersham Biosciences, Piscataway, NJ ). The colonies that accumulated phenol appeared dark in the image and were streaked for purity. A total of 16 pure isolates were obtained by this method.

Evaluation of the phenol-mineralization activity of isolates. To confirm that the purified isolates were able to degrade phenol, we examined each one for 14 14 its ability to release CO2 via C-phenol mineralization. Bacteria were grown in M9 medium supplemented with 5 mM sodium succinate at 28°C, washed thrice in M9 medium without carbon and resuspended in M9 medium without carbon to an OD590 of 0.2-0.3. One 43

hundred µl of this bacterial suspension was inoculated into screw-capped glass tubes containing 5 ml of M9 medium with 2 mM of non-labeled phenol. Caps with mininert valves were used to make the tubes airtight. Ten µl of 14C-phenol (0.01 µCi) was introduced into the tubes using a gas-tight syringe and the tubes were incubated at 28°C on a shaker. M9 medium without bacteria was included as a control to determine the amount of 14C-phenol 14 being trapped along with CO2. One ml of the tube headspace was sampled using a gas-tight 14 14 syringe immediately after adding C-phenol, i.e., at 0 h, and again at 48 h. The CO2 was trapped in 1 ml of 5 M KOH, which was then mixed with 15 ml of Scintisafe Plus 50% cocktail (Fisher Scientific, Fair Lawn, NJ) to measure the radioactivity.

Genomic DNA extraction and amplification of 16S rDNA for identification of bacterial isolates. Bacteria were grown in ½ strength TSA broth at 28°C and genomic DNA was isolated using the Wizard genomic DNA purification kit (Promega, Madison, WI). The 16S rDNA was amplified using the primers 8F (5′-AGAGTTTGATCCTGGCTC-3′) and 1404R (5′- GGGCGGWGTGTACAAGGC-3′; where W=A or T). PCR was performed by using a 50-µl reaction mixture containing 1X PCR buffer (New England Biolabs, Ipswich, MA), each dNTP at a concentration of 200 µM, each primer at a concentration of 0.4 µM, 100 ng genomic DNA, and 2 U Ex Taq (TaKaRa, Madison, WI). The PCR conditions were 1 min at 94°C and then 35 cycles of 30 sec at 95°C, 30 sec at 45°C, and 1 min at 72°C, and a final 5 min extension at 72°C. The 1396-bp PCR product was visualized by electrophoresis using a 1% agarose gel and was sequenced using the 8F and 1404R primers.

Characterization of the phenol catabolic genes. To test for the presence of the gene encoding the first enzyme in the phenol catabolic pathway, phenol hydroxylase (PH), PCR was performed using the universal primers to PH, PheUf/PheUr (Futamata et al., 2001a) for Gram-negative isolates, and primers that we designed, Pheh-3F (5′-CGKATGACSTACGGCTGGATGGGCG-3′; where K=G or T and S=G or C) and Pheh-3R (5′-ACGTCCTGTTCGATGATCTCCTTGATCCGC-3′) for Gram- positive isolates. The latter primers were based on pheA1 of Rhodococcus RHA1 (Accession 44

# NC008268), pheA1 of Rhodococcus erythropolis CCM2595 (Accession # AJ973228), and a gene encoding a putative phenol hydroxylase large subunit of Nocardia farcinia IFM 10152 (locus tag nfa43650, Accession # AP006618). The PCR program used to amplify a 620-bp band with the primer set PheUf/PheUr was a touchdown program as follows: 1) 94°C for 10 min, 2) five cycles of 1 min at 94°C, 1 min at 58°C, and 1 min at 72°C, 3) five cycles of 1 min at 94°C, 1 min at 57°C, and 1 min at 72°C, 4) 25 cycles of 1 min at 94°C, 1 min at 56°C, and 1 min at 72°C, and 5) a 10-min extension at 72°C. The PCR program used for amplification of a 872-bp band with Pheh-3F/Pheh-3R primers was 94°C for 10 min, 30 cycles of 94°C for 1 min, 60.5°C for 1 min, and 72°C for 1 min followed by a final extension step at 72°C for 5 min.

To determine the type of phenol catabolic pathway employed by the Gram-negative isolates, PCR was performed using the degenerate C23O-specific primers C23OF (5’ – GGTCTGATYGAAATGGAYCGCGA-3’) and C23OR (5’- CGTTCGTTSAGCACCCGGTCGTG-3’) and the catechol 1,2-dioxygenase (C12O)-specific primers C12OF (5’-CCTGARCBGTHGGYTTTGCNCGTATGGATGA-3’) and C12OR (5’- TCACGRGTWGCRWARGCAAAGTC-3’). The C23OF/C23OR primers were designed based on the C23O gene sequence of two Acinetobacter sp. strains, PC19 (Accession # AY887962) and YAA (atdB gene, Accession # AB008831), and Alcaligenes sp. KF711 (Accession # S77084). The C12OF/C12OR primers were designed based on the C12O gene sequences of Acinetobacter sp. ADP1 (catA, Accession # CR543861), Acinetobacter lwoffii K24 (catA3, Accession # DQ874401), Acinetobacter calcoaceticus (Accession # Z36909) and Acinetobacter radioresistens (isoA, Accession # AF380158). The 816-bp and 480-bp products for C23O and C12O, respectively, were visualized by electrophoresis using a 1% agarose gel and were sequenced using the C23OF/C23OR and C12OF/C12OR primers, respectively.

Qualitative assay for catechol 2,3-dioxygenase activity. The catechol 2,3-dioxygenase activity of each isolate was examined using a modified version of an assay described by Pankhurst et al. (1965). The phenol-degrading bacterial isolates 45

were streaked onto M9 minimal medium plates with 2.5 mM phenol as the carbon source and grown for 4-5 days at 28°C. A bacterial strain, Pseudomonas sp. strain CF600 harboring a C23O gene on a plasmid, was included as a positive control. The colonies from the plates were lifted onto filter paper (Whatman no. 3) and sprayed with 0.1 M of catechol. The appearance of yellow color around colonies exhibiting C23O activity occurred within 30 min.

Total leaf community DNA extraction and amplification of phenol hydroxylase and catechol dioxygenase genes. To determine the proportion of the total genetic diversity in the leaf community represented by the phenol-degrading isolates obtained in this study, total microbial community DNA was isolated from 3 communities recovered from leaves by using the PowerSoil DNA isolation kit (MoBio Labs, Carlsbad, CA); however, the DNA yield from only one set of leaves was sufficiently high to use in subsequent studies. An aliquot of the DNA from this community was amplified using the GenomiPhi kit (GE Healthcare, Amersham Place, UK). PCR was performed with the unamplified and amplified total community DNA samples using primers PheUf/PheUr and Pheh-3F/Pheh-3R to amplify the phenol hydroxylase genes in the Gram- negative and Gram-positive components of the communities, respectively. PCR products of the appropriate sizes were gel purified and cloned into pCR2.1-TOPO vector (Invitrogen, Carlsbad, CA). Twenty clones resulting from the unamplified community DNA, and ten from the GenomiPhi-amplified community DNA, were selected, the plasmids were isolated, and the cloned fragments were sequenced using the primers R1 (Reverse M13 primer 5' - CAGGAAACAGCTATGACC-3') and T7-1 (5’-AATACGACTCACTATAG-3’).

Sequence analysis. The taxonomic identification of the isolates was determined by comparing the 16S rDNA sequences obtained to the Ribosomal Database (Maidak et al., 2001) to identify the closest matches. The phenol hydroxylase sequences were translated and the amino acid sequences were aligned using ClustalW (Thompson et al., 1994). The identity of the specific residues 46

was used to place each phenol hydroxylase gene into one of the three groups described by Futamata (2001). Results

Confirmation of phenol degradation activity of naturally occurring microbial communities on green ash leaves. The degradation activity of the microbial community on green ash leaves was demonstrated by exposing surface-sterilized and unsterilized leaves to radiolabeled phenol in gastight chambers for 24 h. The amount of radioactivity in KOH in chambers with the unsterilized leaves was significantly greater than that in chambers with the sterilized leaves (P < 0.005) (Fig. 1), indicating that organisms that were susceptible to the surface-sterilization treatment were capable of mineralizing airborne phenol. Similar results were observed in a previous study (Sandhu et al., 2007). Collectively, these studies performed with leaves sampled in Aug and Oct 2005 and Sept 2006 suggest that the degradation activity of these leaf- associated microbial communities is a common characteristic of these communities under these conditions.

12 Sterilized leaves 10 Unsterilized leaves 8 6 4 2 Mean radioactivity % 0 KOHKOH (14CO2) Leaf washing WaterWater 14 14 14 ( CO2) (14C-labeled( C-labeled (Dissolved(Dissolved CO2 14 cells) 14CO2or C-phenol) or 14C- phenol)

Fig. 1. Phenol mineralization activity of naturally occurring microbial communities on green ash leaves sampled in September 2006 from an area rich in airborne organic compounds. 47

Values represent the mean percentage ± SEM of the total radioactivity introduced into the chambers at 0 h (n = 2). Isolation of phenol-utilizing bacteria from green ash leaves. The phenol-utilizing bacteria from green ash leaf microbial communities were isolated using two methods: enrichment and colony autoradiography. Direct selection of phenol utilizers on M9 minimal media plates with 2.5 mM phenol as the sole carbon source was attempted but was not successful because bacterial colonies developed on plates without any carbon source, indicating that bacteria were able to use residual nutrients in the agar to support growth. Also, the phenol-utilizing bacteria may need to adapt to lower phenol concentrations than those used in plates to induce enzymes involved in phenol degradation. Leaves harbored total communities with at least 105-107 cfu g-1 bacteria based on culturable counts on 1/2TSA with cycloheximide (200 µg/ml). Enrichments using minimal media with phenol provided as the sole carbon source exhibited visible turbidity within a week, whereas tubes without phenol had no detectable turbidity. After enrichment, a total of 17 colonies with distinct morphologies were selected from three independent leaf replicates. In addition to enrichments, colony autoradiography was employed to screen for phenol-utilizing bacteria within the microbial communities on green ash leaves. Colony autoradiography is a technique used to identify cells that can utilize or incorporate 14C substrates. Of two hundred bacterial samples that were transferred from plates supporting densely-packed colonies from green ash leaves to 1/10 TSA plates (Fig. 2A), seven samples exhibited evidence of 14C accumulation (Fig. 2B). These samples were streaked for purity, and all were found to be mixtures; 16 isolates with distinct colony morphologies were identified. These isolates were subsequently screened for 14C-phenol mineralization (described below).

A B

Fig. 2. Identification of phenol-utilizing bacteria by colony autoradiography. Bacterial samples were patched onto two 1/10 TSA plates (A) that were exposed to 14C- phenol for 48 h and were lifted onto filter disks (B), which were subsequently exposed to a phosphorscreen to indicate those samples that accumulated 14C.

Evaluation of the phenol-mineralizing activity of the isolates. A total of 33 bacterial isolates were screened for phenol-mineralizing activity in a minimal medium containing 14C-phenol. Bacterial isolates were first grown in a minimal medium with unlabeled phenol as the sole carbon source. After introduction of 14C-phenol into the tubes, 14 air in the headspace was sampled at 0 and 48 h and was injected into KOH to trap CO2, and the KOH was evaluated for radioactivity. A total of 10 isolates, 7 from the enrichments and 3 identified by colony autoradiography, exhibited significantly higher radioactivity in the air sampled at 48 h than at 0 h (Fig. 3A and B), demonstrating that these bacterial isolates were 14 14 in fact phenol degraders based on their ability to mineralize C-phenol to CO2 . A negative control with no bacterial cells in the medium was included to determine the amount of 14 14 airborne C-phenol sampled along with the CO2, and it showed that the amount of radioactivity in air either did not change or decreased 48 h following 14C-phenol introduction (Fig. 3). Tubes with the bacterial strain Pseudomonas syringae DC3000, which does not utilize phenol, were also included and did not show an increase in radioactivity at 48 h, as expected. These tubes also showed that a negligible amount of radiolabel was incorporated into the non-phenol degrading bacterial cells as compared to in the phenol degraders.

B 1600 0 h A 700 1400 48 h 600 1200 500 1000 400 800 300 600 400 200

200 100 DPM per ml of headspace of mlDPM per headspace of perml DPM 0 0 E1 E3 E4 E6 E9 T-2 T-5 E10 E12

T-10 Isolate # Isolate # DC3000 Nobacteria Nobacteria

Fig. 3. 14C-phenol mineralization ability of selected isolates from the enrichments (A) and as identified by colony autoradiography (B). Isolates were grown in M9 minimal medium with 2 mM of non-labeled phenol and 0.1 µCi of 14C-phenol as the carbon source. The amount of 14 14 CO2 released due to bacterial C-phenol mineralization was measured by sampling 1 ml of the tube headspace and trapping in KOH, and the counts per minute (DPM) in the one-ml headspace samples were determined. Values represent the mean DPM ± SEM (n = 2). The isolates that did not show phenol degradation are not shown

Taxonomic identification of the phenol-degrading isolates . The 16S rDNA sequences were determined for the ten bacterial isolates that degraded 14C- phenol and were compared to those of known organisms in the Ribosomal Database (Maidak et al., 2001). While molecular phylogeny based on the ribosomal genes is standard for bacterial classification, there are no universally accepted criteria for the level of similarity among 16S rDNA sequences that indicates that two isolates are from the same genus or species. While 8 out of 10 of the 16S rDNA sequences showed greater than 97% homology to the closest related sequence (Table 1) (Drancourt et al., 2000; Kim et al., 2006), two sequences (E-9 and 10) shared ≤ 95% with known organisms (Table 1). Interestingly, all enrichment isolates were Gram-negative, out of which five were most closely related to the 50

genus Acinetobacter and two to the genus Alcaligenes, whereas all isolates identified by colony autoradiography were Gram-positive and belonged to the genus Rhodococcus (based on a sequence similarity of >99.5%). Previous studies have shown that Acinetobacter, Alcaligenes and Rhodococcus strains can be metabolically versatile (Yang et al., 2001; Barbe et al., 2004; Larkin et al., 2005; Rivas et al., 2007). These species have been previously shown to be members of phyllosphere communities (Jurkevitch and Shapira, 2000; Cohen et al., 2004; Ilori et al., 2006).

Table 1. Identity of phenol-degrading bacterial isolates obtained from leaves

Phenol Closely related organism based Isolate Sourcea Identity hydroxylase Class on 16S rDNA gene sequence identity Acinetobacter E-1 E Acinetobacter lwoffii 99.4% calcoaceticus, γ-Proteobacteria mphN A. E-9 Acinetobacter sp. 95.4% γ-Proteobacteria E calcoaceticus Acinetobacter E-10 Acinetobacter sp. 93.1% γ-Proteobacteria E radioresistens Acinetobacter E-12 E Acinetobacter seohaensis 97% γ-Proteobacteria sp. PD12 A. E-4 E A. lwoffii 99.4% calcoaceticus, γ-Proteobacteria mphN Alcaligenes E-3 E Alcaligenes sp. CC-ESB2 99.7% faecalis β-Proteobacteria BC2001 A. faecalis E-6 E Alcaligenes sp. A72 or CC-ESB2 100% β-Proteobacteria BC2001 Rhodococcus T-2 CA Rhodococcus sp. CHNTR32 100% Actinobacteria sp. RHA1 Rhodococcus T-5 CA Rhodococcus sp. CHNTR32 100% Actinobacteria sp. RHA1 Rhodococcus T-10 CA Rhodococcus sp. CHNTR32 99.5% Actinobacteria sp. RHA1 a Method used for selection of phenol-utilizing bacterial isolates. E, Enrichment; CA, Colony autoradiography.

Genetic characterization of the phenol catabolic genes in the phenol-degrading isolates. To characterize the 7 Gram-negative isolates for the presence of phenol hydroxylase, the first enzyme involved in the degradation of phenol, we used universal primers to amplify the gene encoding the largest subunit of the multi-component phenol hydroxylase. This gene has been 51

used in other studies as a molecular marker to assess the diversity of phenol-degrading bacteria in the environment (Zhang et al., 2004). A 620-bp band was amplified from each isolate and sequenced. Futamata et al. (2001a) aligned the amino acid sequences of the large subunit of phenol hydroxylases from several Gram-negative bacteria and grouped them into three classes, designated I, II and III, which correspond to the previously determined kinetic groups, low-Ks, moderate-Ks, and high-Ks, exhibited by phenol-degrading bacteria during TCE degradation (Futamata et al., 2001b). Based on the sequences of the PH genes in the Gram-negative isolates, all of these isolates had class II PHs. Phenol hydroxylase sequences belonging to isolates E-1 and E-4 were 86% similar to the phenol hydroxylase gene mphN of Acinetobacter calcoaceticus; the genes in E-3 and E-6 were 96% similar to an Alcaligenes faecalis strain BC2001 PH gene; the E-9 PH gene was 88% similar to the A. calcoaceticus PH gene; E-10 PH was 92% similar to the gene for the oxygenase component of the PH of Acinetobacter radioresistens, and the E-12 PH gene was 98% similar to the PH gene of Acinetobacter sp. PD12. The phenol hydroxylase gene sequences of all three Gram-positive isolates were 84 to 86% similar to the Rhodococcus sp. RHA1 phenol hydroxylase gene.

The type of phenol degradation pathway, ortho or meta, in our isolates was determined using PCR and an enzymatic assay for C23O. PCR indicated the presence of the C12O gene, but not the C23O gene, in all of the Gram-negative isolates, suggesting that either all of the Gram-negative isolates harbored only the C120 gene or that the C23O primers were not degenerate enough to amplify the C23O gene sequences from the isolates obtained in this study. The C12O gene sequences of the isolates were similar to those of the Alcaligenes and Acinetobacter sp., suggesting the presence of at least an ortho pathway for phenol degradation in the Gram-negative isolates. The type of pathway present in the Rhodococcus sp. isolates could not be determined by PCR because the available sequences of the Rhodococcus sp. C12O and C23O genes were too dissimilar to design primers effective for multiple strains. None of the Gram-negative or Gram-positive isolates tested positive for C23O activity in an enzymatic assay specific for C23O activity.

Efforts to characterize the phenol hydroxylase gene diversity present in a leaf microbial community. To compare the phenol hydroxylase gene diversity in the total bacterial community to that in the cultured isolates obtained in this study, we amplified and cloned the phenol hydroxylase genes from the DNA isolated from a microbialcommunity recovered from a green ash leaf, with and without subsequent GenomiPhi genomic DNA amplification. Twenty clones obtained with unamplified total community DNA and 10 obtained with amplified community DNA were sequenced. All of the sequences were identical and exhibited 86% identity to an Acinetobacter calcoaceticus phenol hydroxylase gene. Reasons underlying the recovery of only a single PH sequence from the community include the possibility that a single phenol- degrading species, perhaps A. calcoaceticus, dominated this phyllosphere bacterial community, that the phenol hydroxylase genes of other phenol-degrading species were present at a sufficiently low concentration that they were not amplified, cloned, and/or selected for sequencing, and that the primers used for amplifying the phenol hydroxylase genes were less effective for other bacterial species. No amplification was obtained with community DNA using the Pheh-3F/Pheh-3R primers specific for the Rhodococcus isolates. This could be due to a low gene copy number in the community DNA, as well as poor lysis of the Gram-positive cells, and degradation of DNA during leaf or DNA storage, thus losing the sensitivity needed to detect minority community members. Furthermore, the community DNA was isolated from a distinct set of leaves from the set used for culturing phenol- utilizing bacteria; thus, the lower phenol hydroxylase gene diversity identified in the total community DNA than in the cultured isolates may reflect differences in the microbial communities on these distinct leaf sets.

Discussion

Leaf surfaces of terrestrial plants near sites rich in airborne pollutants provide an excellent habitat for pollutant-degrading bacterial communities. Although phenol-degrading bacteria have been isolated from a number of ecosystems, such as polluted river waters and 53

contaminated soils, leaf surfaces remain an unexplored habitat for these organisms. Here, we isolated and characterized phenol-degrading bacteria from the natural microbial communities residing on leaves that were taken from a site rich in airborne volatile organic compounds, including phenol (Zahn et al., 1997). Using a culture-based approach, we found ten phenol- degrading bacterial isolates related to three different genera, the Gram-negative genera Acinetobacter and Alcaligenes and the Gram-positive genus Rhodococcus. Although bacteria belonging to these genera have been isolated from phenol-contaminated waters (Xu et al., 2000) and soils (Saadoun, 2002; Mandri and Lin, 2007), they have also been shown to be a part of phyllosphere communities. For example, Acinetobacter sp. has been shown to be present in the phyllosphere of variety of plant species such as Citrus sinesis (Yang et al., 2001), chestnut trees (Valverde et al., 2005), and field-grown maize (Kadivar and Stapleton, 2003), and Alcaligenes and Acinetobacter have been found to be a part of the phyllosphere community of tropical plants (Ilori et al., 2006) and tomato plants (Jurkevitch and Shapira, 2000), with some isolates exhibiting the ability to degrade hydrocarbons, such as diesel, kerosene and lubricating oil (Ilori et al., 2006). Ilori et al (2006) found that these hydrocarbon-utilizing bacteria isolated from leaves belonged to the genera Acinetobacter, Alcaligenes, Flavobacterium, Micrococcus, Bacillus, Pseudomonas and Corynebacterium. The Rhodococcus genus has also been found on the leaves the aquatic fern Azolla pinnata (Cohen et al., 2004). Collectively, these studies indicate that pollutant-degrading bacteria are a part of naturally occurring microbial communities on leaves.

We identified ten phenol-degrading bacterial isolates from natural phyllosphere communities of green ash leaves. Five of these bacterial isolates were most closely related to the genus Acinetobacter, two to Alcaligenes, and three to Rhodococcus. Bacteria of these genera have been isolated from numerous contaminated natural or engineered environments (Watanabe et al., 1998; Abd-El-Haleem et al., 2002; Arutchelvan et al., 2005) and have been shown to be capable of degrading a variety of aromatic compounds, including phenol and its derivatives (Lenke et al., 1992; Paller et al., 1995; Hill et al., 1996; Abd-El-Haleem et al., 2002). In fact, bacteria belonging to the genus Rhodococcus are among the most metabolically versatile of the Gram-positive bacteria (Karlson et al., 1993; van der Geize and 54

Dijkhuizen, 2004; Larkin et al., 2005). Phenol-degrading isolates may be common in natural communities based on several reports documenting the presence of these genera on leaves, as described above (Jurkevitch and Shapira, 2000; Yang et al., 2001; Ilori et al., 2006), although how common pollutant degradation activities are among phyllosphere isolates of these genera are not known.

Characterization of the phenol catabolic genes of the isolates obtained from leaves did not reveal much genetic diversity. All of the Gram-negative isolates possessed an ortho- pathway for phenol degradation and phenol hydroxylases belonging to group II, the moderate affinity PHs. A few studies (Eichner et al., 1999; Boon et al., 2000; Iwamoto et al., 2000; Ogino et al., 2001) looking at the quantitative changes in overall microbial community structure during pollutant degradation processes have shown a decrease in the microbial diversity and domination by a few functionally important bacterial populations. Different types of PHs, class I, II and III, are present in different bacterial genera or strains and a previous study showed that one strain can even harbor two different multi-component PH genes (Zhang et al., 2004), and this was suggested to help it to adapt to a highly fluctuating environment. Also, in our study none of the isolates tested positive for C23O, an enzyme required for the meta pathway of phenol degradation. This supports a finding in previous studies that bacterial populations possessing C12O are the primary phenol degraders that are dominant under low phenol-conditions (Muller and Babel, 1996; Sei et al., 2004), assuming that leaves have a low phenol content. For example, Sei et al. (2003) demonstrated that in low-phenol seawater microcosms, bacterial populations possessing C12O were the primary degraders, whereas bacterial populations possessing C23O increased under high phenol concentrations. We were not able to test phenol-utilizing Rhodococcus isolates for the presence of C23O and C12O genes by PCR because the sequence similarity of C23O and C12O genes among Rhodococcus strains was too low to design general PCR primers; however, like the Gram-negative isolates, these Rhodococcus strains appeared to lack the C23O gene based on the absence of C23O activity.

Phylloremediation of volatile organic compounds has been suggested by several previous studies and one study has demonstrated pollutant-degrading ability of leaf- associated bacteria (Van Aken et al., 2004). In this study by Van Anken et al., a Methylobacterium sp. strain that was isolated from leaves and other hybrid poplar tissues was able to degrade xenobiotic compounds, demonstrating that bacteria capable of pollutant degradation are present in some phyllosphere communities. In our previous study (Sandhu et al., 2007), we provided direct evidence of active phenol degradation by the natural phyllosphere communities on green ash leaves that were taken from trees near a swine production facility. This site was rich in airborne VOCs, based on the level of odor, and likely included phenol (Zahn et al., 1997). So far few studies have explored leaf surfaces as a habitat for the isolation and identification of pollutant-degrading bacteria.

Characterization of the genes involved in the phenol catabolic pathway revealed that all of the Acinetobacter- and Alcaligenes-like isolates have group II phenol hydroxylase genes and that the three Rhodococcus isolates have phenol hydroxylase genes similar to that of Rhodococcus sp. RHA1. The phenol-degrading Gram-negative bacteria are diverse and their PHs have been categorized based on specific amino acid residues into three groups with distinct kinetics, classes I, II and III, which exhibit high (low Ks), moderate and low (high

Ks) affinity for phenol, respectively (Futamata et al., 2001a). Examples of organisms with PHs in these classes are Comamonas testosteroni R5, Pseudomonas sp. strain CF600, and Pseudomonas putida P-2, for class I, II, and III, respectively (Futamata et al., 2001a). Phylogenetic classification of PHs corresponds to the kinetic classification of phenol- degrading bacteria (Futamata et al., 2001a), thus facilitating the group-specific monitoring of the different types of phenol-degrading bacteria in microbial communities. It has been suggested in previous studies that bacteria with a high affinity for phenol are present at sites with low phenol concentrations (Massol-Deya et al., 1997) and those with low affinity are present at sites with high phenol concentrations (Watanabe et al., 1996; Jiang et al., 2004; Jiang et al., 2006), as would be predicted based on the kinetics. Thus, the phenol available to bacteria on leaves at a site rich in airborne organic compounds was likely high enough to select bacteria with moderate affinity for phenol. Group-specific monitoring of phenol 56

hydroxylase genes that can cometabolize trichloroethylene (TCE) in contaminated soils has been shown to be important in predicting the catabolic potential of indigenous TCE- degrading bacterial populations (Futamata et al., 2001a; Watanabe et al., 2002) and thus could help in developing effective bioremediation strategies.

The proportion of the phenol hydroxylase gene diversity represented by the phenol- degrading isolates in the total leaf community was determined by amplification of the PH genes within the total community DNA. No amplification was obtained with the primers specific for Rhodococcus PH genes, suggesting that Rhodococcus species were a minor population on leaves and thus their PH genes were below the limit of detection by PCR. In contrast, the PheUf/PheUr primers successfully amplified PH genes, the sequences of which were identical for all of the twenty cloned genes and were 86% similar to a known Acinetobacter PH gene. This indicates either that the organism with this gene was a dominant population in the phyllosphere of green ash leaves, possibly because its growth was favored by the airborne phenol, or that the PCR amplification was biased toward this Acinetobacter- like PH gene. The PheUf/PheUr primers were designed based on the phenol hydroxylase gene sequences of pseudomonads, Comamonas species and Burkholderia species (Futamata et al., 2001a) and therefore may not have been efficient in amplifying genes of other bacterial species. In such cases, where degenerate primers cannot be used to capture the total bacterial diversity in environmental samples, other approaches, such as using several PCR primers (Llyod-Jones et al., 1999), need to be employed.

Bacterial phenol degradation occurs mainly via two pathways, initiated by either the ortho or meta cleavage (Muller and Babel, 1996). The ortho pathway is catalyzed by the enzyme C12O, which cleaves the catechol ring at the ortho position, while the meta pathway is catalyzed by C23O that cleaves the ring at the meta position. Bacteria can have one or both of these enzymes (Song et al., 2000; Jiang et al., 2006), which may be expressed under different conditions. Catechol dioxygenase genes have been used as targets in previous studies to monitor types of and changes in bacterial populations in environmental samples contaminated with aromatic compounds. In the present study, we used catechol dioxygenase 57

gene-specific primers to determine gene diversity in phenol-degrading isolates obtained from the phyllosphere of green ash leaves. We found that all phenol-degrading Gram-negative isolates possessed C12O genes based on PCR while no amplification was obtained with the Gram-positive Rhodococcus species. None of the isolates tested positive for C23O in an assay for C23O activity, suggesting little diversity in the type of phenol degradation pathways that are possessed by phyllosphere bacteria. It is also possible that the low phenol concentrations (approximately 0.25 µM) around the swine facility selected for bacteria possessing C12O genes, similar to the finding in a previous study (Tay et al., 2005) that high phenol concentrations (3mM) were associated with the dominance of C23O-possessing bacteria in a bioreactor, consistent with the lower affinity of C23O than C12O for phenol (Muller and Babel, 1996; Filonov et al., 1997). In another study (Sei et al., 2004), behavior of aromatic compound-degrading bacterial populations and changes in the microbial community structure in sea water microcosms revealed that the bacterial populations possessing C12O genes were the primary degraders of phenol and benzoate and that the C23O-harboring bacteria only dominated under higher substrate concentrations.

In conclusion, we examined an unexplored habitat, the phyllosphere, to identify and characterize the phenol-degrading bacterial species. Although ecosystems such as polluted seawaters, rivers and soils have been examined for phenol-degrading bacteria and monitored for changes in bacterial populations, the diversity of phenol-degrading taxa in the phyllosphere was unexplored prior to this study. We isolated ten phenol-degrading bacterial species most closely related to the genera Acinetobacter, Alcaligenes and Rhodococcus from leaves at a site rich in airborne organic compounds. We expected a wide diversity of the phenol-degrading bacteria on leaves, in part because the diversity of bacteria associated with leaves is known to be high (Wilson and Lindow, 1992; Yang et al., 2001). Approaches such as RNA-SIP (stable isotope probing) may be employed in future studies to determine the total diversity of phenol-degrading bacterial populations and to characterize the catabolic genes present in these populations. This technique would thus help in linking microbial phylogeny with function, which is important in developing strategies for bioremediation of polluted environments. 58

References Abd-El-Haleem, D., Moawad, H., Zaki, E.A., and Zaki, S. (2002) Molecular characterization of phenol-degrading bacteria isolated from different Egyptian ecosystems. Microb Ecol 43: 217-224. Arutchelvan, V., Kanakasabai, V., Nagarajan, S., and Muralikrishnan, V. (2005) Isolation and identification of novel high strength phenol degrading bacterial strains from phenol-formaldehyde resin manufacturing industrial wastewater. J Hazard Mater 127: 238-243. Agency for Toxic Substances and Disease Registry (ATSDR). (1998) Toxicological profile for phenol. Atlanta, GA: U. S. Department of Health and Human Services, Public Health Service. Barbe, V., Vallenet, D., Fonknechten, N., Kreimeyer, A., Oztas, S., Labarre, L., Cruveiller, S., Robert, C., Duprat, S., Wincker, P., Ornston, L.N., Weissenbach, J., Marliere, P., Cohen, G.N., and Medigue, C. (2004) Unique features revealed by the genome sequence of Acinetobacter sp. ADP1, a versatile and naturally transformation competent bacterium. Nucleic Acids Res 32: 5766-5779. Bastos, A.E., Moon, D.H., Rossi, A., Trevors, J.T., and Tsai, S.M. (2000) Salt-tolerant phenol-degrading microorganisms isolated from Amazonian soil samples. Arch Microbiol 174: 346-352. Boon, N., Marle, C., Top, E.M., and Verstraete, W. (2000) Comparison of the spatial homogeneity of physico-chemical parameters and bacterial 16S rRNA genes in sediment samples from a dumping site for dredging sludge. Appl Microbiol Biotechnol 53: 742-747. Brakstad, O.G., and Lodeng, A.G. (2005) Microbial diversity during biodegradation of crude oil in seawater from the North Sea. Microb Ecol 49: 94-103. Cohen, M.F., Meziane, T., and Yamasaki, H. (2004) A photocarotogenic Rhodococcus sp. isolated from the symbiotic fern Azolla. Endocytobiosis Cell Res. 15: 350-355. DeRito, C.M., Pumphrey, G.M., and Madsen, E.L. (2005) Use of field-based stable isotope probing to identify adapted populations and track carbon flow through a phenol- degrading soil microbial community. Appl Environ Microbiol 71: 7858-7865. Drancourt, M., Bollet, C., Carlioz, A., Martelin, R., Gayral, J.P., and Raoult, D. (2000) 16S ribosomal DNA sequence analysis of a large collection of environmental and clinical unidentifiable bacterial isolates. J Clin Microbiol 38: 3623-3630. Eichner, C.A., Erb, R.W., Timmis, K.N., and Wagner-Dobler, I. (1999) Thermal gradient gel electrophoresis analysis of bioprotection from pollutant shocks in the activated sludge microbial community. Appl Environ Microbiol 65: 102-109. Filonov, A.E., Duetz, W.A., Karpov, A.V., Gaiazov, R.R., Kosheleva, I.A., Breure, A.M., Filonova, I.F., van Andel, J.G., and Boronin, A.M. (1997) Competition of plasmid- bearing Pseudomonas putida strains catabolizing naphthalene via various pathways in chemostat culture. Appl Microbiol Biotechnol 48: 493-498. Futamata, H., Harayama, S., and Watanabe, K. (2001a) Group-specific monitoring of phenol hydroxylase genes for a functional assessment of phenol-stimulated trichloroethylene bioremediation. Appl Environ Microbiol 67: 4671-4677. 59

Futamata, H., Harayama, S., and Watanabe, K. (2001b) Diversity in kinetics of trichloroethylene-degrading activities exhibited by phenol-degrading bacteria. Appl Microbiol Biotechnol 55: 248-253. Garcia, M.T., Ventosa, A., and Mellado, E. (2005) Catabolic versatility of aromatic compound-degrading halophilic bacteria. FEMS Microbiol Ecol 54: 97-109. Geng, A., Soh, A.E., Lim, C.J., and Loke, L.C. (2006) Isolation and characterization of a phenol-degrading bacterium from an industrial activated sludge. Appl Microbiol Biotechnol 71: 728-735. Ghiorse, W.C., Herrick, J.B., Sandoli, R.L., and Eugene, L.M. (1995) Natural selection of PAH-degrading bacterial guilds at coal-tar disposal sites. Environ Health Perspect. 103: 107-111. Hill, G.A., Milne, B.J., and Nawrocki, P.A. (1996) Cometabolic degradation of 4- chlorophenol by Alcaligenes eutrophus. Appl Microbiol Biotechnol 46: 163-168. Ilori, M.O., Amund, O.O., Ezeani, C.K., Omoijiahina, S., and Adebusoye, S.A. (2006) Occurrence and growth potentials of hydrocarbon degrading bacteria on the phylloplane of some tropical plants. Afr J Biotechnol 5: 542-545. Iwamoto, T., Tani, K., Nakamura, K., Suzuki, Y., Kitagawa, M., Eguchi, M., and Nasu, M. (2000) Monitoring impact of in situ biostimulation treatment on groundwater bacterial community by DGGE. FEMS Microbiol Ecol 32: 129-141. Jiang, H.-L., Tay, S.T.-L., Maszene, A.M., and Tay, J.-H. (2006) Physiological traits of bacterial strains isolated from phenol-degrading aerobic granules. FEMS Microbiol Ecol 57: 182-191. Jiang, H.L., Tay, J.H., and Tay, S.T.L. (2004) Changes in structure, activity and metabolism of aerobic granules as a microbial response to high phenol loading. Appl Environ Microbiol 63: 602-608. Jurkevitch, E.J., and Shapira, G. (2000) Structure and colonization dynamics of epiphytic bacterial communities and of selected component strains on tomato (Lycopersicon esculentum) leaves. Microb Ecol 40: 300-308. Kadivar, H., and Stapleton, A.E. (2003) Ultraviolet radiation alters maize phyllosphere bacterial diversity. Microb Ecol 45: 353-361. Karlson, U., Dwyer, D.F., Hooper, S.W., Moore, E.R., Timmis, K.N., and Eltis, L.D. (1993) Two independently regulated cytochromes P-450 in a Rhodococcus rhodochrous strain that degrades 2-ethoxyphenol and 4-methoxybenzoate. J Bacteriol 175: 1467- 1474. Kim, J.-S., Dungan, R.S., Kwon, S.-W., and Weon, H.-Y. (2006) The community composition of root-associated bacteria of the tomato plant. World J Microbiol Biotechnol 22: 1267-1273. Larkin, M.J., Kulakov, L.A., and Allen, C.C. (2005) Biodegradation and Rhodococcus-- masters of catabolic versatility. Curr Opin Biotechnol 16: 282-290. Lenke, H., Pieper, D.H., Bruhn, C., and Knackmuss, H.J. (1992) Degradation of 2,4- dinitrophenol by two Rhodococcus erythropolis strains, HL 24-1 and HL 24-2. Appl Environ Microbiol 58: 2928-2932. Llyod-Jones, G., Laurie, A.D., Hunter, D.W.F., and Fraser, R. (1999) Analysis of catabolic genes for naphthalene and phenanthrene degradation in contaminated New Zealand soils. FEMS Microbiol Ecol 29: 69-79. 60

Maidak, B.L., Cole, J.R., Lilburn, T.G., Parker, C.T., Jr., Saxman, P.R., Farris, R.J. et al. (2001) The RDP-II (Ribosomal Database Project). Nucleic Acids Res 29: 173-174. Mandri, T., and Lin, J. (2007) Isolation and characterization of engine oil degrading indigenous microorganisms in Kwazulu-Natal, South Africa. Afr J Biotechnol 6: 23- 27. Massol-Deya, A., Weller, R., Rios-Hernandez, L., Zhou, J.Z., Hickey, R.F., and Tiedje, J.M. (1997) Succession and convergence of biofilm communities in fixed-film reactors treating aromatic hydrocarbons in groundwater. Appl Environ Microbiol 63: 270-276. Muller, R.H., and Babel, W. (1996) Growth rate-dependent expression of phenol-assimilation pathways in Alcaligenes eutrophus JMP134-the influence of formate as an auxiliary energy source on phenol conversion characteristics. Appl Microbiol Biotechnol 46: 156-162. Ogino, A., Koshikawa, H., Nakahara, T., and Uchiyama, H. (2001) Succession of microbial communities during a biostimulation process as evaluated by DGGE and clone library analyses. J Appl Microbiol 91: 625-635. Okuta, A., Ohnishi, K., and Harayama, S. (1998) PCR isolation of catechol 2,3-dioxygenase gene fragments from environmental samples and their assembly into functional genes. Gene 212: 221-228. Paller, G., Hommel, R.K., and Kleber, H.P. (1995) Phenol degradation by Acinetobacter calcoaceticus NCIB 8250. J Basic Microbiol 35: 325-335. Rivas, R., Garcia-Fraile, P., Mateos, P.F., Martinez-Molina, E., and Velazquez, E. (2007) Characterization of xylanolytic bacteria present in the bract phyllosphere of the date palm Phoenix dactylifera. Lett Appl Microbiol 44: 181-187. Saadoun, I. (2002) Isolation and characterization of bacteria from crude petroleum oil contaminated soil and their potential to degrade diesel fuel. J Basic Microbiol 42: 420-428. Sandhu, A., Halverson, L.J., and Beattie, G.A. (2007) Bacterial degradation of airborne phenol in the phyllosphere. Environ Microbiol 9: 383-392. Sei, K., Sugimoto, Y., Mori, K., Maki, H., and Kohno, T. (2003) Monitoring of alkane- degrading bacteria in a sea-water microcosm during crude oil degradation by polymerase chain reaction based on alkane-catabolic genes. Environ Microbiol 5: 517-522. Sei, K., Inoue, D., Wada, K., Mori, K., Ike, M., Kohno, T., and Fujita, M. (2004) Monitoring behaviour of catabolic genes and change of microbial community structures in seawater microcosms during aromatic compound degradation. Water Res 38: 4405- 4414. Shafer, W.E., and Schonherr, J. (1985) Accumulation and transport of phenol, 2-nitrophenol, and 4-nitrophenol in plant cuticles. Ecotoxicol Environ Saf 10: 239-252. Song, J., Sung, J., Kim, Y.M., Zylstra, G.J., and Kim, E. (2000) Roles of the meta-and the ortho-cleavage pathways for the efficient utilization of aromatic hydrocarbons by Sphingomonas yanoikuyae B1. J Microbiol 38: 245-249. Tay, S.T., Moy, B.Y., Maszenan, A.M., and Tay, J.H. (2005) Comparing activated sludge and aerobic granules as microbial inocula for phenol biodegradation. Appl Microbiol Biotechnol 67: 708-713. 61

Thompson, J.D., Higgins, D.G., and Gibson, T.J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucl Acids Res 22: 4673- 4680. Valverde, A., Igual, J.M., Santa Regina, I., Salazar, S., and Rodriguez-Barruenco, C. (2005) Preliminary diversity studies of culturable phylloshere bacteria on chestnut (Castanea Sativa). Acta Hort. (ISHS) 693: 263-270. Van Aken, B., Yoon, J.M., and Schnoor, J.L. (2004) Biodegradation of nitro-substituted explosives 2,4,6-trinitrotoluene, hexahydro-1,3,5-trinitro-1,3,5-triazine, and octahydro-1,3,5,7-tetranitro-1,3,5-tetrazocine by a phytosymbiotic Methylobacterium sp. associated with poplar tissues (Populus deltoides x nigra DN34). Appl Environ Microbiol 70: 508-517. Van der Geize, R., and Dijkhuizen, L. (2004) Harnessing the catabolic diversity of rhodococci for environmental and biotechnological applications. Curr Opin Microbiol 7: 255-261. Van der Meer, J.R. (2006) Environment pollution promotes selection of microbial degradation pathways. Front Ecol Environ. 4: 35-42. Van Schie, P.M., and Young, L.Y. (1998) Isolation and characterization of phenol-degrading denitrifying bacteria. Appl Environ Microbiol 64: 2432-2438. Watanabe, K., Futamata, H., and Harayama, S. (2002) Understanding the diversity in catabolic potential of microorganisms for the development of bioremediation strategies. Antonie van Leeuwenhoek 81: 655-663. Watanabe, K., Teramoto, M., Futamata, H., and Harayama, S. (1998) Molecular detection, isolation, and physiological characterization of functionally dominant phenol- degrading bacteria in activated sludge. Appl Environ Microbiol 64: 4396-4402. Watanabe, K., Hino, S., Onodera, K., Kajie, S., and Takahashi, N. (1996) Diversity in kinetics of bacterial phenol-oxygenating activity. J Ferment Bioengin 81: 560-563. Wilson, M., and Lindow, S.E. (1992) Relationship of total viable and culturable cells in epiphytic populations of Pseudomonas syringae. Appl Environ Microbiol 58: 3908- 3913. Xu, Y., Zhang, W., Chen, M., Lin, M., Li, J.M., and Fang, X.J. (2000) Isolation and identification of a phenol-degrading bacterial strain. Acta Sci Circumstantiae 20: 450- 455. Yang, C.H., Crowley, D.E., Borneman, J., and Noel, T.K. (2001) Microbial phyllosphere populations are more complex than previously realized. Proc Natl Acad Sci USA 98: 3889-3894. Zahn, J.A., Hatfield, J.L., Do, Y.S., DiSpirito, A.A., Laird, D.A., and Pfeiffer, R.L. (1997) Characterization of volatile organic emissions and wastes from a swine production facility. J Environ Qual 26: 1687-1696. Zahn, J.A., DiSpirito, A.A.,Do, Y.S., Brooks, B.E., Cooper, E.E., and Hatfield, J.L. (2001) Correlation of human olfactory responses to airborne concentrations of malodorous volatile organic compounds emitted from swine effluent. J Environ Qual 30: 624- 634. 62

Zhang, X., Gao, P., Chao, Q., Wang, L., Senior, E., and Zhao, L. (2004) Microdiversity of phenol hydroxylase genes among phenol-degrading isolates of Alcaligenes sp. from an activated sludge system. FEMS Microbiol Lett 237: 369-375.

Chapter 4. Effect of phenol on Pseudomonas syringae as determined by a transcriptome analysis of the response of P. syringae to phenol and other abiotic stresses

Introduction The bacterial genes or proteins induced in response to a chemical agent may reveal the mechanism of toxicity of that compound. For example, the induction of the oxidative stress response involving anti-oxidative enzymes of the OxyR and SoxRS regulons in response to hydrogen peroxide (H2O2) suggests that H2O2 causes oxidation of cellular components, such as the DNA, lipids and proteins (Zheng et al., 2001). Similarly, the exposure of E. coli to potassium tellurite resulted in the induction of anti-oxidative enzymes, namely superoxide dismutases, suggesting that tellurite causes oxidative stress, possibly via the production of superoxide radicals in the cell cytoplasm. Yet another study showed that E. coli exposure to selenium oxides induced manganese and iron superoxide dismutases (sodA and sodB) and went further to show increased sensitivity of superoxide dismutase mutants to selenium oxide, demonstrating that superoxide production is a major mechanism of selenium toxicity in these cells (Bebien et al., 2002). These studies illustrate how identifying the types of proteins induced in response to toxic compounds can be helpful in understanding the modes of action of the compounds and may provide insight into the mechanisms that bacteria employ to tolerate the stresses imposed by these toxic compounds.

Phenol is an organic solvent that is considered hydrophobic enough to partition into the bacterial cell membrane (Heipieper et al., 1994). Previous studies have suggested that organic solvents affect the cell membrane by disrupting its structural integrity and function, resulting in increased membrane fluidity and permeabilization (Heipieper and de Bont, 1994; Weber and de Bont, 1996; Segura et al., 1999). For example, in E. coli, phenol was found to result in the leakage of potassium ions and ATP (Heipieper et al., 1991) and in Pseudomonas putida, the two major mechanisms of phenol tolerance included changes in cell membrane 64

composition and the level of saturation of membrane fatty acids (Diefenbach et al., 1992; Heipieper et al., 1992).

Surprisingly, little is known about the toxic effects of phenol on constituents in the bacterial cell cytoplasm once phenol has passed through the cell membrane. This is particularly unexpected given that phenol has been suggested to cause oxidative stress in bacteria. For example, degradation of aromatic substrates, including phenol and 4- chlorophenol by mono-and dioxygenases, produces reactive oxygen species in Ochrobactrum anthropi (Tamburro et al., 2001). Similarly, exposure of P. putida to phenol also resulted in up-regulation of some oxidative stress proteins (Santos et al., 2004). The exposure of Bacillus subtilis to phenol and catechol caused an induction of the Spx (suppressor of clpP and clpX) disulfide stress regulon and a thiol-specific oxidative stress response, respectively. These results suggest that phenol generates non-native disulfide bonds and catechol generates oxygen species, both of which reflect the imposition of oxidative stress by these organic solvents (Tam le et al., 2006).

In the present study, we aimed to understand the mechanism of phenol tolerance and toxicity in bacteria during its entry through cell membranes and after it enters the bacterial cell. Cells were exposed not only to phenol but also to H2O2, with the extent of overlap between these stress regulons used to evaluate the extent of oxidative stress caused by phenol in bacteria. Bacterial cells were also exposed to water stress, to evaluate the effects of a stress known to impact the bacterial cell membrane. The cells were exposed to these stresses for 15 min to understand the short-term adaptation response in bacteria upon exposure to phenol. Lack of availability of water to a bacterial cell imposes two kinds of stress (Halverson and Firestone, 2000), osmotic stress (Csonka, 1989) and matric stress (van de Mortel and Halverson, 2004). Osmotic stress occurs when bacteria are present in an environment high in solutes, such as sodium chloride, whereas matric stress occurs when a cell is present in an environment where water is physically absent or is bound to particles such as soil or high molecular weight polyethylene glycol (PEG) (MW ≥ 8000), a compound that cannot cross the outer membrane. Osmotic stress is known to cause changes in the bacterial cell 65

membrane, including a shift in the cis/trans ratio of unsaturated fatty acids in the membrane to maintain membrane fluidity (Heipieper et al., 1996; Halverson and Firestone, 2000). The extent of overlap between genes affected by water stress and phenol will provide insight into the extent to which phenol damages the cell membrane.

The majority of the previous studies examining the effect of toxic compounds in bacteria have been done with bacterial strains that can degrade these compounds. For example, the effects of the aromatic compounds phenol, catechol, and benzoate have been examined in Acinetobacter species (Benndorf et al., 2001; Pessione et al., 2003), a soil bacterium that can metabolize various xenobiotic compounds. A proteomic study with Pseudomonas putida strain KT2440, a model bacterial strain with diverse metabolic capabilities, was performed to gain insight into global mechanisms underlying solvent toxicity and tolerance (Santos et al., 2004). In terms of the mechanisms of tolerance, degradation of toxic organic solvents may reduce solvent toxicity, thus increasing bacterial tolerance (Ferrante et al., 1995). The modes of toxicity of a compound can be different in bacteria capable of degrading a toxic compound versus a non-degrading strain, as suggested by the finding that polychlorinated biphenyl (PCB) toxicity in Burkholderia xenovorans resulted from the production of toxic intermediates produced during metabolism under PCB- degrading conditions (Parnell et al., 2006). To better understand the mode of toxicity of phenol, we examined its effects on Pseudomonas syringae pv. tomato strain DC3000, a non- phenol degrading bacterial strain. P. syringae cells commonly reside on leaf surfaces where they may be exposed to a variety of stresses, such as airborne pollutants and rapidly fluctuating temperatures, humidity, and solar radiation. These stressful conditions on leaf surfaces can affect bacterial survival and thus the effective removal of airborne pollutants by leaf-associated bacteria.

Global gene expression analysis is a powerful tool for exploring bacterial responses to chemical or environmental stresses, providing insights into the mechanisms that bacteria use for stress tolerance. A number of such studies have been performed with the model bacterial species E. coli and B. subtilis to identify the mechanisms that bacteria employ for adaptation 66

or tolerance to antibiotics, H2O2, and organic solvents (Zheng et al., 2001; Phadtare et al., 2002; Hayashi et al., 2003; Tam le et al., 2006). In this study, we investigated the mechanism of phenol tolerance by comparing the genes induced in response to phenol to those induced by oxidative stress and low water availability, including both osmotic and matric stress. To accomplish this, we performed a transcriptome analysis of exponentially-growing P. syringae pv. tomato DC3000 cells exposed to phenol, H2O2, NaCl, and PEG, and proposed a model for phenol toxicity in bacteria based on the induction of heat shock genes.

Material and Methods

Bacterial strains and growth conditions. The strains used in this study were Pseudomonas syringae pv. tomato strain DC3000 (Whalen et al., 1991) for microarray experiments and E. coli strain DH5α for standard cloning procedures (Sambrook et al., 1989). DC3000 was grown in either ½ 21C medium (Halverson and Firestone, 2000; Axtell and Beattie, 2002) supplemented with 0.3% glucose and 5 mM of sodium succinate (designated ½ 21C) or MinA minimal medium (Miller, 1972) supplemented with 5 mM of sodium succinate (designated MinA) and plated on King’s medium B (KB) (King et al., 1954) agar medium for cell counts when determining EC50 values. E. coli strains were grown in LB (Luria-Bertani) medium (Miller, 1972). The antibiotics used were rifampicin (50 µg/ml) and kanamycin (50 µg/ml).

Determination of EC50 of phenol and hydrogen peroxide.

The effective concentration (EC50) of phenol and hydrogen peroxide, defined as the concentration of a compound required to cause a 50% decrease in the growth rate, was determined in test tubes containing 5 ml of ½ 21C, various concentrations of phenol or hydrogen peroxide, and DC3000 cells at an initial density of 108 cells/ml. The cultures were incubated for 9-12 h at 28°C and the optical density (OD600) was measured over time. The colony forming units (CFU) per ml were determined from the standard curve and the growth rates were determined from the slope of the curve obtained by plotting mean log CFU/ml versus time for each inhibitor concentration. These growth rates were used to calculate the 67

percent rate of growth of treated cells relative to that of the untreated cells. The EC50 values were determined for phenol and hydrogen peroxide by plotting the % rate of growth versus inhibitor concentration.

DNA microarray experiments and data analysis. On each of four days, two replicate cultures were prepared. For each, a colony of DC3000 from KB agar was inoculated into ½ 21C medium and grown until dense for 24 h. The cells were sub-cultured twice in fresh medium, grown to mid-log phase (108 cells/ml), centrifuged at 5000 x g for 10 min, and resuspended in pre-warmed ½ 21C medium. For each culture, two replicate tubes were prepared in which 2.5x108 cells were exposed to either 5 mM of phenol, 1 mM of H2O2, 220 mM of NaCl, 16% PEG or no stress in a 1.1-ml final volume. The cells for each of the five treatments were incubated on a shaker for 15 min at 28°C, immediately after which they were amended with RNAprotect bacterial reagent (Qiagen, Valencia, CA) to stabilize the RNA. The cells from replicate tubes for each treatment were combined and the RNA was isolated using the manufacturer’s instructions. After repeating the treatments and the RNA isolation each day for four days, the RNA isolated for a given treatment on days 1 and 3 were pooled, and that collected on days 2 and 4 were pooled, resulting in two RNA samples that represented each treatment. The quality of the RNA samples was analyzed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA).

RNA samples were sent to NimbleGen Systems Inc. (Madison, WI) for cDNA synthesis, hybridization and scanning. The DC3000-specific DNA microarray was designed and manufactured by NimbleGen Systems Inc., and included 5,608 ORFs, each represented by at least 17 unique 24-mer probe pairs, which were synthesized in situ; each ORF was represented in duplicate, designated as blocks, on the glass slide. Each probe pair consisted of a sequence perfectly matched to the ORF and a sequence that included two mismatched bases to determine background and cross-hybridization. After labeled cDNA synthesis and hybridization, the microarrays were scanned to obtain the signal intensity for each probe. The average value across the probe sets for a single gene was determined for each block, and the 68

values for the two blocks were averaged. The values were normalized with the quantile normalization method and analyzed with the RMA algorithm (Bolstad et al., 2003; Irizarry et al., 2003) by NimbleGen Systems, Inc. For each gene, a multiple comparisons test using a blocked design with 5 treatments and 2 replicates was performed using SAS (SAS Statistical Institute, Cary, NC). This yielded a P-value and Q-value (False Discovery Rate, FDR) for each gene. The extent of induction of a gene by a stress was calculated relative to its expression level in non-stressed cells. Genes were identified as responding to a treatment based on the following criteria: FDR of ≤ 2%, P-value of ≤ 0.05, and at least a two-fold change in response to phenol, NaCl or PEG, and at least a ≥1.3-fold change in response to hydrogen peroxide. A two-fold change cut-off value has been most commonly used in microarray studies to identify genes with altered expression levels; however, the two-fold cut-off value was too stringent for H2O2 treatment based on the number of genes obtained that were significantly up-or down-regulated and the level of change in genes commonly associated with oxidative stress.

Construction of Pst DC3000 heat shock gene mutants. An internal fragment of each of the four induced heat shock genes ftsH, htpX, clpA and lon2 was PCR amplified from DC3000 using the following forward and reverse primers: FtsH-F (5'-GAACAGCAGAGCATCT GGACC-3'), FtsH-R (5'-AATCTGCTCGCGACCACGAAT- 3'), HtpX-F (5'-AATGAG CACCGGTACTCAGATC-3'), HtpX-R (5'-ATACGGGCGAAGA AACATCAC-3'), ClpA-F (5'-ATCTTCCTCAGGCAATCCTCTG-3'), ClpA-R (5'-ATAGG CACCCGCT TCG TCA AT-3'), Lon2-F (5'-AAGGAAGTCGAAGTCGCACGTT-3') and Lon2-R (5'-GGCGCAGGTCCAGGTAATGAT-3'). Each of these fragments was ligated into XcmI-digested T-vector pTOK2 (Kitten and Willis, 1996) into which a lacZ gene containing an XcmI site had been introduced (Chiliang Chen, unpublished data). The plasmids containing these gene fragments were electroporated into E. coli strain DH5α and were then transferred to DC3000 by conjugation in a triparental mating using a helper strain containing the plasmid pRK2073 (Better and Helinski, 1983). The plasmids were inserted into the DC3000 heat shock genes by homologous recombination, causing insertional inactivation of the heat shock genes and resulting in htpX, clpA and lon2 mutants of DC3000. We were not 69

able to obtain a DC3000 ftsH mutant, possibly because ftsH is an essential gene in P. syringae as it is in E. coli (Tomoyasu et al., 1993; Akiyama et al., 1994). The mutations were confirmed by PCR using the forward primer carol-1 (5'- CCCAGTCACGACGTTGTAAAACG-3'), which targets the pTOK2 plasmid, and the reverse primer for the appropriate heat shock gene.

Phenol tolerance and toxicity assays. Phenol tolerance was evaluated by identifying the maximum concentration at which DC3000 cells could initiate growth based on an increase in turbidity over time. The cells were grown in MinA medium until log phase (OD600=0.3-0.4), washed and used as inoculum for fresh MinA containing various phenol concentrations. The growth was measured spectrophotometrically over time. Phenol toxicity was assessed by identifying the concentration at which DC3000 cells exhibited significant death based on culturable counts after phenol exposure for 2-3 h. Phenol toxicity in DC3000 was determined in both exponential (OD600 of 0.2-0.3 ) and stationary phase (OD600 of 1.0-1.2) cells grown in MinA medium. The cells were exposed to various phenol concentrations for 2 to 3 h, appropriate dilutions were plated on KB plates, and cfu/ml were calculated.

Results and Discussion

Selection of stressful conditions used in the microarray studies. To study the gene expression profile of DC3000 in response to phenol and hydrogen peroxide at sub-lethal concentrations, we determined the EC50 value, i.e., the concentration that caused a 50% reduction in the growth rate relative to the rate of growth in the absence of each compound. The EC50 values for phenol and H2O2 were 5 mM and 1 mM, respectively

(Fig. 1), and thus these concentrations were used to evaluate the stress response of DC3000 to phenol and H2O2. DC3000 cells were also exposed to 220 mM NaCl and 16% PEG as these concentrations result in a water potential of approximately -1 MPa in ½ 21C, which was estimated as the amount of water stress that was sensed by DC3000 cells during a hypersensitive response in Arabidopsis thaliana leaves (Wright and Beattie, 2004). 70

A B 110 110 100 100 90 90 80 80 70 70 60 60 50 50 40 40

% Growth rate 30 30 rate % Growth 20 20 % Growth rate rate Growth % 10 10 0 0 0 2 4 6 8 10 12 0 0.5 1 1.5 2 2.5 3 Phenol concentration (mM) Phenol conc (mM) Hydrogen peroxide conc (mM)

Fig. 1. Dose-response curve of P. syringae pv. tomato DC3000 growth rate in the presence of phenol (A) and hydrogen peroxide (B) in ½ 21 C medium. The growth rate was determined from the slope of the curve obtained by plotting the mean log CFU/ml versus time for each inhibitor concentration. These growth rates were used to calculate the percent rate of growth of treated cells relative to that of the untreated cells.

Overview of gene expression changes. To determine the extent of membrane and oxidative stress caused by phenol, DC3000 cells were exposed to phenol, NaCl, PEG and H2O2. A total of 288 genes were significantly induced (Fig. 2A) and 372 were significantly repressed (Fig. 2B) by at least one treatment based on a P-value ≤ 0.05, a Q-value ≤ 0.021, and a change in expression that was at least 2- fold for all treatments and at least 1.3-fold for H2O2. Out of the 288 induced genes, 171 genes (35%) exhibited induction by phenol and at least one other stressor, indicating that these genes might be involved directly or indirectly in tolerance to a variety of stresses and not just phenol. Of the 171 genes, 70 genes (24% of the total genes induced) were induced only by phenol. Of the 103 genes that were repressed by phenol and at least one other stressor, 23 genes (10% of the total genes repressed) were repressed only by phenol, suggesting that these genes are involved in responding specifically to phenol. To better understand the effects of phenol in bacteria and the mechanisms underlying phenol tolerance or protection, the genes that were induced or repressed by phenol are discussed here.

Phenol Phenol (171) (103) A B 23 70 21 10 14 2 17 16 NaCl

H O 26 (110) 59 NaCl 2 2 81 4 H2O2 16 2 (91) (157) (121) 94 8 4 60 1 12 112 8 PEG (243) PEG (110)

Fig. 2. Venn diagrams showing the number of genes induced (A) and repressed (B) by phenol, H2O2, NaCl and PEG and the extent of overlap in the number of genes whose expression changed in response to multiple stresses. The numbers next to the stress in parentheses indicates total number of genes induced (A) or repressed (B) by that stress irrespective of the induction or repression by other stresses.

Genes that were induced by phenol. The genes induced by phenol were categorized based on the cellular processes they are involved in according to the Pseudomonas website (www.pseudomonas.com).

Heat shock response. Heat shock proteins (hsps) are general stress proteins induced by a variety of stresses in which proteins are denatured, including high temperature and exposure to toxic substances such as heavy metals, alcohols, and organic solvents. Denatured proteins in the form of incomplete peptides, unfolded and misfolded proteins, and protein aggregates can result from exposure to these stresses and induce the transcription of hsps. Their induction suggests that the hsps play a role in removing abnormal proteins, which may help the cell adapt to or tolerate the stresses. Our study showed that the majority of the genes known to be involved in 72

the heat shock response in E. coli were induced by phenol and not by other stresses in DC3000 (Fig. 3).

Phenol 5 4 3 2 1 -2 -3

H2O2 5 4 3 2 1 -2 -3

NaCl 5

Fold or Fold induction repression 4 3 2 1 -2 -3

5 PEG 4 3 2 1 -2 -3 2 K * L S - a 0 E aJ 15 ftsH n o E n slO htpX clpA lo lon-1 HslV dn grpE r ro slVU d clpX htpG sp clpP yegD h hsp2 g g H h

Fig. 3. Expression of heat shock genes after exposure of DC3000 to phenol, H2O2, NaCl and PEG. The genes htpX, clpA, ftsH and lon-2 were induced more than 2-fold by phenol. 73

Positive values indicate the fold induction and negative values indicate the fold repression relative to unstressed cells.

Among the heat shock genes induced by phenol in DC3000 were genes that were induced by phenol in other species. For example, grpE was induced by phenol in P. putida KT2440, and moreover was induced to similar levels in DC3000 (1.8-fold) and P. putida (1.5- to 1.7-fold) (Santos et al., 2004). The heat shock genes dnaK, groEL, and groES were induced by phenol in B. subtilis (Tam le et al., 2006), as they were in DC3000. In fact, the study by Tam le et al. (2006) showed that phenol induced a classical heat shock response in a 10-20 min exposure in B. subtilis based on transcriptome and proteome analyses. Similarly, phenol induced a heat shock response in Acinetobacter calcoaceticus whereas catechol, which differs from phenol only in the number of hydroxyl groups, did not, although it induced an oxidative stress response (Benndorf et al., 2001).

Among the heat shock genes induced by phenol in DC3000 were genes encoding a chaperone, ClpA, and two proteases, Lon-2 and FtsH, all of which were induced approximately 2-fold (Table 1 and Fig. 3). Although the protease HtpX was induced by both phenol and H2O2 (Fig. 3), the induction was much higher by phenol (5-fold) than by H2O2 (1.4-fold). FtsH and HtpX are membrane proteases while Lon-2 is a cytoplasmic protease. FtsH is a membrane-bound ATP-dependent zinc metalloprotease known to degrade unstable membrane proteins (Akiyama et al., 2004; Ito and Akiyama, 2005), including SecY (Kihara et al., 1995), a subunit of protein translocase, and the F0 subunit of the proton ATPase (Akiyama et al., 1996) when they fail to assemble. FtsH also degrades some unstable cytoplasmic proteins such as the heat shock regulator σ32 (Herman et al., 1995; Tomoyasu et al., 1995). HtpX is another membrane-bound zinc metalloprotease induced by phenol. It is an energy-independent endopeptidase that can also act on both membrane and cytoplasmic proteins. The induction of genes for membrane proteases by phenol suggests that while crossing the cell membrane phenol may denature some membrane proteins, resulting in the induction of FtsH and HtpX to remove or repair damaged membrane proteins. Two other genes induced by phenol encode ClpA, a chaperone, and Lon-2, a protease that acts primarily 74

on damaged cytoplasmic proteins. The ClpA chaperone forms a multisubunit complex with the ClpP protease to recognize, unfold, and degrade damaged cytoplasmic proteins, with the unfolding requiring ATP hydrolysis. Lon is an ATP-dependent cytosolic protease involved in the degradation of various regulatory and abnormal proteins (Gottesman, 1989; Goldberg, 1992; Goldberg et al., 1994). Lon was shown to be a major protease degrading abnormally folded proteins in E. coli based on that E. coli mutants lacking Lon, although viable, contained three times more aggregated proteins than wild type (Rosen et al., 2002). The induction of cytoplasmic proteases by phenol suggests that upon entry into the cell cytosol, phenol causes the production of abnormal proteins by unfolding or misfolding them. These abnormal proteins may aggregate in the cytoplasm causing the induction of chaperones and proteases to either unfold the misfolded proteins and refold them into their proper conformation or remove them by degradation.

Our transcriptome data suggests that the induction of heat shock genes is important to phenol tolerance in DC3000 cells. This has also been suggested in another study where transcriptional analysis of Clostridium acetylbutylicum exposed to butanol resulted in Lon induction (Tomas et al., 2004) and thus may be involved in solvent tolerance. Based on the finding that all of the heat shock genes that were induced by phenol were induced specifically by phenol and not by other environmental stresses, we propose a model for the effect of phenol in which protein damage (protein misfolding, unfolding, aggregation, etc.) occurs as phenol enters the cell through the outer membrane, into the periplasmic space and through the inner membrane into the cytosol (Fig. 4). The induction of both membrane and cytoplasmic proteases by phenol suggests that periplasmic or membrane proteins as well as cytoplasmic proteins are damaged by phenol.The damaged proteins cause the induction of heat shock proteins including chaperones, such as ClpA and proteases FtsH, HtpX, and Lon- 2 to remove the damaged proteins or refold them into proper conformation. 75

Phenol OM

FtsH

HtpX

ClpA PP

Lon-2

Fig. 4. Schematic diagram. Model for the effect of phenol on bacterial cell protein quality. The entry of phenol (represented by hexagons) through the cell outer membrane (OM), periplasmic space (PP), cell membrane (CM) into the cytosol results in protein denaturation by misfolding, unfolding, and aggregation of proteins. The denatured proteins sensed by extracytoplasmic and cytosolic sigma factors result in increased transcription of chaperones, such as ClpA and/or proteases, FtsH, HtpX, and Lon-2, to either degrade or refold the proteins into their native conformation.

Table 1. Induction of the putative heat shock genes in P. syringae pv. tomato DC3000 to 15 min of exposure to phenol, H2O2, NaCl and PEG based on a microarray analysis. Genes with P ≤ 0.05, Q ≤ 2.1% and at least a 2-fold induction in response to at least one treatment, or at least a 1.3-fold induction in response to H2O2, were considered significantly induced and are shown in bold. Gene symbol Gene # Fold induction relative to unstressed cells Annotated Function

Phenol H2O2 NaCl PEG htpX PSPTO1778 5.0 1.4 0.9 1.1 Zn-dependent protease with chaperone function ftsH PSPTO4497 2.1 1.1 0.8 0.8 ATP-dependent Zn protease present in cytoplasmic membrane, involved in protein turnover clpA PSPTO3353 2.1 1.2 0.9 0.6 ATPase with chaperone activity, ATP-binding subunit lon-2 PSPTO4210 2.0 1.2 1.2 1.0 ATP-dependent Lon protease involved in degradation of intracellular proteins lon-1 PSPTO3724 1.8 1.3 0.6 0.6 ATP-dependent Lon protease involved in degradation of intracellular proteins grpE PSPTO4506 1.8 1.1 0.9 0.7 Cytoplasmic protein chaperone hslV PSPTO5140 1.8 1.4 1.1 1.0 ATP-dependent protease, peptidase subunit dnaK PSPTO4505 1.8 1.1 1.0 0.9 Molecular chaperone belonging to Hsp70 family heat shock protein, PSPTO2170 1.7 1.2 0.7 0.5 Small heat shock protein, acts as a chaperone Hsp20 family groEL PSPTO4376 1.5 1.0 0.6 0.8 chaperonin belonging to Hsp60 family groES PSPTO4377 1.5 1.1 0.6 0.9 Co-chaperonin GroE uspA PSPTO3520 1.5 1.2 0.7 0.6 universal stress protein family; small cytoplasmic protein that shows expression upon exposure to stress. hslU PSPTO5141 1.4 1.2 1.1 1.1 ATP-dependent protease, ATPase subunit dnaJ PSPTO4504 1.4 0.9 0.7 0.6 Molecular chaperone clpP PSPTO3409 1.3 1.1 0.5 0.4 Protease subunit of ATP-dependent Clp proteases clpX PSPTO3725 1.2 1.2 0.5 0.5 ATP-dependent Clp protease, ATPase subunit 76 hslR PSPTO0237 1.1 1.0 1.0 1.0 Ribosome-associated heat shock protein implicated in the recycling of the 50S subunit

htpG PSPTO2208 1.1 1.0 0.7 0.7 Molecular chaperone belonging to Hsp90 family yegD PSPTO4898 0.8 0.8 0.8 0.7 ATP-dependent protein chaperone belonging to Hsp70 family hslO PSPTO0238 0.8 0.9 0.8 0.7 Disulfide bond chaperone of Hsp33 family tig PSPTO3727 0.6 0.9 0.6 0.5 ATP-dependent chaperone with peptidyl-prolyl-cis- trans-isomerase activity; involved in cell division 77 78

Response to low water availability. Phenol and water stress strongly induced the extracytoplasmic sigma factor AlgU, an E. coli RpoE homolog, and its regulators MucA and MucB (Table 2). AlgU/RpoE is crucial in the adaptation of Pseudomonas fluorescens to dry conditions and hyperosmolarity (Schnider-Keel et al., 2001), as well as is induced in E. coli in response to stresses that result in the accumulation of misfolded proteins in the periplasmic space (Mecsas et al., 1993; Raina et al., 1995; Missiakas et al., 1996; Rouviere and Gross, 1996). AlgU is also known to upregulate the expression of extracytoplasmic proteases and chaperones as well as genes involved in the synthesis of the extracellular polysaccharide alginate. The induction of extracytoplasmic proteases by AlgU in response to the accumulation of abnormal proteins in the periplasmic space suggests that phenol may cause damage to outer membrane and or periplasmic proteins, and that this damage results in the induction of AlgU to repair or degrade the damaged proteins. There were 91 genes that were induced by phenol and water stress (Fig. 2) whereas only 26 genes were induced by both phenol and hydrogen peroxide. Based on the relative number of genes induced, the effect of phenol is more similar to that of water stress than H2O2, suggesting greater membrane damage than oxidative damage within the cytosol.

The genes involved in the synthesis of the compatible solutes N- acetylglutaminylglutamine amide (NAGGN) and trehalose were also highly induced by phenol, NaCl and PEG (Table 3). The two genes PSPTO1632 and PSPTO1633 are 68% and 75% identical to acetyltransferase PA3460 and glutamine amidotransferase PA3459, respectively, which have been shown to be involved in NAGGN synthesis in P. aeruginosa (Aspedon et al., 2006). The genes of three putative trehalose biosynthetic operons and those involved in periplasmic glucan biosynthesis were also induced by water and phenol stress (Table 3). These results further support a similarity in the responses of DC3000 to phenol and water stress.

Table 2. Induction of DC3000 genes putatively involved in the synthesis or regulation of alginate by phenol, H2O2, and osmotic and matric stress. Gene symbol Gene # Fold induction relative to Annotated Function unstressed cells Phenol H2O2 NaCl PEG Alginate biosynthesis algU PSPTO4224 5.4 1.0 5.5 8.5 RpoE, DNA-directed RNA polymerase specialized sigma subunit mucA PSPTO4223 5.5 1.2 4.4 7.1 algU negative regulator mucB PSPTO4222 5.4 1.0 6.6 9.8 algU regulatory protein mucD PSPTO4221 1.3 1.1 0.9 1.0 serine protease rpoH PSPTO0430 1.8 1.0 1.0 1.4 DNA-directed RNA polymerase specialized sigma subunit algR PSPTO0127 2.4 1.2 2.2 2.1 alginate biosynthesis regulatory protein algZ/fimS PSPTO0126 1.3 1.0 1.2 1.3 alginate biosynthesis protein algC PSPTO0083 2.6 1.3 2.7 3.9 phosphomannomutase algQ/algR1 PSPTO0134 0.8 1.3 0.8 0.7 transcriptional regulator PSPTO0135 0.9 1.0 0.8 0.7 protein folding and export; peptidyl-prolyl cis-trans isomerase, FKBP-type algP/algR3 PSPTO0136 0.9 1.1 0.6 0.5 alginate regulatory protein algB PSPTO0334 1.7 1.1 1.1 1.0 alginate biosynthesis transcriptional regulatory protein algD PSPTO1243 2.4 1.1 3.4 4.8 GDP-mannose 6-dehydrogenase alg8 PSPTO1242 1.1 1.0 1.4 1.9 alginate biosynthesis protein alg44 PSPTO1241 1.2 1.1 2.2 2.4 alginate biosynthesis protein algK PSPTO1240 0.9 1.0 1.4 1.6 alginate biosynthesis protein algE PSPTO1239 1.0 1.0 1.3 1.5 alginate biosynthesis protein algG PSPTO1238 1.0 1.0 1.7 1.8 alginate biosynthesis protein algX PSPTO1237 1.0 1.0 1.3 1.4 alginate biosynthesis protein algL PSPTO1236 1.1 1.0 1.5 1.5 alginate lyase algI PSPTO1235 1.0 1.0 1.2 1.3 alginate biosynthesis protein algJ PSPTO1234 0.9 1.0 1.1 1.2 alginate biosynthesis protein algF PSPTO1233 1.0 1.0 1.1 1.1 alginate biosynthesis protein 79 algA PSPTO1232 1.0 1.0 1.1 1.0 alginate biosynthesis protein

Table 3. Induction of DC3000 genes putatively involved in the biosynthesis of osmoprotectants and periplasmic glucan by phenol,

H2O2, and osmotic and matric stress.

Gene symbol Gene # Fold induction relative to Annotated Function unstressed cells Phenol H2O2 NaCl PEG Compatible solute synthesis N- acetylglutaminylglutamine amide asnB PSPTO1633 5.7 1.1 10.9 23.6 asparagine synthetase, glutamine-hydrolyzing PSPTO1632 3.6 1.2 9.8 21.6 acetyltransferase, GNAT family PSPTO1631 1.8 0.9 5.6 10.6 Cellulase M and related proteins PSPTO1630 1.1 1.1 4.6 8.1 conserved hypothetical protein Trehalose PSPTO3125 6.9 1.1 3.0 4.6 glycogen synthase PSPTO3129 4.2 1.2 2.4 2.9 Maltooligosyl trehalose synthase PSPTO3127 3.0 1.0 2.2 2.4 4-alpha-glucanotransferase PSPTO3126 2.7 1.1 2.4 3.4 alpha-amylase family PSPTO2760 2.7 1.1 3.0 4.1 alpha-amylase family PSPTO3128 1.9 1.1 1.4 1.8 glycosyl hydrolase, family 13 PSPTO2761 1.5 1.0 1.8 2.0 alpha-amylase family PSPTO2762 1.3 1.1 1.3 1.4 1,4-alpha-glucan branching enzyme glgX PSPTO3130 2.2 1.1 1.9 2.2 glycogen operon protein PSPTO3131 1.1 1.0 1.4 1.4 Acyl-CoA dehydrogenases PSPTO3132 1.0 1.0 1.2 1.3 conserved domain Periplasmic glucan biosynthesis PSPTO5162 2.2 1.0 1.8 2.3 periplasmic glucan biosynthesis protein PSPTO5161 1.7 1.0 1.8 1.9 periplasmic glucan biosynthesis protein 80 81

Oxidative stress response. Phenol has been suggested to cause an oxidative stress in P. putida KT2440 due to the increased expression of some genes that encode antioxidants, including AhpC (Alkyl hydroperoxide reductase, subunit C), SodB (Superoxide dismutase), Tpx (Thiol peroxidase), DsbA (Thiol:disulfide interchange protein) and Tig (Trigger factor, presumably required for protein folding or accelerated protein export) (Santos et al., 2004). None of these genes were induced by phenol in DC3000 and there was little overlap between the genes induced by

H2O2 and those induced by phenol (Table 4), suggesting that, unlike in P. putida, phenol probably does not result in an oxidative stress response in DC3000. Also, the genes that were most strongly induced by H2O2, the catalase-encoding katB and katG genes and ahpC and ahpF genes, were not induced by phenol (Fig. 5). AhpC and AhpF are the antioxidant enzymes involved in the catalysis of organic peroxide reduction and KatG is a catalase- peroxidase involved in the detoxification of H2O2. KatG, AhpF and AhpC are regulated by the transcription factor OxyR (VanBogelen et al., 1987; Tartaglia et al., 1989) and have been shown to be strongly induced by H2O2 (Farr and Kogoma, 1991; Zheng et al., 2001; Zeller and Klug, 2004). Our results are consistent with a previous proteomic study that showed induction of AhpF and AhpC upon exposure to H2O2 but not to phenol (Benndorf et al., 2001). Hydrogen peroxide causes oxidative stress by producing reactive oxygen species that can damage cytoplasmic and membrane lipids and proteins, and DNA by peroxidation, thereby, resulting in the upregulation of catalases and hydroperoxidases involved in the removal of reactive oxygen radicals (Demple, 1991). Our results indicate that the toxic effects of phenol in DC3000 are different than those of H2O2.

Table 4. Induction of the putative oxidative stress genes in DC3000 by phenol, H2O2, NaCl, and PEG. Gene symbol Gene # Fold induction relative to Annotated Function unstressed cells Phenol H2O2 NaCl PEG Catalases (convert H2O2 to H2O) katE PSPTO5263 1.5 1.1 1.5 1.9 Catalase katB PSPTO3582 1.2 3.7 0.9 0.9 Catalase katG PSPTO4530 1.1 2.6 0.5 0.4 Catalase/hydrogen peroxidase (HP) I Superoxide dismutases (SOD) (convert superoxide to H2O2) sodC PSPTO1338 1.7 1.1 1.3 1.5 Cu Zn-containing SOD sodA PSPTO4459 1.0 1.1 1.0 1.1 Mn-containing SOD sodB PSPTO4363 0.9 0.8 0.8 0.7 Fe-containing SOD Regulators of the oxidative stress response oxyR PSPTO0066 1.4 0.9 1.1 1.0 Oxidative stress regulatory protein fur PSPTO4508 1.0 1.0 1.0 0.8 Ferric uptake regulatory protein Peroxidases PSPTO1719 1.1 1.1 0.8 0.9 Glutathione peroxidase family protein catalyzes reduction of hydroperoxides with glutathione (GSH) PSPTO1172 0.8 0.9 0.7 0.6 Glutathione peroxidase family protein PSPTO1781 0.8 1.0 1.0 0.8 Glutathione peroxidase family protein PSPTO3598 0.8 1.0 1.0 1.0 dyp-type peroxidase family protein tpx PSPTO2752 0.7 1.0 0.5 0.3 Thiol peroxidase Glutathione reductases gor-1 PSPTO3112 1.1 1.2 0.9 0.9 Glutathione disulfide reductase maintains levels of antioxidant glutathione, GSH gor-2 PSPTO4254 0.9 1.0 1.0 1.1 Glutathione disulfide reductase maintains levels of antioxidant GSH Peptide methionine sulfoxide reductases msrA PSPTO0405 1.4 1.1 0.7 0.4 Repairs damaged proteins by converting methionine sulfoxide to methionine dsbA PSPTO0341 1.1 1.1 0.9 0.9 Thiol:disulfide interchange protein ahpF PSPTO3107 1.1 1.7 0.9 0.9 Alkylhydroperoxide reductase, subunit F ahpD PSPTO2816 1.0 1.0 1.0 1.1 Alkylhydroperoxidase domain protein 82 ahpC PSPTO3108 0.8 1.9 0.7 0.4 Alkylhydroperoxide reductase, subunit C 83

3 Phenol

-2

-3 4

H2O2 3

-2

-3 4 NaCl 3 Fold induction or repression or induction Fold

-2

-3 4 PEG 3

-2

-3 E tB tG pC rA dA dC t fur a odB k ka ah ahpF ms so so ka ahpD oxyR s

Fig. 5. Expression of genes involved in oxidative stress response 15 min after exposure to phenol, H2O2, NaCl and PEG in DC3000. The genes encoding the antioxidant enzymes,

KatB, KatG, AhpC and AhpF were induced > 1.3-fold only by H2O2. Positive values indicate the fold induction and negative values indicate the fold repression relative to unstressed cells.

Functions of cell membrane proteins. Despite the apparent induction of membrane stress suggested by the overlap between the phenol- and water stress-responsive regulons, as described above, the expression of many membrane proteins was not affected by phenol (Table 5). For example, the cti gene encoding the cis/trans isomerase, an enzyme involved in fatty acid composition changes during bacterial adaptation to organic solvents including phenol, was not induced; however, activation of this enzyme likely occurs post-transcriptionally (Kiran et al., 2005) and cti expression in P. putida is known to be constitutive (Heipieper et al., 1992). Consistent with the finding that efflux transporters are often induced by organic solvents, several of the efflux transporter proteins in DC3000 were induced by phenol (Table 4). Interestingly, the tolC gene (PSPTO4977), which encodes a homolog of an outer membrane efflux protein that was induced by phenol in P. putida (Santos et al., 2004), was not induced.

Table 5. Induction of DC3000 genes putatively encoding cell membrane proteins by phenol, H2O2, NaCl, and PEG. Gene Gene # Fold induction relative to unstressed cells Annotated Function symbol Phenol H2O2 NaCl PEG PSPTO4302 1.4 1.0 1.3 1.2 transcriptional regulator, TetR family PSPTO4303 3.0 1.0 1.3 1.5 efflux transporter, RND family, MFP subunit PSPTO4304 1.7 1.0 1.2 1.5 AcrB/AcrD/AcrF family protein PSPTO4305 1.3 1.1 1.0 1.0 outer membrane efflux protein PSPTO4442 0.9 0.7 1.3 1.1 toluene tolerance protein, putative PSPTO4443 1.3 0.8 1.2 1.1 STAS domain protein PSPTO4444 1.9 0.9 1.3 1.2 toluene tolerance protein, putative PSPTO4445 2.1 0.9 1.5 1.3 mce-related protein PSPTO4446 3.0 0.9 1.8 1.9 membrane protein, putative PSPTO4447 2.0 1.0 1.7 1.6 toluene tolerance ABC transporter, ATP-binding protein, putative mexE PSPTO3099 1.0 1.2 0.9 1.0 multidrug efflux RND membrane fusion protein mexF PSPTO3100 1.0 1.1 1.0 1.0 multidrug efflux RND transporter oprN PSPTO3101 0.9 1.1 0.9 1.0 outer membrane efflux protein PSPTO2592 1.7 1.0 0.9 1.1 AcrB/AcrD/AcrF family protein PSPTO2593 1.9 0.9 0.9 1.0 multidrug resistance protein, AcrA/AcrE family tolC PSPTO4977 0.8 1.0 0.8 0.7 outer membrane efflux protein 85

General metabolic functions. Although there was a general decrease in metabolism in DC3000 as expected based on the slowed growth of DC3000 when exposed to 5 mM phenol (Fig. 1) and based on the repression of various metabolic genes described below, some genes involved in general metabolism were induced in response to phenol (Table 6). These included several genes involved in carbohydrate catabolism, such as gpmA, gnl and gcd, which encode a putative 2,3-bisphosphoglycerate-independent phosphoglycerate mutase, gluconolactonase, and glucose dehydrogenase, respectively. The gene gpmA is homologous to pgm of P. putida, which was also induced in response to phenol (Santos et al., 2004). The two genes gnl and gcd encode enzymes that are involved in the pentose phosphate pathway, a pathway that contributes to the production of energy and ribose sugars for biosynthesis. This may suggest that in DC3000 some genes involved in energy production are upregulated to produce energy that could be used to provide phenol tolerance to the cells. Other genes that were induced by phenol and are involved in energy production encode a putative formate dehydrogenase, which allows the use of formate during anaerobic respiration, and azurin, which allows electron transfer during anaerobic growth in bacteria. The induction of these genes adapted for anaerobic growth suggests that the phenol-exposed cells may have encountered oxygen limitation.

Consistent with the induction of some genes involved in amino acid metabolism in P. putida in response to phenol (Santos et al., 2004), a few DC3000 genes related to amino acid metabolism were induced, although most were not. For example, the argA and leuD genes involved in arginine degradation and leucine biosynthesis, respectively, were induced in P. putida in response to toluene, whereas argB involved in arginine biosynthesis was induced in DC3000 and argA and leuD were not (Table 5).

Table 6. Induction of DC3000 genes putatively involved in general metabolic functions following exposure to phenol, H2O2, NaCl, and PEG. Gene Gene # Fold induction relative to Annotated Function symbol unstressed cells Phenol H2O2 NaCl PEG Carbohydrate organic acid metabolism PSPTO5407 3.8 1.2 0.9 1.0 Molybdopterin-binding oxidoreductase fdhD PSPTO 5406 3.6 1.0 1.0 1.0 An accessory protein required for formate dehydrogenase activity gloA PSPTO3106 2.7 1.0 1.9 1.7 lactoylglutathione lyase with ring-cleavage dioxygenase activity, also involved in carbohydrate metabolism fumC-1 PSPTO1731 2.5 1.1 1.4 1.5 Fumarase, involved in TCA cycle

gnl PSPTO1345 2.5 1.1 1.2 1.3 Gluconolactonase gpmA PSPTO5327 2.5 1.1 0.9 0.8 phosphoglycerate mutase, 2,3-bisphosphoglycerate-independent azu PSPTO4923 2.2 1.0 1.3 1.1 Copper-binding protein involved in electron transport gcd PSPTO4196 2.0 1.0 1.1 1.1 glucose dehydrogenase gabD PSPTO0300 1.4 1.3 0.9 0.7 succinate-semialdehyde dehydrogenase Amino acid metabolism PSPTO4171 2.6 1.0 0.6 0.4 Amino acid transporter, binds amino acid in the periplasm PSPTO1343 2.1 1.1 0.9 0.9 Putative dipeptidase involved in protein degradation argB PSPTO0082 2.1 1.2 1.4 1.4 Acetyleglutamate kinase involved in amino acid biosynthesis argA PSPTO0324 1.0 0.8 0.8 1.0 N-acetylglutamate synthase leuD PSPTO2173 0.9 1.1 1.0 1.0 3-isopropylmalate dehydratase, large subunit 87

Transcriptional regulators. At least five genes encoding proteins involved in transcriptional regulation were induced specifically in response to phenol exposure (Table 7). These were psrA (a TetR family Pseudomonas sigma regulator), tvrR (for TetR-like virulence regulator), SigW, and two others belonging to the TetR and LysR family of regulators. The tetracycline repressor (TetR) family of bacterial regulatory proteins allow bacteria to sense and respond quickly to their environment, thus modulating the expression of genes in response to environmental signals, such as hyperosmolarity (Lamark et al., 1991) and antibiotics. In this study, the induction of DC3000 TetR-type regulators by phenol suggests that they may regulate genes important in adaptation and tolerance to phenol. In P. aeruginosa, PsrA positively regulates the rpoS gene at the onset of stationary phase (Kojic and Venturi, 2001; Kojic et al., 2002; Kojic et al., 2005), as well as other novel targets such as an acyl-CoA dehydrogenase, an electron transfer flavoprotein, and an electron transfer flavoprotein-ubiquinone oxidoreductase (Kojic et al., 2005). The acyl-CoA dehydrogenases, in particular, may promote the degradation of fatty acids for energy production to help bacteria cope with stresses such as starvation during the stationary phase. In DC3000, PsrA may also be involved in the induction of acyl-CoA dehydrogenase (PSPTO0500) based on that it was induced 5.5-fold by phenol. TvrR, another TetR family regulator, has been identified as a novel transcriptional regulator that is required for virulence of DC3000 in Arabidopsis thaliana plants and has been speculated to be involved in sensing environmental signals (Preiter et al., 2005). Lastly, the RNA polymerase sigma factor SigW is an extracytoplasmic sigma factor (ECF) of B. subtilis, and thus is known to respond to signals from the extracytoplasmic environment, such as the presence of misfolded proteins in the periplasm. In a previous study (Steil et al., 2003), the SigW regulon was shown to be induced in B. subtilis within 10 min of exposure to high salinity. Of the 75 B. subtilis genes induced immediately after salt shock, 31 were assigned to the SigW regulon, which included transporters for osmoprotectants and enzymes catalyzing the synthesis of compatible solutes. The induction of SigW in DC3000 suggests that phenol may be causing damage to the periplasmic proteins and/or cell membrane, sending signals to this SigW ECF sigma factor to promote the induction of other genes. 89

Table 7. Induction of putative transcriptional regulator genes of DC3000 following exposure to phenol, H2O2, NaCl, and PEG. Gene Gene # Fold induction relative to Annotated Function symbol unstressed cells Phenol H2O2 NaCl PEG Transcriptional regulators psrA PSPTO3508 3.4 0.9 0.9 1.1 Transcriptional regulator PSPTO5462 3.3 1.2 0.9 0.9 LysR family transcriptional regulator tvrR PSPTO3576 2.4 1.0 0.9 0.8 TetR-like virulence regulator PSPTO0378 2.1 1.1 0.6 0.7 DNA-binding heavy metal response regulator sigW PSPTO2298 2.0 1.3 1.3 1.6 Putative RNA polymerase sigma factor

Other cellular functions. Five of the six genes of an operon involved in tellurium resistance, terA, B, D, Z, and a membrane protein belonging to terC family were induced more than 2-fold by phenol, with terD and terZ induced some by H2O2 (Table 8). Tellurite causes toxicity in bacteria by oxidizing and replacing sulfur in cellular proteins, rendering them nonfunctional (Summers and Jacoby, 1977; Garberg et al., 1999). A fivefold induction in tellurium resistance genes as well as significant induction of oxidative stress-responsive genes upon exposure to the antibiotic 4,5-dihydroxy-2-cyclopenten-1-one has also been shown in E. coli (Phadtare et al., 2002), suggesting that some kind of oxidative stress might be responsible for up-regulating tellurium resistance genes. Another study (Turner et al., 1995) has suggested that after tellurium enters a cell through a phosphate transport uptake system it is acted upon by nitrate reductase, which causes reduction of tellurite in the cytoplasm as a basal level of defense against tellurium toxicity. The induction of these genes by phenol suggests that phenol may cause some oxidation of cellular components.

Table 8. Induction of DC3000 genes involved in tellurite resistance following exposure to phenol, H2O2, NaCl, and PEG. Gene Gene # Fold induction relative to Annotated Function symbol unstressed cells Phenol H2O2 NaCl PEG terZ PSPTO0940 3.9 1.3 1.2 1.5 Bacterial stress protein terB PSPTO0942 3.6 1.2 1.0 1.3 Tellurium resistance protein PSPTO2691 3.4 1.1 1.0 1.1 Membrane protein TerC family terA PSPTO0941 3.0 1.2 1.2 1.4 Tellurium resistance protein terD PSPTO0944 2.1 1.3 1.0 1.1 Tellurium resistance protein

Proteins of unknown function. The ten genes that were most induced by phenol (22- to 106-fold) encoded hypothetical proteins with no known function (Table 9). Of the 33 genes that were induced ≥ 6-fold, 24 genes encoded hypothetical proteins. Future studies should be aimed at identifying these genes and their role in DC3000 stress tolerance.

Table 9. Induction of DC3000 genes with unknown functions following exposure to phenol,

H2O2, NaCl, and PEG. Gene Gene # Fold induction relative to unstressed cells symbol Phenol H2O2 NaCl PEG Hypothetical proteins PSPTO2049 106 1.5 61.1 77.9 PSPTO0332 69.0 1.3 46.1 57.5 PSPTO0330 47.3 1.5 27.2 29.0 PSPTO0507 43.7 1.5 17.0 25.3 PSPTO4546 38.1 1.4 10.9 17.2 PSPTO4484 33.6 1.7 24.9 30.9 PSPTO1848 33.3 1.1 22.2 21.9 PSPTO1304 33.2 1.1 32.7 39.1 PSPTO4516 29.8 1.4 12,4 17.4 PSPTO1507 22.3 1.8 12.8 22.3 PSPTO1596 15.4 1.3 4.5 5.4 PSPTO0318 13.5 1.3 3.7 5.7 PSPTO1485 13.2 1.3 4.3 4.6 PSPTO0156 12.7 1.9 11.2 17.5 PSPTO2377 12.5 1.3 8.6 11.3 PSPTO0155 9.8 1.2 7.1 10.8 PSPTO1417 9.0 1.8 2.9 3.2 PSPTO2895 8.7 1.1 2.8 2.5 91

Genes that were repressed by phenol The majority of the genes that were repressed by phenol were predicted to be involved in general metabolism, flagellar biosynthesis and chemotaxis, or protein synthesis. Phenol seemed to result in some membrane stress as suggested by the repression of the genes involved in flagellar biosynthesis (Table 10). This is consistent with the repression in P. putida of the genes involved in chemotaxis and motility as well as the fliC gene encoding the flagellin FliC, which is essential to the formation of flagella, in response to phenol (Santos et al., 2004). The flagellar biosynthesis genes were also repressed by water stress, consistent with water stress causing membrane changes.

A previous proteomic study has shown that exposure of P. putida to toluene results in the upregulation of proteins involved in glucose metabolism to provide energy to extrude toluene out of the cell through efflux pumps (Segura et al., 2005). These proteins were succinyl coenzyme A synthetase (SucC), fructose-1,6-bisphosphate aldolase (Fda), glucokinase (Gck), and oxoglutarate dehydrogenase, lipoamide dehydrogenase component (LpdG). In contrast, DC3000 showed significant repression of the genes in the operons involving SucC and LpdG (Table 10), suggesting a decrease in metabolic activity in response to phenol. In addition, the majority of the genes encoding ribosomal proteins were repressed by phenol but not by other stresses, suggesting a slowed cell growth as a result of phenol exposure. This was expected because the level of phenol used in this study was chosen based on the fact that it resulted in slowed growth without killing of the cells.

Table 10. Repression of various DC3000 genes following exposure to phenol, H2O2, NaCl, and PEG. Only genes specifically repressed by phenol are shown. Gene Gene # Fold repression relative to Annotated Function symbol unstressed cells Phenol H2O2 NaCl PEG Flagellar biosynthesis and chemotaxis proteins flgC PSPTO1934 3.0 0.6 4.3 4.9 flagellar basal-body rod protein flgB PSPTO1933 2.7 0.6 3.7 3.8 flagellar basal-body rod protein flgD PSPTO1935 2.6 0.7 3.2 3.0 basal-body rod modification protein flaG PSPTO1950 3.2 0.9 3.8 6.1 flagellin, putative cheZ PSPTO1981 2.1 0.7 2.4 4.8 chemotaxis protein Proteins involved in general metabolism PSPTO2199 4.3 1.4 1.2 1.8 2-oxoglutarate dehydrogenase, E1 component sucD PSPTO2203 3.4 1.1 1.8 2.4 succinyl-CoA synthase, alpha subunit sucC PSPTO2202 3.2 1.0 1.6 1.7 succinyl-CoA synthase, beta subunit lpdG PSPTO2201 2.9 1.1 1.2 1.1 2-oxoglutarate dehydrogenase, E3 component, lipoamide dehydrogenase PSPTO2200 2.8 1.1 1.2 1.4 2-oxoglutarate dehydrogenase, E2 component, dihydrolipoamide succinyltransferase gck PSPTO1289 1.7 1.1 1.4 1.7 Glucokinase fda PSPTO0390 0.6 1.0 1.1 1.2 fructose-bisphosphate aldolase, class II Translation, ribosomal proteins rpmF PSPTO3835 3.2 1.6 1.1 1.5 50S ribosomal protein L32 rpsU PSPTO0539 2.9 1.2 1.1 0.9 30S ribosomal protein S21 tuf PSPTO0624 2.6 1.1 1.6 1.2 Translation elongation factor Tu rpsT PSPTO0802 2.5 1.2 1.4 1.6 30S ribosomal protein S20 rpmE-2 PSPTO5136 2.4 1.3 1.4 1.6 50S ribosomal protein L31 rpsA PSPTO1750 2.3 1.4 1.0 1.1 30S ribosomal protein S1 rplI PSPTO4930 2.3 1.9 1.4 2.1 ribosomal protein L9 rplL PSPTO0618 2.2 1.2 1.5 1.4 ribosomal protein L7/L12 Hypothetical proteins PSPTO2729 3.2 0.6 0.7 1.0 PSPTO5423 2.3 1.2 1.4 1.9 PSPTO0307 2.2 0.5 1.0 0.7 PSPTO1429 2.1 1.0 0.9 1.4

Phenol tolerance and toxicity in DC3000 heat shock mutants. The role of phenol-induced heat shock proteins, HtpX, ClpA and Lon-2, in providing tolerance to phenol in DC3000 was determined by exposing heat shock mutants to various phenol concentrations in minimal media. We were not able to obtain a DC3000 ftsH mutant, possibly because ftsH is essential and the mutation in this gene in lethal, as has been shown for E. coli (Ogura et al., 1999). In phenol tolerance assays, the cells were grown in phenol and their OD600 was measured over time whereas in toxicity assays, the mutants were exposed to a sudden shock of a high phenol concentration for 2 to 3 h and their survival was determined from viable plate counts. Under the experimental conditions used in this study, we did not see a difference between wild-type DC3000 and any of the mutants in their ability to tolerate phenol (Fig. 6.). However, this does not rule out the role that heat shock proteins may have in protecting the cells from phenol toxicity. Because heat shock proteins have redundant functions it is possible that in the absence of one protease the other takes over the function of removing abnormal proteins from the cell and helping bacterial cells survive under stress.

0 mM phenol 0.5 mM phenol 1 mM phenol 1.4 1.4 1.4 1.2 1.2 1.2 1.0

1.0 1.0 0.8 0.8

600 0.8 600 HtpX 600 0.6 0.6 0.6 ClpA 0.4 0.4 OD OD

0.4 OD Lon2 0.2 0.2 0.2 DC3000 0.0 0.0 0.0 0 10 20 30 0 10 20 30 0 10 20 30 Time (h) Time (h) Time (h)

2.5 mM phenol 5 mM phenol 1.4 1.4 1.2 1.2 1.0 1.0 600 0.8 600 0.8 0.6 0.6 OD 0.4 OD 0.4 0.2 0.2 0.0 0.0 0 10 20 30 0 10 20 30 Time (h) Time (h)

Fig. 6. Phenol tolerance of DC3000 and the heat shock mutants htpX, clpA, and lon-2 in MinA minimal medium amended with phenol at the concentrations indicated.

Conclusion Based on our gene expression data, we proposed a model for the mode of action of phenol in a bacterial cell. Upon entry into the cell through cell membrane, phenol causes denaturation of outer and inner membrane proteins, and periplasmic proteins by misfolding and/or unfolding, resulting in the activation of envelope stress response. The misfolding of proteins in the outer membrane is sensed by AlgU (5.9-fold induced), an E. coli sigma E homologue and an extracytoplasmic sigma factor, which is induced by various agents that cause envelope stress, and the transcription of sigma H, an AlgU regulon member and a positive transcriptional regulator of heat shock genes, is upregulated. Sigma H, in turn induces the transcription of heat shock genes, such as clpA, lon, and ftsH producing proteases, that degrade denatured cytosolic and membrane proteins. AlgU also upregulates the expression of extracytoplasmic proteases and chaperones directly that help degrade or refold the damaged periplasmic and membrane proteins. This constitutes the AlgU signal transduction pathway, which is regulated by two negative regulatory proteins, MucA (induced 5.5-fold) and MucB (induced 5.4-fold), homologues of RseA and RseB of E. coli. MucA is an inner membrane anti-sigma factor that binds to AlgU in the absence of envelope stress. MucB is a periplasmic protein that binds to MucA and stabilizes it. Upon sensing the presence of misfolded periplasmic proteins MucB leaves MucA and binds to the abnormal proteins, thereby destabilizing MucA, which releases AlgU to turn on the transcription of stress response genes.

In conclusion, the transcriptome analysis of P. syringae pv. tomato DC3000 in response to phenol revealed a mechanism of toxicity in bacteria, which was different from the modes of action of other stress agents, including H2O2, PEG, and NaCl. It was interesting to observe that although heat shock response is considered a general stress response that is induced in bacteria by various stresses, including ethanol, antibiotics, and heavy metals, in our study only phenol exposure to DC3000 resulted in the up-regulation of heat shock genes. This suggested that phenol caused denaturation of cytoplasmic, membrane, and/or periplasmic proteins resulting in the activation of sigma 32 regulated-heat shock response. The phenol tolerance and toxicity assays suggested that the heat shock proteins induced due 95

to the appearance of abnormal proteins in the cell have overlapping functions, where one protease takes over for, or complements, the function of the other in its absence under stress.

References Akiyama, Y., Ogura, T., and Ito, K. (1994) Involvement of FtsH in protein assembly into and through the membrane. I. Mutations that reduce retention efficiency of a cytoplasmic reporter. J Biol Chem 269: 5218-5224. Akiyama, Y., Kihara, A., and Ito, K. (1996) Subunit a of proton ATPase F0 sector is a substrate of the FtsH protease in Escherichia coli. FEBS Lett 399: 26-28. Akiyama, Y., Ogura, T., and Ito, K. (eds) (2004) Handbook of proteolytic enzymes. London: Academic Press. Arvidsson, R.H., Nordling, M., and Lundberg, L.G. (1989) The azurin gene from Pseudomonas aeruginosa. Cloning and characterization. Eur J Biochem 179: 195- 200. Aspedon, A., Palmer, K., and Whiteley, M. (2006) Microarray analysis of the osmotic stress response in Pseudomonas aeruginosa. J Bacteriol 188: 2721-2725. Axtell, C.A., and Beattie, G.A. (2002) Construction and characterization of a proU-gfp transcriptional fusion that measures water availability in a microbial habitat. Appl Environ Microbiol 68: 4604-4612. Bebien, M., Lagniel, G., Garin, J., Touati, D., Vermeglio, A., and Labarre, J. (2002) Involvement of superoxide dismutases in the response of Escherichia coli to selenium oxides. J Bacteriol 184: 1556-1564. Benndorf, D., Loffhagen, N., and Babel, W. (2001) Protein synthesis patterns in Acinetobacter calcoaceticus induced by phenol and catechol show specificities of responses to chemostress. FEMS Microbiol Lett 200: 247-252. Better, M., and Helinski, D.R. (1983) Isolation and characterization of the recA gene of Rhizobium meliloti. J Bacteriol 155: 311-316. Bolstad, B.M., Irizarry, R.A., Astrand, M., and Speed, T.P. (2003) A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 19: 185-193. Csonka, L.N. (1989) Physiological and genetic responses of bacteria to osmotic stress. Microbiol Rev 53: 121-147. Demple, B. (1991) Regulation of bacterial oxidative stress genes. Annu Rev Genet 25: 315- 337. Diefenbach, R., Heipieper, H.J., and Keweloh, H. (1992) The conversion of cis- into trans- unsaturated fatty acids in Pseudomonas putida P8: evidence for a role in the regulation of membrane fluidity. Appl Microbiol Biotechnol 38: 382-387. Farr, S.B., and Kogoma, T. (1991) Oxidative stress responses in Escherichia coli and Salmonella typhimurium. Microbiol Rev 55: 561-585. Ferrante, A.A., Augliera, J., Lewis, K., and Klibanov, A.M. (1995) Cloning of an organic solvent-resistance gene in Escherichia coli: the unexpected role of alkylhydroperoxide reductase. Proc Natl Acad Sci U S A 92: 7617-7621. 96

Garberg, P., Engman, L., Tolmachev, V., Lundqvist, H., Gerdes, R.G., and Cotgreave, I.A. (1999) Binding of tellurium to hepatocellular selenoproteins during incubation with inorganic tellurite: consequences for the activity of selenium-dependent glutathione peroxidase. Int J Biochem Cell Biol 31: 291-301. Goldberg, A.L. (1992) The mechanism and functions of ATP-dependent proteases in bacterial and animal cells. Eur J Biochem 203: 9-23. Goldberg, A.L., Moerschell, R.P., Chung, C.H., and Maurizi, M.R. (1994) ATP-dependent protease La (lon) from Escherichia coli. Methods Enzymol 244: 350-375. Gottesman, S. (1989) Genetics of proteolysis in Escherichia coli. Annu Rev Genet 23: 163- 198. Halverson, L.J., and Firestone, M.K. (2000) Differential effects of permeating and nonpermeating solutes on the fatty acid composition of Pseudomonas putida. Appl Environ Microbiol 66: 2414-2421. Hayashi, S., Aono, R., Hanai, T., Mori, H., Kobayashi, T., and Honda, H. (2003) Analysis of organic solvent tolerance in Escherichia coli using gene expression profiles from DNA microarrays. J Biosci Bioeng 95: 379-383. Heipieper, H.J., and de Bont, J.A. (1994) Adaptation of Pseudomonas putida S12 to ethanol and toluene at the level of fatty acid composition of membranes. Appl Environ Microbiol 60: 4440-4444. Heipieper, H.J., Keweloh, H., and Rehm, H.J. (1991) Influence of phenols on growth and membrane permeability of free and immobilized Escherichia coli. Appl Environ Microbiol 57: 1213-1217. Heipieper, H.J., Diefenbach, R., and Keweloh, H. (1992) Conversion of cis unsaturated fatty acids to trans, a possible mechanism for the protection of phenol-degrading Pseudomonas putida P8 from substrate toxicity. Appl Environ Microbiol 58: 1847- 1852. Heipieper, H.J., Meulenbeld, G., van Oirschot, Q., and de Bont, J. (1996) Effect of environmental factors on the trans/cis ratio of unsaturated fatty acids in Pseudomonas putida S12. Appl Environ Microbiol 62: 2773-2777. Heipieper, J., Weber, F.J., Sikkema, J., Keweloh, H., and de Bont, J.A.M. (1994) Mechanisms of resistance of whole cells to toxic organic solvents. Trends Biotechnol 12: 409-415. Herman, C., Thevenet, D., D'Ari, R., and Bouloc, P. (1995) Degradation of sigma 32, the heat shock regulator in Escherichia coli, is governed by HflB. Proc Natl Acad Sci U S A 92: 3516-3520. Irizarry, R.A., Hobbs, B., Collin, F., Beazer-Barclay, Y.D., Antonellis, K.J., Scherf, U., and Speed, T.P. (2003) Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4: 249-264. Ito, K., and Akiyama, Y. (2005) Cellular functions, mechanism of action, and regulation of FtsH protease. Annu Rev Microbiol 59: 211-231. Kihara, A., Akiyama, Y., and Ito, K. (1995) FtsH is required for proteolytic elimination of uncomplexed forms of SecY, an essential protein translocase subunit. Proc Natl Acad Sci U S A 92: 4532-4536. King, E.O., Ward, M.K., and Raney, D.E. (1954) Two simple media for the demonstration of pyocyanin and fluorescin. J Lab Clin Med 44: 301-307. 97

Kiran, M.D., Annapoorni, S., Suzuki, I., Murata, N., and Shivaji, S. (2005) Cis-trans isomerase gene in psychrophilic Pseudomonas syringae is constitutively expressed during growth and under conditions of temperature and solvent stress. Extremophiles 9: 117-125. Kitten, T., and Willis, D.K. (1996) Suppression of a sensor kinase-dependent phenotype in Pseudomonas syringae by ribosomal proteins L35 and L20. J Bacteriol 178: 1548- 1555. Kojic, M., and Venturi, V. (2001) Regulation of rpoS gene expression in Pseudomonas: involvement of a TetR family regulator. J Bacteriol 183: 3712-3720. Kojic, M., Aguilar, C., and Venturi, V. (2002) TetR family member psrA directly binds the Pseudomonas rpoS and psrA promoters. J Bacteriol 184: 2324-2330. Kojic, M., Jovcic, B., Vindigni, A., Odreman, F., and Venturi, V. (2005) Novel target genes of PsrA transcriptional regulator of Pseudomonas aeruginosa. FEMS Microbiol Lett 246: 175-181. Lamark, T., Kaasen, I., Eshoo, M.W., Falkenberg, P., McDougall, J., and Strom, A.R. (1991) DNA sequence and analysis of the bet genes encoding the osmoregulatory choline- glycine betaine pathway of Escherichia coli. Mol Microbiol 5: 1049-1064. Mecsas, J., Rouviere, P.E., Erickson, J.W., Donohue, T.J., and Gross, C.A. (1993) The activity of sigma E, an Escherichia coli heat-inducible sigma-factor, is modulated by expression of outer membrane proteins. Genes Dev 7: 2618-2628. Miller, J.H. (1972) Experiments in molecular genetics: Cold spring harbor laboratory, NY. Missiakas, D., Betton, J.M., and Raina, S. (1996) New components of protein folding in extracytoplasmic compartments of Escherichia coli SurA, FkpA and Skp/OmpH. Mol Microbiol 21: 871-884. Nomura, M. (1999) Regulation of ribosome biosynthesis in Escherichia coli and Saccharomyces cerevisiae: diversity and common principles. J Bacteriol 181: 6857- 6864. Ogura, T., Inoue, K., Tatsuta, T., Suzaki, T., Karata, K., Young, K., Su, L.H., Fierke, C.A., Jackman, J.E., Raetz, C.R., Coleman, J., Tomoyasu, T., and Matsuzawa, H. (1999) Balanced biosynthesis of major membrane components through regulated degradation of the committed enzyme of lipid A biosynthesis by the AAA protease FtsH (HflB) in Escherichia coli. Mol Microbiol 31: 833-844. Palma, M., DeLuca, D., Worgall, S., and Quadri, L.E. (2004) Transcriptome analysis of the response of Pseudomonas aeruginosa to hydrogen peroxide. J Bacteriol 186: 248- 252. Parnell, J.J., Park, J., Denef, V., Tsoi, T., Hashsham, S., Quensen, J., 3rd, and Tiedje, J.M. (2006) Coping with polychlorinated biphenyl (PCB) toxicity: Physiological and genome-wide responses of Burkholderia xenovorans LB400 to PCB-mediated stress. Appl Environ Microbiol 72: 6607-6614. Pessione, E., Giuffrida, M.G., Prunotto, L., Barello, C., Mazzoli, R., Fortunato, D. et al. (2003) Membrane proteome of Acinetobacter radioresistens S13 during aromatic exposure. Proteomics 3: 1070-1076. Phadtare, S., Kato, I., and Inouye, M. (2002) DNA microarray analysis of the expression profile of Escherichia coli in response to treatment with 4,5-dihydroxy-2- cyclopenten-1-one. J Bacteriol 184: 6725-6729. 98

Preiter, K., Brooks, D.M., Penaloza-Vazquez, A., Sreedharan, A., Bender, C.L., and Kunkel, B.N. (2005) Novel virulence gene of Pseudomonas syringae pv. tomato strain DC3000. J Bacteriol 187: 7805-7814. Raina, S., Missiakas, D., and Georgopoulos, C. (1995) The rpoE gene encoding the sigma E (sigma 24) heat shock sigma factor of Escherichia coli. EMBO J 14: 1043-1055. Rosen, R., Biran, D., Gur, E., Becher, D., Hecker, M., and Ron, E.Z. (2002) Protein aggregation in Escherichia coli: role of proteases. FEMS Microbiol Lett 207: 9-12. Rouviere, P.E., and Gross, C.A. (1996) SurA, a periplasmic protein with peptidyl-prolyl isomerase activity, participates in the assembly of outer membrane porins. Genes Dev 10: 3170-3182. Sambrook, J., Fritsch, E.F., and Maniatis, J. (eds) (1989) Molecular Cloning: a Laboratory Manual. NY, USA: Cold Spring Harbor Laboratory Press. Santos, P.M., Benndorf, D., and Sa-Correia, I. (2004) Insights into Pseudomonas putida KT2440 response to phenol-induced stress by quantitative proteomics. Proteomics 4: 2640-2652. Schnider-Keel, U., Lejbolle, K.B., Baehler, E., Haas, D., and Keel, C. (2001) The sigma factor AlgU (AlgT) controls exopolysaccharide production and tolerance towards desiccation and osmotic stress in the biocontrol agent Pseudomonas fluorescens CHA0. Appl Environ Microbiol 67: 5683-5693. Segura, A., Duque, E., Mosqueda, G., Ramos, J.L., and Junker, F. (1999) Multiple responses of gram-negative bacteria to organic solvents. Environ Microbiol 1: 191-198. Segura, A., Godoy, P., van Dillewijn, P., Hurtado, A., Arroyo, N., Santacruz, S., and Ramos, J.L. (2005) Proteomic analysis reveals the participation of energy- and stress-related proteins in the response of Pseudomonas putida DOT-T1E to toluene. J Bacteriol 187: 5937-5945. Steil, L., Hoffmann, T., Budde, I., Volker, U., and Bremer, E. (2003) Genome-wide transcriptional profiling analysis of adaptation of Bacillus subtilis to high salinity. J Bacteriol 185: 6358-6370. Summers, A.O., and Jacoby, G.A. (1977) Plasmid-determined resistance to tellurium compounds. J Bacteriol 129: 276-281. Tam le, T., Eymann, C., Albrecht, D., Sietmann, R., Schauer, F., Hecker, M., and Antelmann, H. (2006) Differential gene expression in response to phenol and catechol reveals different metabolic activities for the degradation of aromatic compounds in Bacillus subtilis. Environ Microbiol 8: 1408-1427. Tamburro, A., Allocati, N., Masulli, M., Rotilio, D., Di Ilio, C., and Favaloro, B. (2001) Bacterial peptide methionine sulphoxide reductase: co-induction with glutathione S- transferase during chemical stress conditions. Biochem J 360: 675-681. Tartaglia, L.A., Storz, G., and Ames, B.N. (1989) Identification and molecular analysis of oxyR-regulated promoters important for the bacterial adaptation to oxidative stress. J Mol Biol 210: 709-719. Teixeira-Gomes, A.P., Cloeckaert, A., and Zygmunt, M.S. (2000) Characterization of heat, oxidative, and acid stress responses in Brucella melitensis. Infect Immun 68: 2954- 2961. Tomas, C.A., Beamish, J., and Papoutsakis, E.T. (2004) Transcriptional analysis of butanol stress and tolerance in Clostridium acetobutylicum. J Bacteriol 186: 2006-2018. 99

Tomoyasu, T., Yuki, T., Morimura, S., Mori, H., Yamanaka, K., Niki, H., Hiraga, S., and Ogura, T. (1993) The Escherichia coli FtsH protein is a prokaryotic member of a protein family of putative ATPases involved in membrane functions, cell cycle control, and gene expression. J Bacteriol 175: 1344-1351. Tomoyasu, T., Gamer, J., Bukau, B., Kanemori, M., Mori, H., Rutman, A.J., Oppenheim, A.B., Yura, T., Yamanaka, K., Niki, H., Hiraga, S. Ogura, T. (1995) Escherichia coli FtsH is a membrane-bound, ATP-dependent protease which degrades the heat-shock transcription factor sigma 32. EMBO J 14: 2551-2560. Turner, R.J., Weiner, J.H., and Taylor, D.E. (1995) The tellurite-resistance determinants tehAtehB and klaAklaBtelB have different biochemical requirements. Microbiology 141: 3133-3140. van de Mortel, M., and Halverson, L.J. (2004) Cell envelope components contributing to biofilm growth and survival of Pseudomonas putida in low-water-content habitats. Mol Microbiol 52: 735-750. VanBogelen, R.A., Kelley, P.M., and Neidhardt, F.C. (1987) Differential induction of heat shock, SOS, and oxidation stress regulons and accumulation of nucleotides in Escherichia coli. J Bacteriol 169: 26-32. Weber, F.J., and de Bont, J.A. (1996) Adaptation mechanisms of microorganisms to the toxic effects of organic solvents on membranes. Biochim Biophys Acta 1286: 225-245. Welsh, D.T., and Herbert, R.A. (1999) Osmotically induced intracellular trehalose, but not glycine betaine accumulation promotes desiccation tolerance in Escherichia coli. FEMS Microbiol Lett 174: 57-63. Whalen, M.C., Innes, R.W., Bent, A.F., and Staskawicz, B.J. (1991) Identification of Pseudomonas syringae pathogens of Arabidopsis and a bacterial locus determining avirulence on both Arabidopsis and soybean. Plant Cell 3: 49-59. Wright, C.A., and Beattie, G.A. (2004) Pseudomonas syringae pv. tomato cells encounter inhibitory levels of water stress during the hypersensitive response of Arabidopsis thaliana. Proc Natl Acad Sci U S A 101: 3269-3274. Zeller, T., and Klug, G. (2004) Detoxification of hydrogen peroxide and expression of catalase genes in Rhodobacter. Microbiology 150: 3451-3462. Zeller, T., Moskvin, O.V., Li, K., Klug, G., and Gomelsky, M. (2005) Transcriptome and physiological responses to hydrogen peroxide of the facultatively phototrophic bacterium Rhodobacter sphaeroides. J Bacteriol 187: 7232-7242. Zheng, M., Wang, X., Templeton, L.J., Smulski, D.R., LaRossa, R.A., and Storz, G. (2001) DNA microarray-mediated transcriptional profiling of the Escherichia coli response to hydrogen peroxide. J Bacteriol 183: 4562-4570.

100

Chapter 5. Conclusion and future directions

Extensive work has been done on bacterial degradation of organic pollutants in soil but the availability and degradation of airborne pollutants by bacteria present on leaf surfaces has not been explored. Pollutants can be translocated from soil to leaves via the transpiration stream and can also be taken up or adsorbed directly from air by the leaves. Leaf surfaces have also been shown to be colonized by bacterial populations as high as 107-108 cfu/ g of leaf. Therefore, it is possible that bacteria present on the leaf surfaces have access to and degrade the pollutants adsorbed onto the leaf surfaces or translocated to leaves from the soil. In the present study, we demonstrated the availability and degradation of airborne organic pollutants on leaves by leaf-associated bacteria using phenol as a model pollutant.

With the help of phenol-responsive A506 biosensor cells constructed in this study, we showed that bacteria present on the leaf surfaces can access airborne phenol as well as the phenol that had been adsorbed by the leaf surfaces; moreover, the concentrations accessible to bacteria were at least 10-fold higher than those in the air, suggesting that leaves accumulate phenol in sites accessible to bacteria (chap 2). The phenol accessible to leaf- associated bacteria was also available for degradation as evident by the degradation of 14C- phenol by phenol-degrading bacteria, Pseudomonas sp. strain CF600, after inoculation on leaf surfaces. A previous study indirectly suggested a role for bacteria inoculated on leaf surfaces in the removal of airborne toluene by measuring air concentrations of toluene (De Kempeneer et al., 2004); however, our study provided the first direct evidence of pollutant degradation by bacteria present on leaf surfaces by measuring the amount of radiolabeled 14 CO2 released due to bacterial degradation of C-phenol. In this chapter, we also showed that naturally occurring bacterial communities on green ash leaves taken from trees around a swine production facility have the ability to degrade airborne phenol.

We also identified and genetically characterized some members of the phenol- degrading naturally occurring bacterial communities from green ash leaves (chap 3). This 101

characterization is important to assess the genetic diversity, functional diversity, and catabolic potential of these organisms (Watanabe et al., 2002) and to identify the main bacterial species responsible for the majority of the pollutant degradation. This information is also important to developing bioremediation strategies (Watanabe et al., 1998; Watanabe et al., 2002). Molecular characterization of pollutant-degrading isolates may also help in developing genetic markers for monitoring the presence of catabolic genes and their specific catabolic activities in situ. In this study, 10 bacterial isolates were obtained, out of which 7 were Gram-negative and 3 were Gram-positive. The Gram-negative isolates were most closely related to the genera Acinetobacter and Alcaligenes, whereas the three Gram-positive isolates belonged to the genus Rhodococcus. Bacterial species belonging to Acinetobacter, Alcaligenes, and Rhodococcus are metabolically versatile and have been shown to degrade a range of aromatic compounds. It is possible that bacteria belonging to these three genera are capable of efficient phenol degradation at a concentration present in the air around the swine production facility.

The phenol-degrading bacterial isolates were characterized genetically to determine the presence and type of phenol catabolic genes. This was done based on the conserved amino acid residues of the large subunit of multi-component phenol hydroxylases (PH), and the presence of catechol dioxygenase genes. Phenol hydroxylase is the first enzyme of the phenol catabolic pathway in which phenol is converted into catechol, a common intermediate in the aerobic degradation of a variety of aromatic compounds. The catechol is then metabolized into tricarboxylic acid cycle intermediates via either the ortho pathway by catechol-1,2 dioxygenase or the meta pathway by catechol-2,3 dioxygenase. PHs from Gram- negative bacteria have been previously classified into class I, II and III PHs, which have low, moderate and high affinity for phenol, respectively. In our study, we found that the PHs from all of the Gram-negative isolates belonged to class II PHs and all of the Gram-negative isolates possessed the ortho pathway for phenol degradation, suggesting that either the phyllosphere or the bacterial isolation techniques used did not support much genetic diversity.

102

In the present study, we also attempted to understand the mechanism of toxicity of phenol in bacteria by transcriptome analysis of a plant pathogen, Pseudomonas syringae pv. tomato strain DC3000, upon exposure to phenol, membrane stress caused by sodium chloride

(NaCl) and polyethylene glycol (PEG), and oxidative stress caused by H2O2. Our analysis indicated that phenol induces a heat shock response and membrane stress in DC3000, but not much oxidative stress. We proposed a model of cellular responses to phenol in which phenol causes denaturation of membrane, periplasmic and cytoplasmic proteins that result in the activation of the sigma factor AlgU. AlgU then activates the transcription of sigma H, a heat shock gene regulator. This regulator then turns on the transcription of heat shock genes that encode proteases and chaperones to repair or remove the damaged cellular proteins. The gene expression analysis also revealed a number of genes with unknown functions that were upregulated more than 20-fold by phenol. Identification of these genes and their functions would be important to further enhance our understanding of the effects of phenol in a bacterial cell.

Future directions In the present study we were able to identify some bacterial species with the ability to degrade phenol among the naturally occurring phyllosphere communities. The techniques we used did not reveal much phylogenetic and genetic diversity; therefore, it seems more work is required in this area to determine the total diversity of phenol-degrading bacteria, including non-culturable bacterial species, in the phyllosphere. One way to determine the phyllosphere community diversity would be to use culture-independent approaches such as Stable Isotope Probing (SIP) (Radajewski et al., 2000) and Reverse transcriptase-PCR (RT-PCR). SIP involves the use of 13C-labeled substrate to label cellular RNA/DNA, which can then be used to identify bacteria that have the ability to use 13C labeled substrate. This will not only help in determining the non-culturable phenol-degrading bacteria among the phyllosphere community, but also in identifying the ecologically relevant bacterial species in the phyllosphere community. The combination of microautoradiography and fluorescent in situ hybridization (FISH) can also be used to link phylogeny with function of bacterial population in situ (Lee et al., 1999). It is also important to characterize the genetic diversity of bacterial 103

populations involved in pollutant degradation as it can help in identifying the functionally dominant bacterial populations in the environment. This information would help in developing effective bioremediation strategies. A number of previous studies have used catabolic genes to identify bacterial genetic diversity. The differences in catabolic gene sequence is important as it may reflect differences in substrate specificity or affinity (Beil et al., 1998). For example, the phylogenetic diversity of functional catabolic genes such as catechol 2,3-dioxygenase has been determined by amplified functional DNA restriction analysis (AFDRA) (Junca and Pieper, 2003), a technique adapted from amplified ribosomal DNA restriction analysis (ARDRA). AFDRA can be applied to define the predominant polymorphism of a functional gene present in the environmental DNA extracts. Determination of the catabolic gene sequence diversity in environmental samples will significantly increase our knowledge of the functional potential of microbial communities.

Future studies aimed at confirming the action of phenol on cellular proteins are required to understand the mechanism of phenol toxicity in bacteria. We initiated this in the present study by constructing three heat shock DC3000 mutants in genes, htpX, clpA, and lon-2, but we did not see any difference between wild type and mutants in phenol toxicity or tolerance based on the measurement of growth. Construction of double mutants or even examining the presence and amount of protein aggregates in the mutants vs. wild type by labeling the proteins would help in understanding the role of heat shock proteins in phenol tolerance. Additionally, several hypotheses were suggested by the microarray data regarding other effects of phenol, such as changes in the membrane and cross-protection provided by exposure to one stress over the other. The cross-protection studies would yield important information because bacterial cells are not exposed to one stress but various stresses simultaneously in the environment. Identification of the function of hypothetical genes that were highly up-regulated by phenol could also provide insights into the mechanism of action of phenol in a bacterial cell, and therefore, need to be examined in order to generate a complete picture of the effect of phenol on bacteria. This information can then be applied to other bacterial species that have important biotechnological and bioremediation applications.

104

References Beil, S., Mason, J.R., Timmis, K.N., and Pieper, D.H. (1998) Identification of chlorobenzene dioxygenase sequence elements involved in dechlorination of 1,2,4,5- tetrachlorobenzene. J Bacteriol 180: 5520-5528. De Kempeneer, L., Sercu, B., Vanbrabant, W., Van Langenhove, H., and Verstraete, W. (2004) Bioaugmentation of the phyllosphere for the removal of toluene from indoor air. Appl Microbiol Biotechnol 64: 284-288. Junca, H., and Pieper, D.H. (2003) Amplified functional DNA restriction analysis to determine catechol 2,3-dioxygenase gene diversity in soil bacteria. J Microbiol Methods 55: 697-708. Lee, N., Nielsen, P.H., Andreasen, K.H., Juretschko, S., Nielsen, J.L., Schleifer, K.H., and Wagner, M. (1999) Combination of fluorescent in situ hybridization and microautoradiography-a new tool for structure-function analyses in microbial ecology. Appl Environ Microbiol 65: 1289-1297. Radajewski, S., Ineson, P., Parekh, N.R., and Murrell, J.C. (2000) Stable-isotope probing as a tool in microbial ecology. Nature 403: 646-649. Watanabe, K., Futamata, H., and Harayama, S. (2002) Understanding the diversity in catabolic potential of microorganisms for the development of bioremediation strategies. Antonie van Leeuwenhoek 81: 655-663. Watanabe, K., Teramoto, M., Futamata, H., and Harayama, S. (1998) Molecular detection, isolation, and physiological characterization of functionally dominant phenol- degrading bacteria in activated sludge. Appl Environ Microbiol 64: 4396-4402.